Watermelon Focused Dietary Inflammatory Index Intervention

Study Objectives: Chronic inflammation underpins the development of numerous chronic diseases that contribute significantly to disability and mortality in the United States (US). Of particular interest, as it relates to inflammation, is diet which is a strong moderator of systemic inflammation. Diets high in fruit and vegetable intake, fish, and whole grains (e.g., Mediterranean and vegetarian diets) have been associated with lower levels of inflammation. On the other hand, diets rich in red meat, high-fat dairy products, and refined grains (Western diet) are typically associated with higher levels of inflammation. Up until 2014, no dietary index has been designed specifically to determine the inflammatory potential of individual's diet. However, researchers in the Cancer Prevention and Control Program, University of South Carolina developed the Dietary Inflammatory IndexTM (DII). The DII quantifies the level of dietary inflammatory potential on a scale from maximally anti- to pro-inflammatory. Watermelon is one of the fruits whose components have been shown to have anti-inflammatory effects. The overall objective of this proposal is to determine if a refined and modified watermelon-focused DII counselling system can reduce levels of chronic inflammation. The DII is a validated measure that has been shown to predict a range of outcomes including inflammation, cancer, CVD, and mortality, among others.

Specifically, the aims of this proposal are to:

1. Refine and modify the DII-based intervention by developing and incorporating numerous watermelon recipes which will score low (anti-inflammatory) on the DII scale.

2. Design and implement a two-arm intervention trial based on the refined watermelon-focused DII-based program for reducing levels of chronic inflammation. Each arm will have 15 participants (plus an additional 15 partners recruited by the intervention-arm participants). The intervention arm will receive the modified DII intervention whereas the control arm will receive general health education.

Justification: Include summary of prior work and expertise related to the topic and references cited: Until 2014 there had been no scientifically valid way to relate what individuals ate to the capacity of foods consumed to modulate inflammation. In response to the absence of an instrument that could summarize diets' ability to influence inflammatory processes, investigators developed the DII which is based on literature (nearly 2,000 peer-reviewed journal articles) on diet and inflammation through 2010.20 The DII categorizes individuals' diets according to their inflammatory potential. Additionally, it was found to predict changes in hs-CRP in the Seasonal Variation in Blood Cholesterol (SEASONS) Study. Subsequently, investigators have used the DII in about 100 datasets from around the world to test the effect of diet-associated inflammation on: 1) Inflammation markers such as CRP, IL-6, and TNF-alpha-R2; 2) CVD, 3) cancers of the breast, colon, prostate, ovary, endometrium, liver and pancreas; 3) all-cause, CVD and cancer-specific mortality; 4) respiratory conditions such as asthma; and 6) cognitive disorders. To date investigators have published more than 80 peer-reviewed DII publications.

Study Design Overview: Investigators have extensive experience with diet interventions spanning over two decades and involving a variety of endpoints including cancer, body weight, and those related to CVD, inflammation, and immune response. Intervention venues have ranged from the community (including worksites) and churches, research institute-based,to clinical practice sites. This study will include two study arms (i.e., intervention and control). Participants will be recruited from the community. Our goal is to recruit 15 participants per arm, in addition to enrolling partners (e.g., spouse, child, sibling) of intervention participants (n total = 45, n intervention = 30, n control = 15). Participants will undergo a baseline clinic, followed by twelve weeks of intervention classes, and then another clinic. The main intervention classes will take place at a state-of the-art demonstration kitchen located in the Cancer Prevention and Control Program at the University of South Carolina's Arnold School of Public Health.

Study Participants, Recruitment, and Randomization: Participants will be recruited from greater Columbia, South Carolina via websites, fliers, newspaper ads, and listserv messages. Eligible participants will: 1) be ≥18 years of age; 2) have no serious, unstable co-morbidity that would make participation in a diet and PA intervention difficult or risky; 3) be willing and able to participate fully in the study for a period of three months, 4) have Columbia, SC area residence and be able to travel to and from the clinic and intervention sites; 5) have access to the Internet; and 6) not be currently enrolled in a weight loss study or actively taking weight loss medications. However, it should be noted that to see a reduction in inflammation levels individuals need to have "room from improvement". Because it is costly to screen all participants for inflammation (the primary outcome of this study) prior to enrollment, a body mass index (BMI, kg/m2) of 25.0-49.9 will be used as an eligibility criterion, as BMI is strongly associated with inflammation.

