Catalysts and Barriers To Increasing The Consumption of Fruits and Vegetables and Reducing The Prevalence of Hypertension

Catalysts and barriers to increasing the consumption of fruits and vegetables and
reducing the prevalence of hypertension

Monitoring metrics, dietary quality, and farmland capacity
PhD Dissertation
Zach Conrad
Agriculture, Food, and Environment Program
Committee Members
Dr. Timothy Griffin (chair)
Dr. Christian Peters
Dr. Kenneth Chui
ProQuest Number: 3720269
All rights reserved
INFORMATION TO ALL USERS

The quality of this reproduction is dependent upon the quality of the copy submitted.
In the unlikely event that the author did not send a complete manuscript
and there are missing pages, these will be noted. Also, if material had to be removed,
a note will indicate the deletion.
ProQuest 3720269
Published by ProQuest LLC (2015). Copyright of the Dissertation is held by the Author.
All rights reserved.

This work is protected against unauthorized copying under Title 17, United States Code
Microform Edition © ProQuest LLC.
ProQuest LLC.
789 East Eisenhower Parkway
P.O. Box 1346
Ann Arbor, MI 48106 - 1346
Table of Contents
List of tables ..................................................................................................................................... i
List of figures ................................................................................................................................... i
Abstract ........................................................................................................................................... ii
Introduction ......................................................................................................................................1
Different fruit and vegetable data collection programs for U.S. adults do not produce the same
results .............................................................................................................................................20
Abstract ......................................................................................................................................22
Introduction ................................................................................................................................23
Methods ......................................................................................................................................25
Results ........................................................................................................................................32
Discussion ..................................................................................................................................35
Nutrient dense fruits and vegetables can help hypertensive adults meet the nutrient targets of the
Dietary Approaches to Stop Hypertension (DASH) diet with fewer servings than is
recommended .................................................................................................................................44
Abstract ......................................................................................................................................46
Introduction ................................................................................................................................47
Methods ......................................................................................................................................49
Results ........................................................................................................................................58
Discussion ..................................................................................................................................66
Agricultural capacity to increase the production of selected nutrient dense fruits and vegetables
for treating hypertension ................................................................................................................73
Abstract ......................................................................................................................................73
Introduction ................................................................................................................................75
Methods ......................................................................................................................................79
Results ........................................................................................................................................84
Discussion ..................................................................................................................................93
Conclusion ...................................................................................................................................106
Appendices ...................................................................................................................................112
i
List of tables
Daily per capita consumption of fruit and vegetable groups, 2007-2010 ......................................35
Daily per capita consumption of fruit and vegetable groups, 2011 ...............................................36
Number of servings and vegetable subgroup proportions of the current diet, USDA Food Pattern,
and DASH diet ...............................................................................................................................58
Nutrient intake among hypertensive adults compared to DASH diet nutrient targets, NHANES
2009-2010 ......................................................................................................................................60
Fruits and vegetables included in Scenario diets ...........................................................................62
Nutrient contribution of Scenario diet series CC (all F&V) compared to current intake ..............63
Nutrient contribution of Scenario diet series CC (consumer preferred F&V) compared to current
intake ..............................................................................................................................................63
Nutrient contribution of Scenario diets, all F&V...........................................................................64
Nutrient contribution of Scenario diets, consumer preferred F&V ...............................................66
National production centers of nutrient dense fruits and vegetables .............................................86
Current agricultural land area and potential agricultural land area suitable for the production of
nutrient dense fruits and vegetables ...............................................................................................92
Current and potential per capita availability of nutrient dense fruits and vegetables for
hypertensive adults.........................................................................................................................93
List of figures
Linkage between nutrition, public health, and agriculture ..............................................................3
Bland-Altman comparison between NHANES and LAFA, daily per capita consumption of fruits
and vegetables, 2007-2010.............................................................................................................34
Attributes of Scenario diets ............................................................................................................56
Suitable agricultural land for dates in Riverside County, California .............................................87
Suitable agricultural land for kiwis in a) Tulare County, California and b) Butte County,
California .......................................................................................................................................87
Suitable agricultural land for broccoli in a) Monterey County, California and b) Santa Barbara
County, California..........................................................................................................................87
Suitable agricultural land for lima beans in Sussex and Ken Counties, Delaware ........................89
Suitable agricultural land for sweet potatoes in a) Johnston, Nash, Sampson, and Wilson
Counties, North Carolina, b) Calhoun County, Mississippi, and c) Merced County, California ..90
Suitable agricultural land for Great Northern beans in Scotts Bluff, Box Butte, and Morrill
Counties, Nebraska ........................................................................................................................91
ii
Abstract
Hypertension is a prominent risk factor for cardiovascular disease (CVD), which is the leading
cause of death in the United States (US). Adequate consumption of fruits and vegetables (F&V)
is associated with a reduced risk of developing hypertension and CVD, yet most Americans do
not meet the recommended daily consumption amounts for F&V. National public health goals
for 2020 include reducing the prevalence of hypertension and increasing the consumption of
F&V, but several barriers stand in the way of achieving these goals. Three different dietary data
collection programs are used to inform nutrition policies, interventions, and messages, but the
degree of agreement among the data generated by these programs has not been examined.
Additionally, although the Dietary Approaches to Stop Hypertension (DASH) diet – which
includes high amounts of F&V – is associated with a reduced risk of hypertension, CVD, and all-
cause mortality, the rate of adherence continues to be low. Finally, most F&V consumed in the
US are produced domestically, but the capacity of the US agricultural system to accommodate
increased consumption of nutrient dense F&V is not well documented. The goal of this
dissertation is to examine several catalysts and barriers to achieving the national public health
nutrition goals of increasing the consumption of F&V and reducing the prevalence of
hypertension; this will be accomplished through three separate studies. In the first study, I
estimate the comparability of daily per capita F&V consumption data generated from the
National Health and Nutrition Examination Survey (NHANES), 2007-2010; the Behavioral Risk
Factor Surveillance System (BRFSS), 2007-2011; and the Loss-Adjusted Food Availability
(LAFA) data series, 2007-2011; using statistical agreement tests and descriptive means
comparisons. In the second study, I develop a food pattern model to examine whether consuming
nutrient dense F&V, instead of those currently consumed, without increasing the amount
consumed, will provide enough fiber, calcium, magnesium, and potassium for hypertensive
adults to meet the nutrient targets of the DASH diet. In the third study, I create geospatial models
to investigate the degree to which the US agricultural system can increase the production of
selected nutrient dense F&V in eleven major production centers. The results of these studies are
as follows. In the first study, I found that NHANES, BRFSS, and LAFA do not produce the same
results and should not be used interchangeably to estimate daily per capita F&V consumption at
the national level. In the second study, I found that consuming nutrient dense F&V instead of
those currently consumed, without increasing the amount consumed, will not provide enough
nutrients for hypertensive adults to meet the nutrient targets of the DASH diet; yet by choosing
more nutrient dense F&V and increasing the number of servings consumed, hypertensive adults
can meet the nutrient targets with fewer servings than is recommended. In the third study, I
found that there is enough suitable agricultural land within the major production centers to
accommodate a substantial increase in demand for selected nutrient dense F&V, but that further
increases in the per capita availability of F&V are needed for hypertensive adults to meet the
DASH diet. This research is important for several reasons. It shows that matching the research
question to the appropriate data source is essential to computing accurate estimates of daily per
capita consumption of F&V. This research also provides clinicians with information that can be
used to help their hypertensive patients adopt more healthy dietary patterns. Finally, this research
highlights the important role that smaller production centers and international food markets will
play if hypertensive adults adopt healthier eating patterns.
1
Introduction
The linkage between public health nutrition and agriculture
The relationship between dietary quality and chronic disease outcomes is driven by dietary
patterns (Figure 1),1, 2 which can be defined as ―the quantities, proportions, variety, or
combination of different foods, drinks, and nutrients in diets, and the frequency with which they
are habitually consumed‖.3 Agricultural production is a response to the dietary patterns of large
populations because, at its most basic level, the purpose of an agricultural system is to produce
the types and amounts of food that are demanded by a given population. While food production
on a large scale can enhance ecosystem services, it also has the potential to deplete or degrade
natural resources, threaten environmental sustainability, and reduce yields, thereby limiting the
agricultural capacity to accommodate dietary patterns. There are also limits to how much, and
how quickly, agriculture can adapt to shifting food demands. Every crop has its own growing
conditions that need to be met in order to thrive, so the availability of suitable cropland can limit
the availability of certain foods for consumers.

2
Dietary data collection

programs
Nutrition policies and

interventions
Agricultural Dietary patterns Agricultural

production capacity
Dietary quality
Environmental
sustainability
Disease outcomes
Figure 1: Linkage between nutrition, public health, and

agriculture
Dietary patterns are also affected by nutrition policies, interventions, and messages, such as the
Dietary Guidelines for Americans,4 USDA MyPlate,5 Dietary Approaches to Stop Hypertension
(DASH) diet,6 and the Fruit & Veggies – More Matters7 campaign. These efforts are informed by
dietary data collection programs, which are used to estimate food consumption of population
groups and to track progress toward achieving public health goals.
The typical American diet is of low quality8 and people are continually urged to change the way
they eat. Yet shifts in dietary patterns depends on the integrity of the dietary data collection
programs that are used to craft nutrition policies, interventions, and messages, and also on the
agricultural capacity to accommodate more healthy dietary patterns. In this dissertation, I
examine the catalysts and barriers to achieving the national public health nutrition goals of
increasing the consumption of F&V9 and reducing the prevalence of hypertension,10
through three distinct studies. The focus is on fruits and vegetables (F&V) because different
3
methods to track progress toward achieving national consumption goals may result in
inconsistent policies, interventions, and messages; the negative health outcomes associated with
under-consuming F&V is a national public health concern; and it is unknown whether achieving
national public health goals related to F&V consumption is limited by the availability of suitable
agricultural land in the US.
In chapter 1, three prominent dietary data collection programs are used to measure F&V
consumption at the national level, and their results are compared with one another using a
combination of statistical methods. These findings will allow us to determine whether it is
appropriate to use these programs interchangeably to track progress toward achieving national
public health nutrition goals. This chapter is titled Different fruit and vegetable data collection
programs for U.S. adults do not produce the same results, and has been formatted to the
specifications of Nutrition Research. In chapter 2, F&V consumption and nutrient intake are
estimated for hypertensive adults, and we examine whether diets that include more nutrient dense
F&V can help this population achieve the nutrient targets associated with normalized blood
pressure. These results will allow clinicians to offer more targeted dietary guidance for their
patients, which could increase nutrient intake and reduce the prevalence of hypertension. This
chapter is titled Nutrient dense fruits and vegetables can help hypertensive adults meet the
nutrient targets of the Dietary Approaches to Stop Hypertension (DASH) diet with fewer servings
than is recommended, and is formatted according to the specifications of Nutrition Research. In
chapter 3, we examine whether there is enough suitable agricultural land in the US to increase
the production of nutrient dense F&V for hypertensive adults. These results will help us to
understand whether agricultural production capacity limits the achievement of national public
4
health nutrition goals relating to F&V consumption and hypertension prevalence. This chapter is
titled Agricultural capacity to increase the production of nutrient dense fruits and vegetables for
treating hypertension, and is formatted according to the specifications of Renewable Agriculture
and Food Systems. Before presenting each of these chapters in turn, I provide some background
on the linkage between public health nutrition and agriculture as it relates specifically to F&V.
Monitoring consumption of fruits and vegetables
It is well established that regular consumption of F&V is associated with a reduced risk of
chronic disease.11, 12 There is particularly strong evidence for a risk reduction of cardiovascular
disease (CVD),11-13 which is the leading cause of death in the United States.14 Wang et al.
(2014)15 found that each serving of F&V consumed per day offered protection against all-cause
mortality, up to two servings of fruit and three servings of vegetables. This is approximately the
same amount that is recommended by the Dietary Guidelines for Americans, yet most people do
not meet these recommendations. Accordingly, national public health nutrition goals for 2020
call for increasing daily F&V consumption to two cup equivalents per 1,000kcal.9 Continuous
monitoring of F&V consumption at the national level is critical for assessing progress towards
this goal, and is used to inform public health messages and dietary interventions.
The first documented attempt to collect dietary data for public health purposes was in 1902 in
Winchester, England, in response to a localized outbreak of enteric fever.16 It was a simple
survey in which guests who had attended a dinner banquet were asked which foods they
consumed from the dinner menu. Today, dietary data is collected as part of carefully planned and
coordinated programs that are designed to be generalizable to large populations. At the national
5
level, F&V consumption is predominantly monitored using the National Health and Nutrition
Examination Survey (NHANES), 17-23 the Behavioral Risk Factor Surveillance System
(BRFSS),24-31 and the Loss-Adjusted Food Availability (LAFA) data series.32-39 NHANES
measures the amount of individual foods consumed in their whole (e.g. tomato), processed (e.g.
tomato juice), and formulated (e.g. tomatoes in pasta sauce) forms; BRFSS measures the
frequency of consumption of F&V subgroups (e.g. dark green, orange, starchy); and LAFA
measures the per capita availability of whole and processed forms of individual foods.
However, there is some indication that these programs may not be producing the same results.
For example, the proportion of adults meeting daily F&V consumption recommendations has
been observed at 13-20% using NHANES data23 and >25% using BRFSS data29. Using LAFA, it
was estimated that mean consumption was 50-70% of the recommended quantity33. These
differences could result in inconsistent public health messaging and dietary interventions, so
testing the comparability of these methods is critical to ensuring consistency among these efforts.
There are no studies that have directly compared F&V consumption estimates across all of these
programs. Studies that have compared nutrient intake40 and fresh F&V consumption41 between
NHANES and the Food Availability data series (which differs from LAFA in that it does not
include adjustments for food losses) observed similarities based on descriptive means
comparison, yet agreement tests are needed to directly compare these programs.
Agreement tests are useful when comparing methods of measurement because they allow for the
quantification of bias, and several methods are well suited for this purpose. The Bland-Altman
method, which was introduced in 198342 and subsequently refined43-45, is the most commonly
6
used method in nutrition studies46. Ordinary least products (OLP) regression is less commonly
used, more complex to execute, and more difficult to interpret47-49, but its utility in detecting bias
may be superior to the Bland-Altman method50. Luiz and Szklo (2005)51 remind us that all
approaches for agreement testing have limitations, and they and others50 suggest that method
comparison studies utilize several approaches to compensate for the limitations of any one
approach. This recommendation has been heeded here.
Fruits, vegetables, and hypertension
The inverse relationship between F&V consumption and CVD risk is partly due to the reduced
risk for hypertension,12 since this is a prominent risk factor for CVD.52, 53 Hypertension is a
condition characterized by systolic blood pressure (SBP) ≥140 mm Hg or diastolic blood
pressure (DBP) ≥90 mm Hg.54 SBP is the force exerted on the arterial walls when the heart beats,
and DBP is the force between the beats. Uncontrolled hypertension can cause small arterial tears
that form scar tissue, and over time this scar tissue collects circulating materials in the blood,
which leads to the development of plaque. As plaque accumulates it stiffens and narrows the
arteries, making it more difficult for blood to reach the heart, causing it to die and leading to
heart failure.55 Approximately 30% of US adults are hypertensive and an additional ~25% of
adults are pre-hypertensive (SBP ≥120mmHg or DBP ≥80mmHg).56 National public health goals
for 2020 include reducing the prevalence of hypertension.10
Hypertension is one of the most modifiable risk factors for CVD, and lifestyle modifications are
recommended for all hypertensive individuals regardless of pharmacotherapy treatment.57 Key
among these lifestyle changes is the adoption of the Dietary Approaches to Stop Hypertension
7
(DASH) diet, which emphasizes fruits, vegetables, whole grains, and low fat dairy.58 The DASH
diet is designed to meet specific daily nutrient targets by increasing the intake of some nutrients
(fiber, calcium, potassium, magnesium) and limiting the intake of others (total fat, saturated fat,
sodium).59 The DASH diet has been consistently associated with normalized blood pressure60-62
and reduced risk of CVD and all-cause mortality.63-66 However, adhering to the DASH diet
requires most people modify their consumption of nearly every food group, and adherence rates
are consistently low.67, 68
Clinical interventions are projected to be one of the most effective means of reducing CVD-
related mortality 69. Clinicians play an important role in their patients’ adoption of hypertension
treatment regimens 70 partly because they are the first to detect the condition, as hypertension is
the most common primary diagnosis by physicians 71. In addition, patients who report receiving
lifestyle modification counseling by their primary care physician are much more likely to modify
their diet than those who report not receiving such counseling 72. However, improved clinical
strategies are needed to increase the rate of adherence to the DASH diet. Developing strategies
for clinicians to help adults adopt CVD related diet recommendations has been identified as a
critical research need by the National Heart, Lung, and Blood Institute 73.
According to some behavior change specialists, people are most likely to adopt simple, well-
defined health and nutrition recommendations.74 There is a growing body of research that shows
that recommendations to make modest lifestyle changes can be readily adopted and maintained,
and can result in positive health outcomes.75 Individuals with hypertension who have difficulty
modifying their consumption of all food groups may still benefit from modifying their
8
consumption of one or several food groups. A focus on fruits and vegetables (F&V) is warranted
given that these foods contain high amounts of nutrients to encourage (fiber, calcium,
magnesium, and potassium) and low amounts of nutrients to limit (total fat, saturated fat,
sodium). 11, 76, 77 If clinicians provided recommendations to their hypertensive patients to increase
their consumption of nutrient dense F&V, these individuals may be able to increase their nutrient
intake and meet the DASH diet nutrient targets for fiber, Ca, Mg, and K without increasing the
amount of F&V currently consumed.
Fruits, vegetables, and agricultural capacity
Given that the majority of the fruits and nuts (~60%) and vegetables (~80%) consumed in the US
(by volume) are produced domestically,78 if hypertensive individuals increased their
consumption of nutrient dense F&V, domestic production of these foods would need to increase
in response. Although recent increases in domestic F&V production are largely attributable to
technological innovations (e.g. improved disease and pest resistance and improved fruit set) and
the adoption of intensive management practices (e.g. high density planting and global positioning
systems), the viability of these approaches is dependent on the availability of suitable agricultural
land. In fact, in some cases yield improvements were the result of relocation to more productive
cropland.79, 80
Although the land use requirements for F&V are small compared to grains and oilseeds,36, 81
many F&V require highly specific growing conditions which can limit the areas in which they
can thrive. Therefore, identifying viable agricultural sites is of primary importance to increasing
production. In order to do this, a methodology known as site suitability analysis is required, and
9
Geographic information systems (GIS) software is the preferred tool for this purpose.82 Site
suitability analysis involves identifying the factors that are required for adequate crop growth,
and then operationalizing these factors so that they can be parameterized in a GIS. A GIS is a
tool that is used to process spatial and geographic data.
The connection between F&V demand and agricultural land use has been examined at a variety
of spatial scales, including sub-state regions (Willamette Valley, Oregon),83 major metropolitan
areas (Philadelphia84 and Detroit85), individual states (New York),86 and nationally.36, 37 Half of
these studies84-86 used a GIS, yet of these only Peters et al.86 used a site suitability analysis. None
of these studies investigated the connection between changes in food demand and agricultural
land use for specific F&V. In this dissertation, I create a crop-specific GIS-based site suitability
model that is parameterized for precipitation, temperature, soil depth, soil type, and agricultural
land availability.
Conclusion
Public health nutrition and agriculture are inherently linked, and this is particularly true for F&V.
Testing the agreement between different dietary data collection programs for F&V is essential
for ensuring consistent nutrition policies, interventions, and messages, which will influence the
adoption of healthy dietary patterns. Examining the effect of small changes to dietary patterns on
daily nutrient intake will allow health care practitioners to provide more targeted dietary advice
for their hypertensive patients, which could increase the prevalence of individuals who adopt
healthy dietary patterns. Investigating the domestic agricultural capacity to accommodate more
healthy dietary patterns can be used to integrate farm policy, international food trade agreements,
10
and public health recommendations. The research presented in this dissertation explicitly bridges
the disciplines of public health nutrition and agriculture to provide a detailed and comprehensive
approach to improving the health of the American population.
References
1. Jacobs, DR and Orlich, MJ. 2014. Diet pattern and longevity: do simple rules suffice? A
commentary. The American Journal of Clinical Nutrition 100: 313S-9S.
2. Jacobs, DR and Steffen, LM. 2003. Nutrients, foods, and dietary patterns as exposures in
research: a framework for food synergy. The American Journal of Clinical Nutrition 78: 508S-
13S.
3. US Departmet of Agriculture, Center for Nutrition Policy and Promotion. 2014. A Series of
Systematic Reviews on the Relationship Between Dietary Patterns and Health Outcomes.
Evidence Analysis Library Division. Available at website: www.nel.gov/publications#DP
(accessed 04 June 2015).
4. US Department of Health and Human Services and US Department of Agriculture. Dietary

Guidelines for Americans. US Government Printing Office, Washington, DC. Available at
website: www.health.gov/dietaryguidelines/ (verified 05 June 2015).
5. USDA Center for Nutrition Policy and Promotion (CNPP). 2015. ChooseMyPlate.gov.
Available at website: http://www.choosemyplate.gov/food-groups/ (verified 18 Mar. 2015).
6. US Department of Health and Human Services. 2006. Your Guide to Lowering your Blood
Pressure with DASH. National Heart, Lung, and Blood Institute. NIH Publication No. 06-4082.
Available at website: www.nhlbi.nih.gov/health/health-topics/topics/dash/ (verified 16 Mar.
2015).
7. Produce for Better Health Foundation. 2015. Fruits and Veggies: More Matters. Available at
website: www.fruitsandveggiesmorematters.org/ (verified 16 Mar. 2015).
8. Miller, PE; Reedy, J; Kirkpatrick, SI, et al. 2014. The United States food supply is not
consistent with dietary guidance: evidence from an evaluation using the Healthy Eating Index-
2010. J Acad Nutr Dietetics 115: 95-100.
11
9. US Department of Health and Human Services. 2010. Healthy People 2020: Understanding
and Improving Health. Nutrition and Weight Status Objectives 14 and 15. 3rd ed. Available at
website: www.healthypeople.gov/2020/topicsobjectives2020/default.aspx (verified 16 Mar.
2015).
10. US Department of Health and Human Services. 2010. Healthy People 2020: Understanding
and Improving Health. Heart Disease and Stroke Objective 5. 3rd ed. Available at website:
www.healthypeople.gov/2020/topicsobjectives2020/default.aspx (verified 17 Mar. 2015).
11. Woodside, JV; Young, IS and McKinley, MC. 2013. Fruits and vegetables: Measuring
intake and encouraging increased consumption. P Nutr Soc 72: 236-45.
12. Boeing, H; Bechthold, A; Bub, A, et al. 2012. Critical review: Vegetables and fruit in the
prevention of chronic diseases. European Journal of Nutrition 51: 637-63.
13. Dietary Guidelines Advosiry Committee. 2015. Scientific Report of the 2015 Dietary
Guidelines Advisory Committee: Advisory Report to the Secretary of Health and Human
Services and the Secretary of Agriculture. Part D, Chapter 2. U.S. Government Printing Office,
Washington, DC. Available at website: www.health.gov/dietaryguidelines/2015-scientific-report/
(verified 16 Mar. 2015).
14. US Department of Health and Human Services, Centers for Disease Control and Prevention.
2012. Ten Leading Causes of Death and Injury. Available at website:
www.cdc.gov/injury/wisqars/leadingcauses.html (verified 17 Mar. 2015).
15. Wang, X; Ouyang, Y; Liu, J, et al. 2014. Fruit and vegetable consumption and mortality
from all causes, cardiovascular disease, and cancer: Systematic review and dose-response meta-
analysis of prospective cohort studies. Brit Med J 349: g4490.
16. Morabia, A and Hardy, A. 2005. The pioneering use of a questionnaire to investigate a food
borne disease outbreak in early 20th century Britain. Journal of Epidemiology and Community
Health 59: 94-9.
17. US Department of Health and Human Services, Centers for Disease Control and Prevention
(CDC). About the National Health and Nutrition Examination Survey. Available at website:
www.cdc.gov/nchs/nhanes.htm (verified 25 Mar. 2013).
12
18. Hajjar, I and Kotchen, T. 2003. Regional variations of blood pressure in the United States
are associated with regional variations in dietary intakes: the NHANES-III data. Journal of
Nutrition 133: 211-4.
19. Kant, AK; Graubard, BI and Kumanyika, SK. 2007. Trends in Black–White differentials in
dietary intakes of U.S. adults, 1971–2002. American Journal of Preventive Medicine 32: 264-
72.e1.
20. Casagrande, SS; Wang, Y; Anderson, C, et al. 2007. Have Americans increased their fruit
and vegetable intake?: the trends between 1988 and 2002. American Journal of Preventive
Medicine 32: 257-63.
21. Kimmons, J; Gillespie, C; Seymour, J, et al. 2009. Fruit and vegetable intake among
adolescents and adults in the United States: percentage meeting individualized recommendations.
Medscape J Med 11:
22. Demydas, T. 2011. Consumer segmentation based on the level and structure of fruit and
vegetable intake: an empirical evidence for US adults from the National Health and Nutrition
Examination Survey (NHANES) 2005–2006. Public Health Nutr 14: 1088-95.
23. Krebs-Smith, SM; Guenther, PM; Subar, AF, et al. 2010. Americans do not meet federal
dietary recommendations. Journal of Nutrition 140: 1832-8.
24. US Department of Health and Human Services, Centers for Disease Control and Prevention
(CDC). Behavioral Risk Factor Surveillance System. Available at website: www.cdc.gov/brfss/
25. Serdula, MK; Coates, RJ; Byers, T, et al. 1995. Fruit and vegetable intake among adults in
16 States: Results of a brief telephone survey. American Journal of Public Health 85: 236-9.
26. Li, R; Serdula, M; Bland, S, et al. 2000. Trends in fruit and vegetable consumption among
adults in 16 US states: Behavioral Risk Factor Surveillance System, 1990-1996. American
Journal of Public Health 90: 777-81.
27. Serdula, MK; Gillespie, C; Kettel-Khan, L, et al. 2004. Trends in fruit and vegetable
consumption among adults in the United States: Behavioral Risk Factor Surveillance System,
1994-2000. American Journal of Public Health 94: 1014-8.
13
28. Blanck, HM; Gillespie, C; Kimmons, J, et al. 2008. Trends in fruit and vegetable
consumption among U.S. men and women, 1994-2005. Prev Chronic Dis 5: 1-10.
29. Grimm, KA; Blanck, HM; Scanlon, KS, et al. 2010. State-specific trends in fruit and
vegetable consumption among adults - United States, 2000-2009. MMWR CDC Surv Summ 59:
1125-30.
30. Troost, JP; Rafferty, AP; Luo, Z, et al. 2012. Temporal and regional trends in the
prevalence of healthy lifestyle characteristics: United States, 1994–2007. American Journal of
Public Health 102: 1392-8.
31. Lutfiyya, MN; Chang, LF and Lipsky, MS. 2012. A cross-sectional study of US rural adults'
consumption of fruits and vegetables: do they consume at least five servings daily? BMC Public
Health 12: 280.
32. US Department of Agriculture, Economic Research Service (ERS). Loss-Adjusted Food

