Randomization to plant-based dietary approaches leads to larger short-term improvements in Dietary Inflammatory Index scores and macronutrient intake compared with diets that contain meat

N U TR IT ION RE S EA RCH 3 5 ( 2 01 5 ) 9 7 –1 06

Available online at www.sciencedirect.com

ScienceDirect www.nrjournal.com

Randomization to plant-based dietary approaches leads to larger short-term improvements in Dietary Inflammatory Index scores and macronutrient intake compared with diets that contain meat☆ Gabrielle M. Turner-McGrievy a,⁎, Michael D. Wirth b , Nitin Shivappa b , Ellen E. Wingard c , Raja Fayad c , Sara Wilcox c , Edward A. Frongillo a , James R. Hébert b a

University of South Carolina, Arnold School of Public Health, Department of Health Promotion, Education, and Behavior, 915 Greene St, Room 529, Columbia, SC 29208 b University of South Carolina, Arnold School of Public Health, Department of Epidemiology and Biostatistics, Cancer Prevention and Control Program, 915 Greene St, Columbia, SC 29208 c University of South Carolina, Arnold School of Public Health, Department of Exercise Science, 921 Assembly St, Columbia, SC 29208 USA

ARTI CLE I NFO

A BS TRACT

Article history:

Studies have examined nutrient differences among people following different plant-based

Received 10 September 2014

diets. However, all of these studies have been observational. The aim of the present study

Revised 21 November 2014

was to examine differences in nutrient intake and Dietary Inflammatory Index (DII) scores

Accepted 27 November 2014

among overweight and obese (body mass index 25.0-49.9 kg/m2) adults randomized to receive dietary instruction on a vegan (n = 12), vegetarian (n = 13), pescovegetarian (n = 13),

Keywords:

semivegetarian (n = 13), or omnivorous (n = 12) diet during a 6-month randomized

Diet

controlled trial. Nutrient intake, nutrient adequacy, and DII score were assessed via two 24-

Vegetarian

hour dietary recalls (Automated Self-Administered 24-Hour Dietary Recall) at baseline and

Vegan

at 2 and 6 months. Differences in nutrient intake and the DII were examined using general

Inflammation

linear models with follow-up tests at each time point. We hypothesized that individuals

Nutrients

randomized to the vegan diet would have lower DII scores and greater improvements in fiber, carbohydrate, fat, saturated fat, and cholesterol at both 2 and 6 months as compared with the other 4 diets. Participants randomized to the vegan diet had significantly greater changes in most macronutrients at both time points, including fat and saturated fat, as well as cholesterol and, at 2 months, fiber, as compared with most of the other diet groups (Ps < .05). Vegan, vegetarian, and pescovegetarian participants all saw significant improvements in the DII score as compared with semivegetarian participants at 2 months (Ps < .05) with no differences at 6 months. Given the greater impact on macronutrients and the DII during the short term, finding ways to provide support for adoption and maintenance of plant-based dietary approaches, such as vegan and vegetarian diets, should be given consideration. © 2015 Elsevier Inc. All rights reserved.

Abbreviations: BMI, body mass index; CI, confidence interval; CRP, C-reactive protein; DII, Dietary Inflammatory Index; DRI, Dietary Reference Intake; Omni, omnivorous; OR, odds ratio; Pesco-veg, pescovegetarian; Semi-veg, semivegetarian; Veg, vegetarian. ☆ Trial registration: ClinicalTrials.gov Identifier: NCT01742572. ⁎ Corresponding author. Tel.: +1 803 777 3932; fax: +1 803 777 6290. E-mail address: [email protected] (G.M. Turner-McGrievy). http://dx.doi.org/10.1016/j.nutres.2014.11.007 0271-5317/© 2015 Elsevier Inc. All rights reserved.

98

1.

N U TR IT ION RE S EA RCH 3 5 ( 2 01 5 ) 9 7 –1 06

Introduction

Appropriately planned vegan and vegetarian diets are considered to be nutritionally adequate and may help prevent and treat certain chronic diseases [1]. The nutritional profile of plant-based diets, such as vegan (exclude all animal products), vegetarian (excludes all meat and seafood), pescovegetarian (excludes meat except seafood), and semivegetarian (limits meat) diets, can vary considerably from one another. Several observational studies have examined differences in dietary intake and health-related outcomes by these varying dietary patterns, finding that participants choosing to follow a vegan diet have better diet quality; higher intakes of fiber; and lower intakes of total fat, saturated fat, protein, and calcium as compared with omnivores, with vegetarians, pescovegetarians, and semivegetarians falling in the middle between these 2 ends [2–4]. Vegetarians and vegans have significantly better metabolic risk factors [5], lower body mass indices [6], and lower prevalence of type 2 diabetes [6] as compared with semivegetarians or omnivores. Tied to many of these chronic diseases is the relationship between dietary intake and inflammation, a risk marker of cancer [7,8] and cardiovascular endpoints [9–11]. For example, studies have found that C-reactive protein (CRP) is lower in vegetarians compared with nonvegetarians [12–14] and that populations that eat very low fat, animal-sparse diets, such as in Japan, have very low CRP values [15,16]. These studies, however, are observational in design and examine the self-selected diets of study participants; and to date, there have been no randomized controlled trials examining the effects of recommending adoption of these varying plant-based diets on dietary intake and inflammatory potential. The goal of this study was to examine the differences in nutrient intake and inflammatory potential of diets among participants randomized to conditions instructing them on how to follow a vegan, vegetarian, pescovegetarian, semivegetarian, or omnivorous control diet. We hypothesized that there would be greater increases in fiber and carbohydrate intake and greater decreases in total fat, saturated fat, protein, and cholesterol intake among participants randomized to a vegan diet as compared with the other 4 diets. Furthermore, we hypothesized that participants randomized to the vegan diet would have lower Dietary Inflammatory Index (DII) scores at both 2 and 6 months as compared with the other 4 diets. Briefly, the DII is a literature-derived, population-based dietary index that was developed to assess the inflammatory potential of an indivi dual’s diet and place it on a continuum from maximal proinflammatory diet to maximal anti-inflammatory diet [17,18]. Therefore, the objective of the present study was to examine differences in dietary intake and DII score at both 2 and 6 months among participants randomized to follow 1 of 5 different dietary approaches.

2.

