Protocol and Preliminary Results of the Nutritarian Women’s Health Study: A Longitudinal Effectiveness-Implementation Hybrid Study Assessing Dietary Intake and Health Outcomes

Abstract

Whole-food plant-rich dietary patterns have been shown to be associated with significant health benefits and disease-risk reduction. One such program, which has been gaining popularity, is the micronutrient-dense plant-rich (mNDPR) “Nutritarian” diet. The goal of this study is to determine the feasibility of implementing an online mNDPR “Nutritarian” intervention program and to determine the effectiveness of this program in reducing risk of chronic disease in women. The Nutritarian Women’s Health Study is a long-term online longitudinal hybrid effectiveness-implementation study. Participants are recruited through social media, medical professionals/offices, and nutrition-related events and conferences. Participants receive online nutrition education and complete regular self-reported questionnaires regarding lifestyle, nutrition practices, and health. The online intervention program appears to be feasible and effective. Some decline in dietary adherence, particularly for certain food types, was observed during the study. For groups at risk, based on body mass index or waist-to-height ratio, there were initial decreases in body mass index and waist-to-height that leveled off over time, in some cases returning to baseline measures. The study suggests the implementation of the Nutritarian dietary pattern, through an online intervention component, may be effective in reducing the risk of chronic disease, with implications for clinical and public health practice.

‘The Nutritarian dietary pattern compared to a whole food plant-based diet is unique in the emphasis of the consumption of nutrient-dense foods.’

In the United States, a significant majority of the population do not consume adequate amounts of whole fruits and vegetables, with only 12.2% and 9.3% meeting recommendations, respectively. ¹ This is a cause for concern because poor diet quality is an independent risk factor for many chronic diseases including cardiovascular disease, cancer, and diabetes. ² Worldwide, 11 million deaths are attributed to poor diet quality including excess sodium and low consumption of whole grains, seeds and nuts, fruits, and vegetables. ³ Furthermore, diets composed of ultra-processed foods contribute to overeating, increasing overall calorie consumption contributing to weight gain and potentially the prevalence of obesity. ⁴

Cancer is a major cause of morbidity and mortality in the United States. ⁵ After skin cancer, breast cancer is the most common form of cancer in women, with nearly 1 in 8 women diagnosed in their lifetime. ⁶ Lifestyle factors such as tobacco use, diet, living a sedentary lifestyle, and obesity have been identified as important contributors to the increasing incidence of breast and other cancers. Specifically, diets low in micronutrients and the overall intake of high glycemic index foods have been shown to influence cancer, cardiovascular disease, and diabetes risk. ⁷

Obesity is the accumulation of excessive body fat; visceral fat (central obesity) may especially induce inflammation, a risk factor for developing cardiovascular disease, insulin resistance, diabetes mellitus, and several other diseases. ⁸ One noninvasive measure for monitoring inflammation is the waist-to-height ratio, an effective screening tool for metabolic syndrome, ⁹ and identifying cardiometabolic risk, ¹⁰ in various populations.

MicroNutrient-Dense, Plant-Rich (mNDPR) dietary patterns, similar to the Nutritarian dietary pattern described in this study, have been shown to be safe and effective in clinical applications and are positively associated with weight reduction,^11-13 lipid management,^11,12,14 glycemic control in diabetes, ¹⁵ inflammation reduction, ¹⁶ perceptions of hunger, ¹⁷ reduction in depressive symptoms and improvement in well-being,^11,18-20 reducing the risk of developing cancer, ²¹ and overall health and longevity. ²² Various plant-based diets such as vegan diets and vegetarian diets have been studied extensively. To date, there have been no studies on the long-term impact of the Nutritarian diet specifically. Short-term outcomes have shown promising results clinically including reductions in weight, cholesterol, and blood pressure. ²³ The Nutritarian dietary pattern compared to a whole food plant-based diet is unique in the emphasis of the consumption of nutrient-dense foods. Unlike a whole food plant-based diet, the Nutritarian dietary pattern requires the daily consumption of greens, beans, onions, mushrooms, berries, and seeds. For the Nutritarian diet, foods that contain higher amounts of micronutrients should be selected over foods that contain fewer nutrients. Similar to a whole food plant-based diet, processed foods should be minimized; preferably whole foods should be consumed. Comparatively, whole food plant-based diets do not indicate the nutrient levels of food choices whereas the Nutritarian diet strives to achieve high nutrient density.