Availability (LAFA) data series. Available at website: www.ers.usda.gov/data-products/food-
availability-(per-capita)-data-system/loss-adjusted-food-availability-documentation.aspx
33. Wells, HF and Buzby, JC. 2008. Dietary Assessment of Major Trends in US Food
Consumption, 1970-2005. USDA Economic Research Service (ERS). Economic Information
Bulletin Number 33. US Government Printing Office, Washington, DC. Available at website:
www.ers.usda.gov/publications.aspx#.Ucmq0zTVDDY (verified 25 June 2013).
34. Krebs-Smith, SM; Reedy, J and Bosire, C. 2010. Healthfulness of the US food supply:
Little improvement despite decades of dietary guidance. American Journal of Preventive
Medicine 38: 472-7.
35. Kantor, LS. 1998. A Dietary Assessment of the US Food Supply: Comparing Per Capita
Consumption with Food Guide Pyramid Serving Recommendations. In E. Frazao (ed.).
America's Eating Habits: Changes and Consequences. USDA Economic Research Service
(ERS). Agriculture Information Bulletin No. 750. US Government Printing Office, Washington,
DC. Available at Website: http://www.ers.usda.gov/Publications/AIB750/ (verified 25 June
2013).
36. Buzby, JC; Wells, HF and Vocke, G. 2006. Possible Implications for US Agriculture from
Adoption of Select Dietary Guidelines. USDA Economic Research Service (ERS). Economic
Research Report Number 31. US Government Printing Office, Washington, DC. Available at
website: www.ers.usda.gov/Publications/ERR31/ (verified 17 Apr. 2012).
14
37. Young, CE and Kantor, LS. 1999. Moving toward the Food Guide Pyramid: Implications
for US agriculture. In E. Frazao (ed.). America's Eating Habits: Changes and Consequences.
USDA Economic Research Service (ERS). Agriculture Information Bulletin No. 750. US
Government Printing Office, Washington, DC. 403-23. Available at Website:
http://www.ers.usda.gov/Publications/AIB750/ (verified 17 April 2012).
38. Buzby, JC; Wells, HF; Kumcu, A, et al. 2010. Canned Fruit and Vegetable Consumption in
the United States: An Updated Report to Congress. USDA Economic Research Service (ERS).
US Government Printing Office, Washington, DC. Available at website:
http://www.ers.usda.gov/publications.aspx#.Ucmq0zTVDDY (verified 25 June 2013).
39. Reedy, J; Krebs-Smith, SM and Bosire, C. 2010. Evaluating the food environment:
application of the healthy eating index-2005. American Journal of Preventive Medicine 38: 465-
71.
40. Ahuja, JK; Juan, WY; Egan, K, et al. 2013. Federal monitoring activities related to food and
nutrition: How do they compare? Procedia Food Science 2: 165-71.
41. Hoelzer, K; Pouillot, R; Egan, K, et al. 2012. Produce consumption in the United States: An
analysis of consumption frequencies, serving sizes, processing forms, and high-consuming
population subgroups for microbial risk assessments. J Food Protect 75: 328-40.
42. Altman, DG and Bland, JM. 1983. Measurement in medicine: The analysis of method
comparison studies. J Roy Stat Soc D-Sta 32: 307-17.
43. Bland, MJ and Altman, DG. 1986. Statistical methods for assessing agreement between two
methods of clinical measurement. Lancet 327: 307-10.
44. Bland, MJ and Altman, DG. 1999. Measuring agreement in method comparison studies.
Statistical Methods in Medical Research 8: 135-60.
45. Bland, MJ and Altman, DG. 2003. Applying the right statistics: Analyses of measurement
studies. Ultrasound Obst Gyn 22: 85-93.
46. Zaki, R; Bulgiba, A; Ismail, R, et al. 2012. Statistical methods used to test for agreement of
medical instruments measuring continuous variables in method comparison studies: A systematic
review. PloS one 7: 1.
15
47. Ludbrook, J. 1997. Comparing methods of measurement. Clin and Exp Pharmacol P 24:
193-203.
48. Ludbrook, J. 2002. Statistical techniques for comparing measures and methods of
measurement: A critical review. Clin Exp Pharmacol P 29: 527-36.
49. Ludbrook, J. 2010. Linear regression analysis for comparing two measurers or methods of
measurement: But which regression? Clin Exp Pharmacol P 37: 692-9.
50. Zaki, R; Bulgiba, A and Ismail, NA. 2013. Testing the agreement of medical instruments:
Overestimation of bias in the Bland–Altman analysis. Preventive Medicine 57: S80-S2.
51. Luiz, RR and Szklo, M. 2005. More than one statistical strategy to assess agreement of
quantitative measurements may usefully be reported. Journal of Clinical Epidemiology 58: 215-
6.
52. Lloyd-Jones, D; Adams, RJ; Brown, TM, et al. 2010. Heart disease and stroke statistics-
2010 update: A report from the American Heart Association. Circulation 121: e46-e215.
53. Lewington, S; Clarke, R; Qizilbash, N, et al. 2002. Age-specific relevance of usual blood
pressure to vascular mortality: A meta-analysis of individual data for one million adults in 61
prospective studies. Lancet 360: 1903-13.
54. Crim, MT; Yoon, SS; Ortiz, E, et al. 2012. National surveillance definitions for
hypertension prevalence and control among adults. Circulation: Cardiovascular Quality and
Outcomes 5: 343-51.
55. American Heart Association. High Blood Pressure or Hypertension. Available at website:
www.heart.org/HEARTORG/Conditions/HighBloodPressure/High-Blood-Pressure-or-
Hypertension_UCM_002020_SubHomePage.jsp (accessed 07 April 2015).
56. Guo, F; He, D; Zhang, W, et al. 2012. Trends in prevalence, awareness, management, and
control of hypertension among United States adults, 1999 to 2010. Journal of the American
College of Cardiology 60: 599-606.
57. James, PA; Oparil, S; Carter, BL, et al. 2014. 2014 evidence-based guideline for the
management of high blood pressure in adults: Report from the panel members appointed to the
eighth joint national committee (JNC 8). JAMA 311: 507-20.
16
58. Karanja, NM; Obarzanek, E; Pao-Hwa, L, et al. 1999. Descriptive characteristics of the
dietary patterns used in the Dietary Approaches to Stop Hypertension trial. American Dietetic
Association. Journal of the American Dietetic Association 99: S19-27.
59. Vogt, TM; Appel, LJ; Obarzanek, EVA, et al. 1999. Dietary Approaches to Stop
Hypertension: Rationale, Design, and Methods. Journal of the American Dietetic Association 99:
S12-S8.
60. Appel, LJ; Moore, TJ; Obarzanek, E, et al. 1997. A clinical trial of the effects of dietary
patterns on blood pressure. New England Journal of Medicine 336: 1117-24.
61. Hikmat, F and Appel, LJ. 2014. Effects of the DASH diet on blood pressure in patients with
and without metabolic syndrome: Results from the DASH trial. Journal of Human Hypertension
28: 170-5.
62. Saneei, P; Salehi-Abargouei, A; Esmaillzadeh, A, et al. 2014. Influence of Dietary

Approaches to Stop Hypertension (DASH) diet on blood pressure: A systematic review and
meta-analysis on randomized controlled trials. Nutrition, Metabolism and Cardiovascular
Diseases 24: 1253-61.
63. Reedy, J; Krebs-Smith, SM; Miller, PE, et al. 2014. Higher Diet Quality Is Associated with
Decreased Risk of All-Cause, Cardiovascular Disease, and Cancer Mortality among Older
Adults. The Journal of Nutrition 144: 881-9.
64. Salehi-Abargouei, A; Maghsoudi, Z; Shirani, F, et al. 2013. Effects of Dietary Approaches

to Stop Hypertension (DASH)-style diet on fatal or nonfatal cardiovascular diseases—Incidence:
A systematic review and meta-analysis on observational prospective studies. Nutrition 29: 611-8.
65. Siervo, M; Lara, J; Chowdhury, S, et al. 2015. Effects of the Dietary Approach to Stop
Hypertension (DASH) diet on cardiovascular risk factors: a systematic review and meta-analysis.
British Journal of Nutrition 113: 1-15.
66. Liese, AD; Krebs-Smith, SM; Subar, AF, et al. 2015. The Dietary Patterns Methods Project:
Synthesis of Findings across Cohorts and Relevance to Dietary Guidance. Journal of Nutrition
145: 393-402.
67. Kwan, MW-M; Wong, MC-S; Wang, HH-X, et al. 2013. Compliance with the Dietary
Approaches to Stop Hypertension (DASH) Diet: A Systematic Review. PloS one 8: 78412.
17
68. Mellen, PB; Gao, SK; Vitolins, MZ, et al. 2008. Deteriorating dietary habits among adults
with hypertension: DASH dietary accordance, NHANES 1988-1994 and 1999-2004. Archives of
Internal Medicine 168: 308-14.
69. Hirsch, G; Homer, J; Trogdon, J, et al. 2014. Using simulation to compare 4 categories of
intervention for reducing cardiovascular disease risks. American Journal of Public Health 104:
1187-95.
70. US Department of Health and Human Services, National Institutes of Health. 2004. The
Seventh Report of the Joint National Committee on Prevention, Detection, Evaluation, and
Treatment of High Blood Pressure. National Heart, Lung, and Blood Institute. NIH Publication
No. 04-5230. Available at website: www.nhlbi.nih.gov/health-pro/guidelines (verified 16 Dec
2014).
71. Hsiao, C-J; Cherry, DK; Beatty, PC, et al. 2010. National Ambulatory Medical Care
Survey: 2007 Summary. National Health Statistics Reports. US Department of Health and
Human Services, Centers for Disease Control and Prevention (CDC). Number 27. Available at
website: http://www.cdc.gov/nchs/products/nhsr.htm (verified 12 Dec 2014).
72. Persoskie, A; Kaufman, AR and Leyva, B. 2014. Receiving and adhering to lifestyle
modification counseling for hypertension: Disparities between smokers and nonsmokers. The
Journal of Clinical Hypertension 16: 429-36.
73. US Department of Health and Human Services, National Institutes of Health. 2013.
Lifestyle Interventions to Reduce Cardiovascular Disease: Systematic Evidence Review From
the Lifestyle Work Group, 2013, p.45. National Heart, Lung, and Blood Institute. Available at
website: www.nhlbi.nih.gov/health-pro/guidelines/ (verified 21 Nov. 2014).
74. Ratner, RK and Riis, J. 2014. Communicating science-based recommendations with

memorable and actionable guidelines. Proceedings of the National Academy of Sciences of the
United States of America 111: 13634-41.
75. Hills, AP; Byrne, NM; Lindstrom, R, et al. 2013. 'Small Changes' to Diet and Physical
Activity Behaviors for Weight Management. Obesity Facts 6: 228-38.
76. Lin, P-H; Aickin, M; Champagne, C, et al. 2003. Food group sources of nutrients in the
dietary patterns of the DASH-sodium trial. Journal of the American Dietetic Association 103:
488-96.
18
77. Dietary Guidelines Advosiry Committee. 2015. Scientific Report of the 2015 Dietary
Guidelines Advisory Committee: Advisory Report to the Secretary of Health and Human
Services and the Secretary of Agriculture. Appendix E-3.2. U.S. Government Printing Office,
78. US Department of Agriculture, Economic Research Service (ERS). 2009. Import Share of
Consumption, US Agricultural Trade. Available at website:
http://www.ers.usda.gov/topics/international-markets-trade/us-agricultural-trade/import-share-of-
consumption.aspx (verified 16 Mar 2015).
79. US Department of Agriculture, Economic Research Service. 2015. Fruit and Tree Nuts:
Background. Available at website: www.ers.usda.gov/topics/crops/fruit-tree-
nuts/background.aspx (verified 16 Mar. 2015).
80. US Department of Agriculture, Economic Research Service. 2015. Vegetables and Pulses:
Overview. Available at website: www.ers.usda.gov/topics/crops/vegetables-pulses.aspx (verified
16 Mar. 2015).
81. Peters, CJ; Wilkins, JL and Fick, GW. 2007. Testing a complete-diet model for estimating
the land resource requirements of food consumption and agricultural carrying capacity: The New
York State example. Renewable Agriculture and Food Systems 22: 145-53.
82. Mathews, AJ. 2013. Applying geospatial tools and techniques to viticulture. Geography
Compass 7: 22-34.
83. Giombolini, KJ; Chambers, KJ; Schlegel, SA, et al. 2011. Testing the local reality: Does the
Willamette Valley growing region produce enough to meet the needs of the local population? A
comparison of agriculture production and recommended dietary requirements. Agriculture and
Human Values 28: 247-62.
84. Kremer, P and Schreuder, Y. 2012. The feasibility of regional food systems in metropolitan
areas: An investigation of Philadelphia's foodshed. Journal of Agriculture, Food Systems, and
Community Development 2: 171-91.
85. Colasanti, KJA and Hamm, MW. 2010. Assessing the local food supply capacity of Detroit,
Michigan. Journal of Agriculture, Food Systems, and Community Development 1: 41-58.
19
86. Peters, CJ; Bills, NL; Lembo, AJ, et al. 2012. Mapping potential foodsheds in New York
State by food group: An approach for prioritizing which foods to grow locally. Renewable
Agriculture and Food Systems 27: 125-37.
20
Title
results
Author names and affiliations
Zach Conrad,a Kenneth Chui,b Christian Peters,a Timothy Griffina*

a
150 Harrison Avenue, Friedman School of Nutrition Science and Policy, Tufts University,
Boston, MA 02111, USA

b
136 Harrison Avenue, School of Medicine, Tufts University, Boston, MA 02111, USA
*
Corresponding author, timothy.griffin@tufts.edu, 617-636-3613
Conflicts of interest
The authors have no conflicts of interest to declare
Sources of funding
There are no funding sources to declare
Ethics of human subject participation
This research does not involve the use or collection of protected health information, and received
exempt status from the Tufts Medical Center/Tufts University Health Sciences Institutional
Review Board.
21
Abbreviations
24HR; 24-hour recall
BRFSS; Behavioral Risk Factor Surveillance System
CDC; Centers for Disease Control and Prevention
CI; confidence interval
Cup-eq.; cup-equivalent
DGA; Dietary Guidelines for Americans
ERS; Economic Research Service
FFQ; food frequency questionnaire
F&V; fruits and vegetables
kcal; kilocalorie
LAFA; Loss-Adjusted Food Availability data series
LOA; limits of agreement
NHANES; National Health and Nutrition Examination Survey
OLP; ordinary least products regression
OLS; ordinary least squares regression
USDA; United States Department of Agriculture
US HHS; United States Department of Health and Human Services

22
Abstract
Continuous monitoring of fruit and vegetable consumption at the national level is critical for
assessing progress towards national public health nutrition goals. Several different dietary data
collection programs are often used interchangeably to monitor fruit and vegetable consumption
at the national level, yet there is some indication that these programs produce different results. It
is important to test the comparability of these programs because they are used to generate public
health messages and conduct dietary interventions, and inconsistent efforts can threaten progress
toward to achieving public health nutrition goals. This study will estimate the comparability of
daily fruit and vegetable consumption data from the National Health and Nutrition Examination
Survey (NHANES), the Behavioral Risk Factor Surveillance System (BRFSS), and the Loss-
Adjusted Food Availability (LAFA) data series. The Bland-Altman method, ordinary least
products regression, and descriptive means comparison were used to compare estimates of adult
daily per capita fruit and vegetable consumption between dietary data collection programs (2007-
2011). Ordinary least products regression revealed proportional bias (95% CI: 0.752, 0.949) but
not fixed bias (95% CI: -0.002, 0.002), and the Bland-Altman method revealed proportional bias
(95% CI= -0.316, -0.058). Descriptive means comparisons yielded inconsistent results. These
programs do not produce comparable estimates of daily per capita fruit and vegetable
consumption, and should not be used interchangeably to monitor progress toward achieving
national nutrition goals. NHANES is best suited for monitoring fruit and vegetable consumption
at the national level, although BRFSS and LAFA retain certain advantages for related research
protocols.
Keywords: Dietary assessment method, comparison, surveillance, diet, food group

23
1. Introduction
Regular consumption of fruits and vegetables (F&V) is associated with a reduced risk of chronic
disease, including hypertension, stroke, and cardiovascular disease [1-3]. Hypertension and
stroke are primary risk factors for cardiovascular disease [4, 5], which is the leading cause of
death in the United States (US) [6]. Recently, Wang et al. [3] reported a protective effect of each
serving of F&V consumed per day on all-cause mortality, up to two servings of fruit and three
servings of vegetables per day. Most Americans, however, do not meet the recommended daily
consumption of F&V [7]. For that reason, national public health goals include increasing daily
fruit and vegetable consumption to 0.9 and 1.1 cup-equivalents per 1,000 kcal, respectively, by
2020 [8]. Continuous monitoring of F&V consumption at the national level is critical for
assessing progress towards this goal.
At the national level, food consumption data are predominantly collected using the National
Health and Nutrition Examination Survey (NHANES) [9], the Behavioral Risk Factor
Surveillance System (BRFSS) [10], and the Loss-Adjusted Food Availability (LAFA) data series
[11]. These dietary data collection programs each utilize a distinct methodology (see below), and
their data can be used for distinct purposes. For example, NHANES is well-suited for nutrient
intake analysis [12, 13], BRFSS for examining state [14] and regional [15] differences in F&V
consumption, and LAFA for tracking the adequacy of the national food supply [16-19].
However, these programs are also often used for the same purpose: estimating F&V consumption
at the national level [7, 14-17, 20-25].

24
Studies using data from NHANES [7], BRFSS [14], and LAFA [25] have shown that Americans
consume fewer F&V than is recommended by the Dietary Guidelines for Americans (DGA). The
prevalence of adults meeting F&V recommendations varies widely depending on the method
used. For example, the proportion of adults meeting these recommendations has been observed
at 13-20% using NHANES data [7] and >25% using BRFSS data [14]. Using LAFA, it was
estimated that mean consumption was 50-70% of the recommended quantity [25]. There are no
studies that have directly compared F&V consumption estimates across all of these programs.
Studies that have compared nutrient intake [26] and fresh F&V consumption [27] between
NHANES and the Food Availability data series (which differs from LAFA in that it does not
include adjustments for food losses) observed similarities based on descriptive means
comparison, yet agreement tests are needed to directly compare these programs.
Dietary data collection programs provide essential information that is used to track progress
toward public health goals. Policymakers and nutrition educators use this information to generate
public health messages and to conduct dietary interventions. If different dietary data collection
programs produce different estimates of food consumption, this could result in inconsistent
messaging and interventions, thereby attenuating progress toward the stated goals. This study
will estimate the comparability of F&V consumption data from NHANES, BRFSS, and LAFA.
This will be achieved by 1) comparing the daily per capita consumption of individual F&V by
processing form using statistical agreement tests, and 2) comparing the daily per capita
consumption of F&V groups using descriptive means comparison.

25
2. Methods
2.1. Dietary assessment data
NHANES is a continuous survey that collects health and nutrition data from a nationally
representative sample of ~5,000 individuals annually [9]. Participants complete a 24-hour recall
(24HR) administered by a trained interviewer using United States Department of Agriculture’s
(USDA) Automated Multiple Pass Method [28], and a subset of the study population completes
a subsequent 24HR by telephone. Data are published in two-year cycles. NHANES is a program
of the US Department of Health and Human Services, Centers for Disease Control and
Prevention (US HHS, CDC).
BRFSS is a continuous survey that collects data on health and nutrition from ~500,000
participants annually, making it the largest telephone survey in the world [10]. State health
departments are responsible for data collection, and technical and methodological assistance is
provided by US HHS, CDC. Participants ≥18 years of age are selected by random-digit dialing,
and F&V consumption is measured by a six-item food frequency questionnaire (FFQ)
administered in odd numbered years. Beginning in 2011, several important changes were made
to the survey design. The weighting methodology was changed from a post-stratification
approach to a raking approach (also known as iterative proportional fitting), and a cell phone
sampling frame was added to account for the nearly one-third of households that rely on a cell
phone instead of a landline telephone [29]. Additionally, changes were made to the F&V module
to better capture consumption data for vegetable groups.

26
In contrast to NHANES and BRFSS, which measure self-reported food intake, LAFA provides
estimates of the per capita availability of food, which is a proxy for individual food consumption
[11]. Food availability for over 200 individual agricultural crops is computed as a residual based
on inflows (imports and domestic production), outflows (exports), and ending stocks. The
residual is divided by the US population and adjusted for food losses that occur throughout the
supply chain. LAFA is maintained by USDA Economic Research Service (ERS), which collects
data from a variety of government agencies as well as non-government sources, and remaining
data gaps are addressed by statistical imputation.
2.2. Data compilation
NHANES, BRFSS, and LAFA report F&V consumption in different measurement units.
NHANES reports consumption of composite dishes (e.g. sandwiches, soups, casseroles) and
individual foods in units of mass, BRFSS reports frequency of servings of F&V subgroups, and
LAFA reports the availability of individual F&V in cup-equivalents. This section describes the
methods used to standardize these units. This study received exempt status from the Tufts
Medical Center/Tufts University Health Sciences Institutional Review Board.
Data on fruit, vegetable, and legume consumption were retrieved from NHANES 2007-2008 [30]
and 2009-2010 [31]. Individuals at least 18 years of age with two days of dietary information
were included in the analysis (2007-2008, n=4,896; 2009-2010, n=5,284). Consumption of F&V
in the form of composite dishes were estimated using the Food and Nutrient Database for Dietary
Studies 4.1 [32] and 5.0 [33], which correspond to the 2007–2008 and 2009–2010 cycles,
27
respectively. Processed forms (i.e. fresh, canned, frozen, dried, and juice) of each F&V were
identified using the USDA National Nutrient Database for Standard Reference, Release 26 [34].
Units of mass were converted to cup-equivalents using the Food Patterns Equivalents Database
2007-2008 [35] and 2009-2010 [36], which provide the mass-to-volume conversions for most of
the foods included in the Food and Nutrient Database for Dietary Studies 4.1 and 5.0.
Data from BRFSS were collected for 2007 (n=430,912) [37], 2009 (n=432,607) [38], and 2011
(n=506,467) [39]. Data on fresh fruit, fruit juice, green salad, potatoes, carrots, and other
vegetables were retrieved from BRFSS 2007 and 2009, and data on fresh fruit, fruit juice, dark
green vegetables, orange vegetables, beans, and other vegetables were retrieved from BRFSS
2011. Data reported as frequency of consumption per year, per month, and per week were
converted to frequency of consumption per day. Frequencies were assumed to be commensurate
with servings based on BRFSS questionnaire language [40, 41], and two servings were assumed
per cup-equivalent in accordance with DGA 2005 [42].
LAFA provided data on 69 F&V, measured in cup-equivalents. Each F&V represented one or
several processing forms, resulting in 119 total F&V processing form combinations (e.g. fresh
apples and canned apples are considered to be distinct F&V processing form combinations). Data
on chili peppers were not included because of small consumption amounts. Frozen potatoes were
not included because the majority are processed into French fries [43], and the nutrient profile of
fried foods is distinct from that of F&V: fried potato items contribute appreciable amounts of
solid fat to the average American diet [44], and solid fat is a food group that is discouraged by
the DGA, which is in contrast to what is recommended for F&V [45]. For these reasons, potato
28
chips and shoestring potatoes were also not included in this analysis. This resulted in 68 F&V
and 116 F&V processing form combinations.
Because LAFA does not report consumption data by age group, we estimated the consumption of
each F&V processing form combination (hereafter referred to as F&V forms) for individuals ≥18
years of age by applying the proportion calories of each F&V form consumed by individuals ≥18
years of age in the NHANES sample to the consumption amounts reported by LAFA.
2.3. Agreement tests
Agreement tests are useful when comparing methods of measurement because they allow for the
quantification of bias, and several methods are well suited for this purpose. The Bland-Altman
method, which was introduced in 1983 [46] and subsequently refined [47-49], is the most
commonly used method in nutrition studies [50]. Ordinary least products (OLP) regression is less
commonly used, more complex to execute, and more difficult to interpret [51-53], but its utility
in detecting bias may be superior to the Bland-Altman method [54]. Luiz and Szklo (2005) [55]
remind us that all approaches for agreement testing have limitations, and they and others [54]
suggest that method comparison studies utilize several approaches to compensate for the
limitations of any one approach. We have heeded that recommendation here.
The agreement between daily per capita consumption estimates of F&V forms, measured in cup-
equivalents, was tested between NHANES (mean of 2007-2008 and 2009-2010) and LAFA
(mean of 2007-2010). Data from BRFSS were not included in the agreement tests because these
29
data are reported as six F&V subgroups, and this number of observations is unlikely to offer
sufficient statistical power for this analysis. Data for F&V forms not included in LAFA were
omitted from the NHANES dataset; thus, 106 F&V forms were included in this analysis
(Appendix A).
Agreement between NHANES and LAFA was examined using OLP regression as well as the
Bland-Altman method supplemented by ordinary least squares (OLS) regression. Although OLS
regression is not a suitable method for specifically testing agreement [53], it retains its
functionality as a test for proportional bias (but not fixed bias) in the Bland-Altman method [52].
Proportional bias means that one method gives values that are progressively different from those
of another method, and fixed bias means that one method gives values that are different from
those of another method by a consistent amount. Agreement was tested based on the absence of
fixed and proportional biases. The Bland-Altman method also offers limits of agreement, which
were used to identify extreme values [48]. We used the equations provided by Ludbrook [56] to
calculate the coefficients for the slope, intercept, and confidence intervals of the OLP regression.
Proportional bias is indicated if the slope is significantly different than one at the P≤0.05 level
(two-sided test), and the presence of fixed bias is indicated if the intercept is significantly
different than zero at the P≤0.05 level.
In the Bland-Altman method, the difference between two data sources is plotted against the mean
of two data sources, and limits of agreement contain 95% of the values around the mean
difference. The difference (D) was estimated as:

30
D = ̅ LAFA/Y – ̅ NHANES/Y, (1)
where ̅ LAFA is the mean per capita consumption of a given F&V and a given processing form in
a given year for individuals at least 18 years of age from LAFA, Y is the number of years for
which data were collected, and ̅ NHANES is the mean per capita consumption of a given F&V and
a given processing form in a given year for individuals ≥18 years of age from NHANES. Mean
consumption of LAFA and NHANES ( ̅ BOTH) was estimated as:
̅ BOTH = ( ̅ LAFA/Y + ̅ NHANES/Y)/2. (2)
Limits of agreement (LOA) were estimated as:
LOA = ̅ ± 1.96s, (3)
where s is the standard deviation of the mean difference ( ̅ ), and ̅ was estimated as:
̅ = ( ̅ LAFA/Y – ̅ NHANES/Y)/n, (4)
where Y is the number of years for which data were collected, and n is the number of F&V forms
included in the analysis.

