ORIGINAL INVESTIGATION

White Rice, Brown Rice, and Risk of Type 2 Diabetes in US Men and Women
Qi Sun, MD, ScD; Donna Spiegelman, ScD; Rob M. van Dam, PhD; Michelle D. Holmes, MD, DrPH; Vasanti S. Malik, MSc; Walter C. Willett, MD, DrPH; Frank B. Hu, MD, PhD

Background: Because of differences in processing and nutrients, brown rice and white rice may have different effects on risk of type 2 diabetes mellitus. We examined white and brown rice consumption in relation to type 2 diabetes risk prospectively in the Health Professionals Follow-up Study and the Nurses’ Health Study I and II. Methods: We prospectively ascertained and updated diet, lifestyle practices, and disease status among 39 765 men and 157 463 women in these cohorts. Results: After multivariate adjustment for age and other

per month) was associated with a lower risk of type 2 diabetes: pooled relative risk, 0.89 (95% CI, 0.81-0.97). We estimated that replacing 50 g/d (uncooked, equivalent to one-third serving per day) intake of white rice with the same amount of brown rice was associated with a 16% (95% CI, 9%-21%) lower risk of type 2 diabetes, whereas the same replacement with whole grains as a group was associated with a 36% (30%-42%) lower diabetes risk.
Conclusions: Substitution of whole grains, including

lifestyle and dietary risk factors, higher intake of white rice ( 5 servings per week vs 1 per month) was associated with a higher risk of type 2 diabetes: pooled relative risk (95% confidence interval [CI]), 1.17 (1.02-1.36). In contrast, high brown rice intake ( 2 servings per week vs 1

brown rice, for white rice may lower risk of type 2 diabetes. These data support the recommendation that most carbohydrate intake should come from whole grains rather than refined grains to help prevent type 2 diabetes. Arch Intern Med. 2010;170(11):961-969 lation, in which white rice consumption was the primary source of carbohydrate (74% of dietary glycemic load).6

Author Affiliations: Departments of Nutrition (Drs Sun, van Dam, Willett, and Hu and Ms Malik), Epidemiology (Drs Spiegelman, van Dam, Holmes, Willett, and Hu and Ms Malik), and Biostatistics (Dr Spiegelman), Harvard School of Public Health; the Channing Laboratory (Drs van Dam, Holmes, Willett, and Hu), Department of Medicine, Brigham and Women’s Hospital and Harvard Medical School; all at Boston, Massachusetts.

in Asian countries for centuries. By the 20th century, the advance of grainprocessing technology made large-scale production of refined grains possible.1 Through refining processes, the outer bran and germ portions of intact rice grains (ie, brown rice) are removed to produce white rice that primarily consists of starchy endosperm. Although findings are not entirely consistent, consumption of white rice, in general, generates a stronger postprandial blood glucose response as measured by the glycemic index (GI) than the same amount of brown rice. A systematic review found that the mean (SD) GI was 64 (7) for white rice and 55 (5) for brown rice.2 Higher dietary GI has been consistently associated with elevated risk of type 2 diabetes (T2D) in prospective cohort studies.3-6 In addition, brown rice consumption may impart beneficial effects on T2D risk by virtue of its high content of multiple nutrients, such as fiber, vitamins, and minerals, the majority of which are lost during refining and milling processes.7 In line with these observations, high intake of white rice was associated with a monotonically elevated risk of developing T2D in a Chinese popu-

R

ICE HAS BEEN A STAPLE FOOD

CME available online at www.jamaarchivescme.com and questions on page 924
Compared with Asian countries, rice consumption is much lower in the United States but is increasing rapidly. According to the US Department of Agriculture 2009 food supply and disappearance data, rice consumption has increased more than 3-fold since the 1930s to reach 20.5 lbs (9.3 kg) per capita, and more than 70% of rice consumed is white rice.8 However, little is known about whether rice intake is associated with diabetes risk in US populations. We therefore evaluated the associations between intake of white rice and brown rice and risk of T2D in 3 large cohort studies with repeated prospective dietary assessments. We have previously observed an inverse association between whole grain consumption and risk of T2D in these cohorts.9,10 In the present study, we extended the follow-up of these previously reported studies and evaluated whether substituting whole grains for white rice is associated with a lower risk of diabetes.

(REPRINTED) ARCH INTERN MED/ VOL 170 (NO. 11), JUNE 14, 2010 961

WWW.ARCHINTERNMED.COM

©2010 American Medical Association. All rights reserved.

