Professional Documents
Culture Documents
Research Article Characteristics of Americans Choosing Vegetarian and Vegan Diets For Health Reasons
Research Article Characteristics of Americans Choosing Vegetarian and Vegan Diets For Health Reasons
ABSTRACT
Objective: Examine the prevalence, patterns, and associated factors of using a vegetarian or vegan diet for
health reasons in the US general population.
Design: Cross-sectional data from the 2012 National Health Interview Survey.
Participants: Nationally representative sample (N ¼ 34,525).
Variables Measured: Prevalence of ever use and 12-month use of vegetarian or vegan diet for health
reasons, patterns of use, and sociodemographic and health-related factor associated with use.
Analysis: Multiple logistic regression analysis.
Results: Prevalence of ever use and 12-month use was 4.0% (n ¼ 1,367) and 1.9% (n ¼ 648), respectively.
Health vegetarians and vegans were more likely aged 30–65 years, female, not Hispanic, from the Western
US region, at least high school educated, chronically ill, and physically active. They were less likely to be in
a relationship, overweight or obese, or smoking, or to have public or private health insurance. Among
health vegetarians and vegans, 6.3% consulted with a practitioner for special diets; 26.1% followed the
diet because of a specific health problem, mainly high cholesterol, overweight, hypertension, and diabetes;
and 59.4% disclosed the diet to their health care provider.
Conclusions and Implications: Less than 2% of participants reported using a vegetarian or vegan diet
for health reasons within the past 12 months. Despite potential benefits of plant-based nutrition, more
research is warranted on the actual use and its effects and safety.
Key Words: vegetarian, vegan, diet, survey, prevalence (J Nutr Educ Behav. 2017;49:561-567.)
Accepted April 21, 2017.
INTRODUCTION food, or products containing these foods,’’1 based nutrition instead of vegan nutrition.
which varies from one that incorporates The diverse and heterogeneous range of
Vegetarianism is a broad term that en- dairy products (lacto-vegetarianism) or dietary practice as well as the nonuni-
compasses a diverse and heterogeneous eggs (ovo-vegetarianism) to one that form use of terminology in both research
range of dietary practices. According to avoids all flesh, dairy, and egg foods and and practice contributes much to the
the Academy of Nutrition and Dietetics, sometimes honey (veganism).2 Several complexity of the topic.4
a vegetarian diet is one that ‘‘does not activists and influential lobby organi- There are a variety of reasons for us-
include meat (including fowl) or sea- zations in the field3 also refer to plant- ing a vegetarian or vegan diet, including
moral, ethical, spiritual, or religious rea-
sons as well as those concerning animal
1
Department of Internal and Integrative Medicine, Kliniken Essen-Mitte, Faculty of Medi- rights and animal welfare,5 or social or
cine, University of Duisburg-Essen, Essen, Germany environmental concerns related to
2
Australian Research Centre in Complementary and Integrative Medicine, Faculty of intensive animal husbandry or sustain-
Health, University of Technology Sydney, Sydney, New South Wales, Australia ability.6,7 A vegetarian or vegan diet
3
Immanuel Hospital Berlin, Department for Complementary Medicine, Berlin, Germany can also be part of certain types of
4
Institute for Social Medicine, Epidemiology and Health Economics, Charite Medical Uni- lifestyle and identity,8 or more health-
versity, Berlin, Germany related and part of a health interven-
5
Department of Neurobiology, Care Sciences and Society (NVS/OMV), Karolinska Insti- tion.4-6 The decision to use a vegetarian
tutet, Stockholm, Sweden or vegan diet for health reasons might be
6
School of Nursing and Midwifery, University of South Australia, Adelaide, Australia strongly influenced by public perceptions
Conflict of Interest Disclosure: The authors’ conflict of interest disclosures can be found online of the health advantages,9 and a
with this article on www.jneb.org. growing number of clinical and epide-
Address for correspondence: Holger Cramer, PhD, Department of Internal and Integrative miological research studies have shown
Medicine, Kliniken Essen-Mitte, Faculty of Medicine, University of Duisburg-Essen, Am health benefits associated with vege-
Deimelsberg 34a, 45276 Essen, Germany; Phone: þ(49-201) 174-25015; Fax: þ(49-201) tarian and vegan diets. Results indi-
174-25000; E-mail: h.cramer@kliniken-essen-mitte.de cated that a vegetarian diet might be
Ó2017 Society for Nutrition Education and Behavior. Published by Elsevier, Inc. All rights associated with a reduction in body
reserved. weight,10 a lower incidence of the
http://dx.doi.org/10.1016/j.jneb.2017.04.011 metabolic syndrome11 or diabetes,11
Journal of Nutrition Education and Behavior Volume 49, Number 7, 2017 561
562 Cramer et al Journal of Nutrition Education and Behavior Volume 49, Number 7, 2017
improvements in blood pressure12 and Data brated to the 2010 Census-based popu-
dyslipidemia13 and a lower incidence lation estimates for age, gender, and
and/or mortality related to ischemic This analysis was based on a nationally ethnicity of the US civilian non-
heart disease14 (indicating particular ben- representative survey monitoring the institutionalized population.
