Parental Activation and Obesity-Related Health Behaviors Among a Racially and Ethnically Diverse Population of Low-Income Pediatric Patients: Protocol for a Cross-Sectional Survey Study

Background Despite a recent decline in the obesity prevalence among preschool-aged children, obesity remains disproportionately high among children from low-income racial or ethnic minority families. Promoting healthy lifestyles (eg, obesity-preventative behaviors) in primary care settings is particularly important for young children, given the frequency of preventative health visits and parent-provider interactions. Higher adoption of specific health behaviors is correlated with increased patient activation (ie, skill, confidence, and knowledge to manage their health care) among adults. However, no published study, to date, has examined the relationship between parental activation and obesity-related health behaviors among young children. Objective The goal of this study is to measure parental activation in low-income parents of preschoolers in 2 large health systems and to examine the association with diet, screen-time, and physical activity behaviors. Methods We will conduct a cross-sectional study of parents of preschool-aged patients (2-5 years) receiving primary care at multiple clinic sites within 2 large health care systems. Study participants, low-income black, Hispanic, and white parents of preschool-aged patients, are being recruited across both health systems to complete orally administered surveys. Results Recruitment began in December 2017 and is expected to end in May 2018. A total of 267 low-income parents of preschool-aged children have been enrolled across both clinic sites. We are enrolling an additional 33 parents to reach our goal sample size of 300 across both health systems. The data analysis will be completed in June 2018. Conclusions This protocol outlines the first study to fully examine parental activation and its relationship with parent-reported diet, physical activity, and screen-time behaviors among low-income preschool-aged patients. It involves recruitment across 2 geographically distinct areas and resulting from a partnership between researchers at 2 different health systems with multiple clinical sites. This study will provide new knowledge about how parental activation can potentially be incorporated as a strategy to address childhood obesity disparities in primary care settings. International Registered Report Identifier (IRRID) RR1-10.2196/9688

Effective obesity management programs in clinical settings are a key component of multi-sector efforts to eliminate child obesity disparities and reverse the childhood obesity epidemic. The American Academy of Pediatrics recommends integrating obesity screening and healthy weight counseling into pediatric well child visits starting in infancy. 5 Counseling should promote healthful diet, physical activity (PA), and screen time behaviors. For overweight and obese preschoolers (BMI≥85 th percentile), clinicians should also perform in-depth behavioral assessments and targeted counseling addressing sugar sweetened beverage intake, playtime, and feeding practices. This should be paired with continual monitoring of weight trajectories. 5 Healthful diet and PA behaviors, and adherence to physician recommendations have been linked to patient activation among adults. 6,7 Patient activation refers to skill, confidence and knowledge in managing one's health and is commonly measured using the Patient Activation Measure (PAM); a 13-item scale with strong psychometric properties. 8 Research suggests that lower-income adults and racial/ethnic minority adults have lower activation than higher-income adults and White adults respectively. 9,10 Most research in this field focuses on activation among adult patients regarding their own health. In fact, no studies have described parental activation among racial/ethnically and socioeconomically diverse populations of parents of preschoolers across primary care settings. Furthermore, describing the relationship between parent activation and child diet and PA behaviors is critical to inform development of child obesity interventions in primary care.
The goal of the proposed pilot project is to inform development of future obesity management interventions addressing obesity disparities in preschoolers in primary care settings . It will be achieved through two Aims: Aim 1: Measure activation among a clinical sample of socioeconomically and racial/ethnically diverse parents of overweight and obese preschoolers seeking primary care in two large health systems, and describe the sociodemographic factors associated with parental activation, using a cross-sectional survey design. Hypothesis 1: Activation levels will be lower for Black and Latino vs. White and low-vs. high-income parents. Aim 2: Examine the association of parental activation with adherence to healthy feeding, screen time and PA behaviors using a cross-sectional survey design. Hypothesis 2: Parental activation is positively associated with adherence to healthy feeding, screen time and PA behaviors.