and treatment was collected. To better represent long-term diet and to minimize the within-person variation. 2006 (HPFS). alcohol intake. the date of diagnosis of T2D. a 116-item FFQ was administered among the NHS I participants to collect information on their usual intake of foods and beverages in the previous year. we replaced missing values with valid ones from a previous FFQ. 2005 (NHS II). we categorized participants into 5 categories ( 1 serving per month. To minimize confounding by other dietary factors.22 STUDY POPULATIONS We used data from 3 prospective cohort studies: the Health Professionals Follow-up Study (HPFS. 2010 962 ©2010 American Medical Association. 32-87 years) and the Nurses’ Health Study (NHS) I (age range. In addition.18 For cases of T2D identified after 1998. By means of the expanded FFQ used in the NHS I. 11. when we first assessed white rice and brown rice consumption in these cohorts). they consumed each food of a standard portion size. dietary data were collected every 4 years during 1986 through 2002 among the HPFS participants and during 1991 through 2003 among the NHS II participants. For NHS I and II participants. on the basis of the distribution of responses to rice intake questions. we excluded HPFS participants who left more than 70 of the 131 food items blank on the baseline food frequency questionnaire (FFQ) or who reported unusual total energy intake levels (ie.METHODS STUDY OUTCOME The study outcome was incident T2D that occurred between the return of the baseline FFQ and January 31. men and women who reported a diagnosis of T2D in the biennial follow-up questionnaires were sent a supplementary questionnaire to confirm the diagnosis. we excluded men and women who had diagnoses of diabetes. All of these covariates are established risk factors for type 2 diabetes and were correlated with white rice or brown rice consumption in these cohorts. JUNE 14. We used the criteria from the National Diabetes Data Group to confirm self-reported diagnosis of T2D. brown rice is a whole grain. cardiovascular disease.53 for white rice and 0. and 5 servings per week) of white rice intake and 3 categories ( 1 serving per month. 2006 (NHS I). 1-4 servings per month. physical activity. and 1991 for NHS II. We used the same approach to examine such an association for whole grains treated as a single food item.26 WWW.11-13 In all 3 cohort studies. data from 39 765 (of 51 530) HPFS participants. and 23. June 30. 97% (57 of 59) of self-reported type 2 diabetes cases were confirmed by means of medical record review. we excluded those who left more than 10 (NHS I) or 9 (NHS II) items blank on baseline FFQs or whose total energy intake was less than 500 or greater than 3500 kcal/d. and endosperm as whole grains. (REPRINTED) ARCH INTERN MED/ VOL 170 (NO. we created cumulative averages of food and nutrient intake from baseline to the censoring events. smoking status. At least 98% of deaths among the study participants were identified. On average. In the current analysis. we applied the American Diabetes Association criteria. By definition. 26-45 years). age range. whichever came first. Among nurses. to collect and update information on lifestyle practice and occurrence of chronic diseases. and 2 servings per week) of brown rice intake to warrant appropriate variation in rice consumption while preserving enough statistical power to make stable estimates for each category. questionnaires were administered at baseline. The corrected Pearson correlation coefficients between these 2 assessments were 0. All rights reserved. and hormone use. The study was approved by the Human Research Committee of Brigham and Women’s Hospital and the Human Subjects Committee Review Board of Harvard School of Public Health. and cancer at baseline for the dietary analyses (1986 for HPFS. hypercholesterolemia.41 for brown rice. germ. similar but expanded FFQs were sent to these participants to update their diet information every 4 years. 12. This method was used in our previous studies.23 in which we stratified the analysis jointly by age in months at baseline and calendar year to control for confounding by these factors as finely as possible. or cancer. in 61 (98%) of them the diagnosis was confirmed after their medical records were reviewed by an endocrinologist blinded to the supplementary questionnaire information. we adjusted for oral contraceptive use (NHS II participants only).3% of NHS II.21 Of a random sample of 62 nurses reporting type 2 diabetes in the supplementary questionnaire. postmenopausal status.1% of HPFS participants had missing data after baseline assessment. Detailed descriptions of these 3 cohorts were introduced elsewhere.14 Assessment of whole grain intake was described in detail elsewhere. as well as biennially after baseline. In all 3 cohorts. 1 serving per week. 1984 for NHS I. cardiovascular disease. 2-4 servings per week. we asked the participants how often.14-17 In a validation study conducted among a subsample of HPFS participants. In the current study. 69 120 (of 81 755) NHS I participants. or June 30. 11).7% of NHS I. 2006 (NHS I). 