A Cluster-Randomized Control Trial Targeting Parents of Pediatric Cancer Survivors with Obesity: Rationale and Study Protocol of NOURISH-T+

Abstract

Approximately 40–50% of pediatric cancer survivors (PCS) are overweight or obese; increasing their risk for metabolic syndrome and other negative long-term physical health complications. Using our successful pilot trial testing the preliminary feasibility and efficacy of NOURISH for Healthy Transitions (NOURISH-T), we refined our intervention, now NOURISH-T+, and will implement these refinements in this larger, multi-site randomized control trial. Parents of PCS with overweight/obesity (BMI ≥85^th%ile), age 5–12, ≥6 months off treatment are randomly assigned to the NOURISH-T+ intervention or Enhanced Usual Care (EUC) comparison. Parents in NOURISH-T+ will participate in a 6-session, manualized intervention, with an additional dietician session and 2 PCS sessions, as well as post-intervention booster sessions. EUC consists of a one-time informational session, nationally available brochures and follow-up check-ins. Both study conditions will be conducted remotely via a videoconferencing platform. Parents and PCS will be assessed on anthropometric measures, physical activity (PA) and dietary behaviors at baseline, 3-, 6-, and 12-months post-intervention. We will enroll a diverse group of 260 parents/PCS dyads from four pediatric oncology clinics with the aim of evaluating the efficacy of our intervention across diverse pediatric oncology clinics. Our main aim is to compare the impact of NOURISH-T+ with EUC on PCS BMI z-score. Secondary aims are to compare intervention impact on PCS PA and eating behaviors and parent BMI and behaviors as well as to explore potential moderators of the intervention. Our longer-term goal is to establish a framework for future translation and dissemination of NOURISH-T+.

1. Introduction

There is an unmet need for interventions addressing overweight/obesity in pediatric cancer survivors (PCS). Estimates of overweight/obesity in PCS are high, ranging from 40–50%, yet few available programs target this high-risk population [1–7]. PCS are at elevated risk of metabolic syndrome, diabetes and cardiovascular diseases [8–12], conditions exacerbated by overweight and obesity. PCS with overweight/obesity are at significantly greater health risk than children with overweight/obesity in the general population [13, 14]. By the time PCS reach young adulthood, they are 10.8 times more likely to have cardiovascular disease (CVD) than healthy siblings; and by middle age, they have a significantly increased risk of a major cardiac event [4, 15].

Healthy eating and physical activity (PA) behaviors in PCS with overweight/obesity play an important preventive function, reducing the risk of adult obesity and other chronic diseases [16]. However, barriers to prevention often arise during treatment when healthy eating and PA dramatically decline, and families report great difficulty reversing these behaviors after cancer treatment is completed [17, 18]. Although treatment related factors contribute [13, 14], evidence suggests that modifiable behaviors, such as sedentary activity and unhealthy dietary intake are key to post-treatment PCS weight gain [17, 19–23]. Targeted interventions aimed at improving PA and eating behaviors in PCS can lower their risk of negative late effects, thereby reducing chronic illness risk over the lifespan [10, 17, 20–24].

Our prior work provides strong evidence that an intervention targeting parents successfully effects change in otherwise healthy children with overweight/obesity. We tailored our intervention, NOURISH (Nourishing Our Understanding of Role modeling to Improve Support and Health) [25–27], to address the specific needs of PCS and their families. Our intervention for PCS, NOURISH for Healthy Transitions (NOURISH-T), targets parents as change agents to promote healthy eating and PA in PCS [27]. The feasibility and acceptability of NOURISH-T was tested in 2 pediatric oncology clinics and yielded positive change among both PCS and their parents [27]. Compared with an Enhanced Usual Care (EUC) group, PCS with a parent in NOURISH-T showed significant decreases in BMI z-score and waist-to-hip ratio (WHR) and increases in PA. NOURISH-T parents also showed significant decreases in BMI, WHR, and daily caloric intake, and significant changes in family eating practices [27]. Based on these strong preliminary results, we are conducting a fully powered randomized controlled trial (RCT), with participants recruited from four pediatric oncology clinics.

1.1. Aims

We aim to determine whether our intervention, with several enhancements to improve parental engagement and adherence (now called NOURISH-T+) is effective in reducing obesity in a highly diverse group of PCS, compared with an EUC condition. Specifically, the primary aim of this study is to evaluate the impact of NOURISH-T+ on PCS’s BMI z-score. The secondary aim is to evaluate the impact of NOURISH-T+ on PCS’s WHR, PA and eating behaviors as well as the impact of NOURISH-T+ on parents’ outcomes, including BMI, WHR, PA and eating behaviors, and child-family eating and PA practices. We also have an exploratory aim to consider potential moderators of the intervention and examine dyadic relations between parents and children.