The aim of the Nutritarian Women’s Health Study (NWHS) is to determine the effectiveness of a Nutritarian dietary pattern on the long-term incidence and progression of chronic diseases such as diabetes, cardiovascular disease, and certain types of cancers. The intent of this article is (1) to describe the protocol of NWHS and (2) to determine the feasibility and effectiveness of an online intervention program designed to educate participants about a Nutritarian dietary pattern. The NWHS is designed to address the following hypotheses:

1. Commitment to a Nutritarian dietary pattern will reduce the incidence and severity of chronic diseases (including nutrition-related cancers, cardiovascular diseases, diabetes, and other lifestyle-related conditions).

2. An online educational program will be effective in educating participants of the value of this lifestyle modification and reduce the risk of chronic, diet-related diseases.

Several other observational longitudinal studies, such as the Nurses’ Health Study and the Adventist Health Study, seek to monitor the health of participants over long periods. The present NWHS recruits participants interested in a Nutritarian dietary pattern and has an intervention component that guides participants to make healthy lifestyle choices and improve adherence to the dietary protocol.

Methods

Study Design

The NWHS was initiated in 2016 and follows a rolling enrollment format. NWHS is a type 2 hybrid effectiveness-implementation study design aimed equally at evaluating the implementation of the online intervention and improvement in health outcomes. ²⁴ This longitudinal study seeks to track adult females over an extended period of time. Study participants answer quarterly online questionnaires in an attempt to measure (1) adherence to the recommended protocol and (2) health outcomes. We are primarily focused on how this online intervention affects participant adoption and adherence to the recommended dietary pattern, as well as long-term disease prevention, management, and progression. Participants will serve as their own controls as the study progresses longitudinally.

Recruitment, Enrollment, Participant Flow, and Consent

Currently, the study has 2 cohorts. Cohort 1 is described in this article and is defined as any participant who enrolled from October 31, 2016, to April 15, 2018 (see Figure 1). Cohort 2 is defined as any participant who enrolled from April 16, 2018, to October 16, 2019. Additional cohorts will be added in 18-month increments until the study enrollment reaches a goal of at least 5000 women. Participants are recruited through community seminar promotions, online media sources including Facebook Ads and Google Ads, numerous health care conferences, health care provider offices and clinics, and by word of mouth. Potential participants are invited to visit a dedicated secure website, which explains information about the study. Participation is limited to English-speaking women, ages 18 years and older, who are US residents, and have internet access. There is no an upper age limit for participation. Volunteers must be willing to complete an online program and follow the Nutritarian dietary guidelines, as outlined, to the best of their ability. During the enrollment process for Cohort 1, pregnant women were excluded from participating. Women’s health, specifically breast cancer, is a primary focus for the study long-term; therefore, men are excluded from participation as breast cancer is very rare among men. The study was approved by the Northern Arizona University Institutional Review Board and all study participants signed informed consent.

Figure 1.

Participant flow.

During registration, participants complete a brief pre-enrollment consent, which provides entry to the 30-day trial, to serve as an instructional period (Figure 1). It is not a requirement that women take all 30 days before committing to the full study. This instructional period has educational tutorials including 4 short videos, created by the study team (~33 minutes total) and online access to the book Super Immunity by Joel Fuhrman, MD. The book was chosen by the research team because it outlines the diet, provides rationale for specific foods included and excluded from the dietary portfolio, provides recipes, and because there are few other resources that specifically feature the Nutritarian diet. On completion of the educational module, participants complete a 15-question quiz to assess their knowledge of the dietary protocol. Participants who decide to continue and enroll in the full study sign the informed consent and complete extensive self-reported medical and dietary history questionnaires. Participant interaction, questionnaires, quizzes, videos, and written material are provided via a designated secure website with electronic data capture tools. After enrollment, participants receive a magnet as shown in Figure 2 to serve as a reminder of the daily Nutritarian principles and a copy of Dr. Fuhrman Nutritarian Handbook & ANDI Food Scoring Guide by Joel Fuhrman, MD. Participants are required to notify their primary care provider of their participation in the study. In addition, the participant Bill of Rights is also provided during the 30-day enrollment process. There is no financial cost to participate in any portion of this study, and all materials are provided without cost to participants.