2005 (NHS II). coffee. In multivariate analysis.25 To estimate the association of substituting brown rice intake for the same amount of white rice. we further adjusted for ethnicity. diagnostic tests. We used the difference between regression coefficients for brown rice and white rice to derive the RR measuring this association of substitution. To address missing values of dietary variables in the follow-up FFQs.21 Deaths were identified by reports from next of kin or postal authorities or by searching the National Death Index. equivalent to one-third serving of white rice per day) in the same multivariate model.20 In another validation study conducted in HPFS participants. information on symptoms. multivitamin use. or January 31. For these participants.ARCHINTERNMED. or June 30. body mass index. We have used this approach in our previous studies to avoid systematic errors in dietary assessment due to potential biased recall after occurrence of chronic diseases. fruits and vegetables. we carried forward the cumulative averages of dietary intake before the occurrence of these diseases to represent diet for later followup. In all FFQs. In this supplementary questionnaire. . and family history of diabetes. The reproducibility and validity of these FFQs have been demonstrated in detail elsewhere. and 88 343 (of 95 452) NHS II participants were available for the analysis.9 We considered any intact or milled form of grain that consisted of the expected proportions of bran. we further adjusted for total energy intake and intake of red meat. During 1986 through 2002. The completion of the self-administered questionnaire was considered to imply informed consent. STATISTICAL ANALYSIS We counted each individual’s person-years of follow-up from the date of return of the baseline FFQ to the date of death. After exclusions. The relative risks (RRs) were estimated by Cox proportional hazards regression. The follow-up rates of the participants in these cohorts are all greater than 90%. we included both white rice and brown rice intake as continuous variables (50 g/d. 2006 (HPFS). June 30. assessments of white rice and brown rice intake were moderately correlated with diet record assessments.20. and whole grains.24. daily energy intake 800 or 4200 kcal/d). 37-65 years) and II (age range. 1-3 servings per month. on average.24 We stopped updating diet when participants first reported having a diagnosis of hypertension.19 The validity of the supplementary questionnaire for the diagnosis of diabetes has been described previously.COM ASSESSMENT OF RICE CONSUMPTION In 1984.

8 17. servings/wk Brown rice.7 293 4.57 121 1.8 7.3 5. Baseline Characteristics of Study Participants by Levels of White Rice and Brown Rice Intake White Rice Intake Characteristic No. All rights reserved.5 34.5 41.0 0.Table 1.3 24.6 16 134 51.3 16.7 2.56 99 1.1 4.9 4. % No.58 125 1.2 795 51.1 24.8 4.4 98.5 21. Texas).2 22.5 1.3 0. % of total energy Cereal fiber. cups/d Multivitamin supplement user.4 27.0 0.61 114 1.1 20.68 109 1.0 16.5 10.6 5.1 44.3 10.1 51 673 50.6 NHS I 17 839 49.2 0. % Postmenopausal.54 99 2.2 22.8 24. Age. servings/wk Fruit and vegetable intake.8 2.9 11.6 0.5 0.3 33.8 1791 8. % High cholesterol.6 17.4 335 11.2 4.2 1943 8.6 1.3 5.4 2042 7.6 19.9 20.8 24. The pooling analysis was conducted with STATA 10.9 351 11.1 51. 11).0 96.0 15. % white Current smoker.4 5.5 0. we conducted a meta-analysis using fixed-effects models. servings/d White rice.2 0.9 7.7 17.5 0.3 0.3 14 690 50.8 0.2 33. g/d P:S ratio Glycemic load Trans fat.6 2229 5.62 143 1.6 1585 8.9 18.2 364 11. % Family history of diabetes. % Hypertension.5 3.6 19.8 426 10.59 126 1.5 2099 8.7 23.2 7.8 0.1 0. North Carolina). version 9.2 23.0 0 0.2 37.0 5.1 7.0 24.2 14. RESULTS We documented 2648 incident T2D cases during 20 years of follow-up in the HPFS.4 43.0 39.0 4.5 20.54 122 1.6 1981 9.9 14.9 40.7 359 8.0 94.2 3. College Station. % of total energy Cereal fiber.8 2.3 296 6.1 19. mg/d Alcohol.8 0 11. g/d Magnesium.5 24. 5500 cases during 22 years in the NHS I.7 25.7 0.9 4.0 5.3 4.8 7.ARCHINTERNMED.9 96.1 (SAS Institute Inc.4 10.6 24.2 0. P values for heterogeneity of study results were calculated by using the Cochran Q test.6 0.8 15.2 17.6 4.5 18.3 39.2 2.5 312 7. g/d Coffee.8 19.6 19. y Physical activity.4 93. MET-hours BMI Race.COM (REPRINTED) ARCH INTERN MED/ VOL 170 (NO. g/d Magnesium.8 23.9 0 16.4 24.7 1.4 0 0.7 44.0 24.6 1. servings/wk Whole grains.2 24. and 2359 cases during 14 years in the NHS II.57 100 1.5 12.4 0.56 98 1.0 9. Data were analyzed with the SAS package.8 25. % Total energy.0 368 6. JUNE 14. servings/wk Brown rice. kcal/d Red meat intake.5 36.7 12.5 14.2 1. g/d Coffee.5 22. % Family history of diabetes.4 24.9 22.0 43.2 13.0 278 6. mg/d Alcohol.27 All P values were 2-sided.4 45.2 6. Men and women who had high white rice intake were less likely WWW.9 19. y Physical activity. MET-hours BMI Race.6 2.6 5.9 44.4 39.0 7.9 8. To summarize the estimates of association across the 3 studies.6 1/wk HPFS 9968 51. servings/wk Fruit and vegetable intake.1 37.8 0.3 20. servings/wk Whole grains.6 1852 9. 2010 963 ©2010 American Medical Association.2 36.1 45. g/d P:S ratio Glycemic load Trans fat. kcal/d Red meat intake.4 13.7 1.9 33.4 2198 51.7 13.6 51.4 1724 9.9 13.4 6.2 49.2 0.8 0.8 1.8 98.8 96.3 2.3 2.5 28.3 24.8 1.7 3.1 1958 6. .9 357 11.3 5.71 138 0.0 51.8 44.7 0.0 10. % Postmenopausal hormone use.3 5.7 7.7 (continued) 5/wk 1/mo Brown Rice Intake 1/mo–1/wk 2/wk Tests for trend were conducted by assigning the median value to each category and modeling this value as a continuous variable. Age. % white Current smoker. % a Hypertension.8 600 49.8 96.0 9.6 1437 50.2 5.7 24. servings/d White rice.4 7.6 49.4 0.8 98.8 25.4 7.3 6. Table 1 describes the distribution of baseline characteristics according to intake of white rice and brown rice. % Total energy.9 97.8 44.6 2200 8. % High cholesterol. % 1/mo 9386 54.3 1851 7.0 12.1 7. cups/d Multivitamin supplement user.3 25.6 5.2 25.7 7.2 17.0 (StataCorp. Cary.5 23.1 0.8 21 433 53.9 43.5 9. We calculated 95% confidence intervals (CIs) for RRs.7 9.0 21.0 13.7 16 010 50.1 288 7.