Overview of design and methods

A cluster randomized, controlled, repeated measures trial will be employed across four pediatric oncology clinics. Parents of PCS (ages 5–12) with overweight and obesity [28], off treatment for ≥6 months, will be allocated randomly to either NOURISH-T+ (n=130) or an Enhanced Usual Care (EUC) comparison group (n=130), stratified by clinic (see Figure 1 for trial design). We will use a permuted block approach to assure successful randomization.

Fig. 1.

Randomized Control Trial Design.

Delivery of the NOURISH-T+ and EUC intervention will be conducted via a web-based videoconferencing platform. This delivery method was requested by parents in our pilot work, who noted that the frequent medical travel required during treatment was burdensome, and a virtual option was much more appealing in this phase of their children’s lives [29, 30]. A centralized data management and bilingual intervention team separate to the clinics will facilitate both conditions for all participants. Outcome measures will be completed at baseline, 3-, 6-, and 12-months post-intervention.

1.2. Changes to the protocol due to COVID-19

In the initial protocol design, we proposed that clinic staff at each study site would: (1) identify patients and their families meeting eligibility criteria; (2) obtain consent and assent in accordance with IRB guidelines, (3) collect anthropometric measurements (BMI and WHR) in the clinic and, (4) distribute accelerometers for measuring PA to participants. However, the COVID-19 crisis has required us to make some changes to our original protocol. It was no longer possible for clinical research staff to conduct enrollment and measures in-person with survivors and their families. Consequently, we therefore adjusted our format. Clinic staff will continue to identify families meeting eligibility through clinician referral/medical record review and obtain informed consent/assent virtually. Our centralized coordinating and intervention site team will conduct all assessments via HIPAA compliant Zoom, including anthropometric assessments and PA and dietary assessments (See Section 4 – Assessment Design and Outcome Measures). Participants will receive Zoom-based training regarding procedures for height, weight, and dietary measurement. In addition to completing all assessments at baseline, 3 months, 6 months and 12 months follow-up to the intervention, parents (both EUC and NOURISH-T+) will also complete exit surveys at the end of the intervention period (after the 6-main NOURISH-T+ sessions are completed) to assess satisfaction with the intervention and their opinion of its impact on their eating and PA behaviors.

2.2. Participants

2.2.1. Inclusion/Exclusion Criteria

Eligible PCS must be: (1) 5–12 years of age at enrollment; (2) off active treatment for ≥ 6 months; (3) at or above the 85^th BMI%ile; (4) able to complete assessments with the help of their parent or a research assistant (RA); (5) residing with the participating parent; (6) able to engage in PA tailored to current medical status; (7) not taking medications that may affect body weight (e.g., steroids) within 6 months of enrollment, and (8) in remission. PCS who experience a relapse of cancer during the intervention will be excused from further involvement. Eligible parents must be either biological, adoptive, or stepparents as well as permanent legal guardians of the PCS. They must also be 18 years or older and self-identify as the main meal preparer at home. Both parents may participate, however, only the primary parent participant (i.e., the main meal preparer) completes the consent and assessments. No families in our pilot [27] had both parents participate beyond the first session (participation will be tracked in NOURISH-T+). Parents are ineligible if they are non-ambulatory and/or do not reside with the PCS at least 50% of the time. Eligibility criteria are based on prior intervention studies [25–27, 31–33]. Both parent and PCS must meet eligibility criteria for the dyad to be enrolled. Only English- or Spanish-speaking parents and children are eligible for participation.

Our age range criteria is the same used in our prior work [25–27, 31] and our time off active treatment criteria is consistent with the point at which families expressed most interest in our NOURISH-T pilot [27]. These criteria also maximize the number of PCS and families eligible for recruitment. Because NOURISH-T+ is delivered at the level of individual families, information is tailored to each PCS’ cognitive and behavioral developmental stage, scaffolding our intervention strategies as a function of age and developmental capacity. We consider both age and time off treatment as covariates in our analyses.

2.2.2. Diversity

We anticipate recruiting a diverse population of participants from the four participating oncology clinics (overall, ~29% Latinx and ~10% African Americans). Two of our study sites have an especially large clinic population of Latinx patients (>60%); we will provide sessions in Spanish using bilingual facilitators as needed. Race/ethnicity will be considered as potential covariates and moderators in the final analyses.

