Prevalence and determinants of menstrual abnormalities among postgraduate students using structural equation model

Abstract

BACKGROUND: In India, the prevalence of menstrual abnormalities is alarmingly high, ranging from 85% to 93.4%, according to documented studies. A significant proportion, of these women fall within the age group of 20–29 years, which includes postgraduate students. This underscores the critical importance of identifying and effectively managing the underlying causes of menstrual abnormalities. MATERIALS AND METHODS: The study involved 272 postgraduate female students from Mangalore University, Karnataka state, India selected through stratified sampling. A questionnaire collected data on the age of menarche, menstrual abnormalities, BMI, hormonal imbalances or thyroid conditions, menstrual cycle characteristics and symptoms, lifestyle and dietary habits, physical health, perceived stress levels, economic factors, attitudes towards menstrual health, and knowledge and behaviors related to menstrual health. The study used a Structural Equation Model (SEM) to evaluate the effects of risk factors on menstrual disorders and developed machine learning models to accurately predict different types of these disorders. RESULTS: The study found that 78% of PG students experienced menstrual abnormalities, with dysmenorrhea being the most common at 24.26%. The SEM model identifies lifestyle, perceived stress, physical health, and practice as significant risk factors for menstrual abnormalities. Mediation analysis reveals that attitude, economic factors, and practice play crucial roles in influencing menstrual abnormalities. The SEM‑based classification model outperforms popular machine learning models, achieving an accuracy of 84.56%. CONCLUSION: Reducing stress and promoting healthy habits are key to managing menstrual abnormalities. Targeted interventions can improve menstrual health, benefiting students’ academic performance, self‑esteem, and long‑term health.