but a trend of increased risk associated with high white rice intake remained.5 94. % Postmenopausal hormone use.8 0.4 0.7 1737 5.40). % a Hypertension. In the pooled analysis.5 44.7 82.810.001) higher risk of developing T2D. % white Current smoker. In addition.9 4744 36.4 6. high white rice intake was associated with high fruit and vegetable intake and low intake of whole grains.6 3.1 81.5 13.9 321 2. In contrast to white rice. All rights reserved.9 6.3 21.8 1.79-0. % Family history of diabetes.6 1.9 4. brown rice intake was not associated with ethnicity but with a more health-conscious lifestyle and dietary profile.5 16.0 23. cups/d Multivitamin supplement user. In all 3 cohorts.5 13. vegetables. Age. in comparison with those in the lowest category of white rice intake. After multivariate adjustment for covariates.2 5.7 385 3. When compared with the participants who ate less than 1 serving of brown rice per month.Table 1. these associations were attenuated. but not germ intake. g/d P:S ratio Glycemic load Trans fat (% of total energy) Cereal fiber.7 24.54 122 1.2 299 2. Table 2 shows the RRs of T2D according to white rice intake.1 13. was associated with a lower risk of developing T2D.1 0 0.7 5. servings/wk Fruit and vegetable intake.3 1.9 0. % Postmenopausal.3 6.4 0.5 0. NHS.0 3. We observed a monotonically decreasing risk of diabetes associated with increasing consumption of whole grains.9 5. participants with higher brown rice intake were more physically active. In contrast.5 5/wk 1/mo Brown Rice Intake 1/mo–1/wk 2/wk Abbreviations: BMI.6 5. each 50-g/d intake of brown rice replacing white rice was associated with an RR (95% CI) of 0.84 (0.70).5 94. 2010 964 ©2010 American Medical Association.0 0. After the multivariate estimates were summarized across the 3 studies. polyunsaturated to saturated fat ratio.0 1. 11). MET-hours BMI Race.8 1. y Physical activity.1 10.6 1/wk NHS II 25 808 36.6 43. In comparison with the lowest quintile.6 51. Bran intake.710.9 3. Nurses’ Health Study.1 1901 36.2 94.5 6.3 30. servings/wk Whole grains.9 32 382 36. metabolic equivalent task.2 18.78.4 11. P:S.4 20. 2%-36%. servings/d White rice. Both white rice and brown rice intake was positively associated with a higher glycemic load in all 3 cohorts.3 18.2 11. g/d Magnesium. participants who ate at least 5 servings of white rice per week had a 17% (95% CI.5 24. and less likely to smoke or have a family history of diabetes and had higher intake of fruits. Baseline Characteristics of Study Participants by Levels of White Rice and Brown Rice Intake (Continued) White Rice Intake Characteristic No.5 24. JUNE 14.3 14. P for trend . .1 0.9 6.001).2 7. to have European ancestry or to smoke and more likely to have a family history of diabetes. white rice intake was associated with an elevated risk of developing T2D across the 3 studies. brown rice intake was associated with a lower risk of T2D in age-adjusted models (Table 3).3 44.3 20. We subsequently examined the RR associated with the replacement of 50 g (one-third serving) of white rice per day with the same amount of brown rice intake.1 13. substituting brown rice for white rice was consistently associated with a lower risk of T2D (Figure).8 1.005. % Total energy.COM (REPRINTED) ARCH INTERN MED/ VOL 170 (NO.1 316 2. the pooled RR (95% CI) for the highest quintile of whole grains was 0.3 0.95 (0.4 5. kcal/d Red meat intake.5 0.1 0 15.6 51 217 36. these associations were attenuated but the statisti- cal significance remained.8 24.ARCHINTERNMED.52 121 1.82.6 5.8 92.1 16. For example. After multivariate adjustment for lifestyle and dietary risk factors. a Current and past use of hormone therapy among postmenopausal women. For germ intake.61 131 1.3 5.8 1829 5. P for trend .7 24.03.76 (0. and whole grains and lower intake of red meat and trans fat.3 14.7 5.6 23.0 72.0 51.50 120 1. mg/d Alcohol.7 4. We further estimated the RRs of T2D associated with bran and germ intake (Table 4).6 16. HPFS.64 (0.58-0. MET.3 15.4 2083 5.57 137 1.001).6 316 3.4 24.0 331 3.4 82.7 1852 5.8 1. In comparison with those in the lowest quintile of bran intake. In age-adjusted models.4 4.88-1.2 24. the corresponding RR was 0. the pooled RR (95% CI) of T2D was 0.7 13.1 3. P for trend .52 121 1.1 3.73 (0.0 5. with a P for trend of . and trans fat.8 1633 4.7 7. We further examined the RR associated with replacing 50 g of white rice intake per day with the same amount of whole grains: the RR (95% CI) was 0.2 3.0 0.7 14.8 4.7 15.6 8. servings/wk Brown rice.0 92. including brown rice (Table 4).3 82.4 1.6 46.2 14. body mass index (calculated as weight in kilograms divided by height in meters squared). WWW.7 2034 3. men and women in the highest quintile had a pooled RR (95% CI) of 0.5 40. leaner. Health Professionals Follow-up Study.5 46.91). % 1/mo 15 753 36. cereal fiber. g/d Coffee.8 1.0 22.89 (0.2 80. % High cholesterol. P for trend=.97) for intake of 2 or more servings per week.68-0.

