22-OR: Exploring the Association between Physical Activity Levels and Sedentary Behaviors with Risk of T2DM from Childhood to Adolescence: A Causal Inference Analysis

Abstract

Background and Objective: Causal inference methods were developed to emulate repeated randomized trials in observational studies, notably by controlling for time-varying confounders. Using these methods, we examined the complex relationship between physical activity, sedentary behaviors and T2DM accounting for interrelations between time-varying lifestyle habits and their reciprocal effects on T2DM across childhood. Methods: We used longitudinal marginal structural models (with inverse probability of treatment and censoring weighting) on prospective data of Caucasian children with a parental history of obesity (QUALITY Cohort) evaluated at 8-10 years (n=630), 10-12 years (n=564) and 15-17 years (n=377). At each visit, participants underwent an OGTT (1.75g/kg glucose, max 75g). ISI-Matsuda estimated insulin sensitivity. Area under the curve of glucose to insulin estimated 1st (0-30 min) and 2nd phase (0-120 min) insulin secretion. ADA cut points determine IFG, IGT and T2DM. Moderate to vigorous physical activity (MVPA) and sedentary behaviors were assessed with 7-day accelerometry and leisure screen time by questionnaire. Confounders included age, sex, BMI z-score, pubertal stage, diet quality index-international, energy intake, sleep, fitness, parental BMI, and income. Results: Every additional 10-min of daily MVPA across 8-10 to 15-17 years was associated with a 5.4% (95% CI: -8.8; -2.0) reduction in 2nd-phase insulin secretion at 15-17 years. Every additional 1-h of daily screen time across 8-10 to 15-17 years was associated with an 8.4% (95% CI: -14.1; -2.7) reduction in insulin sensitivity and a 6.4% (95% CI: 0.3; 12.5) and 8.1% (95% CI: 1.5; 14.7) increase in 1st- and 2nd-phase insulin secretion at 15-17 years. No associations were observed with IFG/IGT/T2DM. Conclusion: Even small increases in MVPA or slight reductions in screen time in childhood/adolescence may play a role in T2DM prevention. Disclosure S. Harnois-leblanc: None. A. Van hulst: None. T. A. Barnett: None. M. Mathieu: Research Support; Self; CAPSANA, FitSpirit, Research Support; Spouse/Partner; Caboma, Medicus. G. Paradis: None. M. Sylvestre: None. M. Henderson: None. Quality research group: n/a. Funding Canadian Institutes of Health Research (OHF-69442, NMD-94067, MOP-97853, MOP-119512); Heart and Stroke Foundation of Canada (PG-040291); Fonds de Recherche du Québec - Santé