31
OLS regression was used to detect proportional bias in the Bland-Altman method by predicting
the differences (D) from the means ( ̅ BOTH). The presence of proportional bias is indicated if the
slope is significantly different than zero at the P≤0.05 level.
2.4. Descriptive means comparison
Daily per capita consumption of F&V groups, measured in cup-equivalents, was compared
between NHANES, BRFSS, and LAFA. Due to the methodological and reporting changes that
were made to the 2011 BRFSS, analyses were conducted for two time frames: 2007-2010
(NHANES 2007-2008 and 2009-2010, BRFSS 2007 and 2009, and LAFA 2007-2010), and
2010-2011 (NHANES 2009-2010, BRFSS 2011, and LAFA 2011). Individual F&V from
NHANES and LAFA were categorized according to BRFSS 2007 and 2009 groups for time
frame 2007-2010 (Appendix B), and were categorized according to BRFSS 2011 groups for time
frame 2010-2011 (Appendix C). The USDA MyPlate [57] categorization scheme was used to
assign individual F&V to groups in cases where BRFSS questionnaires [40, 41, 58] lacked
sufficient specificity.
Data management, statistical analyses, and complex weighting of NHANES and BRFSS data
were performed using Stata 12. Complex survey weights were incorporated into the analysis
whenever appropriate.
32
Results
OLP regression revealed proportional bias (95% CI: 0.752, 0.949) but not fixed bias (95% CI: -
0.002, 0.002), and the Bland-Altman method revealed proportional bias (95% CI= -0.316, -
0.058). Approximately 7% (7 out of 106) of the F&V forms were outside of the limits of
agreement (Figure 1) in the Bland-Altman method. These included dried potatoes, leaf lettuce,
dry beans, orange juice, fresh potatoes, fresh bananas, and fresh apples.
Total F&V consumption from NHANES was 14% higher than from BRFSS and 34% higher than
from LAFA in 2007-2010 (Table 1). Most of this difference was due to fruit, which accounted
for 52% of total F&V consumption in NHANES, compared to 39% and 42% of total F&V
consumption in BRFSS and LAFA, respectively. Although NHANES and BRFSS produced
identical fruit juice estimates, the percent of fruit juice consumption out of fruit consumption was
most similar between NHANES (26%) and LAFA (25%). NHANES and BRFSS yielded similar
vegetable consumption estimates. None of the programs produced estimates for any of the
vegetable groups that were within 10% of one another, except for green salad and other
vegetables between NHANES and BRFSS.

33
.10
0
-.10
-.20
0 .05 .10 .15 .20

Cup-eq. per day, (LAFA+NHANES)/2
Fig. 1 - Bland-Altman comparison between NHANES and

LAFA, daily per capita consumption of fruits and vegetables,
2007-2010
Solid line represents mean difference
Dashed lines represent Limits of Agreement at +/- 1.96 standard deviations from
mean difference
Cup eq., cup-equivalent
NHANES, National Health and Nutrition Examination Survey
LAFA, Loss-Adjusted Food Availability Data Series
BRFSS and LAFA produced nearly identical F&V consumption estimates for 2011, and the
estimate produced by NHANES was ~28% higher (Table 2). Fruit accounted for 54%, 43%, and
31% of F&V consumption from NHANES, BRFSS, and LAFA, respectively. Although
discordant fruit consumption estimates were observed across the programs, the percent of fruit
juice out of fruit consumption was similar (~30%). A 10% difference was observed for
consumption estimates of vegetables, as well as of each vegetable group, between NHANES and
LAFA.
34
Table 1 Daily per capita consumption of fruit and vegetable groups, 2007-2010
a b c
Food group NHANES BRFSS LAFA
Cup-eq. (95% CI)d

Fruit and vegetables 2.22 (2.15-2.30) 1.94 (1.94-1.95) 1.66 (1.61-1.71)
Fruit 1.15 (1.10-1.20) 0.76 (0.76-0.77) 0.69 (0.67-0.70)
Fruit juice 0.30 (0.29-0.32) 0.30 (0.30-0.31) 0.17 (0.15-0.19)
Fruit, non-juice 0.84 (0.80-0.88) 0.46 (0.46-0.46) 0.52 (0.50-0.54)
Vegetables 1.08 (1.04-1.12) 1.18 (1.18-1.18) 0.97 (0.93-1.02)
Green salad 0.24 (0.22-0.25) 0.26 (0.25-0.26) 0.20 (0.19-0.22)
Carrots 0.06 (0.05-0.07) 0.14 (0.13-0.14) 0.04 (0.04-0.04)
Potatoese 0.18 (0.17-0.20) 0.13 (0.13-0.14) 0.32 (0.30-0.34)
Other 0.60 (0.57-0.63) 0.65 (0.65-0.66) 0.41 (0.40-0.42)
BRFSS, Behavioral Risk Factor Surveillance System
LAFA, Loss-Adjusted Food Availability data series
Cup-eq., cup equivalent
CI, confidence interval
a
Mean, 2007-2008 and 2009-2010
b
Mean, 2007 and 2009
c
Mean, 2007-2010
d
Confidence intervals for NHANES and BRFSS represent respondent differences between pooled data,
and confidence intervals for LAFA represent year-to-year variation
e
Non-fried
35
Table 2 Daily per capita consumption of fruit and vegetable groups, 2011
Food group NHANESa BRFSSb LAFAb
Cup-eq. (95% CI)
Fruit and vegetables 2.31 (2.22-2.40) 1.66 (1.66-1.67) 1.67 (NA)
Fruit 1.06 (1.02-1.11) 0.72 (0.72-0.73) 0.51 (NA)
Fruit juice 0.30 (0.28-0.32) 0.22 (0.22-0.23) 0.16 (NA)
Fruit, non-juice 0.76 (0.73-0.79) 0.50 (0.50-0.50) 0.35 (NA)
Vegetables 1.25 (1.17-1.32) 0.94 (0.94-0.95) 1.16 (NA)
Dark green 0.15 (0.14-0.16) 0.27 (0.27-0.27) 0.16 (NA)
Orange 0.08 (0.07-0.09) 0.13 (0.13-0.14) 0.08 (NA)
Beans 0.09 (0.07-0.10) 0.15 (0.15-0.15) 0.08 (NA)
Otherc 0.93 (0.86-1.00) 0.39 (0.39-0.39) 0.85 (NA)
BRFSS, Behavioral Risk Factor Surveillance System
LAFA, Loss-Adjusted Food Availability data series
Cup-eq., cup equivalent
CI, confidence interval
a
2009-2010
b
2011
c
Does not include fried potato items
Discussion
The presence of proportional bias in the agreement tests indicates that daily per capita
consumption estimates of F&V forms between NHANES and LAFA are not in agreement.
Means comparisons yielded similar estimates for some F&V groups between NHANES, BRFSS,
and LAFA, but these similarities were inconsistent. Overall, we cannot conclude that these
programs produced comparable estimates of daily per capita F&V consumption.
On average, NHANES estimates of F&V forms are 19% higher than LAFA estimates. The
difference between these programs increases as consumption amounts increase, such that all of
36
the F&V forms that are outside of the limits of agreement are within the 90th percentile of
consumption amounts. However, removing these extreme values did not improve the agreement,
suggesting that the presence of proportional bias is not a product of the extreme values.
Ahuja and others [26] observed higher nutrient intake estimates using the Food Availability data
series 2006 compared to NHANES 2007-2008, but estimates from the Food Availability data
series were not adjusted for food losses and were not particular to F&V, so these results are not
comparable to the results presented here. Hoelzer and others [27] compared daily per capita
estimates of fresh F&V consumption between LAFA 2006 and NHANES 1999-2006 using
descriptive means comparison. The authors cautioned against interpreting data based on relative
differences reported with a high level of precision, and favored using absolute differences
rounded to the nearest whole number; it was concluded that the programs were comparable. To
the best of our knowledge, there are no studies that have used agreement testing to examine the
comparability between dietary data collection programs.
Examining the consumption estimates of F&V groups between programs for 2007-2010 revealed
similar estimates between some programs for some F&V groups, but these similarities were
inconsistent. For 2010-2011, BRFSS and LAFA yielded nearly identical estimates for total F&V,
but there were no consistent similarities except between NHANES and LAFA for vegetable
groups.
A number of factors inherent in the design of the programs used in this study may account for the
observed differences. Participants in surveys may alter their reported consumption patterns to
37
simplify the interview or to impress the interviewer [2, 59]. The accuracy of dietary surveys
relies partly on respondents’ memory, and the time frame from which respondents are asked to
recall their food consumption is different between BRFSS, which uses a food frequency
questionnaire (represents a typical diet pattern) [10], and NHANES, which uses a 24 hour recall
(represents a diet pattern for a specific 24 hour period) [9]. NHANES can be paired with
supplementary datasets to identify foods consumed as part of mixed dishes [60] and to compute
cup-equivalents [61], yet no such datasets are currently available for BRFSS. Indeed, our method
of converting frequency of consumption to cup-equivalents in the BRFSS datasets may be overly
simplistic, and we look forward to the development of more substantive methods (Grimm, K.
Centers for Disease Control and Prevention. Pers. Comm. April 23, 2014).
LAFA does not include all F&V that are available to consumers. Additionally, it does not
capture foods purchased directly from producers [11], such as those sold at farmers’ markets,
community supported agriculture programs, pick-your-own operations, and others, although
these channels account for less than 1% of total food sales [62]. The conversion factors used by
USDA ERS to estimate food losses are applied in a stepwise fashion as foods move from being
agricultural crops to consumer goods, yet losses at the consumer level (inedible portions and
uneaten portions) are combined rather than applied independently, which may result in an
overestimation of food loss. We explored this possibility by applying these conversion factors
independently instead of combined, and observed an increase in consumption estimates for some
popular F&V forms (fresh bananas and leaf lettuce increased by 16% and 9%, respectively), but
these nonetheless remained far below corresponding estimates from NHANES.

38
To the best of our knowledge, this is the first study to use statistical agreement tests
supplemented with means comparisons to determine whether NHANES, BRFSS, and LAFA
produce similar estimates of daily per capita F&V consumption at the national level. This
endeavor has important public health implications because data from each of these programs are
used to develop public health messages and to conduct dietary interventions, and inconsistent
efforts can threaten progress toward public health nutrition goals. We found that F&V
consumption data from NHANES, BRFSS, and LAFA do not agree with one another, which
means that the public health messages and dietary interventions that are based on these data are
at risk of being inconsistent. While it is true that each of these programs show that adults
consume far fewer F&V than what is recommended, this is due to the paucity of F&V in the
typical American diet and should not be interpreted as representing congruity between these
programs (in other words, the bar is low). The nature of public health messages and dietary
interventions is informed by whether Americans are meeting goals and how close they are to
meeting those goals. Additionally, if F&V consumption were to increase in the future, the
differences we observed between these programs could result in a situation where one of these
programs indicates that public health nutrition goals were achieved while the other programs
indicate otherwise.
We recommend that these programs should not be used interchangeably to monitor progress
toward achieving national nutrition goals. NHANES is best suited for this purpose because it
allows for the amount of individual F&V (as well as F&V groups) to be discerned from
composite foods with multiple ingredients, which allows for a complete accounting of F&V
consumption from all sources in the diet. NHANES is also structured for subpopulation analysis
39
and covariate control, which can be used to estimate consumption patterns for specific
population groups while adjusting for the effects of other variables. By comparison, BRFSS does
not collect consumption data from composite dishes and does not collect data on individual
F&V. LAFA does not collect data on all of the F&V included in NHANES, does not collect data
on all processing forms of each F&V, and does not allow for subpopulation analysis and
covariate control. However, BRFSS and LAFA retain advantages for related research protocols.
BRFSS is the only program that collects F&V consumption data at the state level, which makes
it useful for state and regional comparisons. LAFA provides the only nationally representative
estimates of food losses at various points in the supply chain, which makes it useful for
examining the adequacy of the food supply at the agricultural, retail, and consumer levels.
Acknowledgement
The authors thank Dr. Kate Clancy for her helpful comments. This research was not supported
by a specific funding agency. The authors do not have any conflicts of interest to declare.
References
[1] Boeing H, Bechthold A, Bub A, Ellinger S, Haller D, Kroke A, et al. Critical review:
Vegetables and fruit in the prevention of chronic diseases. Eur J Nutr 2012; 51:637-63.
[2] Woodside JV, Young IS, McKinley MC. Fruits and vegetables: Measuring intake and
encouraging increased consumption. P Nutr Soc 2013; 72:236-45.
[3] Wang X, Ouyang Y, Liu J, Zhu M, Zhao G, Bao W, et al. Fruit and vegetable consumption
and mortality from all causes, cardiovascular disease, and cancer: Systematic review and dose-
response meta-analysis of prospective cohort studies. Brit Med J 2014; 349:g4490.
[4] Lloyd-Jones D, Adams RJ, Brown TM, Carnethon M, Dai S, De Simone G, et al. Heart
disease and stroke statistics-2010 update: A report from the American Heart Association.
Circulation 2010; 121:e46-e215.
[5] Lewington S, Clarke R, Qizilbash N, Peto R, Collins R. Age-specific relevance of usual
blood pressure to vascular mortality: A meta-analysis of individual data for one million adults in
61 prospective studies. Lancet 2002; 360:1903-13.
40
[6] US Department of Health and Human Services, Centers for Disease Control and Prevention.
2010. Ten Leading Causes of Death and Injury. Available at website:
www.cdc.gov/injury/wisqars/leadingcauses.html (verified 04 Feb. 2013).
[7] Krebs-Smith SM, Guenther PM, Subar AF, Kirkpatrick SI, Dodd KW. Americans do not
meet federal dietary recommendations. J Nutr 2010; 140:1832-8.
[8] US Department of Health and Human Services. 2010. Healthy People 2020: Understanding
and Improving Health. Nutrition and Weight Status Objectives 14 and 15. 3rd ed. Available at
2015).
[9] US Department of Health and Human Services, Centers for Disease Control and Prevention
(CDC). About the National Health and Nutrition Examination Survey. Available at website:
www.cdc.gov/nchs/nhanes.htm (verified 25 Mar. 2013).
(CDC). Behavioral Risk Factor Surveillance System. Available at website: www.cdc.gov/brfss/
[11] US Department of Agriculture, Economic Research Service (ERS). Food Availability (per
Capita) Data System. Available at website: www.ers.usda.gov/data-products/food-availability-
(per-capita)-data-system.aspx (verified 11 Nov. 2014).
[12] Fulgoni III VL, Keast DR, Auestad N, Quann EE. Nutrients from dairy foods are difficult
to replace in diets of Americans: Food pattern modeling and an analyses of the National Health
and Nutrition Examination Survey 2003-2006. Nutr Res 2011; 31:759-65.
[13] O'Neil CE, Keast DR, Fulgoni VL, Nicklas TA. Food sources of energy and nutrients
among adults in the US: NHANES 2003-2006. Nutrients 2012; 4:2097-120.
[14] Grimm KA, Blanck HM, Scanlon KS, Moore LV, Grummer-Strawn LM. State-specific
trends in fruit and vegetable consumption among adults - United States, 2000-2009. MMWR
CDC Surv Summ 2010; 59:1125-30.
[15] Troost JP, Rafferty AP, Luo Z, Reeves MJ. Temporal and regional trends in the prevalence
of healthy lifestyle characteristics: United States, 1994–2007. Am J Public Health 2012;
102:1392-8.
[16] Buzby JC, Wells HF, Vocke G. 2006. Possible Implications for US Agriculture from
Adoption of Select Dietary Guidelines. USDA Economic Research Service (ERS). Economic
[17] Krebs-Smith SM, Reedy J, Bosire C. Healthfulness of the US food supply: Little
improvement despite decades of dietary guidance. Am J Prev Med 2010; 38:472-7.
[18] Reedy J, Krebs-Smith SM, Bosire C. Evaluating the food environment: application of the
healthy eating index-2005. Am J Prev Med 2010; 38:465-71.
[19] Miller PE, Reedy J, Kirkpatrick SI, Krebs-Smith SM. The United States food supply is not
consistent with dietary guidance: evidence from an evaluation using the Healthy Eating Index-
2010. J Acad Nutr Dietetics 2014; 115:95-100.
[20] Demydas T. Consumer segmentation based on the level and structure of fruit and vegetable
intake: an empirical evidence for US adults from the National Health and Nutrition Examination
Survey (NHANES) 2005–2006. Public Health Nutr 2011; 14:1088-95.
[21] Kimmons J, Gillespie C, Seymour J, Serdula M, Blanck HM. Fruit and vegetable intake
among adolescents and adults in the United States: percentage meeting individualized
recommendations. Medscape J Med 2009; 11:
41
[22] Casagrande SS, Wang Y, Anderson C, Gary TL. Have Americans increased their fruit and
vegetable intake?: the trends between 1988 and 2002. Am J Prev Med 2007; 32:257-63.
[23] Blanck HM, Gillespie C, Kimmons J, Seymour J, Serdula M. Trends in fruit and vegetable
consumption among U.S. men and women, 1994-2005. Prev Chronic Dis 2008; 5:1-10.
[24] Serdula MK, Gillespie C, Kettel-Khan L, Farris R, Seymour J, Denny C. Trends in fruit and
vegetable consumption among adults in the United States: Behavioral Risk Factor Surveillance
System, 1994-2000. Am J Public Health 2004; 94:1014-8.
[25] Wells HF,Buzby JC. 2008. Dietary Assessment of Major Trends in US Food Consumption,
1970-2005. USDA Economic Research Service (ERS). Economic Information Bulletin Number
33. US Government Printing Office, Washington, DC. Available at website:
www.ers.usda.gov/publications.aspx#.Ucmq0zTVDDY (verified 25 June 2013).
[26] Ahuja JK, Juan WY, Egan K, Buzby JC, Trumbo P, Moshfegh AJ, et al. Federal monitoring
activities related to food and nutrition: How do they compare? Procedia Food Science 2013;
2:165-71.
[27] Hoelzer K, Pouillot R, Egan K, Dennis S. Produce consumption in the United States: An
analysis of consumption frequencies, serving sizes, processing forms, and high-consuming
population subgroups for microbial risk assessments. J Food Protect 2012; 75:328-40.
[28] Moshfegh AJ, Rhodes DG, Baer DJ, Murayi T, Clemens JC, Rumpler WV, et al. The US
Department of Agriculture Automated Multiple-Pass Method reduces bias in the collection of
energy intakes. Am J Clin Nutr 2008; 88:324-32.
(CDC). 2011. BRFSS 2011 Survey Data: Comparability of Data. Available at website:
www.cdc.gov/brfss/technical_infodata/surveydata/2011.htm (verified 05 Feb. 2013).
(CDC). NHANES 2007-2008, Dietary interview: individual foods file, food code description
file. Available at website: wwwn.cdc.gov/nchs/nhanes/search/nhanes07_08.aspx (verified 14
Oct. 2014).
(CDC). NHANES 2009-2010, Dietary interview: Individual foods file, food code description
file. Available at website: wwwn.cdc.gov/nchs/nhanes/search/nhanes09_10.aspx (verified 27 Jan.
2015).
[32] US Department of Agriculture, Agricultural Research Service. Food and Nutrient Database
for Dietary Studies 4.1. Available at website:
www.ars.usda.gov/Services/docs.htm?docid=12068 (verified 14 Oct. 2014).
[34] US Department of Agriculture, Agricultural Research Service. USDA National Nutrient
Database for Standard Reference, Release 26. Nutrient Data Laboratory. Available at website:
www.ndb.nal.usda.gov/ (verified 31 Aug. 2013).
[35] US Department of Agriculture, Agricultural Research Service. Food Patterns Equivalents
Database 2007-2008. Available at website: www.ars.usda.gov/Services/docs.htm?docid=23871
(verified 14 Oct. 2014).
(14 October 2014).
42
(CDC). Behavioral Risk Factor Surveillance System 2007. Available at website:
www.cdc.gov/brfss/annual_data/annual_2007.htm (verified 17 Oct. 2014).
(CDC). Behavioral Risk Factor Surveillance System 2007 Questionnaire. Section 9. Available at
website: www.cdc.gov/brfss/questionnaires.htm (verified 03 Nov. 2014).
[42] US Department of Health and Human Services and US Department of Agriculture. 2005.
Dietary Guidelines for Americans 2005, Table 1. 6th ed. US Government Printing Office,
Washington, DC. Available at website: http://health.gov/dietaryguidelines/ (verified 03 Nov.
2014).
[43] US Department of Agriculture, Economic Research Service. Vegetables and Pulses: Fresh
and Processed potatoes. Available at website: http://www.ers.usda.gov/topics/crops/vegetables-
pulses/potatoes.aspx#potatoes (verified 14 Oct. 2014).
Dietary Guidelines for Americans 2010, Figure 3-5. 7th ed. U.S. Government Printing Office,
Washington, DC. Available at website: http://health.gov/dietaryguidelines/ (verified 25 Mar.
2013).
Dietary Guidelines for Americans 2010. 7th ed. U.S. Government Printing Office, Washington,
DC. Available at website: www.health.gov/dietaryguidelines/ (verified 25 Mar. 2013).
[46] Altman DG,Bland JM. Measurement in medicine: The analysis of method comparison
studies. J Roy Stat Soc D-Sta 1983; 32:307-17.
[47] Bland MJ,Altman DG. Statistical methods for assessing agreement between two methods of
clinical measurement. Lancet 1986; 327:307-10.
[48] Bland MJ,Altman DG. Measuring agreement in method comparison studies. Stat Methods
Med Res 1999; 8:135-60.
[49] Bland MJ,Altman DG. Applying the right statistics: Analyses of measurement studies.
Ultrasound Obst Gyn 2003; 22:85-93.
[50] Zaki R, Bulgiba A, Ismail R, Ismail NA. Statistical methods used to test for agreement of
medical instruments measuring continuous variables in method comparison studies: A systematic
review. PLoS One 2012; 7:1.
[51] Ludbrook J. Comparing methods of measurement. Clin and Exp Pharmacol P 1997;
24:193-203.
[52] Ludbrook J. Statistical techniques for comparing measures and methods of measurement: A
critical review. Clin Exp Pharmacol P 2002; 29:527-36.
[53] Ludbrook J. Linear regression analysis for comparing two measurers or methods of
measurement: But which regression? Clin Exp Pharmacol P 2010; 37:692-9.
43
[54] Zaki R, Bulgiba A, Ismail NA. Testing the agreement of medical instruments:
Overestimation of bias in the Bland–Altman analysis. Prev Med 2013; 57:S80-S2.
[55] Luiz RR,Szklo M. More than one statistical strategy to assess agreement of quantitative
measurements may usefully be reported. J Clin Epidemiol 2005; 58:215-6.
[56] Ludbrook J. Comparing methods of measurement. Appendix 1, equations 12-15. Clin Exp
Pharmacol P 1997; 24:193-203.
[57] USDA Center for Nutrition Policy and Promotion (CNPP). 2015. ChooseMyPlate.gov.
[59] Roark RA,Niederhauser VP. Fruit and vegetable intake: Issues with definition and
measurement. Public Health Nutr 2013; 16:2-7.
for Dietary Studies. Available at website: www.ars.usda.gov/News/docs.htm?docid=12089
(verified 11 Nov. 2014).
Database Available at website: www.ars.usda.gov/Services/docs.htm?docid=23871 (verified 14
Oct. 2014).
[62] US Department of Agriculture, National Agricultural Statistics Service (NASS). 2012.
Census of Agriculture. Volume 1, Chapter 1, Table 2. US Government Printing Office,
Washington, DC. Available at website: www.agcensus.usda.gov/Publications/2012/ (verified 04
Nov. 2014).
44
Title
Nutrient dense fruits and vegetables can help hypertensive adults meet the nutrient targets of the
Dietary Approaches to Stop Hypertension (DASH) diet with fewer servings than is
recommended
Author names and affiliations
Zach S. Conrad,a Kenneth K.H. Chui,b Christian J. Peters,a Timothy S. Griffina*

a

b
*
Corresponding author, zach.conrad@tufts.edu, 1-617-987-0658
Conflicts of interest
The authors have no conflicts of interest to declare
Sources of funding
There are no funding sources to declare
Ethics of human subject participation
This research does not involve the use or collection of protected health information, and received
exempt status from the Tufts Medical Center/Tufts University Health Sciences Institutional
Review Board.
45
Abbreviations
24HR; 24-hour recall
Ca; calcium
CVD; cardiovascular disease
DASH; Dietary Approaches to Stop Hypertension
FNDDS; Food and Nutrient Database for Dietary Studies
F&V: fruits and vegetables
kcal; kilocalorie
Mg; magnesium
NHANES; National Health and Nutrition Examination Survey
NDB; National Nutrient Database for Standard Reference
K; potassium
USDA; US Department of Agriculture