Anyone with one or more of the following criteria will be ineligible to participate in this study: 1) taking any medications that are known to influence CRP levels such as systemic steroids (e.g., prednisone); 2) diagnoses of congestive heart failure, chronic renal failure, chronic liver disease (including alcoholic cirrhosis), cancer within the past year (except for non-melanoma skin cancer); 3) have had any major surgery in the past 3 months; 4) life expectancy <3 years, on hospice, nursing home or other institutionalized care; or 6) currently pregnant or planning on becoming pregnant in the next year.

Recent attention to behavioral interventions has pointed out limitations in traditional randomization schemes among these types of randomized controlled trials (e.g., inability to truly blind participants, lack of decision making by participants). Other methods of allocation, which take into account perceived ability to comply with intervention requirements, are much more appropriate for this kind of intervention. Therefore, investigators propose that only individuals who are motivated to make a healthy diet change will receive the intervention.

Intervention Design: It should be noted that the cognitive and behavioral skills that will be incorporated into the intervention are based largely on social cognitive theory and the transtheoretical model. These include self-regulation (self-monitoring, setting goals, problem solving, and reinforcement), enhancing behavioral competence, increasing self-efficacy (through performance accomplishment, persuasion or encouragement, role modeling, and making the experience pleasant), enhancing social support, stress management, time management, relapse prevention, and incorporating changes into everyday life.

The two arms of this intervention study will be the DII-Based Counseling System, including access to our Imagine Healthy Online Portal, and a general health education control. Group-based behavior change programs provide a way for members to share successes and obtain and provide support to one another. Both arms will be composed of 1 wave of 15 individuals. Participants will be required to invite a partner of their choosing to join them in the intervention activities to improve social support for behavioral change. Investigators have experience conducting studies where participants invite partners to participate. Both arms will undergo clinics for data collection at baseline and immediate post-intervention (i.e., at 12 weeks in both groups). Measurements include lipids (very low density lipoprotein [VLDL], low density lipoprotein [LDL], high density lipoprotein [HDL], total cholesterol, and triglycerides), insulin, and high-sensitivity c-reactive protein (hs-CRP). Anthropometric measurements that will be obtained including height and weight (from which investigators can compute BMI), and waist and hip circumferences.

The standard of care arm will receive general health information (e.g., general guidelines of healthy eating and physical activity, stress management, cancer screening, etc.). This information will be provided through email, which will include a weekly newsletter, healthy recipes that are not focused on reducing the DII scores, links to online content, and health-related event announcements in and around Columbia, SC. Any participants in the standard of care arm who does not use email regularly may request for items to be mailed to their mailing address.

The DII-based counseling group will utilize our 12-week IMAGINE Counseling System curriculum that was developed as part of our previous SBIR award (R44DK103377).The DII-based counseling group will meet once a week for 12 weeks to cook, and to engage in exercise and stress-reduction activities. Along with the initial clinic, participants will meet one-on-one for a nutrition counseling session with a registered dietitian. Participants will have access to online material through our Imagine Healthy Online Portal and will be given take-home activities, referred to as IMAGINE actions. These homework assignments will focus on goal setting for nutrition, physical activity, and stress reduction. The nutrition components of this intervention will be based on the DII and will focus on anti-inflammatory foods and components. Some more specific aspects of the diet include increasing consumption of whole foods (fruits and vegetables generally, legumes, and whole grain intake) and decreasing fat (especially saturated and trans fats) that are known to influence inflammation. Examples of strongly anti-inflammatory foods include dark green leafy vegetables, berries, spices and herbs, and sweet potatoes, among others. Watermelon is an anti-inflammatory food (i.e., has a low DII score) and will be incorporated into our nutrition component.