efits for cardiovascular outcomes).11,15-19 health of the US population in 2012.31 Prevalence of ever use and 12-
Other studies also linked vegetarianism For this analysis, data from the Family month use of vegetarian or vegan diet
and veganism to a reduced incidence Core, the Sample Adult Core, and the for health reasons were analyzed
of cancer14; however, a vegetarian, Adult Complementary and Alternative descriptively, as were details on vege-
and especially vegan diet, might also be Medicine questionnaire were used. tarian and vegan diet, reasons for prac-
related to health risks including nutri- The 2012 National Health Interview tice, and outcomes. Results were
tional deficiencies such as vitamin Survey was approved by the Research reported as means and SDs, medians
B12,20 zinc,21 and iron.22 Ethics Review Board at the National and ranges, weighted frequencies, and dis-
To date few studies have investigated Center for Health Statistics. tributions, as were reasonable.
the prevalence of vegetarian and vegan Sociodemographic characteristics were
nutrition. The reported prevalence rates compared between those who had
have been highly variable, ranging Measures used a vegetarian or vegan diet ever in
from 0.77% in China23 to 0.79% in their life or within the prior 12 months
The Family Core and the Sample Adult
Italy,24 2.4% to 3.3% in the US,25,26 and those who had not, using chi-
Core questionnaire collected data
3% to 8% in South Australia,27 3.8% square test. Factors independently asso-
regarding participants’ sociodemo-
to 15.6% in Scandinavia,28 up to 33% ciated with vegetarian or vegan diet
graphic characteristics, including age,
in South Asia,25 and 36% in India.29 use (ever used, used in the prior
gender, ethnicity, region, marital status,
Whereas cultural and/or religious fac- 12 months) were identified using mul-
education, annual household income,
tors might substantially influence prev- tiple logistic regression analysis. The
and self-perceived general health sta-
alence, methodological issues cannot following sociodemographic variables
tus. The Adult Complementary and
be ruled out as contributing factors. were considered: age in years (18–29,
Alternative Medicine questionnaire
Such issues relate to the definition and 30–39, 40–49, 50–64, or $65), gender
collected data on the use of a number
measurement of vegetarian diets (eg, (female, male), ethnicity (non-His-
of interventions including special
self-identified, analyses of food fre- panic white, Hispanic, African Amer-
diets. Lifetime prevalence of health
quency), or diversity of study samples ican, Asian, or other), region (west,
vegetarian or vegan diet use was deter-
(eg, age, area, socioeconomic status) northeast, midwest, or south), marital
mined with the following question:
among others, whereas some reports status (not in relationship, in rela-
Have you ever used any of the following
also lack a clear report of methodol- tionship), education (less than college,
special diets for two weeks or more for
ogy.26 Last but not least, much research some college or more), and annual
health reasons: vegetarian, including
is older than a decade,25,27,28 although household income (<$20,000, $20,000
vegan? Those who answered yes were
(primarily) plant-based forms of nutri- to $34,999, $35,000 to $64,999 or
presented with an additional question
tion seem to have gained increasing $$65,000). In addition, health-related
asking whether they had also used a
popularity.30 These issues highlight factors such as general health status
vegetarian including vegan diet for health
the need for more recent, robust, and (excellent or very good, good, fair, or
reasons during the past 12 months.
generalizable data on the prevalence poor), body mass index (<18.5–<25,
They were further queried about whether
of vegetarian forms of nutrition. 25–<30, or $30), health behav-
they had ever seen a practitioner for
Given the potential implications of iors such as smoking (nonsmoker,
special diets, and provided information
vegetarian and vegan nutrition on smoker), alcohol consumption (alcohol
on the frequency, costs, and health in-
health and well-being, a better under- abstainer, light drinker, regular, or heavy
surance coverage for those consultations.
