Quantifying parent engagement in the randomized Fuel for Fun impact study identified design considerations and BMI...

Parent participation in children's health interventions is insufficiently defined and measured. This project quantified parent participation to enable future examination with outcomes in an intervention focused on 4th graders, aged 9-11 years, and their families living in northern Colorado.

Indices were developed to measure type (Parent Participation Profile; PPP) and intensity (Parent Engagement Intensity; PEI) of engagement in Fuel for Fun (FFF), an asymmetric school-and family-based intervention for 4th graders. Study arm-specific participation opportunities were catalogued and summed to calculate the PPP. An algorithm considered frequency, effort, convenience, and invasiveness of each activity to calculate PEI. Indices were standardized (0-100%) using study arm-specific divisors to address asymmetric engagement opportunities. Parents who completed ≥75% of the PPP were defined as Positive Deviants. Youth height and weight were measured. Youth BMI percentile change was compared with parent Positive Deviant status using general linear modeling with repeated measures that included the participation indices.

Of 1435 youth, 777 (54%) had parent participation in at least one activity. Standardized means were 41.5 ± 25.4% for PPP and 27.6 ± 20.9% for PEI. Demographics, behaviors or baseline FFF outcomes did not differ between the Positive Deviant parent (n = 105) and non-Positive Deviant parents (n = 672); but more Positive Deviant parents followed an indulgent feeding style (p = 0.015). Standardized intensity was greater for Positive Deviant parents; 66.9 ± 20.6% vs 21.5 ± 12.7% (p < 0.001) and differences with non-Positive Deviant parents were related to activity type (p ≤0.01 for six of eight activities). Standardized participation intensity was associated with engagement in a greater number of standardized activity types. Among participating parents, standardized intensity and breadth of activity were inversely related to the youth BMI percentile (n = 739; PEI r = -0.39, p < 0.001; PPP r = -0.34, p < 0.001). Parent engagement was not associated with parent BMI change.

An activity-specific intensity schema operationalized measurement of parent engagement in a complex, unbalanced research design and can serve as a template for more sensitive assessment of parent engagement. Positive deviance in parent engagement was not a function of personal, but rather activity characteristics. PPP and PEI increased with fewer requirements and convenient, novel, and personalized activities. Parent engagement indices affirmed lower engagement by parents of overweight/obese youth and concerns about target reach.