2.2.3. Sample Size and Power

To determine the sample necessary to address our aims, we estimated recruitment rates based on the prevalence of PCS and power to detect a clinically relevant BMI difference between the NOURISH-T+ and the EUC groups. Previous work indicates that we may experience an overall attrition rate of 15% over the 12-month course of the study. (Note: we track attrition due to relapse separately from our overall attrition estimation). We therefore tested our models assuming an average linear attrition rate of 1.25% of participants per month, which yields approximately 250, 240, and 221 participants at 3-, 6-, and 12-months. With Full Information Maximum Likelihood (FIML) estimation, the effect of this attrition rate was negligible, and the Minimum Detectable Effect Size (MDES) were identical in both cases to the hundredths place. The base analysis strategy is latent growth curve modeling, and the two primary outcomes are a difference in slopes between intervention groups and the correlation between slopes of parents and children. Therefore, we powered the study to these two analytic goals (see Statistical Analyses section below for full details). With this sample size and power ≥ .80, the MDES for group comparison of fixed slopes was .35, and the minimum detectable correlation was .18. To arrive at these estimates, we first conducted chart reviews of our sites, considering clinic size, and using a conservative estimated rate of overweight/obesity among PCS (37.5%), we estimated our ability to recruit a total sample of 260 dyads, or 130 per condition. We enrolled 41 PCS in one of our pilot sites over 20 months [27] with a retention rate from post-assessment to 4 months of .85. This rate is consistent with attrition in other pediatric intervention studies that generally range 10–15% annually [34, 35].

We therefore are confident that our target sample size is attainable, given our more efficient recruitment methods, including booster and check-in sessions, increased contacts, and incentives. A sample size of 260 offers adequate power to detect a clinically relevant difference in the primary child outcome as well as each secondary PCS and parent outcome, even after accounting for the potential of 15% attrition.

2.2.4. Recruiting and Screening Eligible Participants

Briefly, staff at each clinic site are responsible for identifying and enrolling PCS and their parents into the study in accordance with eligibility criteria. Letters describing the study will be sent to eligible parents in accordance with NIH’s single IRB policy regulations. Eligible parents will be informed of the study during outpatient appointments and via phone by clinic staff. Other mediums including, but not limited to, email, text, websites, and social media platforms may be used specific to site/institutional accepted recruitment methods to augment outreach efforts using IRB approved language. Clinic staff will obtain virtual consent and assent. This consent/assent includes permission from participants to be contacted by the centralized-based intervention and research team. The centralized research team will then make arrangements to contact parents and PCS to go over study details, including how we will conduct the baseline assessment session via Zoom. Participants will also be informed that, prior to the baseline session, they will first receive in the mail the scales and tape measures needed to obtain anthropometric assessments and the accelerometers. We then will set up a baseline assessment session for the following week and provide instructions via Zoom to obtain parents’ and child’s height, weight and WHR. Detailed instructions on how the parent and child must wear the accelerometers for one week will then be provided at this baseline assessment session, including the need to mail the accelerometers back to our team in a self-addressed envelope that is provided to all participants. Detailed instructions for completion of the dietary logs to optimize data quality for input into the computer-based ASA-24 will also be included in the assessment session (and during the follow-up assessment sessions at 3-, 6-, and 12-months). Our research team will be making separate appointments to go over the ASA24 log with both the parent and the child to ensure that correct information is provided (See Section 4 – Assessment Design and Outcome Measures). A link to the self-report surveys (via REDCap – Research Electronic Data Capture) will also be sent to participants by the centralized research team. This team will follow-up with families to troubleshoot any technological difficulties/questions.

Families will then be assigned to NOURISH-T+ or EUC, consistent with randomization procedures. Once all the baseline assessments are conducted, the centralized staff will then contact families, arrange to assist them with completing measures, returning accelerometers and scheduling intervention sessions for either NOURISH-T+ or EUC. Similarly, follow-up assessments will be arranged in accordance with the protocol timeline.

2.2.5. Retention

Strategies effective in past work will be used to enhance retention. Briefly, to maximize retention, reminders, monetary incentives, and regular phone, texting and email contact will be used [26, 27]. In our NOURISH-T pilot [27], families perceived our project as highly relevant to them. Once enrolled, retention was 85% from post intervention to follow-up assessment at our pilot study clinic and families were likely to follow-through with all components of the project; we therefore expect high retention in this study. However, because of the unique challenges of research with PCS, relapse or unexpected health changes can occur, making some attrition due to relapse possible (this will be tracked as it was in our pilot; [27]).

2.2.6. Incentives

We will use a graduated incentive structure to optimize retention, as study adherence is more difficult as time elapses. PCS and their families will be compensated for completing study assessments, not the intervention. They will also be permitted to retain the anthropometric measurement tools supplied at baseline.