Methods and materials

The New Dietary Interventions to Enhance the Treatments for weight loss (New DIETs) study was a 2-month weight-loss intervention with a 4-month follow-up period conducted at a

large university in the southeast. The methods have been described elsewhere [19,20]. Briefly, overweight and obese (body mass index [BMI] 25-49.9 kg/m2) adults between the ages of 18 and 65 years were recruited for a 6-month weight loss intervention. Before randomization, participants were instructed on how to complete baseline questionnaires (all completed online), which included demographics and dietary intake from 2 days of unannounced 24-hour dietary recalls (1 weekday and 1 weekend day) collected using the Automated Self-Administered 24-Hour Dietary Recall [21]. Analyses included only intake from foods and did not include supplements (eg, multivitamins or mineral supplements). Nutrition adequacy was assessed by the percentage of participants within each group that met US Dietary Reference Intakes (DRI) or Dietary Guidelines for sodium [22] at each time point. The reference group for comparison with DRI consisted of women 48 years old, reflecting the average age and sex of participants in the study. The DRI do not specify a level of saturated fat or cholesterol but state that intake should be “as low as possible while consuming a nutritionally adequate diet” [23]. For saturated fat and cholesterol, a level of less than or equal to 7% of energy from saturated fat and a level of less than or equal to 300 mg/d of cholesterol was used based on American Heart Association recommendations [24]. All measures were assessed at baseline, 2 months, and 6 months. A university institutional review board approved the study, and all participants provided written informed consent. Participants received a $20 incentive payment for completion of all 2-month assessment activities but did not receive any incentives for completion of baseline or 6-month assessments.

2.1.

Dietary Inflammatory Index

Various micro- and macronutrients, as well as several individual food items (known as food parameters), were used to calculate the DII. These food parameters used in the present study included energy; carbohydrates; protein; total fat; unsaturated, monounsaturated, and polyunsaturated fat; omega 3 and omega 6 fatty acids; grams of alcohol consumption; fiber; cholesterol; vitamins B-1, B-2, B-6, B-12, A, C, D, and E; iron; magnesium; zinc; selenium; folate; β-carotene; and caffeine. The development and validation of the DII have previously been described [17,18]. In short, the food parameters were assigned scores based on research summarizing findings from 1943 articles describing the relationship between the food parameters and inflammation. The DII calculation is linked to a regionally representative world database (ie, food consumption from 11 populations around the world) that provided a mean and standard deviation for each food parameter. The “standard mean” was subtracted from the actual food parameter value and divided by its standard deviation. This z score was then converted to a percentile (to minimize the effect of outliers or right skewing) and centered by doubling the value and subtracting 1. The centered percentile score for each food parameter for each individual was then multiplied by the respective food parameter effect score, which is derived from the literature review, to obtain a food parameter–specific DII score for an individual. All of the food parameter–specific DII scores are then summed to create the overall DII score for every

N U TR IT ION RE S EA RCH 3 5 ( 2 01 5 ) 9 7 –1 06

99

Table 1 – Description of the 5 intervention diets and example dinner meals for each diet Dietary group

Definitions of diet patterns

Example dinner meal

Vegan

Does not contain any animal products (meat, fish, poultry, eggs, or dairy) but emphasizes plant-based foods, such as fruits, vegetables, whole grains, and legumes/beans. Does not contain meat, fish, or poultry but does contain eggs and dairy, in addition to plant-based foods, such as fruits, vegetables, whole grains, and legumes/beans.

• Red beans and brown rice with chopped tomatoes and roasted peppers. • Fruit salad

Veg

Pesco-veg

Semi-veg

Omni

Does not contain meat or poultry but does contain fish and shellfish, eggs, and dairy, in addition to plant-based foods, such as fruits, vegetables, whole grains, and legumes/beans. Contains all foods, including meat, poultry, fish and shellfish, eggs, and dairy, in addition to plant-based foods, such as fruits, vegetables, whole grains, and legumes/beans. However, red meat is limited to 1 time per week; and poultry is limited to 5 times per week or less. Contains all food groups.

participant in the study. Dietary Inflammatory Index values have not been found to exceed the limit of −10 to +10, with more negative scores indicating more anti-inflammatory diets and more positive scores representing more proinflammatory diets. The DII has been shown to be associated with CRP greater than 3 mg/L in a study conducted in Massachusetts, USA (odds ratio [OR] =1.08; 95% confidence interval [CI], 1.01-1.16) [25] and among police officers in Buffalo, NY (OR for DII in quartile 3 vs 1 = OR = 2.17; 95% CI, 1.19-3.95) [26]. In addition, plasma interleukin-6 concentrations were positively associated with DII score (β = 0.13; 95% CI, 0.05-0.21; P = .002) in a study conducted in Australia [27]; and a study in Belgium revealed significant positive associations between DII and interleukin-6 greater than 1.6 pg/mL (OR = 1.19; CI, 1.04-1.36) and homocysteine greater than 15 μmol/L (OR = 1.56; CI, 1.251.94) [28]. Apart from inflammatory markers, the DII also has been shown to be associated with the glucose intolerance component of metabolic syndrome, increased odds of asthma, shift work, colorectal cancer among women from the Iowa Women’s Health Study, and prostate cancer among Italian men [26,27,29–31].

2.2.

Intervention diets

Table 1 provides an overview of the intervention diets used in the New DIETs study (vegan, vegetarian [veg], pescovegetarian [pesco-veg], semivegetarian [semi-veg], or omnivorous [omni]), as well as sample dinner menus. Participants in all diet groups were instructed to follow diets that favored low-glycemic index [32] and low-fat foods. There was no recommended restriction on energy intake for any of the groups. Participants also were provided with recipes for their diet assignment and a handout with information on the diet. All plant-based diet groups (vegan, veg, pesco-veg, and semi-veg) attended weekly 1-hour meetings for 2 months (intensive phase), followed by monthly

• Red beans and brown rice with chopped tomatoes, roasted peppers, and reduced-fat cheddar cheese. • Fruit salad • Red beans and brown rice with shrimp, chopped tomatoes, roasted peppers, and reduced-fat cheddar cheese. • Fruit salad • Red beans and brown rice with low-fat turkey sausage, chopped tomatoes, roasted peppers, and reduced-fat cheddar cheese. • Fruit salad • Red beans and brown rice with low-fat sausage, chopped tomatoes, roasted peppers, and reduced-fat cheddar cheese. • Fruit salad

meetings through month 6 (maintenance phase). In addition to the plant-based diet groups, there also was a control group that was instructed to follow an omni diet. The omni group, which received a less intensive dietary intervention (weekly e-mail lessons during the first 2 months; monthly meetings over all 6 months), allowed for the examination of consuming a usual diet (as all participants were following an omni diet at baseline) while at the same time controlling for the selection made by all participants to participate in a study that involved losing weight. Two registered dietitians with graduate degrees and expertise in all the study diets led the classes. All participants were required to take a multivitamin or other form of vitamin B-12 each day (because vegan diets require supplementary B12) [1]. Topic sessions for all the group meetings were informed by the Diabetes Prevention Program [33] and were grounded in Social Cognitive Theory [34]. Each class included food samples or a cooking demonstration. Each group received instruction about how to plan their assigned diets to meet nutrient needs. All group sessions covered identical topics among the 5 groups (eg, grocery shopping, meal planning tips), with the only difference being the type of diet discussed. Participants met with only their assigned diet group, which corresponded to a unique day of the week. Meals were not provided to participants, and they were required to prepare their own food or purchase meals from restaurants.