Figure 2.

Daily checklist.

Nutrition Intervention

The Nutritarian Diet, the dietary protocol for this study has been described previously, ²⁴ and it has 4 primary principles designed to promote foods shown to support longevity and to contain anticancer properties: (1) it emphasizes the consumption of an assortment of high micronutrient-dense foods; (2) it strives to be hormonally favorable (the use of “hormonally favorable” foods in this dietary protocol is intended to minimize the use of dietary choices, such as dairy products and high-glycemic foods, that have an impact on hormonal and inflammatory markers; such dietary factors tend to elevate IGF-1 and serum insulin levels, respectively, and high IGF-1 and insulin levels have been shown to promote breast and prostate cancer ²⁵ ); (3) it strives to be nutritionally adequate; and (4) it seeks to avoid foods containing toxins and contamination. ²⁶ Environmental toxins such as polychlorinated biphenyls, dioxins, and mercury can be found in our food supply. These substances are chemicals that have been associated with causing cancer and may be found in contaminated animal products.^27,28 Pesticide/herbicide-derived toxins can also be found in food; for this study, however, organic food consumption was not compared with conventional foods, and participants were encouraged to purchase food from any source they chose.

The calorie breakdown of the diet consists of approximately 30% to 60% vegetables (excluding white potatoes), 10% to 40% beans/legumes, 10% to 40% fruit, 10% to 40% seeds/nuts, 20% or less whole grains, and no more than 10% naturally raised and wild animal products, poultry, eggs, fish, dairy, oil, and white potatoes. The consumption of micronutrients, especially antioxidants and phytochemicals, are emphasized. In addition, participants are strongly encouraged to include the following supplements: 200 µg vitamin B₁₂, 1000 to 2000 IU vitamin D₃, 150 µg iodine (unless diagnosed with hyperthyroidism or overactive thyroid), and 200 to 300 mg DHA-EPA. Participants were provided with a resource list for various brands of supplements they could purchase. Supplement usage, including multivitamins, B₁₂, D₃, iodine, and DHA-EPA, were collected in the Medical and Diet History questionnaires directly following entry into the full study only. Participants were also notified that supplements were not a requirement for participation in the study and they would not be withdrawn if they chose to omit supplements.

Compliance to the diet involves consumption of key foods as follows: greens: 3 to 4 servings daily; beans: 1 to 2 servings daily; onion: 1 or more servings daily; mushrooms: 1 or more servings daily; berries: at least 4 to 6 servings daily; seeds: 1 or more servings daily; tomatoes: at least 4 to 6 servings weekly; high-glycemic index food such as sugar/refined flour: 0 servings daily; animal products: no more than 2 servings weekly; alcohol: none. As the Nutritarian diet encourages the daily consumption of greens, beans, onions, mushrooms, berries, and seeds, participants are provided with the acronym GBOMBS to remember these specific foods. The specific foods listed are featured because of their attributes in relation to the principles and benefits outlined above. Greens are defined as leafy greens and other green vegetables such as broccoli, brussel sprouts, and bok choy that belong to the cruciferous family of vegetables. Greens are an important component of the diet because of their extremely high micronutrient density, and also because of their anticancer properties. Participants are encouraged to consume a large leafy green salad each day that includes raw onion, cruciferous vegetables, and tomatoes. In addition, participants are encouraged to consume at least 2 servings of cooked greens with mushrooms and onions each day.

Beans are emphasized in this dietary protocol because of their glycemic favorability, weight reduction benefits, and anticancer effects. ²⁴ Beans offer increased fiber while decreasing the glycemic load, especially when beans are the primary carbohydrate source. It is important to acknowledge the benefits of whole grains for fiber consumption as well. The Nutritarian diet recommends 20% or less of overall calories to be from whole grains. When considering the diet ratio, a higher proportion of the overall diet is typically recommended to be from bean or legume sources rather than whole grains. Participants are encouraged to consume whole unrefined grains and avoid white grains. White grains are defined as refined grains that have been milled, stripping the bran and germ. Foods such as white rice and white flour are considered white grains and are recommended to be avoided.

