Abstract
Objectives
This study examined whether past-month alcohol use after the lockdown was higher or lower than expected based on the pattern observed before the lockdown in 13 Indian states and Union Territories, with particular attention to differences by gender, socioeconomic status, and state alcohol-policy context.
Methods
We analysed National Family Health Survey data from 13 Indian states and union territories surveyed before and after the lockdown period. The primary outcome was any past-month alcohol use vs. no alcohol use. To address sparse alcohol-use categories, especially among women, we used Bayesian aggregated binomial logistic regression models estimated separately for women and men. Models included time trend, post-lockdown period, subgroup interactions, and state-level random intercepts. Subgroup-specific posterior odds ratios, predicted probabilities, 95% credible intervals, and posterior probabilities were estimated for age, residence, education, wealth, marital status, religious alcohol norms, scheduled tribe status, and state alcohol-policy category.
Results
Among women, the odds of past-month alcohol use after the lockdown were higher than expected based on the earlier pattern. The posterior odds ratio for women was 1.83, with a 95% credible interval of 1.42–2.20, and the posterior probability of a positive association was approximately 1.00. Among men, the estimated association was smaller but still had high posterior probability, with a posterior odds ratio of 1.09, a 95% credible interval of 1.01–1.16, and a posterior probability of positive association of 0.99. Subgroup results suggested that positive associations among women were more evident among older women and socioeconomically disadvantaged groups, including rural women and women with lower education or wealth. Among men, subgroup patterns were smaller and more heterogeneous.
Conclusion
The association between the post-lockdown period and past-month alcohol use differed between women and men in the 13 included states and union territories. The strongest positive association was observed among women, particularly women in socioeconomically disadvantaged subgroups such as rural women and women with lower education or wealth. These findings represent adjusted associations and should not be interpreted as causal effects of the national lockdown.