46
Abstract
Hypertension is a prominent risk factor for cardiovascular disease, which is the leading cause of
mortality in the US. Over one-half of adults are hypertensive or pre-hypertensive, and adherence
to the Dietary Approaches to Stop Hypertension (DASH) diet continues to be low. In order to
meet the nutrient targets of the DASH diet most individuals must modify their consumption of
nearly every food group, which is difficult for many people to achieve. A better understanding of
whether moderate changes to dietary patterns can help hypertensive adults meet the DASH diet
nutrient targets is needed. To address this research gap, we developed and combined a nutrient
density index and a food pattern model to investigate whether the DASH diet nutrient targets for
fiber, calcium, magnesium, and potassium can be met by consuming nutrient dense fruits and
vegetables (F&V) and not making any modifications to other food groups. Data on food
consumption and nutrient intake among hypertensive adults were retrieved from the National
Health and Nutrition Examination Survey (NHANES), 2009-2010 (n=738). Our results show
that diets that include nutrient dense F&V can meet the DASH diet nutrient targets with fewer
F&V servings than is recommended by the USDA Food Pattern and the DASH diet. These
findings provide clinicians with the information needed to recommend achievable diet patterns
for their hypertensive patients, which could increase the prevalence of individuals meeting
recommended nutrient intakes.
Keywords
Diet pattern, intervention, nutrient gap, number of servings

47
1. Introduction
Hypertension is defined as systolic blood pressure equal to or greater than 140mm Hg and
diastolic blood pressure equal to or greater than 90mm Hg [1], and is a prominent risk factor for
cardiovascular disease (CVD) [2, 3], the leading cause of mortality in the US [4]. Approximately
30% of US adults are hypertensive, and an additional ~25% of adults are pre-hypertensive
(systolic blood pressure at or over 120mm Hg and diastolic blood pressure at or over 80mm Hg)
[5]. However, hypertension is one of the most modifiable risk factors for CVD. Lifestyle
modifications are recommended for all hypertensive individuals regardless of pharmacotherapy
treatment[6], and key among these lifestyle modifications is the adoption of the Dietary
Approaches to Stop Hypertension (DASH) diet, which emphasizes fruits, vegetables, whole
grains, and low fat dairy products [7]. The DASH diet has been consistently shown to normalize
blood pressure [8-10], and is associated with a reduced risk of CVD and all-cause mortality [11-
13]. However, the proportion of hypertensive adults adhering to the DASH diet decreased from
29% in 1988-1994 to 22% in 1999-2004 [14], and continues to be low according to a review
published in 2013[15].
Clinical interventions are projected to be one of the most effective means of reducing CVD-
related mortality [16]. Clinicians play an important role in their patients’ adoption of
hypertension treatment regimens [17] partly because they are the first to detect the condition, as
hypertension is the most common primary diagnosis by physicians [18]. In addition, patients
who report receiving lifestyle modification counseling by their primary care physician are much
more likely to modify their diet than those who report not receiving such counseling [19].
48
However, improved clinical strategies are needed to increase the rate of adherence to the DASH
diet. Developing strategies for clinicians to help adults adopt CVD related diet recommendations
has been identified as a critical research need by the National Heart, Lung, and Blood Institute
[20].
The DASH diet is designed to increase the intake of some nutrients (fiber, calcium, potassium,
magnesium) and limit the intake of others (total fat, saturated fat, sodium), with the goal of
meeting specific daily targets [21]. In order to achieve these nutrient targets, most individuals
must modify their consumption of most or all of the food groups [7, 22], which is difficult for
many people to achieve beyond the trial phase [15].
Behavior change specialists posit that people are most likely to adopt simple, well-defined health
and nutrition recommendations [23]. Hills et al. [24] summarize a growing body of research that
shows that recommendations to make modest lifestyle changes can be readily adopted and
maintained, and can result in positive health outcomes. Research is needed to better understand
whether moderate changes to one or several food groups can help hypertensive individuals meet
the DASH diet nutrient targets. A focus on fruits and vegetables (F&V) is warranted given that
these foods contain high amounts of nutrients to encourage (fiber, calcium, magnesium, and
potassium) and low amounts of nutrients to limit (total fat, saturated fat, sodium) [25]. If
hypertensive individuals consumed nutrient dense F&V, they may be able to increase their
nutrient intake and meet the DASH diet nutrient targets for fiber, Ca, Mg, and K without
increasing the number of servings of F&V currently consumed.

49
We develop a nutrient density index to estimate the amount of fiber, Ca, Mg, and K in a serving
size of individual F&V (objective #1). We use the output of the nutrient density index to
construct a food pattern model to investigate the degree to which diets that include nutrient dense
F&V contribute higher amounts of fiber, Ca, Mg, and K compared to a typical consumption
pattern (objective #2), and to examine the effects of increasing the number of servings of F&V
and modifying the vegetable subgroup proportions, on meeting the DASH diet nutrient targets
for fiber, Ca, Mg, and K (objective #3). These findings will allow health care practitioners to
provide more targeted dietary advice for their hypertensive patients, which could increase the
prevalence of individuals meeting recommended nutrient intakes.
2. Methods
2.1 Study design
This is a secondary data analysis using multiple publically available data sources. To develop the
nutrient density index we collected data on the amount of fiber, Ca, Mg, and K per serving size
of over 400 F&V, and ranked these F&V by their nutrient density. The most nutrient dense F&V
were used to develop forty-eight modeled F&V diets (Scenario diets) (Figure 1), which differed
from one another by the number of F&V servings, the relative proportion of vegetable
subgroups, and consumer preference for individual F&V. To investigate the degree to which
diets that include nutrient dense F&V contribute higher amounts of fiber, Ca, Mg, and K
compared to a typical consumption pattern, we collected daily nutrient intake data for
hypertensive adults (Sample diet), and compared this to the nutrient content of a Scenario diet
50
that included the same number of servings of F&V and the same relative proportion of vegetable
subgroups as the Sample diet. To examine the effects of increasing the number of servings of
F&V and modifying the vegetable subgroup proportions on meeting the DASH diet nutrient
targets for fiber, Ca, Mg, and K, we compared the Scenario diets to one another on the basis of
closing the nutrient gaps between daily nutrient intakes from non-F&V and the DASH diet
nutrient targets. For all Scenario diets the nutrient contribution from all non-F&V food groups
was held constant. The details are provided below.
2.2 Sample diet
Data on food consumption and nutrient intake were acquired from the National Health and
Nutrition Examination Survey (NHANES) 2009-2010, which collects health and nutrition data
from a nationally representative sample of ~5,000 individuals annually. Data are collected
continuously and published in two year cycles [26]. In the dietary component of the survey,
trained interviewers administer a 24-hour recall (24HR) using US Department of Agriculture’s
(USDA) Automated Multiple Pass Method [27], and a subset of the study population completes
a subsequent 24HR by telephone.
The amount of F&V in composite dishes (e.g. sandwiches, soups, casseroles), as well as their
nutrient contents, were estimated using the Food and Nutrient Database for Dietary Studies
(FNDDS) 5.0 [28]. Cup-equivalents were estimated using the Food Patterns Equivalents
Database 2009-2010 [29]. Fried potato items, such as French fries and potato chips, were not
classified as vegetables because the saturated fat content of these foods makes their nutrient
51
profile distinct from that of vegetables. USDA MyPlate [30] classifies mature legumes (i.e. those
harvested after maturing on the plant) as beans and peas within the vegetable category if
consumption of protein foods is sufficient. We estimated sufficient protein intake among our
sample (39 grams per 1,000 kcal) and therefore classified mature legumes as beans and peas.
Immature legumes (i.e. those harvested from the plant before maturation) were classified as
starchy vegetables [30]. Peanuts and processed soy products were not classified as vegetables
because many consumers do not consider these foods to be vegetables [31]. Fruits and vegetables
in composite dishes that were not included in FNDDS 5.0 could not be disaggregated into
individual F&V, and were therefore not included in this analysis. Stata 12 was used to account
for the complex sampling design of the survey.
Non-pregnant, non-lactating individuals [32] greater than 18 years of age [33], with two or more
systolic blood pressure readings greater than 120mm Hg and two or more diastolic blood
pressure readings greater than 80mm Hg [34] (includes pre-hypertensive and hypertensive
individuals) [17], with two days of dietary information [26] were included in the analysis
(n=738). Mean daily F&V consumption from this sample represented the Sample diet. Nutrient
supplements were not included in this analysis because these are not recommended for
hypertensive patients due to their inconsistent effects on lowering blood pressure and potential
complications with prescription medications [35].
2.3 DASH diet nutrient targets and nutrient gaps

52
The DASH diet nutrient targets [7] were indexed to the mean daily kilocalorie (kcal) intake of
the study population by solving:
Tn = (Tnh – Tnl) / (Eh - El) (1)
where T is the target for a given nutrient (n) for the highest (h) and lowest (l) energy levels, and
E is the energy level. This method was derived from Lin et. al. [36]. The nutrient gap was
estimated for each nutrient as the difference between the daily nutrient intake from non-F&V and
the DASH diet nutrient target for a given nutrient.
2.4 Nutrient density index
Nutrient data for all fruits, vegetables, and legumes (n=1480) were acquired from USDA
National Nutrient Database for Standard Reference (NDB) Release 26 [37], and downloaded into
Microsoft Excel 2010. Legumes were appended to the vegetables list. Foods prepared with added
ingredients (e.g. oil, salt, sweeteners) or foods with ambiguous or subjective processing forms
(e.g. dried foods subsequently stewed, undiluted beverages) were excluded (52%). Foods not
listed in the 2007 USDA Census of Agriculture [38] or the USDA Loss-Adjusted Food
Availability data series [39] (e.g. oheloberries, sapote, soursop), were assumed to be unfamiliar
to most people and would not represent common food choices for most consumers, so these were
also excluded (13%). Foods listed without a primary volume measurement and those lacking data
on fiber, Ca, Mg, or K were excluded (12%). In total, 407 F&V were included in the analysis.
We assumed two servings per cup-equivalent, in accordance with the Dietary Guidelines for
53
Americans 2005 [40] and the National Heart Lung, and Blood Institute’s outreach materials for
the DASH diet [41]. Nutrient density of individual F&V was defined as the nutrient content per
serving (e.g. mg Ca per serving).
Foods were grouped into juices and non-juices (fresh, frozen, canned, dried) according to their
NDB description. The nutrient densities of all varieties of the same foods (e.g. Fuji apples and
Gala apples) within these categories were averaged. Juices were not included with the other
processing forms because doing so would result in artificially low nutrient index rankings for
foods which have juice forms: the nutrient index reflects the relative nutrient densities of each
nutrient of interest and most juices contain very little (if any) fiber [37].
To compute an index that measured the nutrient densities of fiber, Ca, Mg, and K for each food,
we adopted the methodology Swain et al. [42] used to evaluate the relative nutrient contents of
the controlled feeding menus in the DASH trial. The nutrient density of each food was assigned a
relative rank for each nutrient of interest, so that the food with the highest nutrient density
received a rank of one, and all others were ranked in descending order. Thus, each food was
accompanied by four rankings, one for each nutrient. The rankings of each nutrient for each food
were summed to generate a single ranking for each food.
2.5 Scenario diets
Each food in the nutrient density index was categorized as a fruit or a vegetable, and vegetables
were categorized into one of the following subgroups, based on the USDA MyPlate [30]
54
classification scheme: dark green, starchy, red and orange, beans and peas, and other vegetables.
Based on this scheme, immature legumes were classified as starchy vegetables, and mature
legumes (also known as dry beans and peas) were classified as beans and peas.
Forty-eight Scenario diets were developed using the most nutrient dense F&V from each
subgroup, as informed by the nutrient density index. These Scenario diets differed from one
another by four factors: the number of servings, the proportion of total vegetables in each
vegetable subgroup, the relative nutrient density of the F&V in each diet, and consumer
preference for the F&V in each diet (Fig. 1). First we describe the naming scheme for the
Scenario diets, then we describe how each of these four factors was used to create the Scenario
diets.
55
USDA Food
Pattern
Vegetable subgroup CUdp UUdp DUdp
proportions
Current diet
CCdp UCdp DCdp
Current diet USDA Food DASH diet

Pattern
Number of F&V servings
Fig. 1 - Attributes of Scenario diet series

All scenario diets include nutrient dense F&V
F&V, fruits and vegetables
USDA, United States Department of Agriculture
DASH, Dietary Approaches to Stop Hypertension
dRepresents nutrient density in the following levels: 1, 2, 3, 4, where 1 = most nutrient dense
F&V, 2 = second most nutrient dense F&V, 3 = third most nutrient dense F&V, 4 = fourth
most nutrient dense F&V (see Table 3 for list of F&V)
pRepresents consumer preference: 1,2, where 1 = all nutrient dense F&V, 2 = consumer
preferred nutrient dense F&V
The naming scheme for the Scenario diets is as follows: the first letter represents the number of
servings (C is the number of servings in the Current diet, U is the number of servings in the
USDA Food Pattern [43], and D is the number of servings in the DASH diet [44]); the second
letter represents the proportion of total vegetables in each vegetable subgroup (C is the vegetable
subgroup proportions in the Current diet and U is vegetable subgroup proportions in the USDA
Food Pattern [43]); d represents the relative nutrient density of the F&V in each diet (four levels
of nutrient density were included), and p represents whether the F&V included in a given diet
56
were all nutrient dense F&V or only nutrient dense F&V with positive consumption trends from
2007-2011[45]. Consumption trends were indicated by the sign of a regression coefficient using
the five time points to predict the consumption trend; we refer to the slope without the associated
P value because the sample size was only five time points within the time period that we
studied). The Scenario diets were grouped into six series, and each series contains eights diets.
For example, series CCdp includes eight Scenario diets with the number of servings and
vegetable subgroup proportions held constant. Four diets in the series (numbered 1 through 4)
include the most nutrient dense F&V (diet 1 is composed of the most nutrient dense F&V in each
subgroup, diet 2 is composed of the next most nutrient dense F&V in each subgroup, and so on);
and four diets (numbered 1 through 4) include only nutrient dense F&V with positive
consumption trends (diet 1 is composed of the most nutrient dense F&V in each subgroup, diet 2
is composed of the next most nutrient dense F&V in each subgroup, and so on). Subsequent
sections describe how each of these factors (the number of servings, the proportion of total
vegetables in each vegetable subgroup, the relative nutrient density of the F&V in each diet, and
consumer preference for the F&V in each diet) was used to create the Scenario diets.
The number of servings for each Scenario diet were modeled after the Current diet (i.e. sample of
hypertensive adults from NHANES, discussed above), the USDA Food Pattern [43], and the
DASH diet [44] (Table 1). The number of servings of the USDA Food Pattern was indexed to the
mean daily energy intake of the sample (2,300 kcal) by adopting the midpoint values of the
ranges reported for 2,200 kcal/day and 2,400 kcal/day. For the DASH diet, the midpoint values
of the ranges reported for 2,000kcal/day and 2,600 kcal/day were adopted, and the recommended
number of servings of nuts, seeds, and legumes in the DASH diet was added to the vegetables
57
group (the USDA Food Pattern already includes legumes in the vegetables group). The vegetable
subgroup proportions were modeled after the Current diet and the USDA Food Pattern [43]
(Table 1).
Table 1 - Number of servings and vegetable subgroup

proportions of the current diet, USDA Food Pattern, and DASH
diet
Food group/ USDA Food
a b c c
subgroup Current diet Pattern DASH diet
Fruit 2.08 (2.85) 4.00 5.00
Choice 1 0.50 (n/a) 0.50 n/a
Choice 2 0.50 (n/a) 0.50 n/a
Vegetables 2.46 (2.02) 6.00 5.86
Dark-green 0.09 (n/a) 0.10 n/a
Red and orange 0.21 (n/a) 0.29 n/a
Beans and peas 0.05 (n/a) 0.10 n/a
Starchy 0.23 (n/a) 0.29 n/a
Other 0.42 (n/a) 0.24 n/a
USDA, United States Department of Agriculture

n/a, not applicable
a
Values for food groups represent number of servings, values for
subgroups represent proportions of parent food group
b
Values are mean (SD)
c
Number of servings indexed to the mean daily energy intake of the
study population (2,300 kcal/day)
Each of the 48 Scenario diets contains two fruits and five vegetables (one vegetable for each
subgroup). The two fruits with the highest and second highest nutrient densities, and the
vegetables with the highest nutrient densities in each subgroup, were assigned to Scenario diet 1
in each series. The two fruits with the third and fourth highest nutrient densities, and the
vegetables with the second highest nutrient densities in each subgroup, were assigned to Scenario
58
diet 2 in each series. This assignment scheme was extended to Scenario diets 3 and 4 in each
series. Although F&V with positive consumption trends do not necessarily represent the most
commonly consumed F&V, they represent those that are most likely to be adopted by consumers
as food patterns change.
3. Results
In this sample of hypertensive adults, nutrient intake from F&V accounted for 16-57% of total
fiber, Ca, Mg, and K intake, with Ca representing the lower bound and fiber representing the
upper bound (Table 2). The nutrient gap represents the difference between the DASH nutrient
target and total nutrient intake from all non-F&V food groups. For example, the nutrient gap for
fiber is 25 grams, which means that nutrient dense F&V consumption would need to provide at
least 25 grams of fiber in order to close the nutrient gap. Expressed as a proportion of the DASH
diet nutrient targets, the nutrient gap was highest for fiber, followed by K, Mg, and Ca, in
descending order.
Table 2 - Nutrient intake among hypertensive adults compared to DASH diet nutrient targets, NHANES 2009-2010
Current intake
F&V nutrient intake, Nutrient gap,
Total nutrient F&V nutrient proportion of total Total nutrient DASH nutrient proportion of DASH
a a b c
Nutrient intake intake nutrient intake (%) intake, less F&V target Nutrient gap diet nutrient target (%)
Fiber (g) 17 (10) 10 (12) 57 7 32 25 77
Calcium (mg) 1059 (639) 173 (280) 16 886 1301 415 32
Magnesium (mg) 321 (148) 83 (95) 26 238 495 257 52
Potassium (mg) 2923 (1363) 1256 (1428) 43 1667 4844 3176 66
Values may not total due to rounding
a
Values are mean (SD)
b
Indexed to the mean daily energy intake of the study population (2,300 kcal/day)
c
DASH nutrient target - Total nutrient intake, less F&V
59
60
Table 3 displays the F&V included in the Scenario diets, which was informed by the nutrient
density index. Date, passion-fruit, spinach, sweet potato, lima bean, daikon, and white bean were
the most nutrient dense foods in their respective F&V subgroups. Of the foods with positive
consumption trends, date, kiwi, spinach, sweet potato, avocado, and great northern bean were the
most nutrient dense in their respective F&V subgroups. There were no F&V with a positive
consumption trend in the starchy vegetables subgroup, so the most nutrient dense starchy
vegetables were included in this subgroup despite having negative consumption trends. In the
green vegetables subgroup, there were no F&V with a positive consumption trend that had a
higher nutrient density than broccoli, so broccoli was listed in Scenario diet 3 and 4. Seven out of
twenty-eight (25%) of the most nutrient dense F&V had positive consumption trends. Prune juice
and tangerine juice were the only juices that were included in the Scenario diets.
Table 4a displays the nutrient contribution of all nutrient dense F&V in Scenario diet series CC
compared to the current intake of nutrients (i.e. nutrient content of the Sample diet). Scenario
diet series CC contains the same number of F&V servings and the same number of vegetable
subgroup proportions as the Sample diet. Scenario diet CC1 was the only diet that contained
higher amounts of all nutrients compared to the Sample diet. Table 4b displays the nutrient
contribution of the F&V with positive consumption trends in Scenario diet series CC compared
to the current intake of nutrients. None of the diets containing F&V with positive consumption
trends contained higher amounts of all nutrients compared to the current intake, although
Scenario diet CC1 was limited only by Ca.

Table 3 - Fruits and vegetables included in Scenario diets
Vegetables
Fruit 1 Fruit 2 Green Red and orange Starchy Other Beans and peas
Scenario Nutrient
diet density All F&V
1 High Date Passion-fruit Spinach Sweet potato Lima bean* Daikon White bean
*
2 Guava Fig Beet green Carrot Cowpea Avocado Navy bean
*
3 Prune Kiwi Turnip green Pumpkin Edamame Okra French bean
*
4 Low Prune juice Currant Chard Tomato Green pea Parsnip Pink bean
F&V with positive consumption trend
1 High Date Kiwi Spinach Swt. potato Lima bean* Avocado Great N. bean
2 Orange Tangerine Collard Tomato Cowpea* Okra Black bean
*
3 Lemon Papaya Broccoli Winter squash Edamame Artichoke Lima bean
*
4 Low Strawberry Tangerine juice Broccoli Red bell pepper Green pea Asparagus Split pea
Great N., Great Northern
*
Immature (harvested from the plant before reaching maturity)
61
62
Table 4a - Nutrient contribution of Scenario diet series CC (all

F&V) compared to current intake
Current Series CC, all F&V
Nutrient intake 1 2 3 4
Fiber (g) 10 29* 18* 12* 10*
Calcium (mg) 173 222* 156 180* 112
Magnesium (mg) 83 166* 110* 115* 92*
Potassium (mg) 1256 2094* 1245 1192 1272*

*Equal to or greater than current intake
Table 4b - Nutrient contribution of Scenario diet series CC

(consumer preferred F&V) compared to current intake
a
Current Series CC, consumer preferred F&V
Nutrient intake 1 2 3 4
Fiber (g) 10 21* 11* 12* 8
Calcium (mg) 173 142 216* 108 84
Magnesium (mg) 83 142* 101* 96* 56
Potassium (mg) 1256 1808* 922 909 786
Subscripts represent diets that include all fruits and vegetables (a) and
diets that include only fruits and vegetables with a positive
a
Consumer preference was determined based on the presence of a
positive consumption trend from 2007-2011
*Equal to or greater than current intake
In Table 5a, the nutrient contribution from all nutrient dense F&V in each Scenario diet was
compared to the nutrient gap for each nutrient. In series CC and CU, diet 1 closed the nutrient
gap for fiber, but there were no diets that closed all of the nutrient gaps. In series UC, UU, DC,
and DU, diet 1 was the only diet that closed all of the nutrient gaps. All of the diets in series UC,
UU, DC, and DU closed at least one nutrient gap except for Scenario diet 4 in Series UC and
UU. Scenario diets composed of the same number of servings but different vegetable subgroup
63
proportions (series CC vs. series CU, series UC vs. series UU, and series DC vs. series DU)
showed no differential effect in closing the nutrient gaps.
Table 5a - Nutrient contribution of Scenario diets, all F&V

Series CC
Nutrient Nutrient gap 1 2 3 4
Fiber (g) 25 29* 18 12 10
Calcium (mg) 415 222 156 180 112
Magnesium (mg) 257 166 110 115 92
Potassium (mg) 3176 2094 1245 1192 1272
Nutrient Nutrient gap Series CU
Fiber (g) 25 28* 17 12 9
Calcium (mg) 415 186 155 161 106
Magnesium (mg) 257 159 101 105 86
Potassium (mg) 3176 1906 1202 1177 1216
Nutrient Nutrient gap Series UC
Fiber (g) 25 61* 40* 28* 24
Calcium (mg) 415 517* 345 416* 251
Magnesium (mg) 257 368* 249 264* 208
Potassium (mg) 3176 4621* 2757 2578 2781
Nutrient Nutrient gap Series UU
Fiber (g) 25 62* 41* 30* 23
Calcium (mg) 415 461* 379 399 258
Magnesium (mg) 257 387* 259* 269* 220
Potassium (mg) 3176 4428* 2881 2759 2845
Nutrient Nutrient gap Series DC
Fiber (g) 25 69* 43* 30* 25*
Calcium (mg) 415 528* 372 430* 266
Magnesium (mg) 257 396* 262* 275* 219
Potassium (mg) 3176 5000* 2972 2849 3038
Nutrient Nutrient gap Series DU
Fiber (g) 25 70* 45* 33* 25*
Calcium (mg) 415 474* 405 414 274
Magnesium (mg) 257 414* 272* 280* 231
Potassium (mg) 3176 4811* 3093 3025 3101
*
Equal to or greater than the nutrient gap
64
Table 5b is similar to 5a, except that the Scenario diets include only nutrient dense F&V with
positive consumption trends. There were no diets in series CC and CU that closed any nutrient
gaps. In series UC, UU, DC, and DU, diet 1 closed all of the nutrient gaps except for Ca.
Scenario diets composed of the same number of servings but different vegetable subgroup
proportions showed no differential effect in closing the nutrient gaps.
Using the most nutrient dense F&V included in Scenario diet 1, the least number of servings
needed to close all of the nutrient gaps while maintaining the daily fruit-to-vegetable servings
ratio and the vegetable subgroup proportions of the Sample diet, was 8.9 (data not shown; 56 g
fiber, 432 g Ca, 323 mg Mg, and 4081 mg K); this was 11% lower than the number of servings
in series UC and UU, and 18% lower than the number of servings in series DC and DU. Using
only the nutrient dense F&V with positive consumption trends included in diet 1, the least
number of servings needed to close all of the nutrient gaps was 14 (data not shown; 64 g fiber,
437 g Ca, 437 mg Mg, and 5579 mg K).

