Associated factors of anxiety symptoms among reproductive married women in Bangladesh: Using a structural equation model with decomposition analysis.
Abstract
Background
Anxiety symptoms can undermine women's wellbeing and family functioning. This study examines the associated factors with moderate-to-severe anxiety symptoms and wealth-related inequalities among currently married women of reproductive age in Bangladesh.
Methods
This cross-sectional study analyzed 13,438 currently married women aged 15-49 years from the Bangladesh Demographic and Health Survey, 2022. Survey-weighted multivariable logistic regression was used as the primary analytical model, supplemented by exploratory path analysis for assessing direct and indirect associations. Wealth-related inequality in anxiety symptoms was evaluated using concentration indices and decomposition analysis.
Results
The weighted prevalence of moderate to severe anxiety symptoms among currently reproductive married women in Bangladesh was 4.1% (95% CI: 0.04-0.05). Survey-weighted regression indicated that high endorsement of attitudes justifying intimate partner violence, older age, sexual inactivity, pregnancy termination, and having no children were associated with higher odds of anxiety symptoms. The concentration index indicated a pro-poor concentration of anxiety symptoms (CCI = -0.03068590), with education, body mass index, and number of children contributing to observed inequality. Exploratory path analysis showed small model-consistent indirect associations involving socioeconomic status, lifestyle, BMI, and number of children, but these estimates were interpreted cautiously as non-causal indirect associations.
Conclusions
Anxiety symptoms were unequally distributed among married women of reproductive age in Bangladesh and were more concentrated among poorer women. The findings support the need for mental health, reproductive health, and gender-sensitive interventions targeting socioeconomically disadvantaged women. The exploratory path findings should be interpreted cautiously because the cross-sectional design does not establish temporal ordering or causal mediation.