the results did not substantially change.16 (1.23) 1183 327 215 1.0).07 (0. and Asian).20 .35 (1. postmenopausal status. Health Professionals Follow-up Study.09 (0.00 (0. 1. Among white participants.96-1.9. 10.12-1.34) 96 18 150 1.27) 1.1-4.72-0.82 (0.08 (0. currently smoke 1-14 cigarettes per day.04 (0.26 .67) 1.08-1. 30. and Asian) were excluded.COM (REPRINTED) ARCH INTERN MED/ VOL 170 (NO. and oral contraceptive use were further adjusted for. Risk of Type 2 Diabetes Mellitus According to White Rice Intake in the HPFS.60) 1.82 (0.43 (1.30) 1.10 (1.01-1.03 (0.12-1.0-9.96) 0.14) 1.82-1. When we restricted our analysis within minority groups only.0-14.82-1.09 (0.08 1/wk 2-4/wk 5/wk P Value for Trend NHS I NHS II Pooled 1.22) 1. no).92-1.11 (1.96) 0.22 (0.45. However.005 .19) for 2 to 4 servings per week.17-1. Because rice consumption for most of our study participants was relatively stable over time (data not shown).01-1. the pooled RRs (95% CIs) of T2D were 0.81-1.08-1.90 (0.9.11) 1.94 (0.95-1. most associations became nonsignificant because of the dramatically diminished power (we identified 624 cases of T2D among 9644 nonwhite participants).39) 1. body mass index. of cases Person-years RR (95% CI) Model 1 a Model 2 b Model 3 c No. RR.96 (0.58-0. the pooled RR (95% CI) was 1.94-1. .710.01 .93-1.37 (1. P for trend=. whereas brown WWW.9.850. fixed-effects model c Pheterogeneity 320 101 899 1 [Reference] 1 [Reference] 1 [Reference] 415 145 718 1 [Reference] 1 [Reference] 1 [Reference] 248 134 505 1 [Reference] 1 [Reference] 1 [Reference] 1 [Reference] 1-3/mo HPFS 1150 299 985 1.22) 1.93-1.001 .0-34.02 (0. 0.00 (0.0-22. and coffee (all in quintiles).99-1.20) . c Based on model 2. of cases Person-years RR (95% CI) Model 1 a Model 2 b Model 3 c No.26) 1.14) 1.30 (1. body mass index (calculated as weight in kilograms divided by height in meters squared) ( 21.13 (1.20) 1. P .17 (1. of Servings 1/mo No.93 (0.11) 0.49) 1.13) for 1 serving per week.07 (0. 5.Table 2.07-1.02-1.0-26. NHS. 1. we found that regular consumption of white rice was associated with higher risk of T2D.64 (0.9.ARCHINTERNMED.71-0. hormone use.99) for 1 to 4 servings per month and 0.07) 0.96-1. relative risk.10) 553 120 197 1. For example. and NHS II White Rice Intake.04 (0. 15-24 cigarettes per day.24 Finally.34 (1.9.95 (0. whole grains.001) for substituting 50 g of whole grains per day for the same amount of white rice. When we used more recent intake of white rice or brown rice instead of the cumulative average in the analyses.36) .28 the observed associations can be a consequence of confounding by ethnicity.76 (0.0-9.0.9.07) for 1 to 3 servings per month.93 (0. we did not find any interactions between rice consumption and other diabetes risk factors. HPFS. Because ethnicity was associated with both white rice consumption and diabetes risk.0-29.94 (0.12 (0.07 (0. and 0.001) for white rice intake of 2 or more servings per week vs less than 1 serving per month. we found similar results.9.20) 585 324 961 0. 15.0-32.09-1.0-14.92-1.9.9.96-1. 33.43) 102 33 267 1. model 3 was further adjusted for total energy (kilocalories per day) and intake of red meat. Hispanic.07) 537 161 667 1.81-1. fruits and vegetables. of cases Person-years RR (95% CI) Model 1 a Model 2 b Model 3 c RR (95% CI).12) 0.87 (0. a Age-adjusted.84 (0. 0.9. NHS I.02 88 19 172 1.71-0.12-1.11) 924 480 405 0.03 . 21. P for trend .41. these results indicated that the cumulative averages could better represent long-term rice consumption because of reduced random within-person measurement errors.0-24.87-1.02 (0.06 (0.27 (1.001) for substituting 50 g of whole grains per day for the same amount of white rice.93-1. COMMENT In these 3 prospective cohort studies of US men and women.08 (0.12 (1.19 (1.77-1.71.23) 1.003) for 2 or more servings per week. 2010 965 ©2010 American Medical Association.83) 1.9. 23. smoking status (never smoked. African American. All rights reserved. For women.95.96-1.50) 1062 221 681 1.74) 1. The pooled RRs (95% CIs) for brown rice intake levels were 0.78-0.93-1.40 (1. and 15. P .82. ethnicity (white.0-29. or 25 cigarettes per day).81-1. and 1. For example. 10.96-1.08) 1. and family history of diabetes.26) 1. past smoker.02 .87 .1-4.81-1.00-1. 25.81-1. after nonwhite participants (African American. Hispanic.92.03 .55) 1.03-1. alcohol intake (0. b Adjusted for age (years). or 35.001) for replacing 50 g of white rice intake per day with the same amount of brown rice. and 30.95 (0.00-1. No. P . confidence interval. 11).95) 0. 0.38 Abbreviations: CI. multivitamin use (yes.11 (0.03-1. including age.12) . and various comorbidities.23) 500 237 765 0.25 (1.0 g/d for men. 27.9.001 .0 g/d for women). the pooled RRs (95% CIs) of T2D associated with white rice intake were 1. P for trend=. 0.96-1.96. P .24) 2744 691 609 1.90-1.001) for replacing 50 g of white rice intake per day with the same amount of brown rice.97-1. 5.79-0.01 (0. Nurses’ Health Study.9.01-1.06 (0.001 .80) 1.56) 1.83 (0. and 0.001) for 5 or more servings per week.001 . physical activity (quintiles). although we observed largely similar results. in secondary analyses when we repeated these associations among white participants only. JUNE 14.30 (1.900.