Onions and other allium vegetables are emphasized because the compounds contained in this family of vegetables have been shown to prevent the development of cancers by detoxifying carcinogens. ²⁹ Furthermore, onions contain substances that have anti-inflammatory traits that when combined with other micronutrients help improve immune function and potentially prevent disease. ²⁹ Participants are encouraged to consume raw onions and other allium vegetables on a daily basis.

Mushrooms are emphasized for their cancer prevention properties including reducing the level of estrogen and prevent estrogen from stimulating breast tissue. ³⁰ Mushrooms contain antigen-binding lectins that bind to abnormal cells and activate the body’s response to abnormal cells. ³⁰ Even a small amount of mushrooms consumed daily is thought to decrease cancer risk.

Berries contain ellagic acid and anthocyanins, both of which are protective against cancer development. ³¹ Berries have particularly high amounts of these compounds when compared with other types of fruits. ³¹ Daily consumption of berries is encouraged to potentially prevent the formation of specific tumors.

Dietary fat increases the absorption of other micronutrients, especially fat-soluble vitamins. Nuts and seeds contain plant protein, bioflavonoids, and minerals that naturally lower cholesterol. When nuts and seeds are consumed with other micronutrient-dense foods, the absorption of the vitamins and minerals increases. Nuts and seeds are therefore emphasized in the Nutritarian diet both because of their fat content and also because of the ability of those fats to increase the absorption of micronutrients.

The Nutritarian diet permits a small amount of meat consumption, but this practice is not encouraged. The rationale of allowing a small amount of unprocessed meat is that empirical evidence is lacking or inconsistent that there are indeed negative health implications for eating a small amount (<10% of total calories) of naturally raised animal products. Furthermore, for some participants who may be transitioning to the Nutritarian diet, including a small amount of meat at the beginning may help improve long-term compliance. Further research with this population may inform future guidelines on including animal products long-term. If participants do consume animal protein, it is recommended to be from naturally raised and wild animal products.

Data Collection and Management

For each of the outcome measures, data for Cohort 1 were compared at the time of enrollment (Time 0) with post-enrollment data obtained in June 2018 (Time 1), September 2018 (Time 2), and December 2018 (Time 3); thus, the interval between the time of enrollment and the first time point post-enrollment (Time 1) varied for different individuals, because of the rolling enrollment format.

Study data are collected and managed using REDCap (Research Electronic Data Capture) electronic data capture tools hosted at Northern Arizona University. REDCap is a secure, web-based application designed to support data capture for research studies, providing (1) an intuitive interface for validated data entry, (2) audit trails for tracking data manipulation and export procedures, (3) automated export procedures for seamless data downloads to common statistical packages, and (4) procedures for importing data from external sources. ³²

Assessment of Dietary Intake

At the time of enrollment, initial dietary intake was determined with a “Diet and Medical History” questionnaire. A brief food frequency questionnaire was developed specifically for this study and was used as part of the Health Indicator Questionnaire to assess subsequent adherence to the Nutritarian diet every 3 months beginning June 2018 (ie, June 2018, September 2018, and December 2018). In this questionnaire, participants reported the total number of servings they consumed over the past 30 days for the following foods: green vegetables, berries, beans, onions and garlic, mushrooms, tomatoes and tomato products, animal products, sugar and refined foods, and alcoholic beverages. Servings were described for each of the various food groups in common measurements such as cups and tablespoons.

Assessment of Anthropometric Measures

In addition to the dietary assessment, participants self-reported anthropometric measures including weight, height, waist circumference, and hip circumference, as part of the Health Indicator questionnaire, on a quarterly basis. For this questionnaire, the participants were provided with a visual figure to describe the location of where to measure their waist and hip measurements. Participants were instructed to measure their waist at the narrowest point, ensuring the tape measure was horizontal. Hip measurement was taken at the widest lateral extension of the hips. Participants were instructed to measure in inches.

Assessment of Covariates

The Health Indicator questionnaires were also used quarterly to document risk factors for chronic disease, such as smoking, alcohol use, body weight, and physical activity. History of chronic diseases was also documented during enrollment using the Medical and Diet History questionnaire and annually thereafter.