standing of vegetarian and vegan diet drinker), and exercise behavior (low-
Respondents who had used a vege-
use for health reasons is warranted. level exerciser, moderate-level exer-
tarian or vegan diet for health reasons
Thus this study aimed to identify preva- ciser, or high-level exerciser); number
in the past 12 months were asked their
lence and patterns of vegetarian and of chronic medical conditions and dis-
reasons for using such a diet, including
vegan diets for health reasons, and fac- eases (none, 1, 2, or $3); and health in-
general reasons and specific medical
tors associated with vegetarian and surance (no health insurance, public
conditions (a total of 88 possible con-
vegan diet use for health reasons in a na- health insurance, or private health in-
ditions), disclosure toward their per-
tionally representative US sample. The surance) were used as potential associ-
sonal health care provider or reasons
findings of this study will help inform ated factors. These variables were chosen
for nondisclosure, perceived benefits
future research, clinical practice, health because they were shown to be associ-
of diet use, and information sources
policy, and public health initiatives. ated with using other complementary
about vegetarian or vegan diet.
therapies in the National Health Inter-
view Survey.33-39
METHODS Statistical Analyses A backward stepwise procedure with
Study Design a likelihood ratio statistic P # .05 was
A total of 42,366 households were chosen and adjusted odds ratios with
The researchers performed a second- eligible and 34,525 adults provided data 95% confidence intervals were calcu-
ary analysis of 2012 US National (response rate of 79.7%).32 Estimates lated. Only those variables associated
Health Interview Survey data. were calculated using weights cali- with vegetarian or vegan diet use at
Journal of Nutrition Education and Behavior Volume 49, Number 7, 2017 Cramer et al 563
The interpretations that can be drawn tions in particular. Given the abun- lipids: a systematic review and meta-
from the findings are strengthened, dance of health advice and dietary analysis of randomized controlled trials.
however, by the regression analysis advice available from the Internet, J Am Heart Assoc. 2015;4:e002408.
that controlled for confounding vari- magazines, and health food stores, 14. Dinu M, Abbate R, Gensini GF,
ables. No adjustment for multiple which is not necessarily based on sci- Casini A, Sofi F. Vegetarian, vegan
testing in the regression analyses was entific evidence,50 the integration of diets and multiple health outcomes: a
used. It has been shown that without a diet and nutrition into health care de- systematic review with meta-analysis
correction, the overall Type I error rate livery warrants further attention. of observational studies. Crit Rev Food
may be inflated, depending on the Sci Nutr. 2017;57:3640-3649.
number of predictors. Nevertheless, REFERENCES 15. Muller DC, Murphy N, Johansson M,
this approach was chosen because cor- et al. Modifiable causes of premature
rections for multiple testing in regres- 1. Craig WJ, Mangels AR, American Di- death in middle-age in Western Europe:
sion analyses commonly are overly etetic Association. Position of the results from the EPIC cohort study.
conservative and inflate Type II error American Dietetic Association: vege- BMC Med. 2016;14:87.
rates.33 It should be kept in mind, tarian diets. J Am Diet Assoc. 2009;109: 16. Orlich MJ, Singh PN, Sabate J, et al.
though, that the results of the regres- 1266-1282. Vegetarian dietary patterns and mortal-
sion analyses might contain some vari- 2. Academy of Nutrition and Dietetics. ity in Adventist Health Study 2. JAMA
ables that are not actually needed in Vegetarian Nutrition. Evidence-based Intern Med. 2013;173:1230-1238.
the model. The R2 values were low, Practice Guideline 2016. http://www. 17. Appleby PN, Crowe FL, Bradbury KE,
which could be interpreted as a rela- andeal.org/topic.cfm?menu¼5271&pcat¼ Travis RC, Key TJ. Mortality in vege-
tively low amount of explained vari- 4023&cat¼5450. Accessed April 20, 2017. tarians and comparable nonvegetarians
ance. However, in contrast to linear 3. NutritionFacts.org j The Latest in in the United Kingdom. Am J Clin
regressions, low R2 values are the norm Nutrition Related Research. http:// Nutr. 2016;103:218-230.
in logistic regression and their use in in- nutritionfacts.org. Accessed April 20, 18. Schmidt JA, Rinaldi S, Ferrari P, et al.
terpreting the goodness of fit of the 2017. Metabolic profiles of male meat eaters,
regression model is generally discour- 4. Kessler CS, Holler S, Joy S, et al. Person- fish eaters, vegetarians, and vegans
aged.34 The survey was collected by ality profiles, values and empathy: differ- from the EPIC-Oxford cohort. Am J
self-report and as such is at risk of recall ences between lacto-ovo-vegetarians and Clin Nutr. 2015;102:1518-1526.
bias or measurement error. Nonethe- vegans. Forsch Komplementmed. 2016;23: 19. Appleby PN, Key TJ. The long-term
less, the US National Health Survey is 95-102. health of vegetarians and vegans. Proc
an internationally recognized epidemi- 5. Ruby MB. Vegetarianism: a blossoming Nutr Soc. 2016;75:287-293.
ological study and its findings provide field of study. Appetite. 2012;58:141-150. 20. Pawlak R, Lester SE, Babatunde T.
useful insights into the use of vegetarian 6. Fox N, Ward K. Health, ethics and The prevalence of cobalamin deficiency
and vegan diets. environment: a qualitative study of among vegetarians assessed by serum
vegetarian motivations. Appetite. 2008; vitamin B12: a review of literature.
