Seasonal variation in BMI outcomes at 6 months: secondary analyses of a multidisciplinary healthy lifestyle programme for children and adolescents with obesity

TL;DR

Children entering a weight management program in spring had smaller BMI improvements, possibly due to summer holidays disrupting their progress.

Contribution

This study identifies seasonal entry into a weight program as a novel predictor of BMI outcomes in children.

Findings

Participants entering in spring had smaller BMI reductions compared to other seasons.

Baseline BMI and younger age were strongest predictors of BMI improvement.

Seasonal variations in program effectiveness were confirmed using both linear models and machine learning.

Abstract

While international evidence suggests seasonal variations may influence outcomes of interventions for pediatric obesity, data for Aotearoa New Zealand are limited. We examined seasonal variations in changes in body mass index standard deviation score (BMI SDS) in young people with obesity enrolled in an intervention programme.

We studied 397 children and adolescents (median = 10.1 years; range 3.7–16.8 years) from Whānau Pakari, a multidisciplinary community-based healthy lifestyle programme (initially a randomised clinical trial that subsequently transitioned into the regional childhood obesity service). Participants were stratified by season at entry and 6-month BMI SDS changes (Δ) were evaluated. Lifestyle factors were also assessed. Data were analysed using traditional linear models and machine learning (random forest).

68% of participants had BMI SDS reductions at 6 months (mean = − 0.16 SDS; P < 0.0001). Linear models showed seasonal variations in programme effectiveness, with BMI SDS reductions among summer (− 0.17 SDS), autumn (− 0.19 SDS) and winter (− 0.18 SDS) but not among spring entrants. Random forest modelling identified higher baseline BMI SDS and younger age as the most influential predictors of greater 6-month reductions in BMI SDS. Season of entry was more important than any single lifestyle factor; spring entrants exhibited attenuated reductions relative to other seasons.

The season at programme entry was an important factor associated with intervention effectiveness. Spring entry was associated with attenuated BMI SDS reductions, likely due to the inclusion of the summer holidays within the 6-month intervention. These findings highlight the need for targeted support during such unstructured periods to improve participant outcomes.

The online version contains supplementary material available at 10.1007/s12519-025-01016-z.

Figures