2.3.

Statistical analyses

All results are reported as means ± SEM. For differences in baseline demographic characteristics, analysis of variance was used with the Tukey test for post hoc analyses of continuous variables; and χ2 test of independence was used for categorical variables. Differences in nutrient intake and the DII were examined using general linear models and Tukey

100

N U TR IT ION RE S EA RCH 3 5 ( 2 01 5 ) 9 7 –1 06

Figure – CONSORT diagram showing the flow of participants through each stage the New DIETs 6-month randomized trial.

test for post hoc analyses at baseline, 2 months (adjusted for baseline), and 6 months (adjusted for baseline). Tukey test was used because, although a priori hypotheses were stated, we were also interested in other pairwise comparisons. Nutrient intake at 2 and 6 months was adjusted for the baseline value of that nutrient. For nutrients not expressed as percentage energy, energy intake (kilocalories) at the examined time point was also included in the model to control for the effect of total energy [35]. Missing dietary data were analyzed by bringing baseline observations forward, assuming no change in dietary intake. This method was used as a conservative method to analyze all participants [36], regardless of study completion. All analyses were conducted using SPSS 22.0 for Windows software (2013; SPSS Inc, Chicago, IL, USA), with a P value of .05 used to indicate statistically significant differences.

3.

Results

Participants were screened in February 2013, and the trial was completed by August 2013. The Figure provides a CONSORT flow diagram of study participation. Of 219 participants who were screened, 63 (29%) met study criteria and were randomly assigned to 1 of the 4 plant-based diets or the omni group. At the 2-month assessment time point, 57 (90%) completed the body weight assessment and questionnaires; and 56 (89%) completed 2 days of dietary recalls. At the 6-month time point, 50 (79%) completed the study (ie, provided a body weight measurement at 6 months); 46 completed the questionnaires (73%); and 49 (78%) completed 2 days of dietary recalls. There were no differences in baseline demographic characteristics, BMI, or energy intake among the 5 groups with the exception of age (Table 2).

101

N U TR IT ION RE S EA RCH 3 5 ( 2 01 5 ) 9 7 –1 06

Table 2 – Baseline demographics and BMI of study participants in New DIETs weight loss studies

na Age, y b Sex Female Male Race Black White Other Education High school or some college College graduate Advanced degree Marital status Married Other BMI (kg/m2) b a b c

Vegan

Veg

Pesco-veg

Semi-veg

Omni

12 48.2 ± 2.0

13 53.0 ± 1.1

13 48.8 ± 2.2

13 42.7 ± 2.7 c

12 51.0 ± 2.5

8 (67%) 4

10 (77%) 3

9 (69%) 4

10 (77%) 3

9 (75%) 3

3 (25%) 9 (75%) 0 (0%)

3 (23%) 9 (69%) 1 (8%)

3 (23%) 10 (77%) 0 (0%)

2 (15%) 11 (85%) 0 (0%)

1 (8%) 11 (92%) 0 (0%)

0 (0%) 8 (67%) 4 (33%)

0 (0%) 6 (46%) 7 (54%)

1 (8%) 6 (46%) 6 (46%)

0 (0%) 8 (62%) 5 (38%)

3 (25%) 5 (42%) 4 (33%)

9 (75%) 3 (25%) 32.5 ± 1.8

7 (54%) 6 (46%) 35.1 ± 1.4

8 (61%) 5 (39%) 35.8 ± 1.4

5 (39%) 8 (61%) 35.1 ± 1.5

10 (83%) 2 (17%) 36.3 ± 1.6

P value for difference among groups P = .02 P = .97

P = .69

P = .20

P = .16

P = .49

Results in table are presented as number (percentage of participants) unless otherwise indicated. Results are presented as means ± SEM. Significantly different from the veg group (P = .01).

3.1.

Dietary intake

3.3.

Micronutrients

Dietary adherence, physical activity, and weight loss by group at 6 months have been reported elsewhere [20]. Briefly, intentional physical activity and adherence did not differ by group (percentage adherent by group: vegan 33%, veg 39%, pesco-veg 39%, semi-veg 46%, and omni 42%). At 6 months, weight loss in the vegan group (−7.5% ± 4.5%) was significantly different from the omni (−3.1% ± 3.6%, P = .03), semi-veg (−3.2% ± 3.8%, P = .03), and pesco-veg (−3.2% ± 3.4%, P = .03) groups (6-month weight loss in the veg group was −6.3% ± 6.6%). Table 3 shows changes in dietary intake and DII score at each time point. There were no differences in baseline DII score or intakes of nutrients with the exception of potassium and percentage energy from fat. There were no significant differences in reported energy intake among groups at any time point; however, there were several differences in consumption of macro- and micronutrients.

Sodium, calcium, and iron intake did not differ by group at any time point. Omnivorous participants had significantly higher zinc intake than the other 4 groups at both 2 and 6 months (Ps all < .05). At 2 months, vegan participants consumed more potassium than semi-veg or omni participants (Ps all < .05); but this was no longer significant at 6 months. There were also no differences in vitamin B-12 intake at any time point with the exception of pesco-veg participants having significantly higher B-12 intake as compared with vegan and semi-veg participants (P < .05). Folate intake differed at 2 months, with vegan and veg participants consuming more than semi-veg participants (P < .05). At 2 months, vegans consumed significantly less vitamin D than pesco-veg participants (P < .01).

3.2.

Table 3 also shows results for the percentage of participants in each diet group meeting current dietary recommendations for macro- and micronutrients. At baseline, very few participants were meeting macronutrient guideline recommendations, with the exception of protein; and fewer than half of participants within each group were meeting fiber, calcium, sodium, potassium, and vitamin D recommendations. At 2 and 6 months, there were few nutrients that had the majority (at least 75%) of individuals meeting specific dietary recommendations, regardless of diet group. Nutrients that had at least 75% of participants meeting dietary recommendations included saturated fat intake within the vegan group at 2 months; cholesterol intake among vegan, pesco-veg, and semi-veg at 2 month, and vegan, veg, semi-veg, and omni at

Macronutrients, fiber, and cholesterol

Compared with most of the other 4 diet groups, the vegan group saw significantly lower percentage energy from fat (2 and 6 months), percentage energy from saturated fat (2 and 6 months), cholesterol (2 and 6 months), and percentage energy from protein (2 months only) and significantly greater intakes of dietary fiber and percentage energy from carbohydrates (both at 2 months only). Intake among the veg group mirrored the direction of some of the macronutrient changes in the vegan including lower percentage energy from fat (6 months), percentage energy from protein (2 and 6 months), and cholesterol (6 months) and higher fiber intake at 2 months as compared with many of the other diets.

