Testing the Feasibility and Impact of a Whole-Food Plant-Based Diet Intervention on College Students

Diet quality was assessed using predicted dietary intakes calculated via the 26-item Dietary Screener Questionnaire (DSQ). This screener asks about the frequency of consumption in the past month of selected foods and drinks. The DSQ captures intakes of fruits and vegetables, dairy/calcium, added sugars, whole grains/fiber, red meat, and processed meat. Responses are on 9-point (for food items) and 11-point (for beverages) Likert scales, ranging from "never" to "two or more times/day" and "never" to "six or more times/day," respectively. Regression- and logistic-based scoring algorithms are used to convert self-reported food frequency data into quantitative daily intake estimates and prevalence probabilities. For this study, intakes more aligned with the 2015 Dietary Guidelines for Americans reflected better diet quality. The researchers assessed differential change in mean predicted intakes descriptively and via effect sizes across the study groups.

Diet quality was assessed via calculation of plant-based natural (PBN) and animal-based/processed (ABP) scores. The researchers assigned scores of 1 to 4 to quartiles of cup intakes of vegetables and fruits and ounce equivalents of whole grains (different cup and ounce equivalent thresholds by sex) to calculate the PBN scores. The PBN scores ranged from 3 to 12. The researchers assigned scores of 1 to 4 to different frequencies of intakes of dairy products and red and processed meats and quartiles of teaspoon per day intakes of added sugars to calculate ABP scores; ABP scores ranged from 5 to 20. Dairy product and red and processed meat frequency thresholds were the same by sex, and added sugar amounts differed by sex. Higher PBN and lower ABP scores suggested better overall diet quality. There are no validated cut-off scores for either. The researchers assessed differential change in group mean PBN and ABP scores descriptively and via effect sizes across the study groups.

Adherence to a whole-food plant-based (WFPB) vegan diet was assessed via calculation of plant-based natural (PBN) and animal-based/processed (ABP) scores. Scores of 1 to 4 were assigned to quartiles of cup intakes of vegetables and fruits and ounce equivalents of whole grains (different cup and ounce equivalent thresholds by sex) to calculate the PBN scores. The PBN scores ranged from 3 to 12. Scores of 1 to 4 were assigned to different frequencies of intakes of dairy products and red and processed meats and quartiles of teaspoon per day intakes of added sugars to calculate ABP scores. ABP scores ranged from 5 to 20. Dairy product and red and processed meat frequency thresholds were the same by sex, and added sugar amounts differed by sex. Higher PBN and lower ABP scores suggested better adherence to the diet. There are no validated cut-off scores. The researchers assessed differential change in group mean PBN and ABP scores descriptively and via effect sizes across study groups.

The researchers measured weight in kilograms using a digital scale. Weight alone was not interpreted as high or low; rather, the researchers assessed differential change (gain or loss) in group mean weight descriptively and via effect sizes across the study groups. This measure was also used to calculate body mass index.

The researchers measured height in centimeters using a stadiometer at baseline only. The researchers entered this height measurement together with the weight in kilograms into a body mass index (BMI) calculator to calculate and report BMI in kilograms/meter^2. The researchers assessed differential change in group mean BMI descriptively and via effect sizes across the study groups. The reference ranges used to interpret group mean BMIs were as follows: BMI < 18.5 = underweight; 18.5 to 24.9 = healthy weight; 25.0 to 29.9 = overweight; 30 or above = obesity.

The researchers used a flexible tape measure to assess waist circumference in centimeters. The researchers assessed differential change in group mean waist circumference descriptively and via effect sizes across the study groups. The researchers used standard western population cut-offs, i.e., for males, a measurement greater than or equal to 102 cm, and for females, a measurement greater than or equal to 88 cm, for high versus low waist circumference.

The researchers used a stethoscope and manual sphygmomanometer to measure systolic and diastolic blood pressure in millimeters of mercury (mmHG). The researchers assessed and reported differential change in group mean systolic and group mean diastolic blood pressure descriptively and via effect sizes across the study groups. Greater than or equal to 130 systolic and 80 diastolic = hypertension; 130-139 systolic/ 80-89 diastolic = stage 1 hypertension; greater than or equal to 140 systolic/ 90 diastolic = stage 2 hypertension.

The researchers assessed lipids using the Cholestech LDX® glucose and lipid analyzer. The researchers collected 40 micrograms of capillary blood to assess and report total cholesterol (TC), high-density lipoprotein cholesterol (HDL), and low-density lipoprotein cholesterol (LDL) and tryglicerides (TG). The researchers calculated non-HDL cholesterol by subtracting HDL from TC. They reported these results in milligrams per deciliter and assessed and reported differential change in group mean TC, HDL, LDL, and non-HDL cholesterol and TG descriptively and via effect sizes across the study groups. Optimal lipid levels: TC equal to or less than 150 mg/dL; LDL equal to or less than 100 mg/dL; HDL at least 40 mg/dL in males and 50 mg/dL in females; TG less than 150 mg/dL; non-HDL less than 130 mg/dL.

The researchers assessed fasting glucose using the Cholestech LDX® glucose and lipid analyzer. The researchers collected 40 micrograms of capillary blood to assess glucose and reported the results in milligrams per deciliter. They assessed and reported differential change in group mean glucose descriptively and via effect sizes across the study groups. Normal fasting glucose = 99 mg/dL or less; Prediabetes = 100-125 mg/dL; Diabetes = 126 mg/dL or above.

The researchers used the 10-item Center for Epidemiological Studies Depression Scale Revised (CESD-R-10) to measure depressive symptom scores. They assessed and reported differential change in group mean depressive symptom scores descriptively and via effect sizes across the study groups. The possible range of scores is 0-30 with higher scores representing greater degrees of depressed mood. The cut-off score for higher risk for clinical depression is greater than or equal to 10.