65
Table 5b - Nutrient contribution of Scenario diets, consumer

preferred F&V
Series CC
Nutrient Nutrient gap 1 2 3 4
Fiber (g) 25 21 11 12 8
Calcium (mg) 415 142 216 108 84
Magnesium (mg) 257 142 101 96 56
Potassium (mg) 3176 1808 922 909 786
Nutrient Nutrient gap Series CU
Fiber (g) 25 18 10 11 7
Calcium (mg) 415 145 197 105 77
Magnesium (mg) 257 136 89 87 53
Potassium (mg) 3176 1684 917 853 723
Nutrient Nutrient gap Series UC
Fiber (g) 25 46* 24 28* 18
Calcium (mg) 415 316 490* 241 184
Magnesium (mg) 257 321* 234 221 125
Potassium (mg) 3176 3991* 2080 2066 1722
Nutrient Nutrient gap Series UU
Fiber (g) 25 44* 26* 28* 20
Calcium (mg) 415 351 475* 255 181
Magnesium (mg) 257 336* 236 232 135
Potassium (mg) 3176 3930* 2308 2175 1722
Nutrient Nutrient gap Series DC
Fiber (g) 25 50* 25* 30* 18
Calcium (mg) 415 338 514* 258 200
Magnesium (mg) 257 338* 240 228 133
Potassium (mg) 3176 4317* 2201 2169 1878
Nutrient Nutrient gap Series DU
Fiber (g) 25 47* 29* 31* 22
Calcium (mg) 415 372 503* 273 232
Magnesium (mg) 257 353* 236 245 158
Potassium (mg) 3176 4258* 2334 2347 2225
*
Equal to or greater than the nutrient gap
66
4. Discussion
Hypertensive adults can substantially increase the nutrient content of their diets by choosing the
most nutrient dense F&V instead of those currently consumed. However, neither choosing more
nutrient dense F&V nor adopting the vegetable subgroup proportions of the USDA Food Pattern
can close all of the nutrient gaps without simultaneously increasing the number of servings
consumed. By consuming only the most nutrient dense F&V, individuals can close all of the
nutrient gaps with 8.9 servings, which is less than what is recommended by the DASH diet (more
than 10 servings). However, increasing consumption from 4.5 servings (Sample diet) to 8.9
servings may be challenging for most people. This would offer very few F&V to choose from,
even though three of the seven F&V with increasing trends of consumption were also the most
nutrient dense overall.
In the Scenario diets that included only nutrient dense F&V with positive consumption trends,
diet 1 in each series (with the exception of series CC and CU) would have closed all of the
nutrient gaps if it were not limited by Ca. Although some F&V are good sources of Ca
(particularly dark green leafy vegetables), F&V accounted for only 18% of total Ca intake in the
Sample diet. Compared to F&V, dairy foods contain more Ca [46] and it is more bioavailable
[47]. Fulgoni III et al. [48] show that meeting Ca recommendations without increasing dairy
consumption is difficult to achieve.
The nutrient contribution from non-F&V was held constant throughout the range of Scenario
diets we explored, in order to isolate the effects of F&V choice on nutrient intake. This approach
67
precluded substituting F&V for other food groups, which would be needed to reduce the
consumption of total fat, saturated fat, and sodium, all of which were higher in this sample (34%
of total kcal, 11% of total kcal, and 3,772mg, respectively) than what is recommended by the
DASH diet (27% of total kcal, 6% of total kcal, and 2,300mg, respectively). Mellen et al. [14]
observed similar nutrient intake patterns among hypertensive adults in 1999-2004.
The limitations associated with using 24-hour recalls (24HR) to estimate food consumption
include participants intentionally misreporting their consumption patterns to simplify the survey
process or to impress the survey administrator [49, 50], and participants’ inability to accurately
recall food consumption from memory. Additional limitations have been described in detail
elsewhere [50, 51].
Nicklas and others [52] have noted the substantial methodological variation in studies examining
nutrient density. Nutrient density can be computed as the amount of nutrients per kcal, per
weight of food, or per serving size, and this decision ultimately reflects how the nutrient density
score will be utilized. We defined nutrient density as the amount of nutrients per serving size
because it is likely the easiest unit of measurement for consumers to understand.
In developing the nutrient density index we assigned equal weights to each nutrient, an approach
that has precedence in DASH diet adherence studies [53, 54]. However, some of the Scenario
diets have very high amounts of fiber (upwards of 50 grams), and the bulky nature of fiber can
limit the amount of food that can be consumed comfortably [55]. Stewart et al. [56] observed
slightly higher self-reported rates of bloating, cramping, flatulence, and stomach noises
68
associated with diets containing ~30 grams of fiber versus those containing less than 20 grams of
fiber. The nutrient density index used in the present study could be improved by reducing the
weight of fiber relative to the other nutrients of interest, although more work is needed to
determine the nature of a more complex weighting scheme.
This research addresses the National Heart, Lung, and Blood Institutes critical research need to
develop strategies for clinicians to help adults adopt CVD related diet recommendations [20].
We show that diets that include nutrient dense F&V can provide substantially more fiber, Ca,
Mg, and K than the Current diet. We also show that although the DASH diet nutrient targets
cannot be met without also increasing the number of servings of F&V consumed, choosing
nutrient dense F&V offers a way to meet the nutrient targets with fewer servings than is currently
recommended by the DASH diet. For hypertensive adults who struggle to increase their
consumption of F&V, consuming nutrient dense F&V will bring them closer to achieving the
DASH diet nutrient targets and may be an achievable first step toward modifying their
consumption of additional food groups. Therefore, we suggest that clinicians recommend that
their hypertensive patients consume nutrient dense F&V as a way to substantially increase their
intake of key nutrients (i.e. fiber, Ca, Mg, and K). For patients who are able to make additional
modifications to their diet, clinicians should emphasize the added benefits of replacing foods that
contain high amounts of total fat, saturated fat, and sodium (such as fast food and snacks) with
low-fat dairy products.

69
Acknowledgement
The authors thank Dr. Edward Saltzman for his helpful comments. This research was not
supported by a specific funding agency. The authors do not have any conflicts of interest to
declare.
References
[1] US Department of Health and Human Services, National Institutes of Health. 2013.
Managing blood pressure in adults: Systematic evidence review from the Blood Pressure Expert
Panel, 2013. National Heart, Lung, and Blood Institute. Available at website:
http://www.nhlbi.nih.gov/health-pro/guidelines (verified 11 Dec 2014).
[2] Lloyd-Jones D, Adams RJ, Brown TM, Carnethon M, Dai S, De Simone G, et al. Heart
disease and stroke statistics-2010 update: A report from the American Heart Association.
Circulation 2010; 121:e46-e215.
[3] Lewington S, Clarke R, Qizilbash N, Peto R, Collins R. Age-specific relevance of usual
blood pressure to vascular mortality: A meta-analysis of individual data for one million adults in
61 prospective studies. Lancet 2002; 360:1903-13.
[4] US Department of Health and Human Services, Centers for Disease Control and Prevention.
2010. Leading causes of death. Available at website:
www.cdc.gov/injury/wisqars/leading_causes_death.html (verified 8 July 2013).
[5] Guo F, He D, Zhang W, Walton RG. Trends in prevalence, awareness, management, and
control of hypertension among United States adults, 1999 to 2010. J Am Coll Cardiol 2012;
60:599-606.
[6] James PA, Oparil S, Carter BL, Cushman WC, Dennison-Himmelfarb C, Handler J, et al.
2014 evidence-based guideline for the management of high blood pressure in adults: Report from
the panel members appointed to the Eighth Joint National Committee (jnc 8). JAMA 2014;
311:507-20.
[7] Karanja NM, Obarzanek E, Pao-Hwa L, McCullough ML, et al. Descriptive characteristics
of the dietary patterns used in the Dietary Approaches to Stop Hypertension trial. J Am Diet
Assoc 1999; 99:S19-27.
[8] Appel LJ, Moore TJ, Obarzanek E, Vollmer WM, Svetkey LP, Sacks FM, et al. A clinical
trial of the effects of dietary patterns on blood pressure. N Engl J Med 1997; 336:1117-24.
[9] Hikmat F,Appel LJ. Effects of the DASH diet on blood pressure in patients with and without
metabolic syndrome: Results from the DASH trial. J Hum Hypertens 2014; 28:170-5.
[10] Saneei P, Salehi-Abargouei A, Esmaillzadeh A, Azadbakht L. Influence of Dietary
Approaches to Stop Hypertension (DASH) diet on blood pressure: A systematic review and
meta-analysis on randomized controlled trials. Nutrition, Metabolism and Cardiovascular
Diseases 2014; 24:1253-61.
[11] Reedy J, Krebs-Smith SM, Miller PE, Liese AD, Kahle LL, Park Y, et al. Higher diet
quality is associated with decreased risk of all-cause, cardiovascular disease, and cancer
mortality among older adults. The Journal of Nutrition 2014; 144:881-9.
70
[12] Salehi-Abargouei A, Maghsoudi Z, Shirani F, Azadbakht L. Effects of Dietary Approaches

to Stop Hypertension (DASH)-style diet on fatal or nonfatal cardiovascular diseases—incidence:
A systematic review and meta-analysis on observational prospective studies. Nutrition 2013;
29:611-8.
[13] Liese AD, Krebs-Smith SM, Subar AF, George SM, Harmon BE, Neuhouser ML, et al. The
Dietary Patterns Methods Project: Synthesis of findings across cohorts and relevance to dietary
guidance. J Nutr 2015; 145:393-402.
[14] Mellen PB, Gao SK, Vitolins MZ, Goff DC, Jr. Deteriorating dietary habits among adults
with hypertension: DASH dietary accordance, NHANES 1988-1994 and 1999-2004. Arch Intern
Med 2008; 168:308-14.
[15] Kwan MW-M, Wong MC-S, Wang HH-X, Liu KQ-L, Lee CL-S, Yan BP-Y, et al.
Compliance with the Dietary Approaches to Stop Hypertension (DASH) diet: A systematic
review. PloS one 2013; 8:78412.
[16] Hirsch G, Homer J, Trogdon J, Wile K, Orenstein D. Using simulation to compare 4
categories of intervention for reducing cardiovascular disease risks. Am J Public Health 2014;
104:1187-95.
[17] US Department of Health and Human Services, National Institutes of Health. 2004. The
Seventh Report of the Joint National Committee on Prevention, Detection, Evaluation, and
Treatment of High Blood Pressure. National Heart, Lung, and Blood Institute. NIH Publication
No. 04-5230. Available at website: www.nhlbi.nih.gov/health-pro/guidelines (verified 16 Dec
2014).
[18] Hsiao C-J, Cherry DK, Beatty PC, Rechtsteiner EA. 2010. National Ambulatory Medical
Care Survey: 2007 summary. National Health Statistics Reports. US Department of Health and
Human Services, Centers for Disease Control and Prevention (CDC). Number 27. Available at
website: http://www.cdc.gov/nchs/products/nhsr.htm (verified 12 Dec 2014).
[19] Persoskie A, Kaufman AR, Leyva B. Receiving and adhering to lifestyle modification
counseling for hypertension: Disparities between smokers and nonsmokers. J Clin Hypertens
2014; 16:429-36.
[20] US Department of Health and Human Services, National Institutes of Health. 2013.
Lifestyle interventions to reduce cardiovascular disease: Systematic evidence review from the
Lifestyle Work Group, 2013, p.45. National Heart, Lung, and Blood Institute. Available at
website: www.nhlbi.nih.gov/health-pro/guidelines/ (verified 21 Nov. 2014).
[21] Vogt TM, Appel LJ, Obarzanek EVA, Moore TJ, Vollmer WM, Svetkey LP, et al. Dietary
Approaches to Stop Hypertension: Rationale, design, and methods. J Am Diet Assoc 1999;
99:S12-S8.
[22] Krebs-Smith SM, Guenther PM, Subar AF, Kirkpatrick SI, Dodd KW. Americans do not
meet federal dietary recommendations. J Nutr 2010; 140:1832-8.
[23] Ratner RK,Riis J. Communicating science-based recommendations with memorable and
actionable guidelines. Proc Natl Acad Sci 2014; 111:13634-41.
[24] Hills AP, Byrne NM, Lindstrom R, Hill JO. 'Small changes' to diet and physical activity
behaviors for weight management. Obesity Facts 2013; 6:228-38.
[25] Lin P-H, Aickin M, Champagne C, Craddick S, Sacks FM, McCarron P, et al. Food group
sources of nutrients in the dietary patterns of the DASH-sodium trial. J Am Diet Assoc 2003;
103:488-96.
(CDC). NHANES 2009-2010, dietary interview: Individual foods file, food code description file.
71
Available at website: wwwn.cdc.gov/nchs/nhanes/search/nhanes09_10.aspx (verified 27 Jan.

2015).
[27] Moshfegh AJ, Rhodes DG, Baer DJ, Murayi T, Clemens JC, Rumpler WV, et al. The US
Department of Agriculture automated multiple-pass method reduces bias in the collection of
energy intakes. Am J Clin Nutr 2008; 88:324-32.
(14 October 2014).
[30] USDA Center for Nutrition Policy and Promotion (CNPP). 2015. Choosemyplate.Gov.
[31] Thompson FE, Willis GB, Thompson OM, Yaroch AL. The meaning of "fruits" and
"vegetables". Public Health Nutr 2011; 14:1222-8.
(CDC). NHANES 2009-2010, reproductive health file. Available at website:
wwwn.cdc.gov/nchs/nhanes/search/nhanes09_10.aspx (verified 27 Jan. 2015).
(CDC). NHANES 2009-2010, demographics file. Available at website:
(CDC). NHANES 2009-2010, examination data: Blood pressure. Available at website:
[35] Appel LJ. ASH position paper: Dietary approaches to lower blood pressure. J Am Soc
Hypertens 2010; 4:79-89.
[36] Lin P-H, Windhauser MM, Plaisted CS, Hoben KP, McCullough ML, Obarzanek EVA.
The linear index model for establishing nutrient goals in the Dietary Approaches to Stop
Hypertension trial. J Am Diet Assoc 1999; 99:S40-S4.
[37] US Department of Agriculture, Agricultural Research Service. USDA National Nutrient
[38] US Department of Agriculture, National Agricultural Statistics Service (NASS). 2007.
Census of Agriculture. US Government Printing Office, Washington, DC. Available at Website:
http://www.agcensus.usda.gov/Publications/2007/index.php (verified 17 April 2012).
[39] US Department of Agriculture, Economic Research Service (ERS). Food Availability (per
Capita) data system. Available at website: www.ers.usda.gov/data-products/food-availability-
(per-capita)-data-system.aspx (verified 11 Nov. 2014).
Dietary Guidelines for Americans 2005, Table 1. 6th ed. US Government Printing Office,
Washington, DC. Available at website: http://health.gov/dietaryguidelines/ (verified 03 Nov.
2014).
[41] US Department of Health and Human Services. 2006. Your guide to lowering your blood
pressure with DASH. Page 8. National Heart, Lung, and Blood Institute. NIH Publication No.
06-4082. Available at website: www.nhlbi.nih.gov/health/health-topics/topics/dash/ (verified 17
Dec 2014).
72
[42] Swain JF, Windhauser MM, Hoben KP, Evans MA, McGee BB, Steele PD. Menu design
and selection for multicenter controlled feeding studies: Process used in the Dietary Approaches
to Stop Hypertension trial. J Am Diet Assoc 1999; 99:S54-S9.
Dietary Guidelines for Americans 2010, Appendix 7. 7th ed. U.S. Government Printing Office,
Washington, DC. Available at website: http://health.gov/dietaryguidelines/ (verified 27 Dec.
2014).
Dietary Guidelines for Americans 2010, Appendix 10. 7th ed. U.S. Government Printing Office,
Washington, DC. Available at website: http://health.gov/dietaryguidelines/ (verified 27 Dec.
2014).
[45] US Department of Agriculture, Economic Research Service (ERS). 2007-2011. Loss-
adjusted Food Availability (LAFA) data series. Available at website: www.ers.usda.gov/data-
products/food-availability-(per-capita)-data-system/loss-adjusted-food-availability-
documentation.aspx (verified 18 Mar. 2015).
[46] Nicklas TA, O'Neil CE, Fulgoni VL. The role of dairy in meeting the recommendations for
shortfall nutrients in the American diet. J Am Coll Nutr 2009; 28:73S-81S.
[47] Institute of Medicine, National Academy of Sciences. Dietary Reference Intakes. Available
at website: http://fnic.nal.usda.gov/dietary-guidance/dietary-reference-intakes (verified 25 March
2013).
[48] Fulgoni III VL, Keast DR, Auestad N, Quann EE. Nutrients from dairy foods are difficult
to replace in diets of Americans: Food pattern modeling and an analyses of the National Health
and Nutrition Examination Survey 2003-2006. Nutr Res 2011; 31:759-65.
[49] Roark RA,Niederhauser VP. Fruit and vegetable intake: Issues with definition and
measurement. Public Health Nutr 2013; 16:2-7.
[50] Woodside JV, Young IS, McKinley MC. Fruits and vegetables: Measuring intake and
encouraging increased consumption. P Nutr Soc 2013; 72:236-45.
[51] Krebs-Smith SM,Kantor LS. Choose a variety of fruits and vegetables daily: Understanding
the complexities. The Journal of Nutrition 2001; 131:487S-501S.
[52] Nicklas TA, Drewnowski A, O’Neil CE. The nutrient density approach to healthy eating:
Challenges and opportunities. Public Health Nutr 2014; 17:2626-36.
[53] Powell-Wiley TM, Miller PE, Agyemang P, Agurs-Collins T, Reedy J. Perceived and
objective diet quality in US adults: A cross-sectional analysis of the National Health and
Nutrition Examination Survey (NHANES). Public Health Nutr 2014; 17:2641-9.
[54] Morton S, Saydah S, Cleary SD. Consistency with the Dietary Approaches to Stop
Hypertension diet among adults with diabetes. J Acad Nutr Diet 2012; 112:1798-805.
[55] National Academy of Sciences, Institute of Medicine. 2006. Dietary Reference Intakes: The
essential guide to nutrient requirements. Part II: Fiber (p.119) Jennifer J. Otten, Jennifer Pitzi
Hellwig, Linda D. Meyers (eds.). National Academies Press, Washington, DC.
[56] Stewart ML, Nikhanj SD, Timm DA, Thomas W. Evaluation of the effect of four fibers on
laxation, gastrointestinal tolerance, and serum markers in healthy humans. Ann Nutr Metab
2010; 56:91-8.
73
Title
Agricultural capacity to increase the production of selected nutrient dense fruits and vegetables
for treating hypertension
Abstract
Hypertension is a prominent risk factor for cardiovascular disease, and over 50% of US adults
are hypertensive or pre-hypertensive. The Dietary Approaches to Stop Hypertension (DASH)
diet, which features high amounts of fruits and vegetables (F&V), is associated with reduced risk
of hypertension, cardiovascular disease, and all-cause mortality. Evidence shows the important
role that nutrient dense F&V can have on achieving the nutrient targets of the DASH diet.
Despite the growing importance of international food markets, most F&V consumed in the US
are produced domestically. Therefore, if hypertensive adults consumed more nutrient dense F&V
domestic production would likely increase to some degree, but the agricultural capacity to meet
increased demand is not well documented. We examine the capacity of the US agricultural land
base to increase the per capita availability of nutrient dense F&V for hypertensive adults by 1)
identifying national production centers of nutrient dense F&V, 2) estimating the amount of
agricultural land suitable for the production of nutrient dense F&V in and around production
centers, and 3) computing the amount of nutrient dense F&V that can be produced from these
suitable land areas. A nutrient density index from a previous study was used to identify nutrient
dense F&V. National agricultural production centers of these F&V were identified using data on
harvested land area at the county level. Geospatial data on precipitation, temperature, soil depth,
soil type, and land use were used to estimate the amount of agricultural land suitable for the
production of these F&V. Data on crop yields were applied to data on the amount of suitable
74
agricultural land to estimate the production potential for each crop, which were adjusted for food
losses that occur throughout he supply chain, and divided by the population size of hypertensive
adults to estimate daily per capita availability. We identified 11 national production centers
located in six states. There is enough suitable agricultural land in (and adjacent to) existing
production centers to substantially increase the availability of nutrient dense F&V for
hypertensive adults in the US. Yet additional production of F&V (nutrient dense and otherwise)
would be needed in order to increase the availability of F&V to the point where hypertensive
adults can meet the DASH diet. These findings highlight the important role that smaller US
production centers that were not identified in this study, as well as international food markets,
will play if individuals adopt healthier eating patterns.
Keywords
Blood pressure, consumption, output, nutrition, GIS, model
Authors
Zach Conrad1, Christian Peters1, Kenneth Chui,2 Timothy Griffin1