Weakly meal plans will be designed to provide participants with 2 meals per day that incorporate watermelon. In each weekly class session, a recipe with watermelon as an ingredient, will be demonstrated and prepared. These hands-on cooking demonstrations will take place so that individuals learn how to prepare healthy, yet tasty, dishes. As is our practice with our other dietary interventions, investigators will focus the experiential part on sensations of taste and smell as well as tactile sensations, including things such as cutting and washing vegetables. Investigators will rely on foods that are familiar to South Carolinians and are available in local markets while exploring different methods of cooking that incorporate spices from other parts of the world. Investigators have had extensive experience with this type of design of dietary intervention. At the end of each session participants will receive 1 whole Watermelon and 1-gallon bag of precut watermelon per week during class. For continued reinforcement, participants in the DII-based counseling group will have access to our Imagine Healthy Online Portal. The online portal will provide weekly blog posts, a recipe database of 84 recipes that will incorporate watermelon into their meal plans, interactive modules, and short vignettes of cooking tutorials. Information will pertain to nutrition, physical activity, stress reduction, and making meaningful behavior change. Some information will remain on the website from the beginning of the intervention and additional information will be added weekly. Notifications will be sent out to participants when new information is added. Additionally, intervention arm participants will receive a weekly newsletter to their email.

The physical activity targets for this intervention will be consistent with physical activity recommendations for adults. Participants will be encouraged to obtain at least 150 minutes of moderate-intensity physical activity (equivalent to brisk walking) per week and to engage in at least two sessions of strength training each week. To ensure safety, participants will be encouraged to progress very gradually in increasing the intensity and duration of physical activity from their baseline. For stress reduction, several techniques will be employed including breathing exercises and mindfulness-based stress reduction (MBSR)-like activities.

Measurements: The primary exposure of interest will be the intervention arm status. However, a focus of IMAGINE is to reduce sources of dietary inflammation. The DII is comprised of various micro and macronutrients, as well as several individual food items, collectively known as food parameters (more negative scores represent more anti-inflammatory diets and more positive scores represent more pro-inflammatory diets). Participants will report their diet using the Automated Self-administered 24-hour Dietary Recall (ASA24™).The ASA24TM was created by the National Cancer Institute and includes a web-site for participant reporting and a corresponding web-site for researchers to obtain the dietary information (http://epi.grants.cancer.gov/asa24/). The DII will be calculated from ASA24TM data. Previously, investigators have calculated the DII from this tool and found associations with CRP.These self-assessments will take place at each data collection time point (i.e., baseline and 12 weeks) where participants will report their dietary intake for two days on the weekend and two during the week. The entire process of developing the DII is described elsewhere.9 In short, nearly 2,000 research articles, published between 1950 and 2010, examining the relationship between 45 different food parameters (mostly micro and macronutrients) and inflammation were reviewed to derive inflammatory effect scores. Additionally, DII calculation is linked to a regionally representative world database. The world database contains standard means and deviations for the 45 food parameters from 11 populations around the world.9 A z-score is created for each food parameter by subtracting the world standard mean from the participants estimated intake, then dividing this by its standard deviation. This is then converted to a percentile and centered by doubling the value and subtracting 1. The product of the literature-derived inflammatory effect score and the centered percentile for each food parameter is summed across all food parameters to create the overall DII score. Higher (i.e., more positive) scores indicate more pro-inflammatory diets and negative values are more anti-inflammatory.9 To control for the effect of total energy intake, the DII was calculated per 1,000 calories of food consumed.

The primary outcome of interest will include the inflammatory marker hs-CRP. All blood samples will be drawn from an antecubital vein or dorsal hand vein by a trained phlebotomist while subjects are in the seated position. Blood will be collected using two serum separator tubes (SSTs) at a capacity of 8 milliliters and one whole blood EDTA tube. The SSTs will be processed and as many one milliliter aliquots (about 5) as possible of serum will be stored for CRP, lipid, and insulin characterization. The whole blood will be for HbA1c%. All blood samples will be collected in the morning after an overnight fast and information on fasting and exact time of blood draw will be obtained. An acute-phase protein, CRP is a serological marker of inflammation that is produced primarily by the liver in response to stimulation by IL-6. CRP correlates with the magnitude and severity of inflammation. CRP has been associated with the pathogenesis of several chronic diseases, including type 2 diabetes, cancer, and CVD. It is one of the most commonly utilized markers of inflammation because of its chemical stability, long half-life, and relative lack of diurnal variation.Total plasma concentrations of inflammatory cytokines will be determined using quantitative sandwich ELISA kits provided by R&D Systems, Inc. (Minneapolis, MN) for c-reactive protein (Catalog #DCRP00); sensitivity = 0.022ng/mL. Samples and provided standards will be analyzed in duplicate, according to manufacturer's instructions. Inter-assay and intra-assay variability will be assessed.