The researchers used the 7-item Generalized Anxiety Disorder (GAD-7) tool to measure anxiety symptom scores. They assessed and reported differential change in group mean anxiety symptom scores descriptively and via effect sizes across the study groups. The possible range of scores is 0-21. Scores from 0-4 = minimal anxiety; 5-9: mild anxiety; 10-14 moderate anxiety; 15-21 severe anxiety. A score of 10 or greater is the cut-off for generalized anxiety disorder with need for further assessment.

The researchers measured eating self-efficacy using an adaptation of the 8-item Eating Self-Efficacy Brief Scale. Scores were calculated by summing the scale responses; higher scores reflected higher eating self-efficacy. The range of scores is 8 to 48. There is not validated cut-off score. The researchers modified the stem used in this scale from 'Below is a list of situations in which people may have difficulties in resisting the urge to eat. For each situation, please answer considering how easy it would be to resist the urge to eat' to 'On a scale of 1 to 6 (1= extremely hard to 6= extremely easy), how would it be to follow a whole-food plant-based diet?' They assessed and reported differential change in group mean eating self-efficacy scores descriptively and via effect sizes across the study groups.

The researchers measured outcome expectations using a 10-item translated and adapted Outcome Beliefs Scale. The researchers altered the stem for this scale from 'If eating at least two pieces of fruit and 200 grams of vegetables a day…' to read, 'If eating a whole-food plant-based diet…' and removed the "neutral" response, resulting in a 6-point Likert scale (1= strongly disagree to 6= strongly agree). Scores were calculated by summing the responses; higher scores reflected more positive outcome beliefs. The range of scores was 10 to 60. There is no validated cut-off score. The researchers assessed and reported differential change in group mean eating outcome belief scores descriptively and via effect sizes across the study groups.

The researchers measured attitude toward eating a whole-food plant-based (WFPB) vegan diet via the 4-item attitude toward behavior (ATB) measure. Originally created to assess this construct in relation to meat-eating, vegetarian, and vegan diets, for this study, the ATB measure was asked only with regard to a WFPB vegan diet. The ATB measure asks participants to rate on a 7-point Likert scale their attitudes towards a diet, i.e. 1= very bad to 7= very good, 1= very harmful to 7= very beneficial, 1= very unpleasant to 7= very pleasant, and 1=very unenjoyable to 7= very enjoyable. Scores were calculated by summing the responses; higher scores reflected more positive ATB. The scores ranged from 4 to 28. There is no validated cut-off score. The researchers assessed and reported differential change in group mean ATB scores descriptively and via effect sizes across the study groups.

The researchers measured perceived behavioral control (PBC) with following a whole-food plant-based (WFPB) vegan diet with the 3-item perceived behavioral control measure. Originally created to assess PBC in relation to meat-eating, vegetarian, and vegan diets, for this study, this measure was asked with regard only to the WFPB vegan diet. The PBC measure inquired, using a 7-point Likert scale, about personal control (1= no control to 7= complete control), capability (1= very incapable to 7= very capable), and difficulty (1= very difficult to 7= very easy) with following a WFPB vegan diet in the future. Scores were calculated by summing the responses; higher scores reflected higher PBC. The scores ranged from 3 to 21. There is no validated cut-off score. The researchers assessed and reported differential change in group mean PBC scores descriptively and via effect sizes across the study groups.

The researchers measured behavioral intention (BI) to continue following a whole-food plant-based (WFBP) vegan diet in the future with the 1-item behavioral intention measure. Originally created to assess BI in relation to meat-eating, vegetarian, and vegan diets, for this study, participants were asked to rate their agreement with a single statement about intending to eat a WFPB vegan diet in the future on a 6-point Likert scale (removed the "neutral" response on the original measure) from 1= strongly disagree to 6= strongly agree. Scores ranged from 1 to 6. There is no validated cut-off score. The researchers assessed and reported differential change descriptively and via effect sizes in group mean BI scores across the study groups.

Behavioral capability (BC) was measured using the confidence measure from the Food and Cooking Skills tool, consisting of 19-items divided into 5 different domains: meal planning and preparing, shopping, budgeting, resourcefulness, and label reading/consumer awareness. The researchers dropped 3 items due to their lack of relevance (e.g., buying cheaper cuts of meat to save money, using best-before dates, and storage use-by information). Participants were asked to rate on a 7-point Likert scale how good they were at particular food/cooking skills: 1=very poor and 7=very good and 0=never/rarely do the skill. The confidence score was derived by summing the 1 to 7 ratings for the skills used. Thus, the range in scores was 1 to 112, with higher scores indicating greater confidence. There is no validated cut-off score. Differential change in group mean BC scores across the study groups was reported descriptively and via effect sizes.

The researchers measured self-regulation of eating behavior (SREB) using an adaptation of the 5-item Self-Regulation of Eating Behaviour Questionnaire (SREBQ). Three screening questions about intentions to eat a healthy diet or avoid tempting foods precede the questionnaire. The researchers dropped these three screening questions and used one prompt that asked participants to rate on a 5-point Likert scale the frequency with which statements about eating intentions were true for them, 1=never to 5=always (3 statements with reverse scoring). The scale is scored by calculating the means of the responses, ranging from 1 to 5; a mean < 2.8 suggests low self-regulation, 2.8 to 3.6 medium self-regulation, and > 3.6 high self-regulation. The researchers assessed and reported differential change in group mean SREBQ scores descriptively and via effect sizes across the study groups.