Longitudinal trajectories of insulin resistance in women with polycystic ovary syndrome: a functional data analysis approach.

Abstract

Introduction: Insulin resistance (IR) plays a central role in the pathophysiology of polycystic ovary syndrome (PCOS), yet its long-term trajectory across different phenotypes and obesity statuses remains unclear. This study aimed to model longitudinal IR patterns in women with PCOS, those meeting an isolated PCOS criterion, and healthy controls.

Methods: This population-based prospective study included 1,759 reproductive-aged women (18-49 years) from the Tehran Lipid and Glucose Study, categorized into PCOS (n = 287), isolated PCOS criterion (n = 536), and healthy control (n = 936) groups, stratified by BMI (<25 kg/m2, 25-29.99 kg/m2, and ≥30 kg/m2) and WC (<88 cm and ≥88 cm). Functional principal component analysis modeled HOMA-IR and fasting insulin level trajectories as sparse functional data.

Results: Women with PCOS were younger (23.5 vs 25 and 28 years, P < 0.05) and had lower BMI than healthy controls (23.38 vs 24.46 kg/m2, P < 0.05), with no baseline differences in HOMA-IR. Both HOMA-IR and fasting insulin trajectories varied by obesity status. Women with PCOS and either obesity or central obesity showed a pronounced upward trend in IR over time, while those without general or central obesity exhibited more stable or improving patterns. In contrast, healthy and isolated criterion groups showed relatively stable or mildly increasing HOMA-IR and fasting insulin levels, with steeper increases in the presence of obesity.

Conclusion: This study provides new insights into the diverse trajectories of IR among women with PCOS, those with an isolated PCOS criterion, and healthy controls. Progressive worsening of IR was observed in women with PCOS and obesity, while those with PCOS without obesity showed improvement. Healthy and isolated criterion groups exhibited more stable patterns.