Improving Health Among Disadvantaged Girls to Slow Pubertal Onset and Reduce Long-term Health Risks

Abundant research shows girls who experience earlier pubertal onset exhibit substantial risk for poor outcomes in multiple areas of psychosocial functioning and physical health. At the same time, abundant research shows prepubertal body mass index (BMI) is a critical determinant of earlier pubertal onset. Integration of these research areas points to the novel focus of the proposed study which seeks to test whether an RCT targeting weight loss and positive health behavior change in the prepubertal period may slow pubertal onset in girls at risk for accelerated pubertal development. At-risk is defined as 1) being overweight/obese and 2) experiencing socioeconomic disadvantage, both factors with strong prospective links to earlier pubertal onset.

Recruitment efforts will target prepubertal girls from disadvantaged families who are overweight/obese (BMI percentile ≥85th) and ages ~6.5-8.0 years at screening with subsequent participation in the baseline assessment and randomization to the intervention/control condition. Disadvantaged families will be defined by low household income. This will be determined based on local income eligibility criteria for subsidized housing. In King County, the largest county in WA, in which the majority of subjects are expected to reside, the income requirement for subsidized housing is 80% of the area median income (AMI) adjusted for family size. This metric accounts for local variations in costing of living, which, for King County, is 58% higher than the US average. Recruitment parameters will also require at least 50% of the girls (balanced across intervention/control) will identify as Black, Latina, or 'multiple' race or ethnicity.

A total of 240 families will be randomized to the Family Based Treatment (FBT) program (n=120) vs. control condition (n=120). During a 1-2 week 'run-in' period, each family will be screened and those families who are eligible will proceed to participate in the full baseline assessment. Families who successfully complete these activities will be randomized to the FBT program vs. control condition. INTERVENTION: FBT intervention will be implemented over a 5.5-month period. Trained interventionists will lead 20 in-home, weekly sessions (30 min each) with girls and their mothers together and 20 corresponding mother-only group sessions (40 min each) online. A subset of 4 group sessions (1, 2, 10, 20) will be held in-person with both girls and their mothers attending concurrent but separate sessions. Remote sessions will be conducted via secure video-conferencing, which was tested during the COVID-19 pandemic and found to enhance session attendance. The in-home sessions (dyads) will provide protocol-based tailored support for behavioral skills related to family eating and physical activity change and will focus on feedback, accountability, and problem solving for skill use and barriers and goal-setting specific to each family. The mother-only group sessions (~6 members each) will provide education focused on behavior change as well as guidance focused on parenting in areas of healthy eating and active living. The 4 parallel child-only group sessions will provide child adapted content focused on behavior change. CONTROL: The control condition will consist of enhanced usual care. Girls randomized to this condition will receive their usual medical care through their pediatric care providers. The occurrence of these appointments will be recorded by study staff as reported through their mothers. In addition, simplified educational handouts modeled after the 20 group sessions will be sent to the families by regular mail approximately 1/month in a standard order. These handouts will resemble the content covered in the FBT program but will be exclusively knowledge-based without connection to any of the action-focused activities (e.g., goal setting) used in the intervention. Use of an enhanced usual care control condition will facilitate engagement with the families but ensure separation from the treatment elements provided through the intervention condition.

The primary intervention targets focus on positive change in weight indexed by reductions in BMI z-score (∆zBMI) and health behaviors in areas of diet quality, activity level, and sleep duration.

It is expected that the frequency of pubertal onset will be lower in girls in the intervention (vs. control) condition at the 18- and 30-month FUs (post-randomization). Pubertal onset will be indexed both by hormones and questionnaire-based pubertal staging methods to characterize the initiation of the main pubertal development processes: 1) gonadarche (i.e., luteinizing hormone [LH], Tanner stage 2 [TS2] breast) and 2) adrenarche (i.e., dehydroepiandrosterone sulfate [DHEAS], TS2 pubic hair). It is also expected that reduced zBMI and improved health behaviors will predict a lower frequency of pubertal onset at the 18- and 30-month FUs, marked by the initiation of gonadarche (LH, TS2 breast) and the initiation of adrenarche (DHEAS, TS2 pubic hair). Finally, it is expected that differences in pubertal onset will be mediated by differences in weight loss (reduced zBMI) and improved health behaviors (diet quality, activity level, sleep duration), as well as - on an exploratory basis - improved cardiometabolic health status (blood pressure, inflammation, insulin resistance), and metabolic hormones linked to adiposity and pubertal onset (leptin, IGF-1).

The clinical implications of this work are enormous. First, if slowing pubertal onset is possible, by extension, the numerous maladaptive outcomes associated with earlier pubertal onset could be reduced if not ameliorated. Moreover, girls most in need may reap the most benefit as girls from disadvantaged socioeconomic backgrounds and minoritized racial and ethnic identities are more likely to experience earlier pubertal onset. In this way, it is plausible that life course linkages between prepubertal obesity, earlier pubertal onset, and negative post-pubertal outcomes could be disrupted among those who are most vulnerable. Second, this work will inform new directions for existing obesity treatment programs in girls. Most obesity intervention and prevention efforts have not considered pubertal stage or the important intersections between obesity and pubertal onset. Findings will determine the value of conducting such interventions in this new context of slowing pubertal onset. As well, more will be learned about the optimal timing of obesity interventions, including whether implementation in the prepubertal period may be most impactful due to the combined benefits of weight loss along with a reduction in risk for earlier pubertal onset and its sequelae.