Implications for Research and 50:422-429. Eur J Clin Nutr. 2014;68:541-548.
Practice 7. Ruby MB, Heine SJ, Kamble S, 21. Foster M, Chu A, Petocz P, Samman S.
Cheng TK, Waddar M. Compassion Effect of vegetarian diets on zinc status:
The findings of this analysis have impli- and contamination: cultural differences in a systematic review and meta-analysis
cations for future studies and health vegetarianism. Appetite. 2013;71:340-348. of studies in humans. J Sci Food Agric.
care. Missing from this survey was the 8. Fox N, Ward KJ. You are what you 2013;93:2362-2371.
assessment of whether individuals mini- eat? Vegetarianism, health and identify. 22. Jackson J, Williams R, McEvoy M,
mally consume meat and/or follow a Soc Sci Med. 2008;66:2585-2595. MacDonald-Wicks L, Patterson A. Is
full plant-based diet for reasons other 9. Lea EJ, Crawford D, Worsley A. Public higher consumption of animal flesh
than health. About 10.0% of the US views of the benefits and barriers to the foods associated with better iron status
population follows a diet focusing on consumption of a plant-based diet. Eur J among adults in developed countries?
plant-based nutrition with minimal Clin Nutr. 2006;60:828-837. A systematic review. Nutrients. 2016;8:
meat consumption.41 The prevalence 10. Huang RY, Huang CC, Hu FB, 89.
of following this form of diet for health Chavarro JE. Vegetarian diets and 23. Mao X, Shen X, Tang W, et al. [Preva-
reasons and its perceived health bene- weight reduction: a meta-analysis of lence of vegetarians and vegetarian’s health
fits warrant further investigation. Thus, randomized controlled rrials. J Gen dietary behavior survey in Shanghai]. Wei
future studies might provide more de- Intern Med. 2016;31:109-116. Sheng Yan Jiu. 2015;44:237-241.
tails on the use of vegetarian and 11. Orlich MJ, Fraser GE. Vegetarian diets in 24. Ponzio E, Mazzarini G, Gasperi G,
vegan diets, including limited use vs the Adventist Health Study 2: a review of Bottoni MC, Vallorani S. The vege-
full-time use and complete meat absti- initial published findings. Am J Clin tarian habit in Italy: prevalence and
nence vs vegetarian-inclined diets Nutr. 2014;100(suppl 1):353S-358S. characteristics of consumers. Ecol Food
among other factors. To exploit the 12. Yokoyama Y, Nishimura K, Nutr. 2015;54:370-379.
vegetarian and vegan diets’ full poten- Barnard ND, et al. Vegetarian diets 25. Jaacks LM, Kapoor D, Singh K, et al.
tial for prevention and condition man- and blood pressure: a meta-analysis. Vegetarianism and cardiometabolic dis-
agement, further studies are warranted JAMA Intern Med. 2014;174:577-587. ease risk factors: differences between
to examine their efficacy and safety in 13. Wang F, Zheng J, Yang B, Jiang J, Fu Y, South Asian and US adults. Nutrition.
general, and in specific medical condi- Li D. Effects of vegetarian diets on blood 2016;32:975-984.
Journal of Nutrition Education and Behavior Volume 49, Number 7, 2017 Cramer et al 567
26. The Vegetarian Resource Group. How Health Interview Survey. Prev Chronic 43. Chiu YF, Hsu CC, Chiu TH, et al.
many adult vegetarians in the U.S.? Dis. 2016;13:E61. Cross-sectional and longitudinal com-
http://www.vrg.org/press/201511press.htm. 36. Rhee TG, Evans RL, McAlpine DD, parisons of metabolic profiles between
Accessed April 20, 2017. Johnson PJ. Racial/ethnic differences vegetarian and non-vegetarian subjects:
27. Worsley A, Skrzypiec G. Teenage vege- in the use of complementary and alter- a matched cohort study. Br J Nutr. 2015;
tarianism: prevalence, social and cognitive native medicine in US adults with 114:1313-1320.
contexts. Appetite. 1998;30:151-170. moderate mental distress: results from 44. Sabate J, Wien M. A perspective on
28. Larsson CL, Klock KS, Astrom AN, the 2012 National Health Interview vegetarian dietary patterns and risk of
Haugejorden O, Johansson G. Food Survey. J Prim Care Community Health. metabolic syndrome. Br J Nutr. 2015;
habits of young Swedish and Norwe- 2017;8:43-54. 113(suppl 2):S136-S143.
gian vegetarians and omnivores. Public 37. Cramer H, Hall H, Leach M, et al. 45. Elorinne AL, Alfthan G, Erlund I, et al.
