Spatial pattern and determinants of tobacco use among females in India – Evidence from a nationally representative survey (NFHS-4)

INTRODUCTION Tobacco use has been steadily increasing among the females in developing countries. It has led to rise in tobacco-related morbidity and mortality among females. Knowing the geographic distribution of the habit is essential to identify high priority areas and direct the healthcare intervention. Hence, this study was done to assess the spatial patterns and determinants of tobacco consumption among females in India. METHODS Univariate and Bivariate Moran's I statistic and local indicators for spatial association (LISA) maps were generated to determine the spatial clustering of tobacco consumption (smoked and smokeless form). Ordinary least-square regression, spatial-lag and spatial-error models were performed to assess the determinants. Poverty (belonging to poorest and poorer quintile of wealth index), illiteracy (no formal education), marital status, ST population, tobacco use by family members and alcohol use were the explanatory variables. RESULTS Univariate Moran's I was 0.691 suggesting positive spatial autocorrelation. High-prevalence clustering (hot-spots) was maximum in the Central, Eastern and North-Eastern states such as Chhattisgarh, Madhya Pradesh, Odisha, Bihar, Manipur, Tripura, Meghalaya, Mizoram and Assam. This pattern was similar for both smokeless and smoked form. Results of spatial-lag and spatial-error model suggested that alcohol use, scheduled tribes, illiteracy, poverty, marital status and tobacco use by family members were significant determinants of female tobacco consumption. The coefficient of spatial association was maximum for alcohol use (β=0.20, p<0.001) followed by widowed/separated/divorced (β=0.12, p<0.001). CONCLUSION Tobacco consumption among females in India is spatially clustered. Multisectoral coordination and targeted interventions are required in the geographical hotspots of tobacco consumption. IMPLICATIONS This is the first study to explore the geospatial pattern of tobacco consumption among females in India. We found that the pattern of tobacco use among females is spatially clustered in India. Clustering was predominantly found in Central, Eastern and North-Eastern regions of the country. Tribal population in these areas and complementarities between alcohol and tobacco use contributed significantly to the high prevalence clustering. These findings will be helpful for policymakers and planners to devise specific intervention package targeting the high-risk regions.