Abstract
OBJECTIVE: Prediabetes and diabetes are complex conditions associated with interrelated factors from multiple domains, including lifestyle, health, and sociodemographic status. We aimed to identify youth subgroups with co-occurring features from these domains to inform prevention and intervention efforts. METHODS: We analyzed data from 1278 adolescents (ages 12-17 years) from the 2011-2018 United States National Health and Nutrition Examination Surveys. Prediabetes/diabetes was defined using hemoglobin A1c (≥5.7 mg/dL) and/or fasting plasma glucose (≥100 mg/dL). We applied latent class analysis of physical activity, diet quality, screen time, and body mass index (BMI) identified subgroups adjusting for sociodemographic factors. Associations between latent class membership and prediabetes/diabetes were assessed using survey-weighted logistic regression. RESULTS: Four classes emerged: High BMI and unhealthy lifestyle (37.5 %), Healthy BMI and physically active (24.7 %), Healthy BMI and lifestyle (16.0 %), and Average BMI and lifestyle (21.8 %). Compared to youth in the High BMI and unhealthy lifestyle class, youth in the other three classes all had lower odds of prediabetes/diabetes, where the adjusted odds ratio was 0.56 (95 % CI: 0.33-0.95) for youth of the Healthy BMI and active class, 0.46 (95 % CI: 0.21-1.01) for youth in the Healthy BMI and lifestyle class, and 0.68 (95 % CI: 0.42-1.10) for youth in the Average BMI and lifestyle class. CONCLUSIONS: The latent class analysis revealed four youth subgroups with different lifestyle, health, and sociodemographic characteristics, and youth prediabetes/diabetes status was associated with the latent class membership. The results suggest potential heterogeneous interventions over one-size-fits-all approaches to mitigate prediabetes/diabetes risk among youth.