2. Interventions

2.1. NOURISH-T+ Intervention

NOURISH-T+ consists of 6 weekly, 1–1.5-hour, manualized sessions delivered using videoconferencing via Zoom (manuals available in English and Spanish). In each session, the interventionist reviews the prior week’s homework, introduces session content, and discusses weekly PA and eating goals and behavioral strategies. Reminders of upcoming sessions and program information will also be sent to parents via email or text, depending on parental preference. There will be 2 additional brief sessions in which the PCS will participate along with his/her parent to promote engagement of the entire family in the program. These sessions were added based on feedback received in our prior work [27]. In our pilot, parents expressed that they wanted children to learn both from the interventionists, as well as from them, to enhance children’s enthusiasm for making eating and PA-related changes. Using age-appropriate strategies throughout (e.g., using activity games and more pictures and coloring with younger PCS), PCS will be encouraged to discuss their preferred PA and foods, dietary and activity changes they are making, and their own health barriers and goals. PCS participation is expected to vary by developmental capacity and will be carefully tracked. To augment the efficacy of our intervention, we added an individualized dietary session (led by a pediatric oncology dietician) to NOURISH-T+. This session, which will focus on personalized nutritional strategies, occurs in addition to the 6 sessions of NOURISH-T+, and will be scheduled at a time convenient for families. Brief action planning booster sessions at 2-, 4-, 8-, and 10- months that center on improving self-efficacy and reinforcing positive eating and PA behaviors will maximize retention and follow-up participation. We will arrange for families who lack access to a computer, tablet, or smart phone with wireless connection to use a project tablet with a cellular plan for the course of the intervention. Informed consent procedures will include information about data security and confidentiality (including limits regarding mandated reporting), in addition to other detailed disclosures about the trial.

2.1.1. Session Content

Table 1 provides brief descriptions of the content of each session. The Triadic Model of Behavioral Change, which considers cultural, social, and individual influences, forms the foundation of NOURISH-T+ [36]. Our interventions are designed with particular sensitivity to families from diverse backgrounds. Personal/cultural factors include discussions of food choices and barriers to PA (e.g., modifying traditional meals with healthy adaptations) [37], interwoven throughout the entire intervention. Consistent with results of our prior work [27, 38], we expect that the influence of culture on families’ eating and PA preferences will be particularly important topics when working with our diverse sample, which will include Latinx and African American families. Because participatory experiences enhance overall intervention efficacy [39], they will be incorporated throughout, including self-assessments and experiential activities. Interventionists will encourage the development of individualized goals to reinforce healthy behaviors. Finally, the critical importance of healthy lifestyles for PCS will be emphasized, and potential challenges resulting from the cancer experience will be explored. For example, sessions will emphasize the unique challenges that having a child with cancer has had on the family. Discussions on such topics as perceptions of vulnerability and the importance of preventing adult obesity in PCS who are at greater risk for cardiovascular disease and secondary malignancies are incorporated into NOURISH-T+ session content. Because we recognize that our study is being conducted during a pandemic, we now also have added to our sessions discussions about the unique challenges of COVID-19 on families with PCS.

Table 1.

NOURISH-T+ Session Content.

Session	Content
Session 1	• Childhood cancer and overweight/obesity.
Overview, effects of cancer treatment on eating & exercise behaviors	• Barriers to healthy nutrition and physical activity.
	• School and peers after treatment.
	• Parental perceptions of vulnerability.
	• Misconceptions to healthy eating.
	• Transitioning to healthy habits for the whole family.
Session 2	• Parenting styles.
Parenting style and role modeling for getting back to healthy	• Parental role-modeling eating & exercise behaviors.
	• Introducing nutrition and healthy eating.
	• MyPlate and portion size.
	• Beverages and healthy eating.
Session 3	• Nutrition and media awareness.
Dietary intake	• Emotional eating.
	• Hunger scale and STOP technique.
	• Healthy eating behaviors and food home environment.
	• Meal planning.
Session 4	• Physical Activity. Guidelines and barriers.
Reducing sedentary behaviors and overcoming barriers to exercise success by scheduling	• Effects of sedentary lifestyles.
	• Screen time.
	• Alternatives to a sedentary lifestyle.
	• Making a specific physical activity plan.
Session 5	• Definition of mindfulness.
Mindfulness towards a healthy lifestyle: nutrition, physical activity, and sleep	• Mindful eating.
	• High risk eating habits.
	• Mindfulness in physical activity.
	• Sleep and mindfulness.
Session 6	• Getting back to healthy.
	• Self-esteem and body image.
Getting back to Healthy. Dealing with outside influences. Wrap-up	• Peers, peer pressure and teasing.
	• The toxic environment: How can parents fight back?
	• Wrap-up. Coping with setbacks.

