Predictors of responses to clinic-based childhood obesity care.

Background/objective: Lifestyle modification is the therapy of choice for childhood obesity, yet the response rate is variable and may be affected by genetic factors. We aimed to investigate predictors of poor response to lifestyle modification obesity treatment in children.

Methods: A prospective cohort study of 434 youths (64.5% females) between 4 and 20 years of age undergoing a standard care of lifestyle modification obesity management for 35.9 ± 20.8 months at Yale Childhood Obesity Clinic, USA. The primary outcome was a "poor response," defined as the quintile with the largest increase in BMI Z-score over time. The secondary outcome was the endpoint BMI Z-score. Covariates investigated were sex, baseline pubertal status and degree of obesity, race, biochemical profile, and family history of overweight. A subsample (n = 214) had FTO genotyping (SNP rs8050136) tested.

Results: Males (hazard ratio [HR] = 5.35, 95% confidence interval [CI] [3.32-8.61], P < 0.0001) and pubertal adolescents (HR = 2.78, [1.40-5.50], P = 0.003) compared to prepubertal children were more prone to respond poorly. Baseline degree of obesity was associated with relative protection from responding poorly (HR per BMI Z-score unit = 0.32, [0.17-0.61], P = 0.0006). Carriers of the FTO obesity-predisposing allele (AA genotype) were protected from responding poorly compared to non-carriers (CC genotype) (HR = 0.33, [0.12-0.88], P = 0.028).

Conclusions: Boys and pubertal adolescents are more prone to respond poorly to standard obesity care while those with greater baseline degree of obesity and carriers of the FTO obesity-predisposing allele are not.