3.4.

Nutrient adequacy

102

N U TR IT ION RE S EA RCH 3 5 ( 2 01 5 ) 9 7 –1 06

Table 3 – Mean macronutrient and select micronutrient intake among 5 different dietary intervention groups at baseline, 2 months, and 6 months

na Energy (kcal/d) Baseline 2 mo 6 mo Fat (% energy) b Baseline 2 mo 6 mo Saturated fat (% energy) b, h Baseline 2 mo 6 mo Cholesterol (mg) c, h Baseline 2 mo 6 mo Protein (% energy) b Baseline 2 mo 6 mo Carbohydrate (% energy) b Baseline 2 mo 6 mo Fiber (g) c Baseline 2 mo 6 mo Calcium (mg) c Baseline 2 mo 6 mo Iron (mg) c Baseline 2 mo 6 mo Zinc (mg) c Baseline 2 mo 6 mo Sodium (mg) c Baseline 2 mo 6 mo Potassium (mg) c Baseline 2 mo 6 mo Vitamin B-12 (μg) c Baseline 2 mo 6 mo

Vegan

Veg

Pesco-veg

Semi-veg

Omni

12

13

13

13

12

P value for difference among groups

2460 ± 239 1563 ± 186 1484 ± 179

2070 ± 230 2001 ± 178 1891 ± 171

2028 ± 230 1711 ± 178 1757 ± 172

2321 ± 230 1792 ± 178 1882 ± 171

2125 ± 239 1705 ± 181 1956 ± 168

40.2 ± 1.6 e (0%) 27.9 ± 2.4 d.g (50%) 29.0 ± 2.4 e,f,g (42%)

39.0 ± 1.5 e (8%) 34.7 ± 2.2 (23%) 29.7 ± 2.4 f,g (38%)

33.2 ± 1.5 (31%) 31.7 ± 2.4 g (54%) 33.3 ± 2.4 (23%)

36.8 ± 1.5(8%) 33.5 ± 2.2 (38%) 37.2 ± 2.4 (23%)

38.1 ± 1.7 e (17%) P = .045 39.2 ± 2.4 (17%) P = .03 37.0 ± 2.0 (25%) P = .02

13.9 ± 0.8 (0%) 4.8 ± 1.3 d,e,f,g (92%) 7.8 ± 1.0 e,f,g (42%)

12.4 ± 0.7 (8%) 10.8 ± 1.2 (8%) 9.9 ± 0.9 (23%)

10.8 ± 0.7 (0%) 9.9 ± 1.3 (31%) 10.2 ± 1.0 (23%)

12.4 ± 0.7 (0%) 11.9 ± 1.2 (15%) 11.1 ± 0.9 (8%)

12.3 ± 0.8 (0%) 11.9 ± 1.2 (17%) 11.7 ± 0.9 (8%)

291.3 ± 43.9 (50%) 22.0 ± 46.6 d,e,f,g (100%) 94.8 ± 45.3 e,f,g (100%)

365.5 ± 41.9 (62%) 185.6 ± 44.9 (69%) 149.9 ± 43.6 f (92%)

322.8 ± 42.0 (54%) 232.5 ± 44.4 (85%) 237.0 ± 43.2 (54%)

291.1 ± 41.8 (77%) 251.6 ± 44.4 (85%) 307.3 ± 43.2 (77%)

297.0 ± 43.8 (58%) P = .73 261.3 ± 40.5 (50%) P = .001 233.8 ± 35.9 (75%) P = .01

16.1 ± 1.0 (100%) 13.8 ± 1.0 e,f,g (100%) 15.0 ± 1.1 (100%)

17.0 ± 1.0 (100%) 13.2 ± 1.0 e,f,g (100%) 14.0 ± 1.1 e (92%)

16.8 ± 1.0 (100%) 18.5 ± 1.0 (100%) 16.3 ± 1.1 (100%)

16.2 ± 1.0 (100%) 17.7 ± 1.0 (92%) 17.4 ± 1.1 (100%)

16.8 ± 1.0 (100%) P = .96 19.2 ± 1.2 (100%) P < .001 18.3 ± 1.2 (100%) P = .07

41.9 ± 2.6 (42%) 56.8 ± 2.7 d,e,f,g (83%) 54.6 ± 2.9 f,g (67%)

43.8 ± 2.5 (46%) 48.9 ± 2.6 (69%) 50.3 ± 2.8 (69%)

45.2 ± 2.5 (54%) 47.3 ± 2.6 (69%) 48.2 ± 2.8 (85%)

42.8 ± 2.5 (31%) 46.1 ± 2.6 (62%) 43.5 ± 2.8 (38%)

46.1 ± 2.6 (67%) 41.5 ± 2.8 (42%) 43.5 ± 2.8 (8%)

P = .80 P < .01 P < .001

18.6 ± 1.8 (17%) 31.0 ± 3.0 d,e,f,g (67%) 22.0 ± 2.6 (42%)

18.2 ± 1.7 (23%) 26.7 ± 2.9 f,g (46%) 20.2 ± 2.5 (31%)

17.9 ± 1.7 (8%) 21.0 ± 2.9 (23%) 19.0 ± 2.5 (23%)

15.8 ± 1.7 (15%) 17.6 ± 2.9 (23%) 16.3 ± 2.5 (15%)

22.9 ± 2.0 (25%) 17.6 ± 2.5 (17%) 19.3 ± 2.6 (17%)

P = .15 P < .001 P = .44

P = .09 P < .001 P = .03

1044.6 ± 80.9 (50%) 653.6 ± 134.9 (8%) 587.8 ± 109.0 (8%)

965.6 ± 77.3 (38%) 1147.8 ± 128.9 (46%) 987.1 ± 104.2 (50%)

15.6 ± 1.5 (17%) 14.6 ± 1.7 (42%) 11.8 ± 2.3 (17%)

15.6 ± 1.5 (31%) 17.2 ± 1.6 (31%) 16.5 ± 2.2 (15%)

18.5 ± 1.5 (38%) 12.3 ± 1.7 (8%) 13.1 ± 2.2 (23%)

15.7 ± 1.5 (31%) 12.7 ± 1.6 (31%) 15.9 ± 2.2 (38%)

15.7 ± 1.5 (33%) 16.8 ± 1.4 (50%) 15.4 ± 1.9 (42%)

11.8 ± 0.6 (58%) 8.9 ± 1.0 g (58%) 8.3 ± 1.0 g (58%)

12.4 ± 0.5 (85%) 10.7 ± 1.0 g (85%) 9.7 ± 1.0 g (85%)

12.1 ± 0.5 (85%) 8.7 ± 1.0 g (62%) 9.5 ± 1.0 g (62%)