Continuous Participant Support

The participants received a twice-monthly email containing a recipe that complies with the Nutritarian principles, nutritional benefits of specific ingredients included in the recipe, as well as other encouraging information. In addition, the email reminded participants of upcoming study questionnaires.

Statistical Analysis

All analyses for Cohort 1 were conducted using SYSTAT 13.2. Primarily descriptive statistics were used including frequency distributions and measures of central tendency and variability when appropriate. Body mass index (BMI) was measured during enrollment (Time 0) and over the course of the first 3 measurement periods after recruitment began (Times 1, 2, and 3), and was analyzed with repeated-measures analysis of variance (ANOVA) for 5 standard BMI categories: underweight (<18.5 kg/m²), normal (18.5-24.999 kg/m²), overweight (25-29.999 kg/m²), class 1 obesity (30-34.999 kg/m²), and class 2 obesity (≥35 kg/m²). Analyses were conducted using these categories in order to identify any differences within groups. Waist circumference and height was self-reported during enrollment and over the course of the first 3 measurement periods after recruitment began. Study staff calculated the waist-to-height ratio of participants. Changes in waist-to-height ratio were addressed with planned comparisons within standard waist-to-height categories (too thin, <0.40; no increased risk, 0.40-0.495; increased risk, 0.50-0.595; very high risk, ≥0.60), using repeated-measures ANOVA. Dietary adherence to the Nutritarian diet was measured across the first 3 measurement periods after recruitment began, in which participants reported frequency of food consumption of specific foods.

Study Population

Participants in Cohort 1 were enrolled between October 31, 2016, and April 15, 2018. During this time, 2137 women enrolled, and they represented all US states and the District of Columbia. Mean participant age was 51.7 years (SD = 11.8). Nearly two thirds (70.5%) of participants reported that they were currently married. Three quarters (75%) of participants reported having had at least one pregnancy. In the initial recruitment, current tobacco use (cigarette smoking) was reported by 1.1% of participants. Past years of smoking ranged from 2 to 40 years (mean = 17.3 years, SD = 11.4, median = 17.5 years). When asked about parental history, participants (n = 1866) reported on maternal health history. Of these, 8.57% reported that their mother had heart disease, and 10.77% reported that their mother had breast cancer. Slightly fewer participants (n = 1864) reported on paternal health history. Of these, 23.71% reported paternal heart disease, and 12.24% reported paternal prostate cancer.

Other characteristics of the study population are shown in Table 1. At baseline, nearly 47% of participants were categorized as underweight or normal categories based on BMI information categories as defined by the US Department of Health and Human Services. ³³ Over 52% were in overweight, obese, or extremely obese categories. Mean participant BMI was 27.17 kg/m² (SD = 7.02), and mean participant waist-to-height ratio was 0.55 (SD = 0.14). Distributions were skewed with median BMI 25.46 kg/m² and median waist-to-height ratio 0.53.

Table 1.

Characteristics of the Study Population ^a .

	n	Percentage
Race
American Indian/Alaska Native	5	0.2
Asian	28	1.4
Native Hawaiian or Other Pacific Islander	2	0.1
Black or African American	66	3.2
White	1887	92.5
More than one race	53	2.6
Unknown/not reported	33	1.62
Total	2041	100.0
Ethnicity
Hispanic or Latino	121	6.0
Not Hispanic or Latino	1912	94.0
Not reported	41	1.98
Total	2033	100.0
Education level
Less than high school	3	0.1
High school	68	3.3
Some college	495	23.8
Bachelor’s degree	743	35.7
Graduate degree	771	37.1
Total	2080	100.0
Annual household income level
$30 000 or less	157	8.3
$30 000 to $50 000	245	13.0
$50 000 to $70 000	294	15.6
$70 000 to $100 000	381	20.2
$100 000 to $250 000	677	35.8
More than $250 000	136	7.2
Not reported	184	9.7
Total	2074	100
Body mass index (kg/m2)
<18.5	62	3.1
18.5-24.99	881	43.9
25-29.99	483	24.1
30-40	455	22.7
>40	124	6.2
Total	2005	100
Breast cancer treatment
Surgery	128	93.3
Chemotherapy	61	45.5
Radiation	73	53.7
Other self-reported health conditions
None reported	852	39.9
Obesity	454	21.2
Thyroid conditions	272	12.7
Heart conditions	44	2.1
Type 2 diabetes	77	3.6
Type 1 diabetes	5	0.23
Various other health conditions	620	29

^aParticipants can report multiple treatments and health conditions.