1

2

Corresponding author: zach.conrad@tufts.edu
75
Introduction
Increased access to international food markets over the past several decades has greatly increased
the variety of foods available to consumers in the United States (US)1. For foods which follow
seasonal production patterns, imports are essential to minimizing temporal fluctuations in food
availability and price2, 3. Additionally, American consumers depend almost entirely on imports
for some of their favorite foods, like bananas, cocoa, and coffee4. Yet despite the growing
importance of international food markets, the US still produces most of the food (by volume) that
the population consumes4. Increased consumer demand for many types of food would likely spur
increased domestic production of agricultural goods. However, the capacity of US agriculture to
increase the output of specific foods to accommodate increased demand is not well documented.
One way that shifts in consumer food demand would occur is if a greater number of Americans
adhered to federal dietary recommendations. Adequate fruit and vegetable (F&V) consumption is
now a mainstay of the Dietary Guidelines for Americans5-9 because of the relationship between
regular consumption of these foods and the reduced risk of chronic disease10, 11; the evidence
base for reducing the risk of cardiovascular disease (CVD), the leading cause of death in the
United States12, is particularly strong10, 11, 13. However, despite federal public health messaging14,
15
and an ongoing national campaign16 to increase F&V consumption, there has not been an
improvement in consumption patterns for decades17 and Americans fall far short of meeting
recommended daily consumption amounts18. It therefore seems unlikely that non-targeted dietary
interventions delivered by federal agencies will encourage Americans to modify their dietary
patterns.
76
Compared to the general population, one group that may be more likely to modify their dietary
patterns is hypertensive adults. Hypertension is the most common primary diagnosis by
physicians19, and those who receive lifestyle modification counseling from their physician are
more likely to make changes to their diet than those who report not receiving such counseling20.
Regular consumption of F&V is strongly associated with the prevention and treatment of
hypertension11, which is a prominent risk factor for CVD21, 22. Clinical interventions are
projected to be one of the most effective means of reducing CVD-related mortality in the near
and long-term23. Clinicians are encouraged to recommend the Dietary Approaches to Stop
Hypertension (DASH) diet24, 25 for their hypertensive patients18, 19, which features F&V and low-
fat dairy because these foods are good sources of fiber, calcium (Ca), magnesium (Mg), and
potassium (K), which are nutrients that control blood pressure26, 27. Consistent evidence shows
the role of the DASH diet in normalizing blood pressure28-30 and reducing the risk of CVD and
all-cause mortality31-33. However, adhering to the DASH diet requires people to modify their
consumption of most (or all) food groups, which is difficult for many people to achieve34, 35.
Fortunately, evidence shows the important role that nutrient dense F&V can have on achieving
the nutrient targets of the DASH diet (Conrad et al., forthcoming), and clinical recommendations
for hypertensive patients to incorporate more nutrient dense F&V into their current dietary
patterns may be warranted.
If hypertensive adults consumed more nutrient dense F&V, domestic production would likely
increase to some degree because most of the fruits and nuts (~60%) and vegetables (~80%)
consumed in the US (by volume) are produced domestically.4 While yield improvements from
77
technological innovations (e.g. improved disease and pest resistance and improved fruit set) and
the adoption of intensive management practices (e.g. high density planting and global positioning
systems) are largely responsible for recent increases in F&V production, the viability of these
approaches is dependent on the availability of suitable agricultural land36, 37. In fact, in some
cases, yield improvements were the result of relocation to more productive cropland36, 37.
Because many F&V crops require highly specific growing conditions, identifying suitable
agricultural sites is of primary importance to increasing production.
A growing body of research has examined the connection between F&V demand (actual and
potential), agricultural land use, and food production at different spatial scales. Giombolini et
al.38 found that the Willamette Valley in Oregon produced enough food to meet approximately
20-30% and 10-30% of the Willamette Valley population’s fruit and vegetable
recommendations, respectively, as part of a complete diet. Kremer and Shrueder39 estimated that
the Philadelphia foodshed (which represents 37 counties around Philadelphia that have farms that
supply food to the city’s residents) produces more than enough F&V to meet the current and
recommended consumption amounts of the city’s residents. In Detroit Michigan, Colasanti and
Hamm40 found that vacant land areas could produce enough F&V to meet 42% and 76% of
current fresh fruit and vegetable consumption, respectively, of the city’s residents, using food
storage and season extension technologies. Peters et al.41 estimated that if the share of the
population in New York State consuming local food increased, it was more likely that land in the
state would be allocated to higher value crops like F&V at the expense of grains and dairy. At
the national level, Young and Kantor42 estimated that an additional 5-7 million acres (2-2.8
million hectares) of land would be need to be cultivated in order for the population to meet F&V
78
consumption recommendations; later, Buzby et al.43 updated this estimate to 13 million acres
(5.3 million hectares). These studies present useful methodologies for estimating the agricultural
capacity to increase the production of F&V for the general population at different spatial scales.
To the best of our knowledge, our study is the first to examine the agricultural capacity to
increase the production of specific F&V for a specific population at the national level:
hypertensive adults. Our approach is innovative in that it uses a series of geospatial models that
are parameterized for precipitation, temperature, soil depth, soil type, and land availability. In
addition, we recognize that supply chain infrastructure is needed to accommodate increased
agricultural production, so we focus on whether agricultural production can be expanded in and
around current production centers because these co-locate with the requisite supply chain
infrastructure.
Increasing F&V consumption is a crucial step toward reducing the risk of hypertension and CVD
among the US population. Evidence shows the important role that nutrient dense F&V can play
in achieving the nutrient targets of the DASH diet (Conrad et al., forthcoming). Increased
demand for nutrient dense F&V would encourage increased domestic production, yet the limits
of production have not been fully examined. In the present study, our goal is to examine the
capacity of the US agricultural land base to increase the per capita availability of selected
nutrient dense F&V for hypertensive adults. This will be achieved by 1) identifying national
production centers of nutrient dense F&V, 2) estimating the amount of agricultural land suitable
for the production of nutrient dense F&V in and around production centers, and 3) computing the
amount of nutrient dense F&V that can be produced from these suitable land areas.
79
Methods
To identify national production centers of nutrient dense F&V, we used a nutrient density index
from a previous study to identify nutrient dense F&V. National agricultural production centers of
these F&V were identified using data on harvested land area at the county level. To estimate the
amount of agricultural land suitable for the production of nutrient dense F&V in and around
production centers, growing conditions for each crop were identified through published reports
and subsequently verified by personal communication with farm advisors and plant scientists
from the Cooperative Extension System. Geospatial data on precipitation, temperature, soil
depth, soil type, and land use were collected, and these data were used to construct geospatial
models to estimate the amount of suitable agricultural land for each F&V. To compute the
amount of nutrient dense F&V that can be produced from these suitable land areas, data on crop
yields were applied to data on the amount of suitable agricultural land to estimate the production
potential for each crop. These data were adjusted for food losses that occur throughout the supply
chain, and divided by the population size of hypertensive adults to estimate daily per capita
availability. Each of these components is described in detail below.
Identifying nutrient dense fruits and vegetables
The F&V included in this analysis were selected based on the following criteria: 1) these
represented the most nutrient dense F&V in each of the MyPlate subgroups44, 2) these displayed
positive trends of consumption during the most recent five-year period at the time the data were
collected (2007-2011),45 and 3) these displayed positive national production trends during the
80
most recent five-year period at the time the data were collected (2009-2013); in some cases a
lack of data availability resulted in data being collected from the most recent four-year period
(2010-2013).46
Nutrient dense F&V were identified using an indexing method developed by Conrad et al.
(forthcoming) that computed a nutrient density score for each of 407 F&V47 on the basis of fiber,
Ca, Mg, and K per serving. For each F&V, a positive consumption trend was identified based on
the slope of a simple regression equation that predicted consumption amount from the associated
year.45 We assumed that F&V with positive consumption trends were more likely to be adopted
by consumers than those with flat or negative consumption trends. For each F&V, a positive
trend of national production was identified based on the slope of a simple regression equation
that predicted production amount from the associated year.46 We assumed that F&V with
positive production trends were more likely to experience an increase in domestic production as
a result of increased demand compared to those with flat or negative production trends. The most
nutrient dense vegetables in each MyPlate subgroup44 with positive trends of consumption and
production were included in this analysis. Because the fruit group does not include subgroups,
the top two most nutrient dense fruits with positive trends of consumption and production were
included in this group in order to provide dietary variety. In total, two fruits and five vegetables
were included in this analysis.

81
Establishing production centers
The complex and specialized nature of F&V supply chains means that supply chain enterprises
co-locate with areas of high agricultural production48 (hereafter, production centers), and the
difficulties of relocating or replicating this infrastructure can be prohibitive. We therefore
recognize that increased cultivation of nutrient dense F&V is most likely to occur near current
production centers, so we focus our analysis on these areas. For each crop included in this
analysis, we identified production centers as counties that represented at least 10% of mean
(2002-2012) national acreage, or represented at least 10% of mean (2002-2012) state acreage if
the state contained the highest proportion of mean (2002-2012) national acreage.46 Production
centers were identified based on acreage rather than output because of limitations on data
availability of output data at the county level. In cases where production centers abutted one
another, the counties were considered to be a single production center. In order to examine the
potential to increase food production under scenarios of increased cultivated land area, we
constructed a series of buffer areas around each production center that extended 5km, 10km,
15km, and 20km from all sides of each production center (in other words, increasing the area
while maintaining the shape).
Mapping suitable land areas
For each nutrient dense F&V in each production center, a site suitability geographic information
system (ArcGIS 10.1) model was created to identify suitable land areas for crop production in
(and around) each production center. For each crop model, data were mapped using a State Plane
82
North American Datum 1983 projected coordinate system that was particular to the given
production center. All data were analyzed in raster format with a resolution of 30m x 30m. In
addition to estimating the amount of suitable land area, cluster analysis (average nearest
neighbor) was performed to estimate whether suitable land areas were clustered, dispersed, or
spaced randomly; statistical significance was assessed at P≤0.05.
Land areas were considered to be suitable for the production of a given crop if all of the crop-
specific growing conditions were satisfied. Growing conditions for each crop were identified
through published reports49-58 and subsequently verified by personal communication with farm
advisors and plant scientists from the Cooperative Extension System (see Appendices D through
M). Growing conditions that are typically overcome by farm management practices were not
included. For example, degree of soil stratification was not included because this is typically
overcome with deep tillage. The following growing conditions were included in this analysis:
mean monthly (1981-2010) precipitation and mean monthly (1981-2010) surface temperature
(minimum and maximum) were collected from the Parameter-elevation Regressions on
Independent Slopes Model, maintained by the PRISM Climate Group at Orgon State
University59, 60; soil depth (minimum) and type (particle size) were collected from the Gridded
Soil Survey Geographic database, maintained by US Department of Agriculture (USDA) Natural
Resources Conservation Service61; and land use (fallow agricultural land in 2013) was collected
from the Cropland Data Layer, maintained by USDA National Agricultural Statistics Service62.
Only the growing conditions relevant to a given crop were included in a given crop model.
83
In cases where geospatial data lacked sufficient temporal resolution to be parameterized
according to the information provided by published reports and personal communication with
experts, the range of values for the production center of a given crop were used to represent the
growing conditions for that crop. For example, monthly lower and upper temperature bounds for
each crop in each production center were established by adopting the monthly minimum and
maximum temperature readings, respectively, in that production center. In cases where
geospatial data were available62, we identified clusters of land within counties in which at least
95% of a given crop was cultivated, and used these clusters to establish the monthly lower and
upper temperature bounds for that crop.
Estimating current and potential food availability
State-level yield data were collected from USDA Census of Agriculture for each crop63. Yield
data were applied to the land area data generated from the site suitability analysis in order to
estimate potential production in and around each production center, on a farm-weight basis.
These data were adjusted for food losses and waste that occur as food moves through the supply
chain, and were ultimately converted to cup-equivalents using the conversion factors provided by
the USDA Loss-Adjusted Food Availability (LAFA) data series45. For the purposes of this
analysis, we assumed that all potential food availability would be distributed amongst
hypertensive adults. Therefore, data on total food availability were divided by the hypertensive
adult population64 in order to estimate per capita consumption. State-level yield data for lima
beans was adopted from Griffin et al.65 due to lack of data availability.
84
Results
Dates, kiwis, broccoli, sweet potatoes, Great Northern beans, and mushrooms (agaricus) were
identified as the most nutrient dense F&V with positive trends of consumption and production
within their respective F&V subgroups (Table 1). In addition, lima beans (immature) were
included in this study to represent the starchy vegetables subgroup even though they did not
display a positive consumption trend and data were unavailable on the production trends of all
processing types; lima beans were included because there were no starchy vegetables with
positive consumption trends, and lima beans represented the most nutrient dense vegetable
within this subgroup. These seven crops were associated with 11 national production centers
located in six states. Three production centers included multiple counties: lima beans in Sussex
and Kent Counties, Delaware; sweet potatoes in Johnston, Nash, Sampson, and Wilson Counties,
Nebraska; and Great Northern beans in Scotts Bluff, Box Butte, and Morrill Counties, Nebraska.
In terms of the proportion of national acreage of a given crop in each production center, this
ranged from a high of 70% (Great Northern beans in Scotts Bluff, Box Butte, and Morrill
Counties, Nebraska) to a low of 11% (sweet potatoes in Merced County, California).

85
Table 1. National production centers of nutrient dense fruits and vegetables.

Production center
MyPlate Proportion of national
Crop subgroup County State acreage (%)
Fruits
Dates n/a Riverside California 55
Kiwis n/a Tulare California 35
Butte California 24
Vegetables
Broccoli Dark green Monterey California 39
Santa Barbara California 20
Lima beans Starchy Sussex and Kent Delaware 25
Sweet potatoes Red and orange Johnston, Nash, North Carolina 27
Sampson, and Wilson
Calhoun Mississippi 12
Merced California 11
Great Northern Beans and peas Scotts Bluff, Box Butte, Nebraska 70
beans and Morrill
Mushrooms, Other Chester Pennsylvania 29
agaricus
Figures 1-7 display the results of the GIS crop models. All models showed statistically
significant (P<0.001) clustering patterns, meaning that the location of suitable agricultural land
was not random or evenly dispersed.

86
Black dots represent suitable agricultural areas

Grey lines represent expansion of 5km, 10km,
15km, and 20km on all sides of county
Data are presented in vector (point) format for
visual purposes
Figure 1. Suitable agricultural land for

dates in Riverside County, California
b)
a)
Black dots represent suitable agricultural areas Black dots represent suitable agricultural areas
Grey lines represent expansion of 5km, 10km, Grey lines represent expansion of 5km, 10km,
15km, and 20km on all sides of county 15km, and 20km on all sides of county
Data are presented in vector (point) format for Data are presented in vector (point) format for
visual purposes visual purposes
Figure 2. Suitable agricultural land for kiwis in a) Tulare County, California and b) Butte
County, California
87
Black dots represent suitable agricultural areas Black dots represent suitable agricultural areas
Grey lines represent expansion of 5km, 10km, Grey lines represent expansion of 5km, 10km,
15km, and 20km on all sides of county 15km, and 20km on all sides of county
Data are presented in vector (point) format for Data are presented in vector (point) format for
visual purposes visual purposes
Figure 3. Suitable agricultural land for broccoli in a) Monterey County, California and b)
Santa Barbara County, California
88

visual purposes
Figure 4. Suitable agricultural land for

lima beans in Sussex and Kent
Counties, Delaware
b)
a)
c)
Black dots represent suitable agricultural areas Black dots represent suitable agricultural areas Black dots represent suitable agricultural areas
Grey lines represent expansion of 5km, 10km, Grey lines represent expansion of 5km, 10km, Grey lines represent expansion of 5km, 10km,
15km, and 20km on all sides of county 15km, and 20km on all sides of county 15km, and 20km on all sides of county
Data are presented in vector (point) format for Data are presented in vector (point) format for Data are presented in vector (point) format for
visual purposes visual purposes visual purposes
Figure 5. Suitable agricultural land for sweet potatoes in a) Johnston, Nash, Sampson, and Wilson Counties, North Carolina,
b) Calhoun County, Mississippi, and c) Merced County, California
89
90

15km, and 20km on all sides of county Grey lines represent expansion of 5km, 10km,
visual purposes Data are presented in vector (point) format for
visual purposes
Figure 6. Suitable agricultural land for Figure 7. Suitable agricultural land for
Great Northern beans in Scotts Bluff, mushrooms (agaricus) in Chester
Box Butte, and Morrill Counties, County, Pennsylvania
Nebraska
For each crop, suitable land area was identified under the no expansion scenario (i.e. additional
land that could be used within the production center), and the amount of suitable land increased
with every 5km expansion of the production centers (Table 2). For example, 6,000 ha of suitable
land were identified for dates under the no expansion scenario, and this increased to 7,300 ha,
8,600 ha, 9,800 ha, and 10,000 ha when the periphery of the production center was expanded by
5 km, 10 km, 15 km, and 20 km, respectively. Compared to the currently harvested land area, the
greatest proportional increase in suitable land area under the no expansion scenario was for kiwis
(590%), followed by dates (200%), sweet potatoes (190%), Great Northern beans (110%),
broccoli (20%), lima beans (20%), and mushrooms (20%).

91
Table 2. Current agricultural land area and potential agricultural land area suitable for the production of
nutrient dense fruits and vegetables.
2
Production center expansion
Currently No
1
Crop harvested expansion 5km 10km 15km 20km
103 ha3
Dates 3.0 6.0 7.3 8.6 9.8 10
Kiwis 1.7 10 15 19 25 31
Broccoli 54 11 12 15 21 28
Lima beans 19 3.8 4.3 5.1 6.3 8.1
Sweet potatoes 44 83 134 189 235 285
Great Northern beans 27 31 37 44 47 50
Mushrooms, agaricus 0.4 0.1 0.2 0.4 0.8 1.4
1
National mean (2002-2012), Census of Agriculture
2
Represents distance from all sides of production center
3
Values are cumulative
Table 3 displays the current daily per capita availability of nutrient dense F&V compared to the
potential daily per capita availability of these F&V if suitable agricultural land circumjacent to
each production center were brought into production. For example, the daily per capita
availability of kiwis, measured in cup-equivalents, was 0.016 under the no expansion scenario,
and this increased to 0.023, 0.031, 0.040, and 0.049 when the production center was expanded by
5 km, 10 km, 15 km, and 20 km, respectively. Compared to the current daily per capita
availability, the greatest proportional increase under the no expansion scenario was for kiwis
(1,432%), followed by sweet potatoes (1,360%), Great Northern beans (1,249%), dates (852%),
broccoli (85%), lima beans (79%), and mushrooms (28%).

Table 3. Current and potential per capita availability of nutrient dense fruits and vegetables for hypertensive
adults.
Production center expansion2
No
Crop Current1 expansion 5km 10km 15km 20km
10-3 cup eq. per person per day3, 4, 5
Dates 2.4 21 25 30 34 35
Kiwis 1.1 16 23 31 40 49
Broccoli 43 37 42 50 72 94
Lima beans 1.5 1.2 1.3 1.6 1.9 2.5
Sweet potatoes 10 141 226 317 395 476
Great Northern beans 3.5 44 53 63 67 72
Mushrooms, agaricus 26 7.4 18 34 63 114
1
National mean (2007-2011), Loss-Adjusted Food Availability (LAFA) data series
2
Represents distance from all sides of production center
3
Population of hypertensive adults adopted from Guo et al., 2012
4
Adjusted for losses at the farm, retail, and consumer levels
5
Values are cumulative
92
93
Discussion
We identified enough suitable land within the existing production centers to substantially
increase the daily per capita availability of nutrient dense F&V for hypertensive adults. We also
identified additional suitable land with every 5 km expansion of the production centers. Under
the no expansion scenario, a total of nearly 144,000 ha of land could be brought into production,
resulting in an additional 0.27 daily per capita cup-equivalents of nutrient dense F&V. This
represents an 11% increase over the 2.25 total daily per-capita cup equivalents of F&V
consumed by hypertensive individuals (Conrad et al., forthcoming). Expanding the production
centers by 5 km, 10 km, 15 km, and 20 km could increase the daily per capita availability of total
F&V by 17%, 23%, 30%, and 37%, respectively. This is a substantial increase over the amount
of F&V that is currently consumed, but even a 37% increase (which would increase daily
consumption to 3.09 cup-equivalents) is far less than what is needed to achieve the Dietary
Approaches to Stop Hypertension (DASH) diet (5 cup-equivalents).
Additional production of F&V (nutrient dense and otherwise) in the US could be achieved by
further expanding the size of the production centers, increasing the number of production
centers, converting land currently used for crops like grains and oilseeds to F&V production, and
technological innovations to increase F&V crop yields. Several factors present challenges to
these approaches, such as barriers to adoption among producers, labor availability, supply chain
capacity, and threats to crop yields from environmental concerns.