10.8 ± 0.5 (77%) 8.4 ± 1.0 g (69%) 10.2 ± 1.0 g (69%)

12.8 ± 0.7 (92%) P = .29 13.2 ± 0.8 (83%) P = .001 12.7 ± 0.9 (100%) P = .03

4246.5 ± 227.4 (8%) 3488.8 ± 392.5 (17%) 2725.3 ± 366.3 (25%)

4126.4 ± 217.1 (15%) 3512.4 ± 373.9 (23%) 3564.0 ± 348.9 (15%)

4350.7 ± 217.7 (15%) 3838.5 ± 216.7 (15%) 3593.0 ± 378.7 (15%) 3364.6 ± 380.9 (31%) 3433.1 ± 353.4 (15%) 3302.8 ± 355.5 (23%)

2582.3 ± 160.1 (8%) 2903.9 ± 245.3 f,g (0%) 2357.2 ± 235.1 (0%)

3072.2 ± 152.9 (8%) 2807.8 ± 247.2 (69%) 2523.2 ± 236.9 (8%)

2749.1 ± 153.3 (8%) 2799.8 ± 233.7 f (0%) 2636.7 ± 223.9 (8%)

4.0 ± 0.8 (92%) 2.5 ± 0.9 e (25%) 2.7 ± 0.7 (58%)

5.5 ± 0.7 (92%) 4.2 ± 0.9 (69%) 3.5 ± 0.7 (85%)

974.6 ± 77.5 (46%) 1042.9 ± 77.1 (46%) 873.8 ± 129.1 (23%) 826.0 ± 128.6 (23%) 912.7 ± 104.3 (15%) 909.6 ± 104.0 (23%)

P = .65 P = .50 P = .28

5.2 ± 0.7 (85%) 5.8 ± 0.9 (100%) 4.4 ± 0.6 (77%)

821.5 ± 83.8 (17%) P = .33 684.9 ± 111.6 (17%) P = .08 819.5 ± 96.5 (23%) P = .14 P = .59 P = .10 P = .74

3720.9 ± 219.9 (8%) P = .24 3120.4 ± 234.6 (8%) P = .17 3173.7 ± 207.9 (8%) P = .58

2396.8 ± 152.6 d,g (8%) 3033.6 ± 174.2 (0%) P = .03 2559.0 ± 242.4 (8%) 2459.0 ± 169.8 (0%) P = .049 2813.9 ± 232.3 (8%) 2579.1 ± 205.6 (0%) P = .82

6.0 ± 0.7 (77%) 3.2 ± 0.9 e (62%) 3.8 ± 0.7 (77%)

6.6 ± 0.9 (83%) 4.2 ± 0.9 (75%) 5.7 ± 0.9 (83%)

P = .29 P = .07 P = .07

103

N U TR IT ION RE S EA RCH 3 5 ( 2 01 5 ) 9 7 –1 06

Table 3 (continued) Vegan

Folate (μg) c Baseline 2 mo 6 mo Vitamin D (μg) c Baseline 2 mo 6 mo DII b, i Baseline 2 mo 6 mo

Veg

Pesco-veg

Semi-veg

383.1 ± 37.6 (33%) 500.1 ± 51.3 f,g (75%) 395.4 ± 42.8 (50%)

410.7 ± 36.0 (54%) 533.5 ± 48.7 f (69%) 403.2 ± 44.8 (62%)

431.7 ± 36.0 (54%) 402.5 ± 49.1 (38%) 382.9 ± 45.1 (46%)

406.8 ± 35.9 (38%) 314.3 ± 48.7 (15%) 399.0 ± 44.8 (46%)

3.2 ± 1.0 (0%) 1.3 ± 1.2 e (0%) 2.6 ± 1.1 (0%)

5.1 ± 0.9 (0%) 3.5 ± 1.2 (0%) 3.4 ± 1.0 (0%)

5.3 ± 0.9 (8%) 5.8 ± 1.2 (8%) 5.2 ± 1.0 (8%)

4.1 ± 0.9 (0%) 3.4 ± 1.1 (0%) 3.8 ± 1.0 (0%)

0.3 ± 0.6 −1.2 ± 0.5 f 0.1 ± 0.6

0.4 ± 0.6 −1.0 ± 0.5 f −0.2 ± 0.7

0.9 ± 0.6 −0.7 ± 0.5 f −0.2 ± 0.6

0.9 ± 0.6 1.3 ± 0.6 0.2 ± 0.7

Omni

P value for difference among groups

444.0 ± 42.2 (50%) P = .88 392.3 ± 42.0 (50%) P < .01 385.9 ± 42.4 (42%) P = .99

6.5 ± 1.3 (8%) 2.9 ± 1.1 (0%) 2.9 ± 1.0 (8%) −0.1 ± 0.6 0.2 ± 0.7 −0.5 ± 0.8

P = .46 P = .049 P = .17 P = .75 P = .04 P = .97

a

All results are means ± SEM. Percentage in parentheses represents the percentage of participants within each group meeting US DRI or Dietary Guidelines for the nutrient. b Adjusted for baseline value of examined variable. c Adjusted for baseline value of examined dietary variable and energy intake (kilocalories) at the examined time point. d Significantly different from the veg group (P < .05). e Significantly different from the pesco-veg group (P < .05). f Significantly different from the semi-veg group (P < .05). g Significantly different from the omni group (P < .05). h The DRI do not specify a level of saturated fat or cholesterol but state that intake should be “as low as possible while consuming a nutritionally adequate diet” [31]. Therefore, a level of less than or equal to 7% of energy from saturated fat and a level of less than or equal to 300 mg/d of cholesterol was used based on American Heart Association recommendations [33]. i More negative scores represent more anti-inflammatory diets, and more positive scores represent more proinflammatory diets.

6 months; protein intake among all groups and all time points; carbohydrate intake for vegans at 2 months and pesco-veg at 6 months; zinc for veg and omni at both 2 and 6 months; vitamin B-12 for pesco-veg and omni at 2 months and for veg, pesco-veg, semi-veg, and omni at 6 months; and folate for vegans at 2 months.

3.5.

Dietary Inflammatory Index

The DII was calculated at each time point (Table 3). Although we hypothesized that vegan participants would have a greater improvement in DII (ie, a lowering) scores as compared with other groups, at 2 months, vegan as well as veg and pesco-veg participants had a significantly lower DII score as compared with semi-veg participants (Ps all < .05). At 6 months, the groups were no longer different from one another (P = .95).

4.

