INTRODUCTION: This paper estimates the effect of women's alcohol use on body mass index (BMI) and related indicators of nutritional outcomes in India. METHODS: Using nationally representative data from the fifth round of the National Family Health Survey, we employ an instrumental variables strategy that leverages peer drinking prevalence within district-age clusters to address the endogeneity of alcohol use. The instrument captures local exposure to social drinking norms while mitigating reflection bias through a leave-one-out construction. RESULTS: The results show a robust and significant negative relationship between drinking and women's nutritional outcomes. After correcting for endogeneity, alcohol use is associated with lower BMI and a reduced likelihood of being overweight. These effects are consistent across caste groups, residence types and policy environments, with stronger impacts observed among women from rural or socioeconomically disadvantaged backgrounds. DISCUSSION AND CONCLUSIONS: The findings suggest that household resource constraints and the social stigma surrounding women's drinking jointly contribute to these outcomes. By providing evidence on the nutritional effects of alcohol use, this study highlights an overlooked dimension of women's health in developing contexts. The analysis also extends peer-based identification strategies to the study of nutrition and health behaviour, offering new insights into how social norms shape gendered health disparities.
Kumar et al. (2026) studied this question.