The Effect of Polycystic Ovarian Syndrome on Development of Diabetes Mellitus by a Multistate Modelling Approach

BACKGROUND: Women with polycystic ovary syndrome (PCOS) are more susceptible to metabolic dysfunction, especially glucose impairment. Previous studies have considered prediabetes (pre-DM) and diabetes mellitus (DM) as single-point outcomes without accounting for the transition states. This study aimed to explore the transition between different blood glucose states using a multistate Markov model among women with and without PCOS. METHODS: This study employed a dataset from the population-based study of the Tehran Lipid and Glucose Study. Women, aged over 20 years, without pre-DM and DM at baseline (1999-2000) and with documented PCOS status, were included; subsequent follow-ups were conducted every 3 years, totaling six follow-up assessments. Three states of 'Normoglycemia', 'Pre-DM', and 'DM' and three possible transitions of (I) normoglycemia to pre-DM, (II) normoglycemia to DM, and (III) pre-DM to DM were defined. The multi-state Markov models were applied to examine the predictive effect of PCOS on the transition rates throughout the progression of DM. RESULTS: Out of 1587 participated women, 782 (49.3%) remained free of pre-DM and DM at the end of the study. Transition from normoglycemia to pre-DM was the most frequent progression observed, affecting 46.6% of the study participants. Women with PCOS significantly had 23% higher hazard of transition I [HR (95% CI): 1.23 (1.01-1.49); p = 0.034] compared to those without PCOS. While being PCOS did not significantly affect the risk of other transitions. CONCLUSIONS: This study underscores the significant impact of PCOS on the dynamic progression from normoglycemia to pre-DM. Future longitudinal studies are essential to support our findings.