Secondary outcomes will include HbA1c%, insulin, and lipids. HbA1c will be measured in whole blood collected in the EDTA tube using a Siemens/Bayer DCA 2000+ Analyzer and Test Kit DCA™ (SEMDIA-10311134), which efficiently performs hemoglobin A1c tests in minutes. Lipid analysis will include very low density lipoprotein (VLDL), low density lipoprotein (LDL), high density lipoprotein (HDL), total cholesterol, and triglycerides using a standardized protocol. Lastly, insulin concentrations will be determined using quantitative sandwich ELISA kits provided by Mercodia, INC (Winston Salem, NC, Catalog #10-1113-01); sensitivity = 1.0 mU/L.

Actigraph GT3X-BT based PA and sleep: The GT3x-BT produced by Actigraph is lightweight (i.e., slightly less than 1 ounce) and is about the size of a wristwatch. Participants will be asked to wear the GT3X-BT for 24 hours on their wrist, except during water-based activities, for 10 days around the clinic dates. Minimum wear time requirements will be 7 days (including Saturday and Sunday) with at least 20 hours/day. During periods the GT3x-BT is not worn, participants will report their activities in a log. Energy expenditure during non-wear time will be estimated based on the Compendium of Physical Activities.168 The following PA parameters are provided using propriety algorithms by the ActiLife 6 software for physical activity: activity counts, energy expenditure, metabolic equivalents (METs), steps, physical activity intensity, activity bouts, and sedentary bouts. The Actigraph GT3X products have been validated in numerous studies. Sleep also will be characterized by the GT3X and has been shown to be valid for sleep when worn on the wrist. Specifically, sleep duration, efficiency, onset (i.e., bedtime), and offset (i.e., wake time) will be used.

Anthropometric measurements taken during each of the two clinic visits will include height (cm), body weight (kg), and body fat percentage will be assessed according to standard procedures using a wall-mounted stadiometer (Model S100, Ayerton Corp., Prior lake, MN) and electronic scale (Healthometer® model 500 KL, McCook, IL) in bare feet. BMI (kg/m2) will be calculated from measured weight and height. Waist circumference (cm) and hip circumference (cm) will be measured with a calibrated, spring-loaded tape measure by trained technicians. Blood pressure will be measured according to standard procedures using a stethoscope and manual sphygmomanometer after the participant has been sitting for 5 minutes.

Covariates measured will include data on basic demographics, income, education, smoking, other health behaviors and current medications. Investigators also will measure biases known as social desirability and social approval using the 33-item Marlowe-Crowne Social Desirability (MCSD) Scale and the 20-item Martin-Larson Approval Motivation (MLAM) scale. Investigators will measure the effect of self-efficacy (confidence in the ability to change across problem situations) beliefs about diet and PA using adaptations of previously validated questionnaires. Depression and stress also will be measured using the Center for Epidemiologic Studies Depression (scale) and the Perceived Stress Scale, respectively.

Statistical Analyses: All analyses will be performed using SAS® version 9.4 (Cary, NC). Differences in descriptive covariates between study arms will be tested using chi-square tests for categorical covariates and t-test (normally distributed) or Wilcoxon rank sums tests (non-normal) for continuous measures. Investigators will employ linear mixed models (LMM) to determine whether the intervention arms are associated with the primary outcome (i.e., CRP). Specifically, outcome values (e.g., CRP) at time point 2 (i.e., 12 weeks) will be modelled as the outcome. Intervention status, additional covariates, and the outcome measure at baseline will be on the right side of the equation. Confounder selections will begin with a series of multivariable analyses (i.e., time point two outcome = exposure + covariate + baseline outcome). If a covariate has a p-value of ≤0.20, it will be added to a full model. Confounders will include those that lead to a 10% change in the beta coefficient of intervention status when the confounder is removed; significant covariates (p≤0.05) also will be included in final models. The residuals for all final models will be examined for their adherence to the assumptions of linear regression. If the assumptions are not met, other analytical approaches can be taken. For example, quantile regression is an analytic technique that allows for an assessment of associations throughout the distribution of the outcome of interest.Additionally, CRP has a standard cut-point of 3.0mg/L for clinical reference. Therefore logistic regression can be used to obtain odds ratios [OR] and 95%CIs for the different intervention arms based on this cut-point.