Caregiver involvement in interventions for improving children's dietary intake and physical activity behaviors.

backgroundPoor diet and insufficient physical activity are major risk factors for non-communicable diseases. Developing healthy diet and physical activity behaviors early in life is important as these behaviors track between childhood and adulthood. Parents and other adult caregivers have important influences on children's health behaviors, but whether their involvement in children's nutrition and physical activity interventions contributes to intervention effectiveness is not known.

objectives• To assess effects of caregiver involvement in interventions for improving children's dietary intake and physical activity behaviors, including those intended to prevent overweight and obesity • To describe intervention content and behavior change techniques employed, drawing from a behavior change technique taxonomy developed and advanced by Abraham, Michie, and colleagues (Abraham 2008; Michie 2011; Michie 2013; Michie 2015) • To identify content and techniques related to reported outcomes when such information was reported in included studies SEARCH

methodsIn January 2019, we searched CENTRAL, MEDLINE, Embase, 11 other databases, and three trials registers. We also searched the references lists of relevant reports and systematic reviews. SELECTION CRITERIA: Randomised controlled trials (RCTs) and quasi-RCTs evaluating the effects of interventions to improve children's dietary intake or physical activity behavior, or both, with children aged 2 to 18 years as active participants and at least one component involving caregivers versus the same interventions but without the caregiver component(s). We excluded interventions meant as treatment or targeting children with pre-existing conditions, as well as caregiver-child units residing in orphanages and school hostel environments. DATA COLLECTION AND ANALYSIS: We used standard methodological procedures outlined by Cochrane. MAIN

resultsWe included 23 trials with approximately 12,192 children in eligible intervention arms. With the exception of two studies, all were conducted in high-income countries, with more than half performed in North America. Most studies were school-based and involved the addition of healthy eating or physical education classes, or both, sometimes in tandem with other changes to the school environment. The specific intervention strategies used were not always reported completely. However, based on available reports, the behavior change techniques used most commonly in the child-only arm were "shaping knowledge," "comparison of behavior," "feedback and monitoring," and "repetition and substitution." In the child + caregiver arm, the strategies used most commonly included additional "shaping knowledge" or "feedback and monitoring" techniques, as well as "social support" and "natural consequences." We considered all trials to be at high risk of bias for at least one design factor. Seven trials did not contribute any data to analyses. The quality of reporting of intervention content varied between studies, and there was limited scope for meta-analysis. Both validated and non-validated instruments were used to measure outcomes of interest. Outcomes measured and reported differed between studies, with 16 studies contributing data to the meta-analyses. About three-quarters of studies reported their funding sources; no studies reported industry funding. We assessed the quality of evidence to be low or very low. Dietary behavior change interventions with a caregiver component versus interventions without a caregiver component Seven studies compared dietary behavior change interventions with and without a caregiver component. At the end of the intervention, we did not detect a difference between intervention arms in children's percentage of total energy intake from saturated fat (mean difference [MD] -0.42%, 95% confidence interval [CI] -1.25 to 0.41, 1 study, n = 207; low-quality evidence) or from sodium intake (MD -0.12 g/d, 95% CI -0.36 to 0.12, 1 study, n = 207; low-quality evidence). No trial in this comparison reported data for children's combined fruit and vegetable intake, sugar-sweetened beverage (SSB) intake, or physical activity levels, nor for adverse effects of interventions. Physical activity interventions with a caregiver component versus interventions without a caregiver component Six studies compared physical activity interventions with and without a caregiver component. At the end of the intervention, we did not detect a difference between intervention arms in children's total physical activity (MD 0.20 min/h, 95% CI -1.19 to 1.59, 1 study, n = 54; low-quality evidence) or moderate to vigorous physical activity (MVPA) (standard mean difference [SMD] 0.04, 95% CI -0.41 to 0.49, 2 studies, n = 80; moderate-quality evidence). No trial in this comparison reported data for percentage of children's total energy intake from saturated fat, sodium intake, fruit and vegetable intake, or SSB intake, nor for adverse effects of interventions. Combined dietary and physical activity interventions with a caregiver component versus interventions without a caregiver component Ten studies compared dietary and physical activity interventions with and without a caregiver component. At the end of the intervention, we detected a small positive impact of a caregiver component on children's SSB intake (SMD -0.28, 95% CI -0.44 to -0.12, 3 studies, n = 651; moderate-quality evidence). We did not detect a difference between intervention arms in children's percentage of total energy intake from saturated fat (MD 0.06%, 95% CI -0.67 to 0.80, 2 studies, n = 216; very low-quality evidence), sodium intake (MD 35.94 mg/d, 95% CI -322.60 to 394.47, 2 studies, n = 315; very low-quality evidence), fruit and vegetable intake (MD 0.38 servings/d, 95% CI -0.51 to 1.27, 1 study, n = 134; very low-quality evidence), total physical activity (MD 1.81 min/d, 95% CI -15.18 to 18.80, 2 studies, n = 573; low-quality evidence), or MVPA (MD -0.05 min/d, 95% CI -18.57 to 18.47, 1 study, n = 622; very low-quality evidence). One trial indicated that no adverse events were reported by study participants but did not provide data. AUTHORS'

conclusionsCurrent evidence is insufficient to support the inclusion of caregiver involvement in interventions to improve children's dietary intake or physical activity behavior, or both. For most outcomes, the quality of the evidence is adversely impacted by the small number of studies with available data, limited effective sample sizes, risk of bias, and imprecision. To establish the value of caregiver involvement, additional studies measuring clinically important outcomes using valid and reliable measures, employing appropriate design and power, and following established reporting guidelines are needed, as is evidence on how such interventions might contribute to health equity.