75-1.01) 0. In contrast.003 210 61 606 0.6. PhD. 11).001 . only 0. c Based on model 2. which was within the lowest reference level ( 200 g/d) in the prospective study of Chinese women.98) NHS II 941 537 974 0.89 (0.10 .72-0. white rice consumption was prospectively associated with developing T2D in Chinese women living in Shanghai. inverse association for brown rice intake. In addition. Consistent with our previous analyses.83 (0. model 3 was further adjusted for total energy (kilocalories per day) and intake of red meat.74-0.7% of total energy intake (Xiao-Ou Shum MD. fruits and vegetables. White rice consumption contributed.17 . but mixed results were observed. fixed-effects model b Pheterogeneity 1142 283 230 1 [Reference] 1 [Reference] 1 [Reference] 3127 785 713 1 [Reference] 1 [Reference] 1 [Reference] 1271 580 179 1 [Reference] 1 [Reference] 1 [Reference] 1 [Reference] 1296 358 085 0. These associations were independent of lifestyle and dietary risk factors for T2D.53 . higher intake of refined grains including white rice was associated with metabolic risks in cross-sectional analyses.2% (NHS II) of total participants reported having 5 or more servings of white rice per week ( 107 g/d).51 1/mo–1/wk 2/wk P Value for Trend Abbreviations: CI. relative risk. we found that substitution of brown rice for white rice was associated with a significantly lower risk of developing diabetes.62 (0. we detected a significant association for white rice intake. Risk of Type 2 Diabetes Mellitus According to Brown Rice Intake in the HPFS.001 . JUNE 14. of Servings 1/mo HPFS No.96 (0.88-1.29.96) .82-1. Because brown rice consumption levels were rather low in our participants.77 (0.92 (0.6 Likewise.77 147 92 750 0. of cases Person-years RR (95% CI) Model 1 a Model 2 b Model 3 c RR (95% CI).67-0.83 (0.04) NHS I 2167 554 634 0.04) Pooled 0. of cases Person-years RR (95% CI) Model 1 a Model 2 b Model 3 c No. and NHS II Brown Rice Intake. confidence interval.63 (0.88 (0.6 We observed a moderate.COM (REPRINTED) ARCH INTERN MED/ VOL 170 (NO.81-0.90) 0.65-0.21 206 64 026 0. of cases Person-years RR (95% CI) Model 1 a Model 2 b Model 3 c No.90-0. white rice accounted for 29% of daily total energy intake in the Japanese population. brown rice was not separated from white rice or other refined grains in these studies. as well as ethnicity.71 (0.97) . white rice consumption accounted for 53. Health Professionals Follow-up Study.92 (0.001 . coffee (in quintiles).9% (NHS I) to 2. and white rice (see Table 2 for intake categories).74-1.94 (0.95 (0.89-1. a Age-adjusted.52-0.87-0.72) 0.77-0.89 (0.83) 0. we could not determine whether brown rice intake at much higher levels is associated with a further reduction of diabetes risk.87-0. in the aforementioned Chinese female population. the current studies are the first prospective investigations conducted among Western populations that have specifically evaluated white rice and brown rice intake in relation to T2D risk. e-mail communication).30 In comparison with Asian populations. HPFS.12) . No. rice intake was associated with lower risk.Table 3. in which white rice intake of 300 g/d or more (equivalent to 2 servings per day in our analysis) was associated with a 78% increased risk of T2D in comparison with intake levels of less than 200 g/d.6 However.87-1. NHS.91) 0.97) 0. NHS I. MPH. less than 2% of total energy intake in our study populations.31 Consumption of rice or food groups consisting of rice in relation to risk of T2D was also evaluated in Western populations.001 .84 (0.95 (0.78) 0. .ARCHINTERNMED. our data suggest that replacing white rice intake with the same amount of brown rice or whole grains was associated with a lower risk.32-34 To our knowledge.96 (0.92 (0.6 In addition. on average. Nonetheless.10 we found a significant inverse association between whole grain consumption and diabetes risk.96 (0. Our data are consistent with the Chinese study. RR.005 .78 (0.45 .54-0. according to the Japanese National Nutrition Survey. by pooling data from the 3 studies. b Adjusted for the same set of covariates as for model 2 in Table 2.32-34 However. white rice intake in Western populations was much lower. All rights reserved. Nurses’ Health Study.85-1.83-1.9. in Asian Indians and Japanese.30 For example. In our cohorts.12) 0.72-0.03) 0.29. This observation may result from the more reliable WWW. higher white rice consumption has been associated with elevated risk of diabetes or metabolic syndrome.05) 0.96) 0. Substitution of whole grains for white rice was more strongly associated with diabetes risk than was the substitution of brown rice. In Asian populations in which rice is a staple food.07) 0.98) .73) 0. 2010 966 ©2010 American Medical Association.