NOURISH-T+ content is theory-based, manualized, and builds upon strengths of our prior work with enhancements made based on the experience and findings of conducting the NOURISH-T pilot [27]. NOURISH-T+ now offers multiple means of communication and delivery of content (via videoconferencing, print, and text messaging). During booster sessions, the focus will be on goal progress, obstacles, and concerns, and use of intervention content and materials.

2.2. Enhanced Usual Care

Parents randomized to EUC will participate in one information session on the role of diet and PA in pediatric overweight/obesity. Session content is taken from the publicly available We Can! Manual [40] and was used in our pilot EUC group [27]. As in our prior trial, EUC parents will receive nationally available brochures on pediatric overweight/obesity on two occasions during the six weeks that NOURISH-T+ families participate in the study, so that similar (although less intensive) attention is provided across study arms. At the end of the intervention period (i.e., length of NOURISH-T+), EUC parents will complete the exit survey. We will also use email or text reminders about availability of web-based information, plus check-ins at 2-, 4-, 8- and 10-months to maximize follow-up assessment participation. During check-ins, participants will be asked generally about their progress in goal accomplishment and reminded of their next assessment.

2.3. Treatment Fidelity, Process Measures, & Staff Training

The NOURISH-T+ and EUC intervention manuals assure implementation consistency and fidelity to treatment implementation (FTI). We will monitor and track program implementation to ensure fidelity and enhance internal validity. Fidelity is optimized via the following strategies:

1. All NOURISH-T+ and EUC interventionists will participate in PI-led training. Interventionists will be trained separately and learn either NOURISH-T+ or EUC to ensure project integrity.

2. All NOURISH-T+ and EUC sessions will be videotaped using Zoom’s recording option; 20% will be randomly selected on a monthly basis to be coded using both behavioral counts and global ratings for consistency across sessions utilizing a fidelity checklist. The checklist was developed using the Fidelity of Implementation Rating System [41], and assesses the 5 elements of implementation fidelity (intervention manual adherence, exposure/dosage, delivery quality, participant responsiveness, and program differentiation). Both verbal and non-verbal cues will be rated in the assessment of participant responsiveness.

3. In line with “best practices,” interventionists, including our dietician for the project, will complete a brief (<2 min) fidelity checklist and survey assessing implementation barriers and perceptions about participant engagement.¹⁰² Checklists are a useful way to remind interventionists of the protocol, ensure uniformity of intervention delivery, and identify aberrant events that require review. All possible deviations will be carefully monitored, investigated, recorded, and intensive training efforts will be implemented to prevent any potential recurrence to ensure consistency and standardization in data collection methods across clinic sites.

2.3.1. . Training in Cultural Competence

All staff will be trained in the delivery of culturally sensitive interventions, including readings on the interplay among culture, food, eating, weight, and PA behaviors, and review of scenarios reflecting issues that may arise and ways of managing them [42]. This training is particularly important for our bilingual facilitators working with our Spanish-speaking participants. We anticipate having ~29% of our sample of PCS and their parents indicating a preference for Spanish intervention delivery.

3. Assessment Design and Outcome Measures

At each assessment session (baseline, 3-, 6-, and 12-months post-intervention), parents and PCS will be instructed in how to record anthropometric measurements of BMI and WHR, complete surveys via REDCap, and use an accelerometer for one week. Accelerometers (and instructions) will be mailed to each family prior to the Zoom assessment appointment at each time point by the centralized research team and returned via prepaid envelope. In addition to outcome measures, at the end of the 6-week period, an exit survey will be distributed to both EUC and NOURISH-T+ families.

3.1. Measures

We assess standard anthropometric measures (BMI, WHR), PA and eating behaviors at each time point. All measures have been used in our prior work [26–28, 43, 44], are psychometrically sound, and available in English and Spanish.

3.1.1. Anthropometric Measurements

Measurements include: BMI z-score (PCS based on CDC Growth Charts) [28], BMI (parents) and WHR. Because of the necessity to adapt our protocol to conduct all assessments via Zoom, parents and PCS will be instructed in how to complete these measurements during each assessment session by the centralized research team. To ensure validity of the measurements being reported, we also will ask parents to take pictures of the measurements during the Zoom assessment session. Standardized tools (e.g., scales and measuring tapes) will be provided to each family prior to obtaining these assessments. Weight will be measured on digital scales to the nearest ¼ pound and height will be measured to the nearest ¼ inch. Hip (maximum girth of the hips, above the gluteal fold), and waist (narrowest part of the torso above the umbilicus and below the xiphoid process) circumferences will be assessed using an anthropometric measuring tape, and these data will be used to calculate WHR.

