Longitudinal BMI Trajectory in Reproductive-Age Women at Risk of Prediabetes: Implications for Preventive Screening.

Abstract

Preventive strategies for dysglycemia typically rely on single BMI measurements or thresholds, despite evidence that gradual weight gain precedes metabolic deterioration. Whether longitudinal BMI trajectories may help identify reproductive-age women at elevated risk for prediabetes remains unclear. A retrospective cohort study was conducted using annual occupational health checkup data collected from 2015 to 2024 at four centers in Japan. The analysis included 2,271 normoglycemic women aged 18-40 years who remained free of prediabetes during the landmark period (2015-2017). Annual BMI change during the landmark period (kg/m²/year) was estimated using within-person regression. Incident prediabetes (HbA1c 5.7%-6.4%) after 2017 was evaluated using Cox proportional hazards models, including time-dependent analyses. During a median follow-up of 7.0 years, 346 women (15.2%) developed prediabetes. Each 0.1 kg/m²/year increase in annual BMI gain was associated with higher risk of incident prediabetes (hazard ratio 1.04, 95% confidence interval 1.02-1.05). Time-dependent analyses incorporating repeated BMI measurements demonstrated that persistent BMI gain remained associated with incident prediabetes. Kaplan-Meier analyses demonstrated earlier risk separation among women with greater BMI gain, whereas women with mild BMI gain showed delayed but progressive separation after approximately 5 years, particularly among those with normal baseline BMI. Monitoring BMI trajectory in routine health checkups may improve early identification of reproductive-age women at risk for dysglycemia and support trajectory-informed preventive strategies.