94
Many F&V crops require specialized knowledge, planning, and management from producers in
order to ensure adequate production, and the time needed to develop this knowledge and
implement these practices could be a barrier to adoption for non F&V producers66. Additionally,
many F&V require specialized equipment for planting and harvesting that cannot be repurposed
for other crops, and capital investment in this equipment is high52, 54. Uncertainty, risk, and
market volatility are prominent features of F&V production because of the highly perishable and
seasonal nature of these crops2, 48. In particular, some orchard crops can take several years to
bear edible fruit and up to nine or ten years to reach full productive potential, during which time
the costs of production are not fully compensated by product sales.57
Access to labor is particularly problematic for F&V firms because of the knowledge and know-
how needed to complete highly specialized tasks, which can include dethorning, pollinating,
pruning, fruit bunch tie down, picking, field grading, weeding, and scouting for pests. Although
mechanization is more common for vegetables than fruit, and for processed produce than for
fresh varieties, even crops grown under a mechanized production system may still require field
grading and packing by hand (such as sweet potatoes)67. (In this analysis the exception is Great
Northern beans, which have a production structure that is almost entirely mechanized, and
resembles that of agronomic crops like soybeans rather than F&V). Mechanization can also
introduce challenges that cannot be easily overcome. Not all crops in a field mature at the same
rate, so nonselective machine harvesters can reduce usable yields when they harvest everything
in a field. Also, machines cannot replicate the dexterity of farmworkers, which is needed to
prevent crop damage that can reduce the market value of harvested crops67. The future
95
availability of farmworkers is also uncertain, given the decline in the number of seasonal, short-
term workers since the late 1990s68.
Supply chain networks for F&V are complex and specialized, which limits their capacity to
accommodate increased volume in the short term. Depending on the type of F&V and whether it
is marketed as fresh or processed, its supply chain can include packinghouses, third party
certifiers, storage facilities, distributors, marketers, brokers, shippers, processors, wholesalers,
and retailers48, 69. Adequate coordination across these entities is essential for moving F&V from
producers to consumers, but this can be challenging because of the diversity of the handling
requirements and market demands of different F&V48, 52, 54. Although storage times for many
F&V have increased over the past several decades due to genetic innovations and improvements
in atmospheric control systems in storage and transport facilities, storage times vary widely
across different types of F&V and between fresh and processed varieties2, 70. Limited storage
space at the retail level often requires store produce managers to place orders from distribution
centers on a near daily basis. Although the adoption of computerized database systems has
facilitated coordination between supply chain entities, maintaining these databases requires
continuous human input at multiple points throughout the supply chain70.
Multi-year drought conditions have reduced water availability in California, threatening the
vitality of the largest F&V industry in the country71. Reduced crop yields were reported on
nearly 1.2 million acres (0.49 million hectares) in 2013 due to a shortage of surface and ground
water72. For some areas in California, aquifer recharge after a prolonged drought could take
many years to reach pre-drought levels73. This is particularly problematic for crops such as
96
broccoli and sweet potatoes, since nearly 100% of their acreage in California is irrigated52. The
prospect for increasing F&V production in California is made even more uncertain by the effects
of salinization of soil and water sources74, 75, as well as the effects of climate change projections
that include altered patterns of precipitation and temperature76-78, all of which have the potential
to reduce yields for some crops in some areas79-81.
Even if the challenges associated with increasing domestic production were overcome, imports
of F&V would still be essential for meeting increased consumer demand. Imports are needed to
smooth out otherwise strong seasonal price fluctuations that often result from climatic
seasonality or unexpected weather events2, 3. In cases where the availability of domestically
produced F&V lags behind demand (as a result of the time it takes to establish productive
orchards, increase labor availability, or increase the capacity of the supply chain), imported F&V
can fill this gap. Some F&V that display positive consumption trends, such as avocados, are
sourced primarily from international markets, and the proportion of consumption that is
represented by imports has increased dramatically over the past several decades. In these cases, it
is likely that further increases in consumption would be met predominantly by imports. Although
the US currently produces the majority of F&V its population consumes, imports for fresh and
all processing types of F&V have increased steadily since 197582, 83.
This study has several limitations. It has been reported that recent drought conditions in
California have encouraged producers to fallow at least 410,000 acres (166,000 hectares) of
cropland84. This land could have been identified as suitable cropland by the geospatial models
we used in this study, and may have contributed to an overestimation of suitable land area for
97
crops with production centers in California. This study may not have included all of the
production centers for each crop because of limited data availability for some crops in some
areas. For example, the Census of Agriculture46 reports that nearly 30% of the national acreage
of dates is located in Arizona, yet data are not disclosed for counties with the highest acreage of
dates in that state. Incomplete data on the location of national production centers would likely
have contributed to an underestimation of suitable land area for some crops. To estimate food
loss, USDA collects data from household surveys, landfill garbage inspections, and plate waste
studies, and each has limitations45. Household survey participants may not report their actual
consumption patterns in order to impress the interview administrator or to simplify the interview
process10, 85, and landfill garbage inspections and plate waste studies are not nationally
representative. Additionally, USDA does not collect data on the temporal variation of food
loss45.
Additional research is needed to examine the availability of suitable agricultural land in and
around smaller production centers throughout the country in order to fully capture the national
potential to increase F&V production. These site suitability analyses could be integrated into
existing modeling frameworks86-88 that have been used to examine the capacity of supply chain
infrastructure to handle additional volume. This research should focus on the potential to
increase production in areas least likely to experience environmental degradation and natural
resource limitations.
There are several key strengths to this study. We included a broad range of parameters that
allowed for a robust analysis of site suitability. Our models were specific to each crop, which
98
allowed for a higher level of specificity when establishing the upper and lower bounds of the
relevant growing conditions than would have been possible if our crop focus was more broadly
defined (for example, field crops vs. specialty crops, or annual vs. perennial crops). By
narrowing our focus to crops with increasing trends of consumption and domestic production, we
were able to identify, based on our assumptions, crops that would be most likely to be adopted
by consumers and producers. Finally, we focused our analyses on areas where supply chain
infrastructure was likely to be present in order to address the reality that agricultural production
requires the co-location of supply chain infrastructure in order for food to reach consumers.
In this study we show that there is enough suitable agricultural land in (and adjacent to) existing
production centers to substantially increase the availability of certain nutrient dense F&V for
hypertensive adults in the US. Yet additional production of F&V (nutrient dense and otherwise)
would be needed in order to increase the availability of F&V to the point where hypertensive
adults can meet the DASH diet. These findings highlight the important role that smaller US
production centers (that were not identified in this study), as well as international food markets,
will play hypertensive adults adopt healthier eating patterns. US agricultural policy and
international food trade agreements could be better informed by considering the interplay
between nutrition (public health recommendations), food consumption (trends in consumer
demand), and agriculture (production capacity).
References
1 Zahniser, S., Angadjivand, S., Herts, T., Kuberka, L. and Santos, A. 2015. NAFTA at 20:
North America's free-trade area and its impact on agriculture. USDA Economic Research
Service (ERS). WRS-15-01. US Government Printing Office, Washington, DC. Available at
99
website: www.ers.usda.gov/publications/wrs-international-agriculture-and-trade-outlook/wrs-15-
01.aspx (verified 21 Apr. 2015).
2 Plattner, K., Perez, A. and Thornsbury, S. 2014. Evolving U.S. Fruit markets and seasonal
grower price patterns. USDA Economic Research Service. FTS-357-01. US Government
Printing Office, Washington, DC. Available at Website: www.ers.usda.gov/publications/fts-fruit-
and-tree-nuts-outlook/fts357-01.aspx (verified 08 May 2015).
3 Huang, S.W. 2013. Imports contribute to year-round fresh fruit availability. USDA Economic
Research Service (ERS). FTS-356-01. US Government Printing Office, Washington, DC.
Available at website: www.ers.usda.gov/publications/fts-fruit-and-tree-nuts-outlook/fts-356-
01.aspx (verified 16 Mar. 2015).
4 US Department of Agriculture, Economic Research Service (ERS). 2009. Import share of
consumption, US agricultural trade. Available at website:
http://www.ers.usda.gov/topics/international-markets-trade/us-agricultural-trade/import-share-of-
consumption.aspx (verified 16 Mar 2015).
5 US Department of Health and Human Services and US Department of Agriculture. 1990.
Dietary Guidelines for Americans 1990, p. 18. 3rd ed. US Government Printing Office,
Washington, DC. Available at website: http://health.gov/dietaryguidelines/ (verified 18 May
2015).
Dietary Guidelines for Americans 1995. Chapter 3. . 4th ed. US Government Printing Office,
Washington DC. Available at website: http://health.gov/dietaryguidelines/ (verified 18 May
2015).
Dietary Guidelines for Americans 2000. Chapter 2. 5th ed. US Government Printing Office,
2015).
Dietary Guidelines for Americans 2005, p.24. 6th ed. US Government Printing Office,
2015).
Dietary Guidelines for Americans 2010, p. 35. 7th ed. U.S. Government Printing Office,
Washington, DC. Available at website: www.health.gov/dietaryguidelines/ (verified 18 May
2015).
10 Woodside, J.V., Young, I.S. and McKinley, M.C. 2013. Fruits and vegetables: Measuring
intake and encouraging increased consumption. P Nutr Soc 72: 236-245.
11 Boeing, H., Bechthold, A., Bub, A., Ellinger, S., Haller, D., Kroke, A., Leschik-Bonnet, E.,
Müller, M., Oberritter, H., Schulze, M., Stehle, P. and Watzl, B. 2012. Critical review:
Vegetables and fruit in the prevention of chronic diseases. European Journal of Nutrition 51:
637-663.
12 US Department of Health and Human Services, Centers for Disease Control and Prevention.
2012. Ten leading causes of death and injury. Available at website:
www.cdc.gov/injury/wisqars/leadingcauses.html (verified 17 Mar. 2015).
13 Wang, X., Ouyang, Y., Liu, J., Zhu, M., Zhao, G., Bao, W. and Hu, F.B. 2014. Fruit and
vegetable consumption and mortality from all causes, cardiovascular disease, and cancer:
100
Systematic review and dose-response meta-analysis of prospective cohort studies. Brit Med J
349: g4490.
Dietary Guidelines for Americans 2010, chapter 4. 7th ed. U.S. Government Printing Office,
Washington, DC. Available at website: www.health.gov/dietaryguidelines/ (verified 16 Mar.
2015).
15 US Department of Health and Human Services. 2010. Healthy People 2020: Understanding
and improving health. Nutrition and weight status objectives 14 and 15. 3rd ed. Available at
2015).
16 Produce for Better Health Foundation. 2015. Fruits and veggies: More matters. Available at
website: www.fruitsandveggiesmorematters.org/ (verified 16 Mar. 2015).
17 Krebs-Smith, S.M., Reedy, J. and Bosire, C. 2010. Healthfulness of the US food supply:
Little improvement despite decades of dietary guidance. American Journal of Preventive
Medicine 38: 472-477.
18 Krebs-Smith, S.M., Guenther, P.M., Subar, A.F., Kirkpatrick, S.I. and Dodd, K.W. 2010.
Americans do not meet federal dietary recommendations. Journal of Nutrition 140: 1832-1838.
19 Hsiao, C.-J., Cherry, D.K., Beatty, P.C. and Rechtsteiner, E.A. 2010. National Ambulatory
Medical Care Survey: 2007 summary. National Health Statistics Reports. US Department of
Health and Human Services, Centers for Disease Control and Prevention (CDC). Number 27.
Available at website: http://www.cdc.gov/nchs/products/nhsr.htm (verified 12 Dec 2014).
20 Persoskie, A., Kaufman, A.R. and Leyva, B. 2014. Receiving and adhering to lifestyle
modification counseling for hypertension: Disparities between smokers and nonsmokers. Journal
of Clinical Hypertension 16: 429-436.
21 Lloyd-Jones, D., Adams, R.J., Brown, T.M., Carnethon, M., Dai, S., De Simone, G.,
Ferguson, T.B., Ford, E., Furie, K., Gillespie, C., Go, A., Greenlund, K., Haase, N., Hailpern, S.,
Ho, P.M., Howard, V., Kissela, B., Kittner, S., Lackland, D., Lisabeth, L., Marelli, A.,
McDermott, M.M., Meigs, J., Mozaffarian, D., Mussolino, M., Nichol, G., Roger, V.L.,
Rosamond, W., Sacco, R., Sorlie, P., Stafford, R., Thom, T., Wasserthiel-Smoller, S., Wong,
N.D. and Wylie-Rosett, J. 2010. Heart disease and stroke statistics-2010 update: A report from
the American Heart Association. Circulation 121: e46-e215.
22 Lewington, S., Clarke, R., Qizilbash, N., Peto, R. and Collins, R. 2002. Age-specific
relevance of usual blood pressure to vascular mortality: A meta-analysis of individual data for
one million adults in 61 prospective studies. Lancet 360: 1903-1913.
23 Hirsch, G., Homer, J., Trogdon, J., Wile, K. and Orenstein, D. 2014. Using simulation to
compare 4 categories of intervention for reducing cardiovascular disease risks. American Journal
of Public Health 104: 1187-1195.
24 US Department of Health and Human Services, National Institutes of Health. 2013. Lifestyle
interventions to reduce cardiovascular disease: Systematic evidence review from the Lifestyle
Work Group, 2013, p.45. National Heart, Lung, and Blood Institute. Available at website:
www.nhlbi.nih.gov/health-pro/guidelines/ (verified 21 Nov. 2014).
25 Eckel, R.H., Jakicic, J.M., Ard, J.D., de Jesus, J.M., Houston Miller, N., Hubbard, V.S., Lee,
I.-M., Lichtenstein, A.H., Loria, C.M., Millen, B.E., Nonas, C.A., Sacks, F.M., Smith, S.C.,
Svetkey, L.P., Wadden, T.A. and Yanovski, S.Z. 2014. 2013 AHA/ACC guideline on lifestyle
management to reduce cardiovascular risk: A report of the American college of
101
cardiology/American heart association Task Force on Practice Guidelines. Circulation 129: S76-
S99.
26 Dietary Guidelines Advosiry Committee. 2015. Scientific report of the 2015 Dietary
Guidelines advisory committee: Advisory report to the Secretary of Health and Human Services
and the Secretary of Agriculture. Part D, Chapter 1. U.S. Government Printing Office,
27 Lin, P.-H., Aickin, M., Champagne, C., Craddick, S., Sacks, F.M., McCarron, P., Most-
Windhauser, M.M., Rukenbrod, F. and Haworth, L. 2003. Food group sources of nutrients in the
dietary patterns of the DASH-sodium trial. Journal of the American Dietetic Association 103:
488-496.
28 Appel, L.J., Moore, T.J., Obarzanek, E., Vollmer, W.M., Svetkey, L.P., Sacks, F.M., Bray,
G.A., Vogt, T.M., Cutler, J.A., Windhauser, M.M., Lin, P.-H., Karanja, N., Simons-Morton, D.,
McCullough, M., Swain, J., Steele, P., Evans, M.A., Miller, E.R. and Harsha, D.W. 1997. A
clinical trial of the effects of dietary patterns on blood pressure. New England Journal of
Medicine 336: 1117-1124.
29 Hikmat, F. and Appel, L.J. 2014. Effects of the DASH diet on blood pressure in patients
with and without metabolic syndrome: Results from the DASH trial. Journal of Human
Hypertension 28: 170-175.
30 Saneei, P., Salehi-Abargouei, A., Esmaillzadeh, A. and Azadbakht, L. 2014. Influence of
Dietary Approaches to Stop Hypertension (DASH) diet on blood pressure: A systematic review
and meta-analysis on randomized controlled trials. Nutrition, Metabolism and Cardiovascular
Diseases 24: 1253-1261.
31 Reedy, J., Krebs-Smith, S.M., Miller, P.E., Liese, A.D., Kahle, L.L., Park, Y. and Subar,
A.F. 2014. Higher diet quality is associated with decreased risk of all-cause, cardiovascular
disease, and cancer mortality among older adults. The Journal of Nutrition 144: 881-889.
32 Salehi-Abargouei, A., Maghsoudi, Z., Shirani, F. and Azadbakht, L. 2013. Effects of Dietary
Approaches to Stop Hypertension (DASH)-style diet on fatal or nonfatal cardiovascular
diseases—incidence: A systematic review and meta-analysis on observational prospective
studies. Nutrition 29: 611-618.
33 Liese, A.D., Krebs-Smith, S.M., Subar, A.F., George, S.M., Harmon, B.E., Neuhouser,
M.L., Boushey, C.J., Schap, T.R.E. and Reedy, J. 2015. The Dietary Patterns Methods Project:
Synthesis of findings across cohorts and relevance to dietary guidance. Journal of Nutrition 145:
393-402.
34 Kwan, M.W.-M., Wong, M.C.-S., Wang, H.H.-X., Liu, K.Q.-L., Lee, C.L.-S., Yan, B.P.-Y.,
Yu, C.-M. and Griffiths, S.M. 2013. Compliance with the Dietary Approaches to Stop
Hypertension (DASH) diet: A systematic review. PloS one 8: 78412.
35 Mellen, P.B., Gao, S.K., Vitolins, M.Z. and Goff, D.C., Jr. 2008. Deteriorating dietary habits
among adults with hypertension: DASH dietary accordance, NHANES 1988-1994 and 1999-
2004. Archives of Internal Medicine 168: 308-314.
36 US Department of Agriculture, Economic Research Service. 2015. Fruit and tree nuts:
Background. Available at website: www.ers.usda.gov/topics/crops/fruit-tree-
nuts/background.aspx (verified 16 Mar. 2015).
37 US Department of Agriculture, Economic Research Service. 2015. Vegetables and pulses:
Overview. Available at website: www.ers.usda.gov/topics/crops/vegetables-pulses.aspx (verified
16 Mar. 2015).
102
38 Giombolini, K.J., Chambers, K.J., Schlegel, S.A. and Dunne, J.B. 2011. Testing the local
reality: Does the Willamette Valley growing region produce enough to meet the needs of the
local population? A comparison of agriculture production and recommended dietary
requirements. Agriculture and Human Values 28: 247-262.
39 Kremer, P. and Schreuder, Y. 2012. The feasibility of regional food systems in metropolitan
areas: An investigation of Philadelphia's foodshed. Journal of Agriculture, Food Systems, and
Community Development 2: 171-191.
40 Colasanti, K.J.A. and Hamm, M.W. 2010. Assessing the local food supply capacity of
Detroit, Michigan. Journal of Agriculture, Food Systems, and Community Development 1: 41-
58.
41 Peters, C.J., Bills, N.L., Lembo, A.J., Wilkins, J.L. and Fick, G.W. 2012. Mapping potential
foodsheds in New York State by food group: An approach for prioritizing which foods to grow
locally. Renewable Agriculture and Food Systems 27: 125-137.
42 Young, C.E. and Kantor, L.S. 1999. Moving toward the Food Guide Pyramid: Implications
for US agriculture. In E. Frazao (ed.). America's Eating Habits: Changes and Consequences.
USDA Economic Research Service (ERS). Agriculture Information Bulletin No. 750. US
Government Printing Office, Washington, DC. 403-423. Available at Website:
http://www.ers.usda.gov/Publications/AIB750/ (verified 17 April 2012).
43 Buzby, J.C., Wells, H.F. and Vocke, G. 2006. Possible implications for US agriculture from
adoption of select Dietary Guidelines. USDA Economic Research Service (ERS). Economic
44 USDA Center for Nutrition Policy and Promotion (CNPP). 2015. Choosemyplate.Gov.
45 US Department of Agriculture, Economic Research Service (ERS). 2007-2011. Loss-
adjusted Food Availability (LAFA) data series. Available at website: www.ers.usda.gov/data-
products/food-availability-(per-capita)-data-system/loss-adjusted-food-availability-
documentation.aspx (verified 18 Mar. 2015).
46 US Department of Agriculture, National Agricultural Statistics Service (NASS). 2002-2012.
Quick Stats 2.0: Census of Agriculture. US national level data. Available at website:
www.nass.usda.gov/Quick_Stats/ (verified 18 Mar. 2015).
47 US Department of Agriculture, Agricultural Research Service. USDA National Nutrient
48 Cook, R.L. 2011. Fundamental forces affecting U.S. Fresh produce growers and marketers.
Choices: the magazine of food, farm and resource issues (Online) 26.
http://go.galegroup.com/ps/i.do?id=GALE%7CA354272486&v=2.1&u=mlin_m_tufts&it=r&p=
AONE&sw=w&asid=7eb7c91df6ea5a74a7c3411a2426bf18.
49 Johnson, G., Ernest, E., Kleczewski, N., Everts, K., Van Gessel, M. and Whalen, J. 2015.
Delaware commercial vegetable production recommendations for 2015. University of Delaware
Cooperative Extension Service. Extension Bulletin 137. Available at website:
www.extension.udel.edu/ag/vegetable-fruit-resources/commercial-vegetable-production-
recommendations/.
50 University of Nebraska-Lincoln. 2014. Dry edible beans. Available at website:
www.cropwatch.unl.edu/drybeans (verified 20 Mar. 2015).
103
51 University of California-Davis, Division of Agriculture and Natural Resources. 2014.

Kiwifruit in California. Available at website:
http://fruitandnuteducation.ucdavis.edu/education/fruitnutproduction/Kiwi/ (verified 20 Mar.
2015).
52 Stoddard, C.S., Davis, R.M. and Cantwell, M.I. 2013. Sweetpotato production in California.
Vegetable Production Series. University of California, Agriculture and Natural Resources.
Publication 7237. Available at website: www.anrcatalog.ucdavis.edu/ (verified 20 Mar. 2015).
53 Dara, S.K., Klonsky, k.M. and Tumber, K.P. 2012. Sample costs to produce fresh market
broccoli: Central coast region - San Luis Obispo County. University of California Cooperative
Extension. Available at website: www.coststudies.ucdavis.edu/current.php (verified 20 Mar.
2015).
54 Beutel, J.A. 2012. Kiwi production in California. University of California, Cooperative
Extension, Small Farm Program. Available at website:
www.sfp.ucdavis.edu/pubs/brochures/Kiwi/ (verified 20 Mar. 2015).
55 LeStrange, M., Cahn, M.D., Koike, S.T., Smith, R.F., Daugovish, O., Fennimore, S.A.,
Natwick, E.T., Dara, S.K., Takele, E. and Cantwell, M.I. 2010. Broccoli production in
California. Vegetable Production Series. University of California, Agriculture and Natural
Resources. Publication 7211. Available at website: www.anrcatalog.ucdavis.edu/ (verified 20
Mar. 2015).
56 Hodel, D.R. and Johnson, D.V. 2007. Imported and American varieties of dates (Phoenix
dactylifera) in the United States. University of California Agriculture and Natural Resources:
Oakland, CA.
57 Takele, E., Mauk, P. and Sharabeen, I. 2007. Sample costs to establish a date palm orchard
and produce dates in the Coachella Valley, Riverside County, 2005-2006. University of
California Cooperative Extension. Available at website:
www.ucanr.edu/sites/Farm_Management/Costs_and_Returns/ (verified 20 Mar. 2015).
58 Beyer, D.M. 2003. Basic procedures for agaricus mushroom growing. Pennsylvania State
University, College of Agricultural Sciences, Agricultural Research and Cooperative Extension.
Available at website: http://pubs.cas.psu.edu/PubTitle.asp?varTitle=mushroom.
59 Oregon State University, PRISM Climate Group. 1981-2010. Parameter-elevation
Regressions on Independent Slopes Model (prism): Monthly precipitation. Available at website:
www.prism.oregonstate.edu/ (verified 20 Mar 2015).
60 Oregon State University, PRISM Climate Group. 1981-2010. Parameter-elevation
Regressions on Independent Slopes Model (prism): Monthly minimum and maximum
temperature. Available at website: www.prism.oregonstate.edu/ (verified 20 Mar 2015).
61 US Department of Agriculture, Natural Resources Conservation Service (NRCS). 2014.
Gridded Soil Survey Geographic (gSSURGO) database state-tile package. Available at website:
http://datagateway.nrcs.usda.gov/ (verified 20 Mar 2015).
62 US Department of Agriculture, National Agricultural Statistics Service. 2013. Cropland Data
Layer. Available at website: http://datagateway.nrcs.usda.gov/ (verified 20 Mar 2015).
63 US Department of Agriculture, National Agricultural Statistics Service (NASS). 2002-2012.
Quick Stats 2.0: Census of Agriculture. US state level data. Available at website:
www.nass.usda.gov/Quick_Stats/ (verified 18 May 2015).
64 Guo, F., He, D., Zhang, W. and Walton, R.G. 2012. Trends in prevalence, awareness,
management, and control of hypertension among United States adults, 1999 to 2010. Journal of
the American College of Cardiology 60: 599-606.
104
65 Griffin, T., Conrad, Z., Peters, C., Ridberg, R. and Tyler, E.P. 2014. Regional self-reliance
of the Northeast food system. Renewable Agriculture and Food Systems FirstView: 1-15.
66 Johnson, D., Krissof, B., Young, E., Hoffman, L., Lucier, G. and Breneman, V. 2006.
Agronomic and economic barriers to expanding fruit and vegetable production in eliminating
fruit and vegetable planting restrictions: How would markets be affected? USDA, Economic
Research Service. Economic Research Report Number 30. US Government Printing Office,
Washington, DC. Available at website: www.ers.usda.gov/publications/err-economic-research-
report/err30.aspx (verified 14 June 2015).
67 Calvin, L. and martin, P. 2010. The U.S. Produce industry and labor: Facing the future in a
global economy. USDA Economic Research Service (ERS). Economic Research Report Number
106. US Government Printing Office, Washington, DC. Available at website:
www.ers.usda.gov/publications/err-economic-research-report/err106.aspx (verified 9 May 2015).
68 Fan, M., Gabbard, S., Alves Pena, A. and Perloff, J.M. 2015. Why do fewer agricultural
workers migrate now? American Journal of Agricultural Economics 97: 665-679.
69 Lucier, G., Pollack, S., Ali, M. and Perez, A. 2006. Fruit and vegetable backgrounder.
USDA Economic Research Service (ERS). VGS-313-01. US Government Printing Office,
Washington, DC. Available at website: www.ers.usda.gov/publications/vgs-vegetables-and-
pulses-outlook/vgs-31301.aspx (verified 11 May 2015).
70 King, R.P., Hand, M.S., DiGiacomo, G., Clancy, K., Gomez, M.I., Hardesty, S., Lev, L. and
McLaughlin, E.W. 2010. Comparing the structure, size, and performance of local and
mainstream food supply chains. USDA Economic Research Service (ERS). Economic Research
Report Number 99. US Government Printing Office, Washington, DC. Available at website:
www.ers.usda.gov/publications/err-economic-research-report/err99.aspx (verified 12 May 2015).
71 USDA Economic Research Service. 2015. California drought: Food and farm impacts.
Available at website: www.ers.usda.gov/topics/in-the-news/california-drought-farm-and-food-
impacts.aspx (verified 9 May 2015).
72 USDA National Agricultural Statistics Service (NASS). 2013. Farm and Ranch Irrigation
Survey, Census of Agriculture. Table 17. US Government Printing Office, Washington, DC.
Available at Website: http://www.agcensus.usda.gov/Publications/Irrigation_Survey/ (verified 08
May 2015).
73 Miller, N.L., Dale, L.L., Brush, C.F., Vicuna, S.D., Kadir, T.N., Dogrul, E.C. and Chung,
F.I. 2009. Drought resilience of the California Central Valley surface-ground-water-conveyance
system. Journal of the American Water Resources Association 45: 857-866.
74 Mukherjee, M. and Schwabe, K.A. 2014. Where's the salt? A spatial hedonic analysis of the
value of groundwater to irrigated agriculture. Agricultural Water Management 145: 110-122.
75 Schoups, G., Hopmans, J.W., Young, C.A., Vrugt, J.A., Wallender, W.W., Tanji, K.K. and
Panday, S. 2005. Sustainability of irrigated agriculture in the San Joaquin Valley, California.
Proceedings of the National Academy of Sciences 102: 15352-15356.
76 Pierce, D.W., Das, T., Cayan, D.R., Maurer, E.P., Miller, N.L., Bao, Y., Kanamitsu, M.,
Yoshimura, K., Snyder, M.A., Sloan, L.C., Franco, G. and Tyree, M. 2013. Probabilistic
estimates of future changes in California temperature and precipitation using statistical and
dynamical downscaling. Climate Dynamics 40: 839-856.
77 Mehta, V.K., Haden, V.R., Joyce, B.A., Purkey, D.R. and Jackson, L.E. 2013. Irrigation
demand and supply, given projections of climate and land-use change, in Yolo County,
California. Agricultural Water Management 117: 70-82.
105
78 Joyce, B.A., Mehta, V.K., Purkey, D.R., Dale, L.L. and Hanemann, M. 2011. Modifying
agricultural water management to adapt to climate change in California's Central Valley.
Climatic Change 109: 299-316.
79 Medellín-Azuara, J., Howitt, R., MacEwan, D. and Lund, J. 2011. Economic impacts of
climate-related changes to California agriculture. Climatic Change 109: 387-405.
80 Hatfield, J.L., Boote, K.J., Kimball, B.A., Ziska, L.H., Izaurralde, R.C., Ort, D., Thomson,
A.M. and Wolfe, D. 2011. Climate impacts on agriculture: Implications for crop production.
Agronomy Journal 103: 351-370.
81 Deschenes, O. and Kolstad, C. 2011. Economic impacts of climate change on California
agriculture. Climatic Change 109: 365-386.
82 USDA Economic Research Service. 2014. Fruit and tree nut yearbook. Tables H3-H7.
Available at website: www.ers.usda.gov/data-products/fruit-and-tree-nut-data/yearbook-
tables.aspx#40989 (verified 12 May 2015).
83 USDA Economic Research Service. 2015. Vegetables and pulses yearbook. Table 9.
Available at website: www.ers.usda.gov/data-products/vegetables-and-pulses-data/yearbook-
tables.aspx (verified 12 May 2015).
84 Howitt, R., Medellin, A., Josue, MacEwan, D.J., Lund, J. and Sumner, D. 2014. Economic
analysis of the 2014 drought for California agriculture. Center for Watershed Sciences,
University of California, Davis. Available at website: https://watershed.ucdavis.edu/2014-
drought-report (verified 12 May 2015).
85 Roark, R.A. and Niederhauser, V.P. 2013. Fruit and vegetable intake: Issues with definition
and measurement. Public Health Nutr 16: 2-7.
86 Etemadnia, H., Goetz, S.J., Canning, P. and Tavallali, M.S. 2015. Optimal wholesale
facilities location within the fruit and vegetables supply chain with bimodal transportation
options: An lp-mip heuristic approach. European Journal of Operational Research 244: 648-661.
87 Atallah, S.S., Gomez, M.I. and Bjorkman, T. 2014. Localization effects for a fresh vegetable
product supply chain: Broccoli in the eastern United States. Food Policy 49: 151-159.
88 Nicholson, C.F., Gómez, M.I. and Gao, O.H. 2011. The costs of increased localization for a
multiple-product food supply chain: Dairy in the United States. Food Policy 36: 300-310.
106
Conclusion
Summary of results
In this dissertation I investigate the capacity of the US agricultural system to accommodate
national public health nutrition goals through three distinct studies. In the first study, titled
results, I found that three prominent dietary data collection methods that are used to estimate
daily per capita F&V consumption at the national level do not produce the same results.
Statistical agreement tests (Bland-Altman method and ordinary least products regression)
indicated the presence of proportional bias between NHANES and LAFA, and descriptive means
comparisons yielded inconsistent similarities between NHANES, BRFSS, and LAFA. Based on
these findings, I recommend that these programs should not be used interchangeably to estimate
daily per capita F&V consumption, and that each should be used for distinct purposes. NHANES
is best suited for estimating daily per capita consumption because it allows for the amount of
individual F&V (as well as F&V groups) to be discerned from composite foods with multiple
ingredients, which allows for a complete accounting of F&V consumption from all sources in the
diet. NHANES is also structured for subpopulation analysis and covariate control, which can be
used to estimate consumption patterns for specific population groups while adjusting for the
effects of other variables. Although BRFSS does allow for subpopulation analysis and covariate
control, it does not include consumption data from composite dishes and does not collect data on
individual F&V. Yet BRFSS is the only program that collects data at the state level and is
therefore best suited for state and regional comparisons. LAFA provides the only nationally
107
representative estimates of food losses at various points in the supply chain, which makes it best
suited for examining the adequacy of the food supply at the agricultural, retail, and consumer
levels.
In the second study, titled Nutrient dense fruits and vegetables can help hypertensive adults meet
the nutrient targets of the Dietary Approaches to Stop Hypertension (DASH) diet with fewer
servings than is recommended, I found that consuming nutrient dense F&V instead of those
currently consumed, without increasing the amount consumed, will not provide enough nutrients
for hypertensive adults to meet the nutrient targets of the DASH diet. However, by choosing
more nutrient dense F&V and increasing the number of servings of F&V consumed, individuals
can meet the nutrient targets of the DASH diet with 1.1 fewer servings than is recommended. I
suggest that clinicians recommend that their hypertensive patients consume nutrient dense F&V
as a way to substantially increase their intake of fiber, Ca, Mg, and K. For patients who are able
to make additional modifications to their diet, clinicians should emphasize the added benefits of
replacing foods that contain high amounts of total fat, saturated fat, and sodium (such as fast
food and snacks) with low-fat dairy products.
In the third study, titled Agricultural capacity to increase the production of nutrient dense fruits
and vegetables for treating hypertension, I found that there is enough suitable agricultural land in
(and adjacent to) existing production centers to substantially increase the availability of certain
nutrient dense F&V for hypertensive adults in the US. Yet additional production of F&V
(nutrient dense and otherwise) would be needed in order to increase the availability of F&V to
the point where hypertensive adults can meet the DASH diet. These findings highlight the
108
important role that smaller US production centers (that were not identified in this study), as well
as international food markets, will play if individuals adopt healthier eating patterns. US
agricultural policy and international food trade agreements could be better informed by
considering the interplay between nutrition (public health recommendations), food consumption
(trends in consumer demand), and agriculture (production capacity).
Importance of this research
This research is important for several reasons. It shows that nutrition policies, interventions, and
messages are at risk of being inconsistent if researchers continue to use different dietary data
collection programs interchangeably to estimate daily per capita F&V consumption. This
research also provides clinicians with the information that can be used to help their hypertensive
patients adopt more healthful dietary patterns. Finally, this research shows that the US
agricultural system is limited in its capacity to accommodate improved dietary patterns for
hypertensive adults, which highlights the importance of international food markets in meeting
public health nutrition goals.
Future research directions
These findings present several avenues for future research. A more complete understanding of
the comparability between NHANES and LAFA could be achieved by testing the agreement for
other food groups (dairy, grains, meat, seafood, oils, and sweeteners), as well as by testing the
comparability between the temporal trends of each food group produced from these programs.
109
Future research should investigate whether calibration methods can be applied to these programs
to remove differences, perhaps by using the method used by Del Gobbo et al.1 With effective
calibration methods it may be possible to blend the data generated from these programs, which
would allow researchers to take advantage of the benefits afforded by each of these programs.
For example, if calibration methods could be used to align data from NHANES and LAFA it
may be possible to apply LAFA food loss coefficients to NHANES data, which would allow
researchers to use NHANES data for food system analyses. Additional research should examine
how the different results generated by these programs have resulted in different policy
formulations and interventions.
Additional research should modify the food pattern model used in this study by removing the
calcium constraint from the F&V group, which would reduce the number of servings of F&V
needed to achieve the DASH diet nutrient targets for fiber, Mg, and K. At the same time, low-fat
dairy options should be included in the model to address Ca intake. While the model used in this
dissertation was useful for examining the impacts on nutrient intake as a result of modifying only
F&V consumption, the proposed changes to the model would be useful for examining the
impacts on nutrient intake if individuals modified their consumption of F&V and diary. These
changes would allow researchers to examine whether small modifications to F&V and dairy
consumption would elicit more favorable impacts on nutrient status than larger modifications to
just F&V consumption.
The geospatial models that were developed as part of this dissertation provide a solid foundation
for examining environmental sustainability, supply chain capacity, and economic outcomes.
110
Environmental sustainability could be addressed by expanding the models to include erosion
potential, soil quality, pesticide application rates, proximity to water bodies, and others. While
some of these parameters could be addressed at the map unit level, others could be addressed at
the county level for specific crops. Supply chain capacity and economic outcomes could be
addressed by linking the models used in this dissertation to other models, such as those used by
Etemadnia et al.2, Atallah et al.,3 and Nicholson et al.4
Importance of food systems perspective for public health nutrition research
This dissertation identifies several barriers (and catalysts) to achieving national public health
nutrition goals at multiple parts of the food system. This was accomplished by integrating the
methodologies and tools from related disciplines into a single research framework, which allows
for a more comprehensive analysis than is customarily afforded by single discipline endeavors.
This approach recognizes that public health nutrition is a key component of a food system, and
therefore affects – and is affected by – agriculture. To be sure, single discipline research
frameworks are useful for addressing highly technical issues within the confines of linear
problem solving frameworks, and should not be discounted; however, they are incomplete
methods for adequately addressing many public health nutrition problems because these are
multidisciplinary by nature. Investigating ways to improve the nutrition-related health of large
populations requires food system analyses that combine the methodologies and tools of nutrition
science, public policy, supply chain analyses, and agricultural science.