At enrollment, 6.5% of participants had received a diagnosis of breast cancer and had been treated (as shown in Table 1). Of those who had undergone surgery, 46.4% had undergone lumpectomy, 18.4% lateral mastectomy, and 35.2% bilateral mastectomy.

At enrollment, 40% of participants reported no health problems. When asked about specific disorders, participants reported the conditions shown in Table 1 (individuals could report more than one health condition).

Results

Preliminary results of this longitudinal hybrid effectiveness-implementation study are presented, including BMI, waist-to-height ratio, and adherence to the dietary protocol for the first cohort of participants (individuals who enrolled between October 31, 2016, and April 15, 2018).

Changes in BMI Over Time

BMI of participants was determined as described in Methods, and the results are summarized in Table 2. Repeated-measures ANOVA indicated that no changes were observed in individuals who were underweight (F[3, 81] = 1.917, P = .133, MSE = 0.194) or normal (ie, BMI <25 kg/m²; F[3, 912] = 2.398, P = .067, MSE = 0.906) at any time point. Individuals who had initial BMI >25 kg/m² classified based on initial BMI as overweight (F[3, 261] = 12.497, P < .001, MSE = 1.582), class 1 obesity (F[3, 129] = 2.718, P = .047, MSE = 3.637) or class 2 obesity (F[3, 93] = 6.316, P = .001, MSE = 9.141, ie, BMI >25 kg/m²). Bonferroni post hoc tests showed significant reductions in BMI between Time 0 and Times 1 and 2 for individuals who were overweight or in the obesity class 2 category. For the obesity class 1 category, the post hoc testing did not show significant differences in BMI across time periods. Bonferroni tests can be overly conservative when differences are small. ³⁴ In the BMI categories overweight, class 1 obesity, and class 2 obesity, a pattern of initial reduction in BMI between baseline and the first 2 measurements is shown followed by a return to baseline.

Table 2.

Body Mass Index (BMI) Changes Over Time.

BMI Category	n	BMI Time 0 (Initial)	BMI Time 1 (June)	BMI Time 2 (September)	BMI Time 3 (December)	F and P	Post Hoc Tests (Bonferroni)
Underweight (<18.5 kg/m2)	28	17.42 ± 0.71	17.50 ± 1.08	17.64 ± 1.01	17.38 ± 0.74	F(3, 81= 1.917, P = .133, MSE = 0.194	No differences
Normal (18.5-24.999 kg/m2)	305	21.62 ± 1.77	21.48 ± 2.18	21.43 ± 2.03	21.58 ± 1.84	F(3, 912) = 2.398, P = .067, MSE = 0.906	No differences
Overweight (25-29.99 kg/m2)	88	26.89 ± 1.34	26.243 ± 2.51	25.93 ± 2.17	26.86 ± 1.35	F(3, 261) = 12.497, P < .001, MSE = 1.582	Time 0 different from Time 1 and Time 2; Time 3 different from Times 1 and 2
Class 1 obesity (30-34.99 kg/m2)	44	31.94 ± 1.37	31.23 ± 3.07	30.99 ± 3.74	31.88 ± 1.77	F(3, 129) = 2.718, P = .047. MSE = 3.637	Bonferroni post hoc testing nonconclusive about differences
Class 2 obesity (≥35 kg/m2)	32	40.74 ± 5.25	38.40 ± 7.37	38.20 ± 7.17	40.50 ± 4.83	F(3, 93) = 6.316, P = .001, MSE = 9.141	Time 0 different from Times 1 and 2

Changes in Waist-to-Height Ratio Over Time

As shown in Table 3, the waist-to-height ratio reduced over time for individuals with ratio >0.5 (increased risk and very high-risk groups). There was a slight increase (1.33% average increase between initial measurement and Time 2) in the “no increased risk” group and a nonsignificant trend toward improvement in the “too thin” group.

Table 3.

Waist-to-Height Ratio Changes Over Time.