Discussion

This study examined dietary changes among overweight adults randomly assigned to receive instruction on 1 of 4 plant-based diets (vegan, veg, pesco-veg, or semi-veg) or a control diet (omni). At baseline, participants’ diets reflected what is commonly observed among US adult populations, including a diet that is high in total fat [37,38], saturated fat [39], cholesterol [38], and sodium [38,40] and low in carbohydrates [37], fiber [41], potassium [40], and iron [42]. As hypothesized, changes in examined macronutrients were greatest among vegan participants as compared with the other diet groups, particularly at 2 months. Differences in

nutrient intake were less pronounced at 6 months, potentially because of low dietary adherence among all groups. Lower DII scores were observed not only among vegan participants but also veg and semi-veg participants as compared with semiveg; however, contrary to our hypothesis, this was only observed at 2 months. Despite the fact that all groups were counseled to choose low-fat foods, mean percentage energy from fat and saturated fat exceeded recommendations with the exception of the vegan group. Vegan participants may have been the most successful with reducing dietary fat because of elimination of 4 (beef, cheese, milk, and poultry) out of the 10 top sources of fat intake in the US diet [43] and limiting cheese, milk, frozen dairy desserts, and butter, which are the top 1, 3, 5, and 7, respectively, sources of saturated fat in the US diet [43]. In contrast to total and saturated fat intakes in this study, which were reduced to a greater degree in vegan participants, carbohydrate and fiber intake increased to a greater degree in vegan participants as compared with the other groups at 2 months. Participants randomized to the vegan diet were instructed to rely on legumes, tofu and tempeh, and meat analogues (which contain fiber) for protein. Although dried beans and lentils are the second most prevalent source of fiber in the US diet, they appear infrequently in the diets of US adults [43]. High fiber intake has been associated with lower risk of cardiovascular disease, obesity, some forms of cancer, and type 2 diabetes [44]; therefore, identifying dietary patterns that can lead to significant increases in fiber is important. Examining minerals, the vegan group was the only group to have mean calcium levels that would not have met DRI for

104

N U TR IT ION RE S EA RCH 3 5 ( 2 01 5 ) 9 7 –1 06

males or females. Calcium intake is often lower in vegans [45], so future studies using a vegan dietary approach for weight loss among adults should emphasize the inclusion of plantbased sources of calcium, such as leafy greens and fortified nondairy milks, or calcium supplements. Iron intake was low among all groups. Both observational [2,3,46] and intervention studies [47–49] have shown higher or similar iron intakes between vegetarians and vegans as compared with omnivores, with the difference being that veg and vegan diets only contain nonheme iron. With the exception of an increase in potassium at 2 months among the vegan group, recommendations to follow any of the diets did not differentially impact potassium intake. Because currently only 2% of US adults meet potassium requirements [40], dietary strategies that can lead to improvements in potassium warrant further consideration. There were few differences in examined vitamin intakes among the groups. Whereas approximately half of participants were meeting vitamin B-12 and folate requirements at 6 months, very few participants were meeting vitamin D requirements at any time point. Vegan diets are typically low in vitamin D [45], and pesco-veg participants may have had higher vitamin D intakes at 2 months because of the recommendation to include fish in the diet [50]. However, all groups would benefit from the inclusion of vitamin D–rich foods or supplements. Inflammation in the body may play a role in cardiovascular disease and cancer [51]. Because dietary intake can have a significant impact on inflammatory markers [7,8] and observational research has shown that vegetarians have lower levels of CRP than nonvegetarians [12,13], adoption of plantbased dietary approaches may be a potential strategy for reducing inflammation. In addition, studies have demonstrated that vegan diets can have beneficial effects on inflammatory diseases, such as rheumatoid arthritis [52,53]. Nutrients that had greater changes in the vegan group, such as lowered saturated fat [54] and increased fiber [55], have also been associated with improvements in inflammation. One hypothesis that has linked dietary patterns among vegans with lower inflammation is differences in gut microbiota between vegans and omnivores. Studies have demonstrated more protective bacterial species in the guts of vegans, which in turn lead to lower inflammation [56]. This study provides some of the first evidence to the potential antiinflammatory properties of varying plant-based diets. The vegan, veg, and pesco-veg diets all yielded more improved DII (ie, lower) scores than the semi-veg diet at 2 months. Future studies should measure objective changes in biological risk factors for cancer (markers of inflammation, growth signal dysregulation, and vascular integrity related factors) as part of randomized controlled trials examining differing plant-based dietary approaches. The present paper has several strengths, including the use of a randomized design and the use of high-quality measures. These measures included use of a validated dietary tool to assess diets that may promote or decrease inflammation and dietary data collected by 2 unannounced, 24-hour recalls, which are considered to be an accurate way to measure overall dietary intake [57–59]. In addition to the modest contact of weekly or monthly group meetings, other aspects

of the study also make the findings applicable outside the research setting, including that participants prepared all their own foods or found meals to eat at restaurants. The study also had a low attrition rate of 21%, particularly considering that no incentive was provided at 6 months. Furthermore, the present study was conducted in the southern United States where obesity rates are high and where there may be greater challenges to adoption of more plant-based eating styles than other regions in light of traditional southern food preferences [60]. There are also limitations, including the short duration and a sample that was mostly white and educated. In addition, neither participants nor study personnel were blinded to diet assignment. Only intake from foods and beverages, but not dietary supplements (eg, multivitamins), was included in the analysis; so it is possible that participants met their nutrient requirements through supplemental means. Participants randomized to follow a vegan diet had greater changes in macronutrient profile during the study with greater decreases in total fat and saturated fat and greater increases in fiber than participants following diets that included animal products. Nutrients of concern for vegan diets, such as protein, vitamin B-12, iron, and zinc, were not significantly different than the other diet groups. All groups consumed diets low in calcium, iron, vitamin D, and potassium and high in sodium. Given the greater impact on macronutrients and the DII during the short term, finding ways to provide support for adoption and maintenance of plant-based dietary approaches, such as vegan or veg diets, should be given consideration as dietary strategies for the prevention and treatment of diseases associated with inflammation, such as type 2 diabetes, cardiovascular disease, and cancer [61,62].

Acknowledgment Funding was provided by internal startup funds of the principal investigator. The authors would like to thank the University of South Carolina’s Office of Public Health Practice for assistance with survey design. GTM, EEW, RF, SW, and EAF have no conflicts of interest to declare. JRH owns controlling interest in Connecting Health Innovations, LLC, a company planning to license the right to his invention of the DII from the University of South Carolina to develop computer and smart phone applications for patient counseling and dietary intervention in clinical settings. MDW and NS are employees of Connecting Health Innovations, LLC. The subject matter of this paper will not have any direct bearing on that work, nor has that activity exerted any influence on this project.

REFERENCES

[1] Craig WJ, Mangels AR. Position of the American Dietetic Association: vegetarian diets. J Am Diet Assoc 2009;109: 1266–82. [2] Clarys P, Deliens T, Huybrechts I, Deriemaeker P, Vanaelst B, De Keyzer W, et al. Comparison of nutritional quality of the vegan, vegetarian, semi-vegetarian, pesco-vegetarian and omnivorous diet. Nutrients 2014;6:1318–32.