3.1.2. PA Behaviors

PA is assessed for both parents and PCS via waist-worn accelerometers (brand name: ActiGraph) which objectively measure daily PA, including average steps per day, time spent in moderate-vigorous activity, and sedentary time. We provide detailed instructions (written and via videos we developed for the study) for wearing accelerometers and regard data as valid when worn for 4/7 days (e.g., 2 weekend and 2 weekdays). Previous research conducted by our team found high compliance rates for returning accelerometers when detailed instructions were provided to families [44]. Total average time (min/wk) spent in moderate/vigorous activity or mean total sedentary time will be examined. Note: we will keep track of whether time of year affects PA activity. There might be some variability due to the geographic location of our four pediatric oncology clinics (mid-Atlantic vs. southeastern states).

3.1.3. Dietary Behaviors

Dietary intake of parents and PCS will be assessed via the Automated Self-administered 24-Hour Dietary Recall-2020 (ASA24; English & Spanish versions) [45]. ASA24 is a computer assisted 24-hour dietary recall interview with animated guides and audio and visual cues to instruct participants and enhance its user-friendliness in low-literacy populations. The process involves asking respondents to provide a list of foods and drinks consumed at each meal during the previous 24-hours and during meal gaps (snacks). Respondents are then asked for details about the foods and drinks they reported, including preparation methods, portion sizes (using visual images as guides), and where they were purchased. Given our experience with the ASA24 [27], we will instruct parents (verbally and via a PowerPoint video we developed for the study) to complete detailed dietary logs for themselves and their child (if <10 yrs) and take pictures of their meals, based on the information required for the ASA24. Research staff will assist PCS with the logs and will input data to the ASA24 website for all participants. Of interest in our analyses will be intake of total calories, fruits and vegetables, and grams of fat and sugar consumed in one day, as these components of dietary intake align with the focus of NOURISH-T+. The ASA24 is comparable in reliability to more expensive interviewer-recall methods [46].

For reliability and triangulation, PCS (with RA assistance) and parents will complete a behavior questionnaire adapted from the Child/Adults Food and Activity Questionnaire of the national Expanded Food and Nutrition Education Program (EFNEP) [47]. Parents will also complete a 28-item survey of family behaviors, the Family Eating and Exercise Behaviors (FEEB) questionnaire that assesses family home environment and behavioral factors related to eating and exercise habits (e.g., frequency of family meals, screen time during meals, family PA time, eating habits, and availability of snacks/sweets, and sugary beverages in the home) [27, 48]. Parents also report their approach to child-feeding practices via the 31-item Child Feeding Questionnaire (CFQ). The CFQ yields reliable and valid scores [49]. USF research staff will assist with all assessments.

3.1.4. Additional measures, Covariates and Potential Moderators

Our choice of key moderators for our exploratory aim are drawn from theory, our prior feasibility trial [27] (which identified variables related to adherence) [50], and the pertinent oncology literature. These potential moderators include personal and survivorship status variables, (e.g., PCS BMI%ile at diagnosis, time off treatment, PCS age, race/ethnicity, and sex, treatment intensity and late effects, and pubertal status).

4.1.3.a. Demographic Information

Sex, race/ethnicity, cancer and treatment information, age, parental education, income, and insurance status will be collected via questionnaire at baseline using REDCap. Medical information will be verified with clinic medical charts. PCS pubertal status will be completed by parents (using Tanner staging for increased standardization); assessed via REDCap at baseline [51]. In addition to this baseline demographic information, several questions about the impact of COVID-19 on family functioning, stress, and food security are included. These questions derive from several COVID-19 measurement protocols catalogued in the NIH-funded PhenX Toolkit. Specifically, items were adapted from the Specific Distress domain in the Impact of the Pandemic and HRQOL in Cancer Patients and Survivors (IPHCPS) questionnaire [52], the Impacts on Dietary Changes domain in the Pennington Biomedical COVID-19 Survey (PBRC) [53], and the COVID-19 Exposure and Family Impact Survey (CEFIS) [54].

4.1.3.b. Pittsburgh Sleep Quality Index (PSQI)

The self-reported questionnaire consists of nineteen individual items that generate seven “component” scores: subjective sleep quality, sleep latency, sleep duration, habitual sleep efficiency, sleep disturbances, use of sleeping medication, and daytime dysfunction. Parents will help complete this assessment for the child participants. Parents will also complete this assessment for themselves.

4.1.4.c. Intensity of Treatment Rating Scale [55]

PCS medical records will be audited by each pediatric oncology clinic staff to derive a level of treatment intensity from 1 (least intensive) to 4 (most intensive) using e.g., diagnosis, stage or risk level, relapse status, and/or type of treatment modality used (e.g., surgery, chemotherapy or radiation).