111
References
1. Del Gobbo, LC; Khatibzadeh, S; Imamura, F, et al. 2015. Assessing global dietary habits: a
comparison of national estimates from the FAO and the Global Dietary Database. American
Journal of Clinical Nutrition 101: 1038-46.
2. Etemadnia, H; Goetz, SJ; Canning, P, et al. 2015. Optimal wholesale facilities location
within the fruit and vegetables supply chain with bimodal transportation options: An LP-MIP
heuristic approach. European Journal of Operational Research 244: 648-61.
3. Atallah, SS; Gomez, MI and Bjorkman, T. 2014. Localization effects for a fresh vegetable
product supply chain: Broccoli in the eastern United States. Food Policy 49: 151-9.
4. Nicholson, CF; Gómez, MI and Gao, OH. 2011. The costs of increased localization for a
multiple-product food supply chain: Dairy in the United States. Food Policy 36: 300-10.
112
Appendices
Appendix A Fruit and vegetable forms included in Bland-Altman method and

ordinary least products regression analysis
Processing form
Food Fresh Canned Frozen Dried Juice
Apple    
Apricot        
Avocado       
Banana         
Blackberry    
    
Blueberry  
       
Cantaloupe        
Cherry   
  
    
Cranberry       
Date       
 
Fig         
Grape  
      

Grapefruit         
Honeydew        
Kiwi         
Lemon         

Lime         
Mango        
Olive  
      
Orange  
      

Papaya        
Peach   
  
   
 
Peara        
Pineapple        

Plum        
Prune      

Raisin       
 
Raspberry     
   
Strawberry 
        
Tangerine        
Watermelon         
Artichoke         
      
113
Asparagus   
Dry bean    
 
Lima bean 
    
  
Snap bean   
     
Bell pepper      
Broccoli     
   
Brusells sprout       
Cabbage        
Carrot   
  
   
Cauliflower       
Celery       
Collard green        
Cucumber        
Eggplant        
Escarole/endive        
Garlic        
Kale        
Lentil and       

split pea
Lettuce, head 
      
Lettuce, leaf        
Mushroom   
     
Mustard green       
Okra        
Onion       
 
Pea, green  
  
  
Pickle       
Potato 
     
 
Pumpkin       
Radish        
Spinach     
   
Squash       
Sweet corn   
  
   
Sweet potato      
Tomato   
     
Turnip green       
      
a
2001-2005 data used for dried pears because data for subsequent years not available
Appendix B Fruit and vegetable group categorization scheme for 2007-2010 means comparison
114
Fruit Vegetables
Fruit Fruit, Green

Food juice non-juice salad Potatoes Carrots Other
Acerola juice 
Apple juice           
Apricot juice           
Blackberry juice           
Grape juice           
Grapefruit juice           
Guanabana juice           
Guava juice           
Lemon juice           
Lime juice           
Mango juice           
Orange juice           
Papaya juice           
Passion-fruit juice           
Peach juice           
Pineapple juice           
Pomegranate juice           
Prune juice           
Tamarind juice           
Tangerine juice           
Tomato juice           
Abiyuch  
        
Acerola           
Apple           
Apricot           
Avacoado           
Banana           
Blackberry           
Blueberry           
Boysenberry           
Breadfruit           
Carissa           
Casaba melon           
Cantaloupe melon           
Chayote           
         
115
Cherimoya 
Cherry           
Clementine           
Cranberry           
Currant           
Date           
Durian           
Elderberry           
Feijoa           
Fig           
Gooseberry           
Grape           
Grapefruit           
Guava           
Honeydew melon           
Horned melon           
Jackfruit           
Java-plum           
Jujube           
Kiwi           
Kumquat           
Lemon           
Lime           
Litchi           
Loganberry           
Longan           
Loquat           
Mango           
Mangosteen           
Mulberry           
Nance           
Naranjillo           
Nectarine           
Oheloberry           
Orange           
Papaya           
Passion-fruit           
Peach           
Pear           
Persimmon           
Pineapple           
         
116
Plantain 
Plum           
Pomegranate           
Prune           
Pummelo           
Quince           
Raisin           
Rambutan           
Raspberry           
Rowal           
Sapodilla           
Sapote           
Soursop           
Starfruit           
Strawberry           
Tamarind           
Tangerine           
Tomato           
Watermelon           
Amaranth leaf    
      
Arugala           
Beet green           
Bitter gourd leafy          
tip 
Brocolli raab           
Butterbur           
Cabbage           
Cardoon           
Celtuce           
Chard           
Chicory green           
Chinese cabbage           
Chrysanthemum green           
Cilantro           
Collard           
Cowpea leafy tip           
Dandelion green           
Dock           
Drumstick leaf           
Endive           
Fireweed           
         
117
Grape leaf 
Head lettuce           
Kale           
Lambs quarter           
Leaf lettuce           
Mustard green           
Pumpkin leaf           
Seaweed           
Spinach           
Sweet potato leaf           
Taro leaf           
Turnip green           
Watercress           
Winged bean leaf           
Potato      
    
Carrot        
  
Arrowhead          

Arrowroot           
Artichoke           
Asparagus           
Bamboo shoot           
Bean and peas in          
pod 
Beet           
Bitter gourd           
Borage           
Broccoli           
Brussels sprout           
Burdock root           
Cabbage           
Cassava           
Cauliflower           
Celeriac           
Celery           
Chicory           
Chive           
Cress           
Cucumber           
Daikon           
Eggplant           
Epazote           
         
118
Eppaw 
Fennel           
Fiddlehead           
Garlic           
Ginger           
Gourd           
Green pea           
Heart of palm           
Jute           
Kohlrabi           
Leek           
Lemon grass           
Lotus root           
Mushroom           
Nopales           
Okra           
Olive           
Onion           
Parsley           
Parsnip           
Pepeao           
Pepper           
Pickle           
Pimento           
Poi           
Pokeberry shoot           
Pumpkin           
Purslane           
Radicchio           
Radish           
Rhubarb           
Roselle           
Rutabega           
Salsify           
Sesbania           
Shallot           
Snap bean           
Sprouted bean, pea,          

lentil, and seed
Squash           
Sweet corn           
         
119
Sweet potato 
Taro           
Tree fern           
Turnip           
Waterchestnut           
Winged bean tuber           
Yam           
         
        
        
Appendix C Fruit and vegetable group categorization scheme for 2011 means comparison
Fruit Vegetables
Fruit Fruit, Dark

Food juice non-juice green Beans Orange Other
Acerola juice 
Apple juice           
Apricot juice           
Blackberry juice           
Grape juice           
Grapefruit juice           
Guanabana juice           
Guava juice           
Lemon juice           
Lime juice           
Mango juice           
Orange juice           
Papaya juice           
Passion-fruit juice           
Peach juice           
Pineapple juice           
Pomegranate juice           
Prune juice           
Tamarind juice           
Tangerine juice           
Abiyuch  
        
Acerola           
Apple           
Apricot           
Avocado           
Banana           
         
120
Blackberry 
Blueberry           
Boysenberry           
Breadfruit           
Starfruit           
Carissa           
Chayote           
Cherimoya           
Cherry           
Clementine           
Cranberry           
Currant           
Date           
Durian           
Elderberry           
Feijoa           
Fig           
Gooseberry           
Grapefruit           
Grape           
Guava           
Horned melon           
Jackfruit           
Java-plum           
Jujube           
Kiwi           
Kumquat           
Lemon           
Lime           
Litchi           
Loganberry           
Longan           
Loquat           
Mango           
Mangosteen           
Cantaloupe melon           
Casaba melon           
Honeydew melon           
Mulberry           
Nance           
Naranjillo           
         
121
Nectarine 
Oheloberry           
Orange           
Papaya           
Passion-fruit           
Peach           
Pear           
Persimmon           
Pineapple           
Plantain           
Plum           
Pomegranate           
Prune           
Pummelo           
Quince           
Raisin           
Rambutan           
Raspberry           
Rowal           
Sapodilla           
Sapote           
Soursop           
Strawberry           
Tamarind           
Tangerine           
Watermelon           
Amaranth leaf    
      
Arugula           
Bitter gourd leafy          
tip 
Beet green           
Broccoli           
Chinese broccoli           
Butterbur           
Chinese cabbage           
Cabbage           
Cardoon           
Celtuce           
Chard           
Chicory green           
Chrysanthemum           
         
122
Collard 
Cilantro           
Cowpea leafy tip           
Dandelion green           
Dock           
Drumstick leaf           
Endive           
Fireweed           
Grape leaf           
Kale           
Lambs quarter           
Leaf lettuce           
Spinach           
Mustard green           
Pumpkin leaf           
Seaweed           
Swamp cabbage           
Sweet potato leaf           
Taro leaf           
Turnip green           
Watercress           
Winged bean leaf           
Adzuki bean      
    
Black bean           
Cranberry bean           
French bean (podded)           
Great northern bean           
Kidney bean           
Lima bean           
Mung mean           
Navy bean           
Pink bean           
Pinto bean           
White bean           
Yellow bean           
Fava bean           
Chickpea           
Black eyed pea           
Cowpea (podded)           
Edamame           
Lentil           
         
123
Split pea 
Pigeon pea           
Soybean           
Tempeh           
Winged bean           
Jicama           
Yardlong bean (podded)           
Carrot       
  
Yam           
Pumpkin           
Winter squash           
Sweet potato           
Arrowhead          

Arrowroot           
Artichoke           
Asparagus           
Bamboo shoot           
Beans and peas in pod           
Beet           
Bitter-gourd           
Borage           
Brusells sprout           
Burdock root           
Cabbage           
Cassava           
Cauliflower           
Celeriac           
Celery           
Chicory           
Chive           
Cress           
Cucumber           
Daikon           
Eggplant           
Epazote           
Eppaw           
Fennel           
Fiddlehead           
Garlic           
Ginger           
Gourd           
         
124
Green pea 
Head lettuce           
Heart of palm           
Jute           
Kohlrabi           
Leek           
Lemon grass           
Lotus root           
Mushroom           
Nopales           
Okra           
Olive           
Onion           
Parsley           
Parsnip           
Pepeao           
Pepper           
Pickle           
Pimento           
Poi           
Pokeberry shoot           
Potato           
Pumpkin flower           
Purslane           
Radicchio           
Radish           
Rhubarb           
Roselle           
Rutabega           
Salsify           
Sesbiana           
Shallot           
Snap bean           
Sprouted bean, pea,          

lentil, and seed
Summer squash           
Sweet corn           
Taro           
Tomato           
Tree fern           
Turnip           
         
125
Waterchestnut 
Winged bean tuber           
   
  
  
  
 

Appendix D. Growing conditions for dates in Riverside County, California, by month
Maximum
precipitation Minimum Maximum
Month (mm)1 temperature (°C)2 temperature (°C)3 Notes4
May 45.24 -3.41 35.77 Minimum precipitation is not incldued because limited
June 45.24 0.66 40.38 moisture is typically overcome with irrigation.
July 45.24 4.72 42.63 Maximum soil depth is 0.91 meters. Acceptible soil
August 45.24 4.53 41.94 types are: sand, loamy sand, sandy loam, sandy clay
September 45.24 1.64 39.40 5
loam, silty clay loam, sandy clay, clay, and clay loam.
October 45.24 -2.50 33.76
1
Represents maximum precipitation in Riverside County, California in August (mean 1981-2010), when dates are most susceptible to
damage
2
Represents minimum temperature in Riverside County, California (mean 1981-2010)
3
Represents maximum temperature in Riverside County, California (mean 1981-2010)
4
Notes apply to all months
5
Proportion of sand, silt, and clay for each soil type was retrieved from: USDA NRCS. Soil Texture Calculator. Available at website:
http://www.nrcs.usda.gov/wps/portal/nrcs/detail/soils/survey/?cid=nrcs142p2_054167 (verified 13 June 2015)
126
127
Appendix E. Growing conditions for kiwis in Tulare County, California, by month

Minimum Maximum
1
Month temperature (°C) temperature (°C)2 Notes3
March -15.11 20.78 Minimum precipitation is not
April -13.50 25.33 incldued because limited moisture is
May -8.96 30.17 typically overcome with irrigation.
June -3.83 34.28 Maximum precipitation is not
July -0.30 37.84 included because excessive
August -0.71 37.29 moisture is uncommon. Maximum
September -3.61 34.28 soil depth is 0.91 meters.
October -7.52 28.26 Acceptible soil types are: sandy
November -11.71 19.70 loam and silt loam.4
1
Represents minimum temperature in Tulare County, California (mean 1981-2010)
2
Represents maximum temperature in Tulare County, California (mean 1981-2010)
3
4
Proportion of sand, silt, and clay for each soil type was retrieved from: USDA NRCS. Soil
Texture Calculator. Available at website:
http://www.nrcs.usda.gov/wps/portal/nrcs/detail/soils/survey/?cid=nrcs142p2_054167
(verified 13 June 2015)
128
Appendix F. Growing conditions for kiwis in Butte County, California, by month

Minimum Maximum
1
Month temperature (°C) temperature (°C)2 Notes3
March -3.85 19.16 Minimum precipitation is not
April -2.34 23.20 incldued because limited moisture is
May 1.16 27.85 typically overcome with irrigation.
June 4.48 32.67 Maximum precipitation is not
July 7.85 36.87 included because excessive
August 7.15 36.61 moisture is uncommon. Maximum
September 5.20 33.38 soil depth is 0.91 meters.
October 0.82 27.26 Acceptible soil types are: sandy
November -3.28 18.09 4
loam and silt loam.
1
Represents minimum temperature in Butte County, California (mean 1981-2010)
2
Represents maximum temperature in Butte County, California (mean 1981-2010)
3
4
129
Appendix G. Growing conditions for broccoli in Monterey County, California, by month

Minimum Maximum
1 2 3
Month temperature (°C) temperature (°C) Notes
January -1.23 19.76 Minimum precipitation is not
February 0.21 19.07 incldued because limited moisture is
March 1.13 20.84 typically overcome with irrigation.
April 1.73 24.25 Maximum precipitation is not
May 4.63 28.84 included because excessive
June 7.31 32.41 moisture is uncommon. Maximum
July 8.69 36.14 soil depth is 0.61 meters. Can be
August 9.24 36.42 grown in all types of soil.4
September 8.19 34.08
October 4.93 28.79
November 0.57 22.70
December -1.43 19.08
1
Represents minimum temperature in Monterey County, California (mean 1981-2010)
2
Represents maximum temperature in Monterey County, California (mean 1981-2010)
3
4
130
Appendix H. Growing conditions for broccoli in Santa Barbara County, California, by month
Minimum Maximum
1 2 3
January -1.75 19.29 Minimum precipitation is not
February -1.06 19.37 incldued because limited moisture is
March 0.11 20.70 typically overcome with irrigation.
April 1.23 24.04 Maximum precipitation is not
May 4.62 27.57 included because excessive
June 7.73 32.46 moisture is uncommon. Maximum
July 10.29 36.12 soil depth is 0.61 meters. Can be
August 10.50 35.96 grown in all types of soil.4
September 8.09 32.48
October 4.36 28.27
November 0.98 22.34
December -1.86 19.03
1
Represents minimum temperature in Santa Barbara County, California (mean 1981-2010)
2
Represents maximum temperature in Santa Barbara County, California (mean 1981-2010)
3
4
Appendix I. Growing conditions for lima beans in Sussex and Kent Counties, Delaware, by month
Minimum Maximum
1 2 3
June 16.24 29.17 Minimum precipitation is not incldued because limited moisture
July 18.78 31.45 is typically overcome with irrigation. Maximum precipitation is
August 17.92 30.31 not included because excessive moisture is uncommon.
September 13.93 26.94 Maximum soil depth is 0.30 meters. Can be grown in all types
October 7.46 21.16 4
of soil.
1
Represents minimum temperature in Sussex and Kent Counties, Delaware (mean 1981-2010)
2
Represents maximum temperature in Sussex and Kent Counties, Delaware (mean 1981-2010)
3
4
Proportion of sand, silt, and clay for each soil type was retrieved from: USDA NRCS. Soil Texture Calculator.
Available at website: http://www.nrcs.usda.gov/wps/portal/nrcs/detail/soils/survey/?cid=nrcs142p2_054167
131
Appendix J. Growing conditions for sweet potatoes in Johnston, Nash, Sampson, and Wilson Counties, North Carolina, by month
Minimum Maximum
1 2 3
June 17.66 31.06 Minimum precipitation is not incldued because limited moisture is typically
July 19.83 32.59 overcome with irrigation. Maximum precipitation is not included because
August 18.91 31.62 excessive moisture is uncommon. Maximum soil depth is 0.20 meters.
September 15.23 28.83 4
Acceptible soil types are: sand, loamy sand, and sandy loam.
October 8.57 24.22
1
Represents minimum temperature in Johnston, Nash, Sampson, and Wilson Counties, North Carolina (mean 1981-2010)
2
Represents maximum temperature in Johnston, Nash, Sampson, and Wilson Counties, North Carolina (mean 1981-2010)
3
4
Proportion of sand, silt, and clay for each soil type was retrieved from: USDA NRCS. Soil Texture Calculator. Available at
website: http://www.nrcs.usda.gov/wps/portal/nrcs/detail/soils/survey/?cid=nrcs142p2_054167
132
Appendix K. Growing conditions for sweet potatoes in Calhoun County, Mississippi, by month
Minimum Maximum
1 2 3
September 15.75 29.86 Acceptible soil types are: sand, loamy sand, and sandy loam.4
October 9.34 24.47
1
Represents minimum temperature in Calhoun County, Mississippi (mean 1981-2010)
2
Represents maximum temperature in Calhoun County, Mississippi (mean 1981-2010)
3
4
website: http://www.nrcs.usda.gov/wps/portal/nrcs/detail/soils/survey/?cid=nrcs142p2_054167 (verified 13 June 2015)
133
Appendix L. Growing conditions for sweet potatoes in Merced County, California, by month
Minimum Maximum
1 2 3
September 13.08 31.98 Acceptible soil types are: sand, loamy sand, and sandy loam.4
October 9.74 26.77
1
Represents minimum temperature in area of Merced County, California, in which 95% of the county's sweet potato land is
cultivated (mean 1981-2010)
2
Represents maximum temperature in area of Merced County, California, in which 95% of the county's sweet potato land is
cultivated (mean 1981-2010)
3
4
website: http://www.nrcs.usda.gov/wps/portal/nrcs/detail/soils/survey/?cid=nrcs142p2_054167 (verified 13 June 2015)
134
Appendix M. Growing conditions for Great Northern beans in Scotts Bluff, Box Butte, and Morrill Counties, Nebraska, by month
Maximum
precipitation Minimum Maximum
1 2 3 4
Month (mm) temperature (°C) temperature (°C) Notes
June 51.60 9.59 28.52 Minimum precipitation is not incldued because limited moisture is
July 51.60 12.86 32.86 typically overcome with irrigation. Maximum soil depth is 0.8 meters.
August 51.60 12.32 31.72 5
Acceptible soil types are: sandy loam, loam, and silt loam.
September 51.60 5.94 26.54
1
Represents maximum precipitation in Scotts Bluff, Box Butte, and Morrill Counties, Nebraska in August and September (mean 1981-2010),
when Great Northern beans are most susceptible to damage
2
Represents minimum temperature in Scotts Bluff, Box Butte, and Morrill Counties, Nebraska (mean 1981-2010)
3
Represents maximum temperature in Scotts Bluff, Box Butte, and Morrill Counties, Nebraska (mean 1981-2010)
4
5
Proportion of sand, silt, and clay for each soil type was retrieved from: USDA NRCS. Soil Texture Calculator. Available at website:
http://www.nrcs.usda.gov/wps/portal/nrcs/detail/soils/survey/?cid=nrcs142p2_054167 (verified 13 June 2015)
135

Catalysts and Barriers To Increasing The Consumption of Fruits and Vegetables and Reducing The Prevalence of Hypertension

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Catalysts and Barriers To Increasing The Consumption of Fruits and Vegetables and Reducing The Prevalence of Hypertension

Uploaded by

Copyright:

Available Formats

Catalysts and barriers to increasing the consumption of fruits and vegetables and

reducing the prevalence of hypertension

All rights reserved

INFORMATION TO ALL USERS

All rights reserved.

The linkage between public health nutrition and agriculture

the availability of certain foods for consumers.

Dietary data collection

Nutrition policies and

Agricultural Dietary patterns Agricultural

Figure 1: Linkage between nutrition, public health, and

groups and to track progress toward achieving public health goals.

agricultural capacity to accommodate more healthy dietary patterns. In this dissertation, I

increasing the consumption of F&V9 and reducing the prevalence of hypertension,10

agricultural land in the US.

combination of statistical methods. These findings will allow us to determine whether it is

than is recommended, and is formatted according to the specifications of Nutrition Research. In

treating hypertension, and is formatted according to the specifications of Renewable Agriculture

Monitoring consumption of fruits and vegetables

Winchester, England, in response to a localized outbreak of enteric fever.16 It was a simple

approach. This recommendation has been heeded here.

Fruits, vegetables, and hypertension

condition characterized by systolic blood pressure (SBP) ≥140 mm Hg or diastolic blood

for 2020 include reducing the prevalence of hypertension.10

recommended for all hypertensive individuals regardless of pharmacotherapy treatment.57 Key

are consistently low.67, 68

amount of F&V currently consumed.

Fruits, vegetables, and agricultural capacity

(by volume) are produced domestically,78 if hypertensive individuals increased their

tool that is used to process spatial and geographic data.

approach to improving the health of the American population.

4. US Department of Health and Human Services and US Department of Agriculture. Dietary

32. US Department of Agriculture, Economic Research Service (ERS). Loss-Adjusted Food

62. Saneei, P; Salehi-Abargouei, A; Esmaillzadeh, A, et al. 2014. Influence of Dietary

64. Salehi-Abargouei, A; Maghsoudi, Z; Shirani, F, et al. 2013. Effects of Dietary Approaches

74. Ratner, RK and Riis, J. 2014. Communicating science-based recommendations with

Author names and affiliations

Zach Conrad,a Kenneth Chui,b Christian Peters,a Timothy Griffina*

Boston, MA 02111, USA

The authors have no conflicts of interest to declare

There are no funding sources to declare

Ethics of human subject participation

24HR; 24-hour recall

BRFSS; Behavioral Risk Factor Surveillance System

CDC; Centers for Disease Control and Prevention

CI; confidence interval

DGA; Dietary Guidelines for Americans

ERS; Economic Research Service

FFQ; food frequency questionnaire

F&V; fruits and vegetables

LAFA; Loss-Adjusted Food Availability data series

LOA; limits of agreement

NHANES; National Health and Nutrition Examination Survey

OLP; ordinary least products regression

OLS; ordinary least squares regression

USDA; United States Department of Agriculture

US HHS; United States Department of Health and Human Services

Keywords: Dietary assessment method, comparison, surveillance, diet, food group

assessing progress towards this goal.

at the national level [7, 14-17, 20-25].

consumption of F&V groups using descriptive means comparison.

2.1. Dietary assessment data

(24HR) administered by a trained interviewer using United States Department of Agriculture’s