N U TR IT ION RE S EA RCH 3 5 ( 2 01 5 ) 9 7 –1 06

[3] Rizzo NS, Jaceldo-Siegl K, Sabate J, Fraser GE. Nutrient profiles of vegetarian and nonvegetarian dietary patterns. J Acad Nutr Diet 2013;113:1610–9. [4] Davey GK, Spencer EA, Appleby PN, Allen NE, Knox KH, Key TJ. EPIC-Oxford: lifestyle characteristics and nutrient intakes in a cohort of 33 883 meat-eaters and 31 546 non meat-eaters in the UK. Public Health Nutr 2003;6:259–69. [5] Rizzo NS, Sabaté J, Jaceldo-Siegl K, Fraser GE. Vegetarian dietary patterns are associated with a lower risk of metabolic syndrome. Diabetes Care 2011;34:1225–7. [6] Tonstad S, Stewart K, Oda K, Batech M, Herring RP, Fraser GE. Vegetarian diets and incidence of diabetes in the Adventist Health Study–2. Nutr Metab Cardiovasc Dis 2011:1–8. [7] Esmaillzadeh A, Kimiagar M, Mehrabi Y, Azadbakht L, Hu FB, Willett WC. Dietary patterns and markers of systemic inflammation among Iranian women. J Nutr 2007;137:992–8. [8] Galland L. Diet and inflammation. Nutr Clin Pract 2010;25: 634–40. [9] Albert MA, Ridker PM. The role of C-reactive protein in cardiovascular disease risk. Curr Cardiol Rep 1999;1:99–104. [10] Sarwar N, Thompson AJ, Di Angelantonio E. Markers of inflammation and risk of coronary heart disease. Dis Markers 2009;26:217–25. [11] Das A. How does race get "under the skin"? Inflammation, weathering, and metabolic problems in late life. Soc Sci Med 2013;77:75–83. [12] Paalani M, Lee JW, Haddad E, Tonstad S. Determinants of inflammatory markers in a bi-ethnic population. Ethn Dis 2011;21:142–9. [13] Chen CW, Lin YL, Lin TK, Lin CT, Chen BC, Lin CL. Total cardiovascular risk profile of Taiwanese vegetarians. Eur J Clin Nutr 2007;62:138–44. [14] Cundiff DK, Agutter PS, Malone PC, Pezzullo JC. Diet as prophylaxis and treatment for venous thromboembolism? Theor Biol Med Model 2010;7:1–24 [31]. [15] Oda E, Kawai R. Tentative cut point of high-sensitivity C-reactive protein for a component of metabolic syndrome in Japanese. Circ J 2009;73:755–9. [16] Kubota Y, Moriyama Y, Yamagishi K, Tanigawa T, Noda H, Yokota K, et al. Serum vitamin C concentration and hs-CRP level in middle-aged Japanese men and women. Atherosclerosis 2010;208:496–500. [17] Shivappa N, Steck SE, Hurley TG, Hussey JR, Hébert JR. Designing and developing a literature-derived, populationbased dietary inflammatory index. Public Health Nutr 2014; 17:1689–96. [18] Shivappa N, Steck SE, Hurley TG, Hussey JR, Ma Y, Ockene IS, et al. A population-based dietary inflammatory index predicts levels of C-reactive protein in the Seasonal Variation of Blood Cholesterol Study (SEASONS). Public Health Nutr 2014; 17:1825–33. [19] Turner-McGrievy GM, Davidson CR, Wilcox S. Does the type of weight loss diet affect who participates in a behavioral weight loss intervention? A comparison of participants for a plant-based diet versus a standard diet trial. Appetite 2014; 73:156–62. [20] Turner-McGrievy G, Davidson C, Wingard E, Wilcox S, Frongillo E. Comparative effectiveness of plant-based diets for weight loss: a randomized controlled trial of five different diets. Nutrition 2014 [in press, http://www.sciencedirect. com/science/article/pii/S0899900714004237]. [21] Subar AF, Crafts J, Zimmerman TP, Wilson M, Mittl B, Islam NG, et al. Assessment of the accuracy of portion size reports using computer-based food photographs aids in the development of an automated self-administered 24-hour recall. J Am Diet Assoc 2010;110:55–64. [22] The 2010 dietary guidelines for Americans. http:// wwwcnppusdagov/Publications/DietaryGuidelines/2010/ PolicyDoc/PolicyDocpdf. [Accessed June 10, 2013].

105

[23] Otten JJ, Hellwig JP, Meyers LD. Dietary reference intakes: the essential guide to nutrient requirements. Washington D.C.: The National Academies Press; 2006 [http://www.nap.edu/ catalog/11537.html]. [24] Lichtenstein AH, Appel LJ, Brands M, Carnethon M, Daniels S, Franch HA, et al. Diet and lifestyle recommendations revision 2006: a scientific statement from the American Heart Association Nutrition Committee. Circulation 2006; 114:82–96. [25] Shivappa N, Steck SE, Hurley TG, Hussey JR, Ma Y, Ockene IS, et al. A population-based dietary inflammatory index predicts levels of C-reactive protein in the Seasonal Variation of Blood Cholesterol Study (SEASONS). Public Health Nutr 2013; 10:1–9. [26] Wirth MD, Burch J, Shivappa N, Violanti JM, Burchfiel CM, Fekedulegn D, et al. Association of a dietary inflammatory index with inflammatory indices and metabolic syndrome among police officers. J Occup Environ Med 2014;56:986–9. [27] Wood LG, Shivappa N, Berthon BS, Gibson PG, Hebert JR. Dietary inflammatory index is related to asthma risk, lung function and systemic inflammation in asthma. Clin Exp Allergy 2014. http://dx.doi.org/10.1111/cea.12323 [n/a-n/a]. [28] Shivappa N, Hebert JR, Rietzschel ER, De Buyzere ML, Langlois M, Debruyne E, et al. Association between dietary inflammatory index and inflammatory markers in Asklepios study. Br J Nutr 2014:1–6 [in press]. [29] Shivappa N, Prizment AE, Blair CK, Jacobs DR, Steck SE, Hebert JR. Dietary Inflammatory Index (DII) and risk of colorectal cancer in Iowa Women's Health Study. Cancer Epidemiol Biomarkers Prev 2014. http://dx.doi.org/10.1158/ 1055-9965.EPI-14-0537 [in press]. [30] Wirth MD, Burch J, Shivappa N, Steck SE, Hurley TG, Vena JE, et al. Dietary inflammatory index scores differ by shift work status: NHANES 2005 to 2010. J Occup Environ Med 2014;56: 145–8. [31] Shivappa N, Bosetti C, Zucchetto A, Montella M, Serraino D, La Vecchia C, et al. Association between dietary inflammatory index and prostate cancer among Italian men. Br J Nutr 2014:1–6. [32] Jenkins DJ, Wolever TM, Taylor RH, Barker H, Fielden H, Baldwin JM, et al. Glycemic index of foods: a physiological basis for carbohydrate exchange. Am J Clin Nutr 1981;34: 362–6. [33] DPP. The Diabetes Prevention Program (DPP). Diabetes Care 2002;25:2165–71. [34] Bandura A. Health promotion by social cognitive means. Health Educ Behav 2004;31:143–64. [35] Willett WC, Howe GR, Kushi LH. Adjustment for total energy intake in epidemiologic studies. Am J Clin Nutr 1997;65: 1220S–8S. [36] Liu-Seifert H, Zhang S, D'Souza D, Skljarevski V. A closer look at the baseline-observation-carried-forward (BOCF). Patient Prefer Adherence 2010;4:11–6. [37] Vadiveloo M, Scott M, Quatromoni P, Jacques P, Parekh N. Trends in dietary fat and high-fat food intakes from 1991 to 2008 in the Framingham Heart Study participants. Br J Nutr 2014;111:724–34. [38] CDC data briefs. Intake of calories and selected nutrients for the United States Population, 1999-2000. http://www.cdc. gov/nchs/data/nhanes/databriefs/calories.pdf. [Accessed June 10, 2014]. [39] Wright JD, Kennedy-Stephenson J, Wang CY, McDowell MA, Johnson CL. Dietary intake of ten key nutrients for public health, United States: 1999-2000. Advance data from vital and health statistics; no. 334. Hyattsville, Maryland: National Center for Health Statistics; 2003. [40] Cogswell ME, Zhang Z, Carriquiry AL, Gunn JP, Kuklina EV, Saydah SH, et al. Sodium and potassium intakes among US adults: NHANES 2003-2008. Am J Clin Nutr 2012;96:647–57.