Waist-to-Height Category	n	Initial	Time 1	Time 2	Time 3	ANOVA Results and Post Hoc Tests
Too thin (<0.40)	18	0.38 ± 0.01	0.39 ± 0.03	0.39 ± 0.02	0.40 ± 0.04	F(3, 51) = 1.724, P = .174, MSE = 0.001 (no differences)
No increased risk (0.40-0.495)	247	0.45 ± 0.03	0.45 ± 0.04	0.46 ± 0.05	0.46 ± 0.05	F(3, 738) = 3.355, P < .019, MSE = 0.001 (initial and Time 2 different)
Increased risk (0.50-0.595)	133	0.54 ± 0.03	0.51 ± 0.05	0.51 ± 0.06	0.51 ± 0.06	F(3, 396) = 19.838, P < .001, MSE = 0.001 (initial different from others)
Very high risk (≥0.60)	72	0.66 ± 0.05	0.62 ± 0.07	0.62 ± 0.08	0.61 ± 0.08	F(3, 213) = 23.938, P < .001, MSE = 0.002 (initial different from others)

Abbreviation: ANOVA, analysis of variance.

Changes in Dietary Intake and Adherence Over Time

Dietary intake data from the Diet and Medical History questionnaire at the start of the study are available for 1855 participants. As described above, key components of the Nutritarian diet are minimizing or eliminating the consumption of meat, dairy, processed food, white grains, soft drinks, alcohol, fat/oil, and snacks. At enrollment, 50.51% participants reported no meat consumption, 43.45% reported no dairy consumption, 42.86% reported no processed-food consumption, 67.71% reported no white-grain consumption, 82.53% reported no soft drink consumption, 54.88% reported no alcohol consumption, 17.95% reported no fats/oil used, and 22.53% reported no snacks consumed.

After the end of the Cohort 1 rolling-enrollment period, participants were asked to report on various aspects of their diet quarterly, using the Health Indicator questionnaire, as described in Methods. Results for compliance for each of the food items are shown in Figure 3. In general, compliance was higher at the first post-enrollment time point (Time 1) than at subsequent time points. There was no Time 0 measure for diet adherence. Across all time periods, 75% of participants adhered to the recommended daily intake of berries; 63% of individuals adhered to the recommendation of no more than 2 servings of animal products per week; 53% adhered to the recommended daily intake of seeds; 53% adhered to the recommended daily intake of tomatoes; 46% adhered to the recommended daily intake of greens, beans (48%), onions (47%), no sugar/refined flour (44%), and no alcohol (48%); and 21% adhered to the recommended daily intake of mushrooms.

Figure 3.

Dietary adherence percent for specific foods across time.

Discussion

The intent of this article is (1) to describe the protocol for the NWHS, which is aimed at determining the effectiveness of a Nutritarian dietary pattern on the long-term incidence and progression of chronic diseases and (2) to determine the feasibility and effectiveness of an online intervention program designed to educate participants about a Nutritarian dietary pattern.

For Cohort 1 participants, we describe the following outcomes: (1) changes in BMI, (2) changes in waist-to-height ratio, and (3) adherence to the dietary components of the protocol.

1. Changes in BMI: Underweight and normal weight participants maintained their BMI from Time 0 to Time 3. Individuals with overweight, class 1 obesity, and class 2 obesity had reduced BMI from Time 0 to Time 1 and Time 2, which returned to baseline by Time 3. It may be significant that Time 3 data were obtained over the holiday period (December 2018); this, or seasonal differences, may explain a change in eating patterns at Time 3. It will be important to determine if there is an ongoing trend toward reduction in BMI for these groups, or whether the initial reduction seen between Time 0 and Times 1 and 2 is not sustained long-term.

2. Changes in waist-to-height ratio: There were no significant differences for the “too thin” waist-to-height category, although a trend was seen toward weight gain in this group. A significant difference was seen between Times 0 and 2 for the “no increased risk” waist-to-height category; this reflected a small increase in waist-to-height ratio for this group. A significant difference was also seen between Time 0 and all other time points for both the “increased risk” and “very high risk” waist-to-height categories; in both cases, this difference reflected a reduction in waist-to-height ratio over time. It is interesting that the increase observed in BMI for overweight, class 1 obese, and class 2 obese groups at Time 3 (see above) is not seen in the waist-to-height comparison for the increased risk and very high risk categories. One possible explanation is that the categories are not split in the same way for BMI as for waist-to-height ratio, but analysis of waist-to-height ratio, using the same groupings as for BMI, indicate that at Time 3 a similar increase is not observed (data not shown). Other possible explanations are that weight measurements are more sensitive to short-term fluctuations than are waist measurements and/or that participants did not gain weight in the abdominal region, which may be favorable as abdominal (visceral) fat has been shown to increase disease risk. ³⁵