4.1.3.d. Rating of Medical Late Effects

Study site clinicians will rate PCS late effects following treatment, from no limitation of activity to significant restriction on daily activity and need for significant medical attention or equipment [56].

4.1.3.e. Exit Survey:

At post-testing, NOURISH-T+ and EUC parents will complete a brief exit survey assessing perceptions of acceptability and satisfaction with the intervention (e.g., likes/dislikes; perceived benefits and barriers to implementing intervention goals; and suggestions for improvement). We will also assess interest in an online social support group to inform possible future research directions. An additional item drafted by the centralized research team was integrated to assess if the increased use of virtual meetings during the COVID pandemic helped participants feel more comfortable with participating in the virtual sessions.

4. Data Management & Security

Briefly, data will be centrally stored in a centralized-based master database. Data will be collected and maintained in accordance with legal and ethical standards. REDCap, a secure web application for research studies in HIPAA compliant environments, will be used to collect all assessment data. REDCap will allow for data to be entered and uploaded to a central, secure database. Families will be emailed a link to the REDCap-based surveys. The centralized research team will download the collected data from RedCap into a secure SPSS file on a monthly basis. Accelerometers will be mailed back to the centralized research team, at which point the team will download and manage the data in a secure database. Zoom was chosen as the virtual platform to deliver the interventions given it is HIPAA compliant and it was recently updated to an Advanced Encryption Standard using a 256-bit algorithm.

5. Statistical Analyses

5.1. Analytical Plan

All variables will be examined for outliers, skewness, missing data, etc., and appropriate procedures will be followed (e.g., non-normal distributions, sensitivity analysis) to address any issues with the data. Because participants will be nested within sites, we will include a fixed effect of site to account for the study site effect [57]. For all outcomes, the first step of the analyses will involve identifying covariates to include in the primary analyses (e.g., age, sex, cancer-related factors) by regressing primary outcomes on these variables and examining effect sizes. These baseline characteristics and survivorship status variables will be included in all subsequent models and examined as potential moderators.

With four time points, the base analysis strategy for these data will be latent growth curve modeling [58]. Time will be coded as number of months since baseline, taking values of 0 for baseline, and then individually varying values of approximately 3, 6, and 12 for the following three time points. The base model will include a fixed intercept and slope, a random intercept and slope, and a random intercept-slope correlation to address whether initial status at baseline is related to growth over time. The correct growth (i.e., linear/quadratic) and covariance structure will be identified using likelihood ratio tests to determine the base growth model. Once the appropriate base growth model and covariance structure has been determined, the intercept and slope factors will be regressed on the intervention group to assess whether PCS in the NOURISH-T+ group change more over time compared to the EUC group. In addition, we will also conduct pairwise mean differences at each follow-up time point to examine the extent of the difference at different time points. Moreover, when assessing outcomes measured on both parents and PCS, we will expand the base model to a multivariate growth curve perspective [59]. The two outcomes will be examined simultaneously in a single model. The test of interest will be the correlation between the random slope terms from the two variables. A correlation indicates that the two processes are dependent, and a difference in the correlation between groups indicates that the intervention has made an impact on the relation between the processes.

Latent growth curve modeling as described above is used for all analyses of our main aims. For our Exploratory Aim, we will explore moderation, multiple mediation, and additional, simultaneous processes of behavioral and anthropometric measures. Given the number of hypothesis tests, we will use the Benjamini-Hochberg [60] correction for False Discovery Rate to control for Type I error.

6. Discussion

6.1. Overview

More than 80% of pediatric cancer patients survive into adulthood but are at an increased risk of weight gain and decreased physical activity. Moreover, obesity rates in PCS are alarmingly high. The negative consequences of obesity that is prevalent among PCS, provide support for the importance of identifying ways to reduce overweight and obesity in this population. Despite the urgent need for interventions to address healthy eating and activity behaviors among PCS, few have been evaluated for this vulnerable group. To address this critical need, we have systematically built upon our prior work and addressed eating and exercise behaviors in PCS by evaluating the efficacy of an intervention focusing on caregivers as agents for change.

The current study extends our prior research which established the feasibility and acceptability of our pilot NOURISH-T intervention [27]. We have made recommended modifications to our current intervention, enhancing the likelihood that this version, NOURISH-T+, will be efficacious. We will test the efficacy of our intervention across four different pediatric oncology clinics, each serving diverse populations. Our findings should provide critical information for future interventions aimed at preventing obesity and related negative health outcomes in PCS and their families.

