ASSOCIATION BETWEEN POLYCYSTIC OVARY SYNDROME (PCOS) AND ANXIETY DISORDERS: A RETROSPECTIVE COHORT STUDY USING HOSPITAL RECORDS

Abstract

Background :Polycystic Ovary Syndrome (PCOS) represents the most prevalent endocrine disorder affecting reproductive-age women globally, with Indian prevalence estimates ranging from 9.13% to 36% across diverse diagnostic frameworks and regional populations. The clinical burden of PCOS extends substantially beyond reproductive dysfunction to encompass significant psychiatric comorbidity — particularly depression and anxiety disorders — yet the magnitude and mechanistic underpinnings of this association remain inadequately characterised in rural North Indian hospital populations. Adesh Medical College and Hospital (AMCH), Mohri, Shahbad, serving the predominantly agrarian Kurukshetra and Kaithal districts of Haryana, encounters a substantial PCOS caseload in its Gynaecology outpatient department. No prior longitudinal data linked PCOS diagnosis to subsequent anxiety disorder incidence within this specific regional cohort.
Objectives :To determine whether PCOS diagnosis is independently associated with increased hazard of incident anxiety disorder over a seven-month follow-up; to characterise anxiety subtype distribution; and to identify hormonal and metabolic predictors of anxiety within the PCOS cohort at AMCH, Mohri, Shahbad.

Methodology: A retrospective hospital-based cohort study using medical records from the AMCH Gynaecology OPD and Psychiatry Department (April–October 2021). PCOS was diagnosed using the revised Rotterdam 2003 criteria in 262 women. Age-matched controls (n=262) without PCOS were identified from the same OPD in a 1:1 ratio. Anxiety disorders were ascertained using DSM-5 criteria, GAD-7 scoring, and psychiatric consultation records. Cox proportional hazards regression estimated crude and adjusted hazard ratios with 95% confidence intervals. Analysis used IBM SPSS v24.0 and R v3.6.1.

Results: Cumulative anxiety incidence was 40.8% in PCOS versus 16.4% in controls (crude HR 2.82; 95% CI: 1.96–4.06; p<0.001). After full covariate adjustment, PCOS-associated hazard remained significant (adjusted HR 2.48; 95% CI: 1.70–3.62; p<0.001). Generalised anxiety disorder was the predominant subtype in PCOS (19.8% vs 7.3% in controls). Independent predictors within PCOS included total testosterone >60 ng/dL (adjusted HR 2.08; p<0.001), insulin resistance (adjusted HR 1.80; p=0.002), and menstrual irregularity duration >6 months (adjusted HR 1.82; p=0.002). A significant testosterone quartile dose-response was demonstrated in PCOS (Cochran-Armitage trend p<0.001) but not controls (p=0.192).

Conclusion: PCOS independently confers a 2.48-fold hazard of incident anxiety disorder in this Kurukshetra-Haryana hospital cohort. The association is substantially driven by hyperandrogenism and insulin resistance — both biologically modifiable — rather than sociodemographic factors. Routine psychiatric screening using GAD-7 at PCOS diagnosis, integrated into AMCH Gynaecology OPD workflow, represents a cost-effective, immediately implementable intervention to reduce unmet psychiatric burden in this rural North Indian population.