Analysis of participants’ adherence to various components of the diet indicates that there is an overall trend toward reduced compliance over time for most components of the dietary recommendations. Overall adherence is higher for berries than other components. Interestingly, adherence to no more than 2 servings/week of animal products is also relatively high (60% to 70% over the course of the 3 time points). These findings suggest it will be important to emphasize the importance of all components of the diet, and incorporating all components into the diet as a lifestyle choice.

Strengths of this study include a geographically diverse population of women who represent all 50 states and the District of Columbia. This intervention is easily managed, as it is delivered via online methods. In addition, the online nature of the intervention makes it very flexible and adaptable for various people. The intervention emphasizes the consumption of unprocessed whole foods that are widely available and require minimal preparation and cooking. The participants are provided regular interaction and support and study team staff are available via phone and email.

Limitations of this study are an underrepresentation of low-income populations as well as ethnically diverse populations. Participants must speak English and have access to the internet; therefore, populations without access to computers and/or the internet are excluded from participation. In addition, all data collected were self-reported and monitoring physical exercise was limited. The nature of self-reported data is a limitation in itself; however, the validity of self-reported height and weight has been assessed previously in a similar population and considered a valid measure. ³⁶ Participants who self-selected for both initial enrollment and continued participation represent a small fraction of the total population at initial enrollment. Because of the absence of a control group for comparison, monitoring variation in dietary adherence and health measures over time is utilized. Participants will serve as their own controls and repeated measures have been utilized. Initial diet measures at time of enrollment suggest that a large proportion of the population have already committed to some aspects of the diet. A limitation of this study is that we did not collect dietary data using the same tool as the other time measures; therefore, we are unable to make direct comparison between dietary data collected initially and at Times 1, 2, and 3. Because it is early in the projected length of the study, the ability to track the progression of chronic diseases is not yet possible.

The aforementioned limitations are being addressed for future cohorts enrolled in this study. In subsequent cohorts, 24-hour diet recalls utilizing the National Cancer Institute Automated Self-Administered 24-Hour Dietary Assessment Tool (ASA-24) will provide the ability to determine a healthy eating index at baseline for incoming participants and twice per year for all current participants. Exercise at baseline and throughout the study will be tracked using self-reported measures to control for exercise. In order to recruit a more diverse group of participants, additional avenues are being considered and explored such as targeting additional health care providers that primarily provide female care services. To improve participant response to the questionnaires, various forms of maintaining contact such as email reminders, text reminders, and postcard reminders will be implemented. A benefit of a longitudinal study is the ability to make necessary changes as the study progresses. Participants in future cohorts will receive questionnaires based on enrollment date, rather than concurrently, to allow for a more informative assessment of comparing compliance over time following the intervention. Future cohorts will have variations in enrollment procedures.

We have identified the initial 30-day trial as a time when participants may lose motivation or become confused on how to resume their participation. Therefore, for future cohorts, the 30-day trial will be eliminated. While making dietary changes, regardless of the diet protocol, can be difficult, a series of support features will be implemented immediately after enrollment. This “barrier sequence” will provide participants with techniques for overcoming common barriers.

Due to the nature of this report, being an introduction, the generalizability of the findings is not yet known. This is an ongoing longitudinal study and periodic reports are anticipated, which will allow for the determination of applicability to various populations and the generalizability of the findings.

Conclusion

In this article, we describe the protocol for the NWHS, which was initiated in October 2016. For this longitudinal study, frequent ongoing measurements of dietary consumption, health measures, and medical history will allow us to assess health and disease trends and to determine the effectiveness of the online program in motivating participants to adopt and/or continue the Nutritarian protocol. Combining the elements of clinical effectiveness and implementation can provide for more effective implementation strategies and rapid translational gains for clinical and public health practice. Judicious use of the proposed hybrid designs could speed the translation of research findings into routine practice.^24,37