Patterns of tobacco or nicotine-based product use and their quitting behaviour among adults in India: a latent class...

The aims of the study were to identify latent classes of adult tobacco or nicotine-based product users in India, compare their sociodemographic distribution and quitting behaviour and explore the association of quitting behaviour and time to first tobacco use with class membership.

This was a nationally representative cross-sectional survey.

Data from the Global Adult Tobacco Survey 2016-2017 in India, which covered adults aged >15 years, were analysed. Latent class analysis was used to examine patterns of tobacco or nicotine-based product use (cigarette, bidi, cigar, e-cigarette, chewable tobacco and snuff) among current tobacco users. Classes were compared across sociodemographic and tobacco use-related characteristics. Various model fit statistics (Akaike, Bayesian and Sample Size-Adjusted Bayesian Information Criteria, Likelihood Ratio Tests and Entropy) and meaningfulness of the classes were used to select the number of latent classes.

Of 21,857 current tobacco users, five latent classes were extracted: 'poly-tobacco use' (103, 0.5%), 'oral chewable products predominantly' (11,306, 51.7%), 'bidi predominantly' (4965, 22.7%), 'cigarette predominantly' (5318, 24.3%) and 'snuff and chewable products' (165, 0.8%). Significant differences between classes emerged on sociodemographics (age, sex, residence, education, wealth quintile, region). 'Bidi predominantly' class was associated with higher likelihood of quit attempts. Compared with 'cigarette predominantly' class, other classes were significantly associated with time to first tobacco use.

We found that people in India could be grouped into five classes based on their tobacco or nicotine-based product use pattern. It may be efficient to tailor messages to different latent classes and address the distinct profiles of these groups of tobacco product users.