6.2. Potential benefits

We expect that our findings might ultimately benefit researchers, practitioners, patients and their families to better understand the risks of obesity for young cancer survivors. Specifically, parents should learn ways in which they can help their children develop healthy habits and prevent obesity after cancer treatment has ended. We fully anticipate families benefiting from their involvement in our study in several specific ways.

Families are anticipated to gain knowledge and skills conducive to supporting healthy lifestyles and behavior change. Specifically, families will become more knowledgeable of the influence of cancer treatment and parental behaviors on obesity as well as children’s eating and PA behaviors. They will learn how modifying their family’s behaviors in relation to food and exercise can help prevent weight gain and subsequent obesity for their child who has survived cancer. Families in NOURISH-T+ will also learn individualized behavioral strategies and participate in an individual family session with a dietician that will address barriers to nutritional changes. NOURISH-T+ will increase parents’ awareness of the critical role they play as role models of healthy eating and exercise, which is likely to lead to important family-wide changes in these behaviors. We also expect that our intervention will be an important step towards prevention of adult obesity in PCS whose parents have participated. Thus, participants are expected to leave the intervention with a new set of skills that can help them improve the wellbeing of their child, themselves, and their entire families.

From a broader perspective, this study will provide important empirical data regarding the effectiveness of an intervention with the potential to change pediatric oncology clinical practice. There are no current standards/recommendations for addressing eating and PA behaviors in PCS. The knowledge gained from this trial has the potential to be generalizable to a variety of practice settings and because we use m-health strategies, eliminates many barriers to widespread adoption.

6.3. Challenges and limitations

Although this trial has many strengths and potential benefits, we must, nonetheless, acknowledge its challenges and limitations. Clearly conducting clinical trials with families who have undergone cancer treatment is always challenging. Families often want to forget the experience, move on with their lives and return to their “normal.” Engaging in additional interventions, even those offering promise to improve the health of their child, might be perceived as too much to manage as they transition to survivorship. Related to this challenge is the issue of attrition that plagues all clinical trials. To address these challenges, we have incorporated several strategies to recruitment and retention that should offset these potential limitations (e.g., using booster/check-in follow-up sessions, graduated incentives, and increased PCS involvement). Conducting the study across four different pediatric oncology clinics with diverse populations, will enhance the generalizability of our results. We are particularly interested in targeting Latinx families and conducting our intervention in Spanish, when preferred by the family.

Another challenge that ultimately might impact our findings is the COVID-19 pandemic. As we noted above, we began planning the implementation of our project just prior to the outbreak of the pandemic. The COVID-19 crisis required significant adjustment to our original protocol compelling us to change our protocol design including the use of virtual assessments of key anthropometric measurements. While not ideal, this shift will ensure greater data capture, increase the ecological validity of our project, and may allow broader adoption of the intervention in the future.

Another limitation pertains to the number of sessions in NOURISH-T+. Although more clinical contacts are often recommended in pediatric obesity treatment [61, 62], families in our original NOURISH project [31] reported that our initial version with 12 sessions was too long; given this feedback we reduced NOURISH to 6 sessions. We reduced attrition while still yielding positive outcomes [27]. Participation of patients with cancer in clinical trials is generally low, and recruitment often difficult [63, 64]. Keeping these practical factors and participant burden issues in mind, we kept the number of NOURISH-T+ core sessions to 6. An additional 7^th session, led by our dietician, focuses on individualized goals and we added PCS sessions and booster/check-in contacts (see Figure 1).

Given our experience in NOURISH and NOURISH-T, we believe this approach will increase both protocol engagement and adherence. Our EUC comparison condition receives more than standard of care, but clearly has less involvement than families in NOURISH-T+. However, we have designed the protocol to provide an equivalent number of core contacts between NOURISH-T+ and EUC PCS and their families over the course of the year.

6.4. Summary

This project is the first to evaluate a theory-based, evidence-informed intervention targeting diet and PA in PCS that includes direct parent involvement, uses a rigorous RCT design, includes a relatively large, racially and ethnically diverse sample, and assesses a range of outcomes including anthropometrics, PA and eating behaviors. NOURISH-T+ parents will learn tailored strategies to promote healthy eating and PA behaviors in their PCS over 14-months (from baseline assessment to 12-months follow-up). In addition to evaluating the intervention across a longer time frame, we also aim to demonstrate that NOURISH-T+ can be implemented across pediatric oncology clinics. In sum, this trial will evaluate this intervention program and ultimately produce findings that can easily be translated to pediatric after care (survivorship clinics) oncology clinics so that patients, their parents, and their health care providers might use our intervention strategies in their clinic and thereby benefit from our findings.