1-48.83-0.4) 403 243 479 1.5 (0.3) 546 143 049 0.89 (0.1-0.8) 446 242 569 0.6) 769 284 412 0.7) 559 143 546 0.0 (1.78 47.9) 475 144 486 0.43 7.6) 592 147 061 1.77) 40.02) 1.4) 667 234 309 1 [Reference] 1 [Reference] 2.98) .81) 9.6 (0.8) 616 144 206 0. of cases Person-years RR (95% CI) b NHS II Intake levels a No.74-0.6 (0.20) 0.83-0.68-0.Table 4.3 (3.1 (1.79-1.90) .8) 1009 285 877 0.83) .91-1.1-1.2 (1.19) 1.0-8.01) 4.3) 779 281 864 0.13) 0.3-1.2 (0.02 (0. of cases Person-years RR (95% CI) b Pooled RR (95% CI).29 2.5) 393 245 347 0.9-7.2-5.5-0. Health Professionals Follow-up Study.81-0.5-0.4 (17. In addition. of cases Person-years RR (95% CI) b Pooled RR (95% CI).01 .002 .7-7. HPFS.04 (0.9 (0.86 29.85-1.87 (0.73-0.3-0.3 (0.9) 563 141 461 1.86 (0.83) 31.7.2) 667 132 594 1 [Reference] 3. relative risk.96) 2.0 (35.001 .63-0.1-0. of cases Person-years RR (95% CI) b Pooled RR (95% CI). NHS I.04 (0.0) 512 244 811 0.94) 0.7-1. .83) 26.2-3.3 (11.9-28.001 . All rights reserved. fixed-effects model b Pheterogeneity HPFS Intake levels a No.0-15.90 (0.1 (10.2 (4.2-0. and 55 (5) WWW. confidence interval.78 (0.07 .3 (3.74-0.84-1.72 (0. possibly various effects on glucose response. 11).96 (0.2-8.85 (0.5) 971 285 978 0.1-3.06 (0.77-0.92 (0.0) 676 236 653 1 [Reference] 1 [Reference] 12. Nurses’ Health Study.1-57.9-3.0) 1346 271 378 1 [Reference] 6.79-0.001 . a Median (interquartile range).78-1. thus.05) 0.95 (0.39 Germ 0.4-14.1 (2.5 (6.9-5.0 (11.88) 19.6 (0.84 (0.68 14.84 (0.8 (3.8 (2.0-21. whole wheat and barley generate lower glucose response than brown rice: the mean (SD) GI values were 41 (3) for whole wheat.9) 1038 288 588 0.1-2.95 (0.09 4. fixed-effects model b Pheterogeneity HPFS Intake levels a No.83 (0.3 (1.98) 0.ARCHINTERNMED.04) 2.01) 1.69-0.02) 0.7-0.86) 12.04 (0.7-10. For example.82-1.77 (0.6 (2.6 (10.65 .85 (0.8-1.79-1.17 . whole grains included multiple grains with various nutrient composi- tions and.4) 617 234 430 1 [Reference] 1 [Reference] 0.001 .8 (0.5 (1.9 (27.1 (41.8-4.88-1.21) 1.0-14.97) 12. of cases Person-years RR (95% CI) b NHS II Intake levels a No. Bran and germ intake were mutually adjusted for in the analysis.69-0.68-0.97) 13.93 (0.94) 6.41 Bran 0.73-0.92) 18.9-20. of cases Person-years RR (95% CI) b NHS I Intake levels a No.70 (0.8 (16.7-12.95-1. of cases Person-years RR (95% CI) b NHS II Intake levels a No.04 (0. 2010 967 ©2010 American Medical Association. b Adjusted for the same set of covariates as for model 3 in Table 3.78 (0.3) 524 241 524 0.5 (7.97) 2.77-0.COM (REPRINTED) ARCH INTERN MED/ VOL 170 (NO.0-32. of cases Person-years RR (95% CI) b NHS I Intake levels a No.67 1.76 (0.91) .78) .79-0.82-0.2-15.99 (0.0) 484 241 509 1.6) 372 139 558 0.4 (17.7 (5.4) 391 245 747 0.71-0.70-0.74-0.88 (0.7) 464 245 739 0.0) 634 133 254 1 [Reference] 0.03) .80-1.87 (0.81 (0.85 (0.84 (0. 0.2 (6.88-1.6) 1237 287 868 0.1) 1262 278 327 0. JUNE 14.96) 0.6 (2.1-1.01) .5 (1.93) . RR.82 (0.3) 577 129 975 1 [Reference] 0. fixed-effects model b Pheterogeneity 2 Whole Grain 5.94) . estimates of the association with diabetes for whole grains than those for brown rice because of the low overall consumption of brown rice.88-1.3-2. Risk of Type 2 Diabetes Mellitus According to Whole Grain.86 (0.91) 4.3-0.65 0.0-38.6 (11.76 (0.4) 462 243 592 0.92 (0.4-0.89 (0.9) 574 142 608 0.90-1.4) 320 244 259 0.94 (0.00 (0.92) 0.77-0.00) 0.90 (0.64-0.2 (23.69 (0.2) 1224 256 119 1 [Reference] 0.90-1.71-0. Bran. and NHS II Intake Quintile 1 HPFS Intake levels a No.90-1.87-0.89-1.71-0.9) 341 243 867 0.1) 838 276 483 0.8 (0.98) 0.40 .02 .78-0.5-2.6) 377 244 875 0.21) 0. or Germ Intake in the HPFS.0) 1246 281 807 0.92 (0.60-0.2 (1.3 (27.83-1.89 (0.89-1.00) 0.09) 0.4) 506 145 060 1.88-1.1) 1158 283 347 0.88 (0.9) 1278 270 994 1 [Reference] 1.70-0.0 (0. of cases Person-years RR (95% CI) b NHS I Intake levels a No.4) 410 139 362 1.001 .3 (0.81 (0.82) .79 (0.6) 373 139 687 0.92) 8.73 (0.8-18.6 (0.04) 0.7-2.2 (0.9-22.8) 480 142 852 0.97) 0.01) .4) 1163 295 315 0.09) .28 20.93 (0. 25 (1) for barley.3-5. NHS.001 .6) 1182 284 764 0.93-1.23 3 4 5 P Value for Trend .56 .21) 0.06 Abbreviations: CI.

Accepted for Publication: December 19. the FFQs used in these studies were validated against multiple diet records.44-0. P values are P for heterogeneity. replacing refined grains such as white rice by whole grains.36-38 which made it difficult to directly compare white rice with brown rice for effects on postprandial glucose response. Author Contributions: Dr Hu had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.0 1. Health Professionals Follow-up Study. high rates of follow-up. However. lignans.75-0.6%. P = .75 (0. Second. First.A Study HPFS NHS I NHS II Overall (I2 = 66. in comparison with whole wheat and barley. 50 g/d.66-0. magnesium and other minerals. 665 Huntington Ave. Department of Nutrition.4 1. The strengths of the current study include a large sample size. residual confounding is still possible.2. including rice.41.51 (0. both white rice and brown rice demonstrated a wide variety of GI values. although we adjusted for established and potential risk factors for T2D.39 The high GI of white rice consumption is likely the consequence of disrupting the physical and botanical structure of rice grains during the refining process. including brown rice. However.2%. we not only excluded participants with a history of major chronic diseases at baseline but also stopped updating dietary intake after participants reported having diagnoses of diseases that might influence their subsequent report of diet.0 1. some measurement error of rice intake assessment is inevitable.42 In addition. Spiegelman. Although the homogeneity of socioeconomic status helps reduce confounding. should be recommended to facilitate the prevention of T2D.05) 0. for the same amount of white rice intake. white rice consumption generates a relatively stronger postprandial glucose response than the same amount of brown rice.14 Because we used a prospective study design.86) 0.7 In both observational and experimental studies.70 (0. therefore. 11).2 Relative Risk (95% CI) 0.2 Despite this inconsistency inherent to rice GI values.79-0. Third. higher magnesium intake has been consistently associated with reduced risk of T2D in cohort studies or improved glucose metabolism in clinical trials. for brown rice. However. Intact rice grains contain nearly exclusively insoluble fiber. JUNE 14.85 (0. the generalizability of the observed associations may be limited to similar populations. and repeated assessments of dietary and lifestyle information. the supplementary questionnaire that we used for the confirmation of self-reported diabetes diagnoses has been demonstrated to be highly accurate. the same amount of brown rice likely bears a higher glycemic load. we calculated cumulative averages of rice intake to minimize the random measurement errors caused by within-person variation.60) 0. MA 02115 (qisun@hsph. we did not perform oral glucose tolerance tests to confirm diabetes diagnoses because this is infeasible in large cohort studies.6 1.74 (0. To minimize the possibility of systemic measurement error incurred by recall bias. which is an established risk factor for T2D.2%. Harvard School of Public Health.40 The other consequence of the refining process includes loss of fiber.83-1. vitamins.8 as seen in our studies. in which almost all the bran and some of the germ are removed.43-45 The combination of these mechanisms may explain the beneficial effects of replacing white rice with brown rice or other whole grains.05) 0.21 Finally. Moreover. Our data suggest that regular consumption of white rice is associated with an increased risk of T2D. phytoestrogens. and phytic acid.70) Figure.harvard.97) 0. The current Dietary Guidelines for Americans identifies grains. 2009.2%) significantly decreased postprandial glucose and insulin levels in a randomized clinical trial.edu).2 This notion was corroborated by the observation that isocaloric replacement of white rice with whole grains (66. primarily composed of brown rice and barley) and legume powder (22. because diet was assessed by FFQs. Holmes.58-0.35 Depending on the botanical structure. our study populations primarily consisted of working health professionals with European ancestry. P = . are likely to attenuate the associations toward the null. ScD.85) 0. any measurement errors of rice intake are independent of study outcome ascertainment and. as one of the primary sources for carbohydrate intake and recommends that at least half of carbohydrate intake come from whole grains. Bars indicate 95% CIs. and reasonable correlation coefficients between these assessments of rice intake were observed.ARCHINTERNMED. whereas replacement of white rice by brown rice or other whole grains is associated with a lower risk. Study concept and design: Sun. Individual associations were controlled for the same set of covariates as for model 3 in Table 3. NHS. Correspondence: Qi Sun. amylase contents. Nurses’ Health Study. From a public health point of view.46 Rice consumption in the US population is increasing. The consistency of the results across all 3 cohorts indicates that our findings are unlikely to be due to chance.5. MD.84 (0. The current study was subject to a few limitations as well. insoluble fiber intake was consistently associated with improved insulin sensitivity and decreased risk of developing T2D.002) 0. the biological mechanisms underlying the positive associations observed in both our study populations and the Chinese study6 are likely to be the same in other populations.2 As a consequence.20.COM (REPRINTED) ARCH INTERN MED/ VOL 170 (NO. 2010 968 ©2010 American Medical Association. Pooled fixed-effects relative risk and 95% confidence interval (CI) of type 2 diabetes mellitus from substituting intake of brown rice (A) or whole grains (B).7 many of which may be protective factors for diabetes risk.4. All rights reserved.93 (0.64 (0.58-0.8 However.85) 0. HPFS.64-0. and processing methods. most rice consumption is refined white rice.91) B Study HPFS NHS I NHS II Overall (I2 = 84. AcquiWWW. . Boston. and Hu.2 Relative Risk (95% CI) 0. in general.