Conceptual Model:
The conceptual model for the proposed research ( Fig. 1) depicts the influence of sociodemographic factors on parental activation and the hypothesized relationship between parental activation and adherence to healthy feeding, screen time and PA behaviors. Specifically, the proposed research posits that sociodemographic factors predict parental activation and that this activation directly influences healthful feeding, screen time and PA practices. C . Methods: Study participants will be recruited from JHM and KPSC pediatric primary care clinics. Two pediatric primary care clinics in East Baltimore serving low-income, racial/ethnic minority patients at JHM will participate. KPSC serves a racial/ethnic and socioeconomically diverse patient population, with more than 10% of pediatric patients insured through Medicaid.

Study Population and Data Collection:
We will recruit a convenience sample of parents of preschoolers presenting for care at the JHM and KPSC sites to participate in a survey. Parents whose preferred healthcare language is English or Spanish will be eligible to participate if they are: 1) ≥ 18 years, 2) the parent of an overweight/obese preschooler (BMI ≥ 85 th percentile) with height and weight measured in the past 3 months, 2) White, non-Hispanic Black or Hispanic/Latino by self-report. Preschool patients' EMR data will be linked with parental survey data at both sites. Surveys will be orally administered at JHM. KPSC will identify potential participants through EMR and administer surveys by phone.
Measures: parent self-reported race / ethnicity and parent self-reported socioeconomic status are the independent variables for Aim 1. P arent activation concerning their child ' s health is the main outcome variable for Aim 1. This will be measured using the Parent Patient Activation Measure (P-PAM), a standardized 13-item survey adapted from the well-validated adult patient activation measure. 11 Parental activation (measured by the P-PAM) is the independent variable for Aim 1. Outcome variables assessed in Aim 2 are: parent self -reported adherence to recommended feeding , screen -t ime and PA behaviors. Parent's child feeding behaviors will be measured using questionnaires assessing parental report of feeding: sugar-sweetened beverages, fruits and vegetables, and fast food. The survey will also include items adapted from the Child Food Frequency Questionnaire; the Timing and Frequency of Infant SSB Consumption Questionnaire; and the Preschool-aged Children's Physical Activity, which includes measures of screen time. [12][13][14] Covariates for parents and preschoolers will be included in analyses: 1) parental healthcare language, English language proficiency, race/ethnicity, age, gender, educational attainment, employment status, and selfreported height and weight; and 2) child race/ethnicity, gender, age, health insurance, and BMI percentile. Analysis Plan: Descriptive statistics will be used to summarize parent activation, parent feeding, screen time and PA behaviors. Multivariate regression analyses will be used to test hypotheses. Specifically, we will determine whether there is a statistically significant difference in parent activation by parent race/ethnicity and socioeconomic status (Aim 1) and determine whether parent activation is associated with parent feeding, screen time, and PA behaviors (Aim 2). These analyses will adjust for key covariates. Sample Size and Power Calculations: Aim 1: Borrowing from the adult literature, we estimate a 20% difference in activation levels between racial/ethnic minority and white parents, with similar effect size expected for differences in activation level between low-and high-income parents. 9,10 Assuming a fixed sample size of 300 parents with equal numbers of Black, Latino, and White parents; and equal numbers of low-and highincome parents, we will have 83% and 96% power respectively to detect a statistically significant difference (p<0.05) in parental activation for Latino vs. White parents and low-vs. high-income parents.

D. Prior Research Experience at Proposed Sites:
Drs. Showell, Thornton and DeCamp have a strong track record of collaborative research involving successful recruitment and retention of English and Spanishspeaking low-income parents of preschoolers from the proposed JHM clinics. Drs. Showell and DeCamp have also used the P-PAM with this population. This work has resulted in cultural adaptations of the P-PAM including item modification to address variations in colloquial language among Spanish-speaking parents. Drs. Young and Koebnick have extensive experience using EMR data to link clinical practice and health behavior change among KPSC pediatric patients. They are also highly experienced in participant recruitment. Table 1 provides a timeline for achieving all key deliverables during the award period. If funded, this project will produce an NIH or foundation grant proposal to develop a clinic-based childhood obesity intervention, and 2-4 coauthored manuscripts including one comparing parent activation levels among sociodemographic subgroups, and a second describing the association between parent activation and parent feeding, screen time, and physical activity behaviors.