Health Nutr. 2001;4:1005-1014. Prevalence, patterns, and predictors of Food and nutrient intake and nutri-
29. Agrawal S, Millett CJ, Dhillon PK, meditation use among US adults: a na- tional status of Finnish vegans and
Subramanian SV, Ebrahim S. Type of tionally representative survey. Sci Rep. non-vegetarians. PLoS One. 2016;11:
vegetarian diet, obesity and diabetes in 2016;10:36760. e0148235.
adult Indian population. Nutr J. 2014; 38. Cramer H, Ward L, Steel A, Lauche R, 46. Schupbach R, Wegmuller R,
13:89. Dobos G, Zhang Y. Prevalence, pat- Berguerand C, Bui M, Herter-Aeberli I.
30. Kessler CS, Michalsen A. Veg* in der terns, and predictors of yoga use: results Micronutrient status and intake in omni-
Medizin—schon langst viel mehr als of a U.S. nationally representative sur- vores, vegetarians and vegans in
nur ein Trend [in German]. Forsch Kom- vey. Am J Prev Med. 2016;50:230-235. Switzerland. Eur J Nutr. 2017;56:283-293.
plementmed. 2016;23:64-65. 39. Lauche R, Wayne PM, Dobos G, 47. Kang SJ, Kim EH, Ko KJ. Effects of
31. Centers for Disease Control and Pre- Cramer H. Prevalence, patterns, and aerobic exercise on the resting heart
vention. National Health Interview predictors of t’ai chi and qigong use in rate, physical fitness, and arterial stiff-
Survey. https://www.cdc.gov/nchs/nhis/. the United States: results of a nationally ness of female patients with metabolic
Accessed April 20, 2017. representative survey. J Altern Comple- syndrome. J Phys Ther Sci. 2016;28:
32. Blackwell DL, Lucas JW, Clarke TC. ment Med. 2016;22:336-342. 1764-1768.
Summary health statistics for U.S. adults: 40. Barnes PM, Powell-Griner E, 48. Lee YM, Kim SA, Lee IK, et al. Effect
national health interview survey, 2012. McFann K, Nahin RL. Complemen- of a brown rice based vegan diet and
Vital Health Stat 10. 2014;(260):1-161. tary and alternative medicine use conventional diabetic diet on glycemic
33. Perrett DJ, Schaffer J, Piccone A, among adults: United States, 2002. control of patients with type 2 diabetes:
Roozeboom M. Bonferroni adjust- Adv Data. 2004;(343):1-19. a 12-week randomized clinical trial.
ments in tests for regression coefficients. 41. Vegetarian Times. Vegetarianism in PLoS One. 2016;11:e0155918.
Multiple Linear Regression Viewpoints. America. http://www.vegetariantimes. 49. Orlich MJ, Singh PN, Sabate J, et al.
2006;32:1-6. com/article/vegetarianism-in-america/. Vegetarian dietary patterns and the
34. Hosmer DW, Lemeshow S. Applied Lo- Accessed April 20, 2017. risk of colorectal cancers. JAMA Intern
gistic Regression.. New York, NY: Wi- 42. Zhang Y, Leach MJ, Bishop FL, Med. 2015;175:767-776.
ley-Interscience; 2000. Leung B. A comparison of the character- 50. Alnemer KA, Alhuzaim WM,
35. Falci L, Shi Z, Greenlee H. Multiple istics of acupuncture- and non-acupunc- Alnemer AA, et al. Are health-related
chronic conditions and use of comple- ture-preferred consumers: a secondary tweets evidence based? Review and anal-
mentary and alternative medicine among analysis of NHIS 2012 data. J Altern ysis of health-related tweets on Twitter. J
US adults: results from the 2012 National Complement Med. 2016;22:315-322. Med Internet Res. 2015;17. e246.
567.e1 Cramer et al Journal of Nutrition Education and Behavior Volume 49, Number 7, 2017
CONFLICT OF INTEREST
The authors have not stated any con-
flicts of interest.