Malik. Rimm EB. et al. Miller JB. Shai I. van Dam. Colditz GA. Am J Clin Nutr. 8. 2000. and Hu. Dietary fiber. Schofl C. Song Y.ARCHINTERNMED.29(7):1585-1590. Quantifying heterogeneity in a meta-analysis. Williams DE. Gao YT. Palmer JR. McMillan-Price J. Rimm EB. Jang Y. 2001. Sugiyama M. Economic Research Service. Sampson L. Liu S. US Department of Agriculture. Whole grain. Hu FB. Am J Clin Nutr. 18. Hu FB.29(4):775-780. 29. Rimm EB. Litin LB. Rasmussen OW.23(5):991-999. 31. 46. Am J Clin Nutr. Hunter DJ. England: Chapman and Hall. and chronic disease risk: a meta-analysis of observational studies. Reproducibility and validity of food intake measurements from a semiquantitative food frequency questionnaire. and DK58845 from the National Institutes of Health. 1990. Liu S. Oakes D. Am J Clin Nutr. 30. Dietary glycemic index and load in relation to metabolic risk factors in Japanese female farmers with traditional dietary habits. Spiegelman D. glycemic load. JAMA. Drafting of the manuscript: Sun and Hu.20(4):545-550. et al. Stampfer MJ. Rimm EB. 32. Consumption of whole grain and legume powder reduces insulin demand. Hu FB. Asp NG.131(6):1068-1071. 1992. 1979. Glycemic index and dietary fiber and the risk of type 2 diabetes. Stampfer MJ. 1997. Dietary guidelines for Americans.23(10):1050-1056. Cereal fiber improves whole-body insulin sensitivity in overweight and obese women. Manson JE.health. Willett WC. Giovannucci E. Brand-Miller JC. Larsen HN. Stampfer MJ.21(11):1539-1558. Björck I. Belanger CF. Effectiveness of various mailing strategies among nonrespondents in a prospective cohort study. Spiegelman. Wareham NJ. Rasmussen OW. 1985. 2009. Schulz M. 15. Feskanich D. Stampfer MJ. 5. 27. Study supervision: Spiegelman and Hu. Willett.50(1):22-27. et al. Thompson SG. Liu S. Dietary fat and coronary heart disease: a comparison of approaches for adjusting for total energy intake and modeling repeated dietary measurements. Accessed June 16. dietary fiber. Willett WC. Salmeron J. Administrative. Diabetes Care. 2005.52(4):329-335. Am J Clin Nutr. 35. Wolf AM. Cox DR. Am J Clin Nutr. and germ intake and risk of type 2 diabetes: a prospective cohort study and systematic review.21(12):2065-2071. O’Dea K. Prospective study of dietary carbohydrates.28(12):1039-1057. Analysis and interpretation of data: Sun. Liljeberg H. 4. 43. London. Sasaki S. Am J Epidemiol. 2. 25. National Diabetes Data Group. 3. 2007. All rights reserved. Stampfer MJ. analysis. Spiegelman. 1996. Arch Intern Med. Hu FB. Wakaki Y. Rice: a high or low glycemic index food? Am J Clin Nutr.sition of data: Spiegelman and Hu. Expert Committee on the Diagnosis and Classification of Diabetes Mellitus. Arch Intern Med.70(3)(suppl):459S-463S.119(5):837-839. Willett WC. et al. Ethnicity. 135(10):1114-1136. Giovannucci EL.20(7):1183-1197. Kim OY. Am J Public Health. Willett WC. Lancet. 2007. Willett. obesity. Rasmussen PH. Schulze MB. Tang AC. He K.27(11):2701-2706. Arterioscler Thromb Vasc Biol. Granfeldt Y.277(6):472-477. et al. Hales CN. Financial Disclosure: None reported. Report of the Expert Committee on the Diagnosis and Classification of Diabetes Mellitus. Colditz GA. Jacobs DR Jr. Stampfer MJ. 1992. 17. et al.58(5):675-681. 11. Test of the National Death Index. 45. Hodge AM. Am J Epidemiol. Physical activity and television watching in relation to risk for type 2 diabetes mellitus in men. Sudha V. Arch Intern Med. et al.ers. Whole-Grain Foods in Health and Disease. Hu FB. Schulze MB. Prakash A. 2009.57(6):743-752. International table of glycemic index and glycemic load values: 2002. 2002. Reproducibility and validity of a semiquantitative food frequency questionnaire. 2006.0040261. Int J Epidemiol. et al. Metabolism. Barclay AW. and Hu. et al. 2004. Stampfer MJ. Rimm EB. van Dam RM. and plasma homocysteine concentrations in patients with coronary artery disease: randomized controlled clinical trial. Boeing H. 22. Am J Clin Nutr. 2007.93(7):790-796. van Dam RM. van Dam. Willett WC. 23. Liu S. 24. 21. US Dept of Health and Human Services and US Dept of Agriculture. Analysis of Survival Data. 36. et al. major food sources. and risk of ´ NIDDM in men. Int J Epidemiol. Funding/Support: This study was supported by research grants CA87969. and incidence of type 2 diabetes mellitus in middleaged Chinese women. Higgins JP. 44. Meyer KA. Diabetes Care. Giles GG. Slavin JL. Hu FB. Plausible mechanisms for the protectiveness of whole grains. The Nurses’ Health Study. Koyama W. 1994. Am J Clin Nutr. 7. Ganesan A. 33. Kushi LH. Glycemic index of single and mixed meal foods among common Japanese foods with white rice as a reference food. Willett WC. 12. 38. Eur J Clin Nutr.18(4):858-867. Effects of oral magnesium supplementation on glycaemic control in type 2 diabetes: a meta-analysis of randomized double-blind controlled trials. 2010 969 WWW. Petocz P. Holt SH. Accessed June 16. Lee JH.32(4):629-634.pmed. Statistical analysis: Sun. Sun Q. and Willett. Rimm E. Am J Nurs. 26. Malik.78(6):1039-1040.149(6):531-540. 2006. Colditz GA. 2002.4(8):e261. Diabetes Care. REFERENCES 1. doi:10. 2001. 2009. Giovannucci EL. glycemic index. 1991. Manson JE. Park HY. Glycaemic index of parboiled rice depends on the severity of processing: study in type 2 diabetic subjects. 42. Manson JE. 1999. Salvini S. Classification and diagnosis of diabetes mellitus and other categories of glucose intolerance. Reproducibility and validity of a self-administered physical activity questionnaire. 1997. and Hu. bran. Hennekens CH. Refined grain consumption and the metabolic syndrome in urban Asian Indians (Chennai Urban Rural Epidemiology Study 57). 1999. van Dam. Halton TL. Am J Epidemiol. Whole-grain intake and the risk of type 2 diabetes: a prospective study in men. Hu FB. and incident type 2 diabetes in older women. Stat Med. 9. et al. 13. 28. and risk of non–insulin-dependent diabetes mellitus in women. Pang E. Martini MC. Manson JE. Day NE. Am J Clin Nutr. Ascherio A. and risk of type 2 diabetes in women: a 20-year follow-up study. 2002. 1994. Stampfer MJ. Speizer FE. Christensen C. Willett WC. Diabet Med. 2006. Tovar J. Murakami K. Slavin J. 6th ed. Diabetes Care. van Dam RM. Franz M.161(12):1542-1548.gov/Data/. Malik. English DR. A prospective study of whole-grain intake and risk of type 2 diabetes mellitus in US women.76(1):5-56. 2003. et al. Diabetes. Eur J Clin Nutr.usda. http://www . et al. Rimm EB. Radhika G. 2006. technical. Mohan V.S. Fiber and magnesium intake and incidence of type 2 diabetes: a prospective study and meta-analysis.COM ©2010 American Medical Association. Willett WC.76(3):535-540. Villegas R. Schienkiewitz A. MN: American Association of Cereal Chemists. Dietary calcium and magnesium. 2000.1371/journal. Holmes. et al. Krishnan S. 40. 2008. Hunter DJ. 20. JUNE 14. 2002. 2004. 2006. Miller G. PLoS Med. Van Dam RM.71(4):921-930. Sellers TA. CA050385. 1989. and dietary fiber intake and incidence of type 2 diabetes in younger and middle-aged women. Diabetes Care. 1984. Eur J Clin Nutr. Carbohydrates. and material support: Sun. Jacobs DR Jr. 1978. Wing AL. Glycemic index. Role of the Sponsor: The National Institutes of Health had no role in the collection. 1993.167(9):956-965. Hoffmann K.gov/dietaryguidelines /dga2005/document/. Lee SY.167(21):2310-2316. glycemic load. Jiang R. and interpretation of the data or in the decision to submit the manuscript for publication. glycemic load. Hu FB. black women. Manson JE. 2000. Physical activity and incidence of non– insulin-dependent diabetes mellitus in women. Data sets. Rosner B. 6. Am J Epidemiol. Sampson L. Reproducibility and validity of an expanded self-administered semiquantitative food frequency questionnaire among male health professionals. 2009. Rimm EB. St Paul. Willett WC. glycemic load. Prospective study of zinc intake and risk of type 2 diabetes in women. et al. 39.56(6):1034-1036. J Clin Epidemiol. Dietary ´ fiber. 14. Decker E. lipid peroxidation. Heidemann C. glycemic load. 41. 10. Pereira MA. Obtained funding: Spiegelman. Leitzmann MF. Fung TT.87(3):627-637. 1997. 16.80(2):348-356. Manson JE. Rosenberg L. Frequent salad vegetable consumption is associated with a reduction in the risk of diabetes mellitus. http://www. Liu S. 19.122(1):51-65. 1984. Food properties affecting the digestion and absorption of carbohydrates. CA055075. Glycemic index. Bramall L. and Hu.338(8770):774778. Influence of parboiling and physico-chemical characteristics of rice on the glycaemic index in non–insulindependent diabetic subjects. Diabetes Care. Colditz GA. Stampfer MJ. Salmeron J. Foster-Powell K. 11). 34. 37. Larsen HN. Weickert MO. Manson JE. Diabetes Care. et al.59(3) (suppl):699S-705S. Byrne CD. Marquart L. 90(9):1409-1415. 2006. Food-based validation of a dietary questionnaire: the effects of week-to-week variation in food consumption. Stampfer MJ. Spiegelman. (REPRINTED) ARCH INTERN MED/ VOL 170 (NO. Am J Epidemiol. Critical revision of the manuscript for important intellectual content: Sun.83(5):1161-1169. 83(2):284-290. Takahashi Y. 1999. et al. Mohlig M. Colditz GA. Holmes. Potato and French fry consumption and risk of type 2 diabetes in women. J Am Diet Assoc. Cox BD. Levitan EB. Folsom AR. de Munter JS. and risk of type 2 diabetes in U. .54(5):380-385. Dr Sun is supported by a postdoctoral fellowship from Unilever Corporate Research. Speizer FE.29(10):2238-2243.

Sign up to vote on this title
UsefulNot useful

Master Your Semester with Scribd & The New York Times

Special offer for students: Only $4.99/month.

Master Your Semester with a Special Offer from Scribd & The New York Times

Cancel anytime.