106

N U TR IT ION RE S EA RCH 3 5 ( 2 01 5 ) 9 7 –1 06

[41] Grooms KN, Ommerborn MJ, Pham DQ, Djoussé L, Clark CR. Dietary fiber intake and cardiometabolic risks among US adults, NHANES 1999-2010. Am J Med 2013;126:1059–1067.e4. [42] Perry GS, Byers T, Yip R, Margen S. Iron nutrition does not account for the hemoglobin differences between blacks and whites. J Nutr 1992;122:1417–24. [43] Cotton PA, Subar AF, Friday JE, Cook A. Dietary sources of nutrients among US adults, 1994 to 1996. J Am Diet Assoc 2004;104:921–30. [44] Slyper AH. The influence of carbohydrate quality on cardiovascular disease, the metabolic syndrome, type 2 diabetes, and obesity—an overview. J Pediatr Endocrinol Metab 2013;26:617–29. [45] Craig WJ. Nutrition concerns and health effects of vegetarian diets. Nutr Clin Pract 2010;25:613–20. [46] Clarys P, Deriemaeker P, Huybrechts I, Hebbelinck M, Mullie P. Dietary pattern analysis: a comparison between matched vegetarian and omnivorous subjects. Nutr J 2013;12:1–6 [82]. [47] Mishra S, Barnard ND, Gonzales J, Xu J, Agarwal U, Levin S. Nutrient intake in the GEICO multicenter trial: the effects of a multicomponent worksite intervention. Eur J Clin Nutr 2013; 67:1066–71. [48] Turner-McGrievy GM, Barnard ND, Cohen J, Jenkins DJ, Gloede L, Green AA. Changes in nutrient intake and dietary quality among participants with type 2 diabetes following a low-fat vegan diet or a conventional diabetes diet for 22 weeks. J Am Diet Assoc 2008;108:1636–45. [49] Turner-McGrievy GM, Barnard ND, Scialli AR, Lanou AJ. Effects of a low-fat vegan diet and a Step II diet on macroand micronutrient intakes in overweight postmenopausal women. Nutrition 2004;20:738–46. [50] Lu Z, Chen TC, Zhang A, Persons KS, Kohn N, Berkowitz R, et al. An evaluation of the vitamin D3 content in fish: is the vitamin D content adequate to satisfy the dietary requirement for vitamin D? J Steroid Biochem Mol Biol 2007;103:642–4. [51] Tete S, Nicoletti M, Saggini A, Maccauro G, Rosati M, Conti F, et al. Nutrition and cancer prevention. Int J Immunopathol Pharmacol 2012;25:573–81.

[52] Peltonen R, Nenonen M, Helve T, Hanninen O, Toivanen P, Eerola E. Faecal microbial flora and disease activity in rheumatoid arthritis during a vegan diet. Br J Rheumatol 1997;36:64–8. [53] Kjeldsen-Kragh J. Rheumatoid arthritis treated with vegetarian diets. Am J Clin Nutr 1999;70:594S–600S. [54] Masquio DC, de Piano A, Campos RM, Sanches PL, Carnier J, Corgosinho FC, et al. Reduction in saturated fat intake improves cardiovascular risks in obese adolescents during interdisciplinary therapy. Int J Clin Pract 2014. http://dx.doi.org/10.1111/ijcp.12573 [Epub ahead of print]. [55] Young RP, Hopkins RJ. High dietary fiber lowers systemic inflammation: potential utility in COPD and lung cancer. Am J Med 2014;127:e13. [56] Glick-Bauer M, Yeh MC. The health advantage of a vegan diet: exploring the gut microbiota connection. Nutrients 2014;6: 4822–38. [57] Field AE, Colditz GA, Fox MK, Byers T, Serdula M, Bosch RJ, et al. Comparison of 4 questionnaires for assessment of fruit and vegetable intake. Am J Public Health 1998;88: 1216–8. [58] Lagerros YT, Mucci LA, Bellocco R, Nyren O, Balter O, Balter KA. Validity and reliability of self-reported total energy expenditure using a novel instrument. Eur J Epidemiol 2006;21:227–36. [59] Kristal AR, Peters U, Potter JD. Is it time to abandon the food frequency questionnaire? Cancer Epidemiol Biomarkers Prev 2005;14:2826–8. [60] Yang Y, Buys DR, Judd SE, Gower BA, Locher JL. Favorite foods of older adults living in the Black Belt Region of the United States. Influences of ethnicity, gender, and education. Appetite 2013;63:18–23. [61] Ridker PM, Lüscher TF. Anti-inflammatory therapies for cardiovascular disease. Eur Heart J 2014;35:1782–91. [62] Xu CX, Zhu HH, Zhu YM. Diabetes and cancer: associations, mechanisms, and implications for medical practice. World J Diabetes 2014;5:372–80.