Background Smoking is a modifiable risk factor and has a well-known relationship with socioeconomic status and mental health. However, research into alcohol harm from others and its association with smoking habits is sparse. Despite a significant minority(20.1%) of the population experiencing some form of harm from other’s alcohol consumption there has been little study into its effects on smoking habits. In this study we aimed to compare the prevalence of smoking in those who experience harm from other’s alcohol consumption and those who don’t.

Methods We conducted cross-sectional analysis on the Alcohol Toolkit Study data from 2015 to 2016. Households were selected using a hybrid system of random probability sampling and simple quota sampling. Respondents aged over 16 were asked about their smoking habits and whether they had experienced harm from another’s drinking in the last year. Aggressive harm was categorised to include more serious harms. We used logistic regression to explore if the characteristics of those who experienced harm including age, ethnicity, work status, disability, housing tenure, life stage, qualifications and personal drinking, account for any of the difference in smoking rates between the people who experience harm and those who don’t.

Results 4881 participants were included in the analysis. 38% of respondents that had experienced aggressive harm from others’ alcohol consumption were smokers compared with 26% of those who had experienced some harm and 17% of those who had experienced no harm. Some of the variation was explained by the respondents’ personal characteristics, but the odds of smoking remained higher for those who experienced harm (OR 1.27 95CI% 1.05–1.56) and for those who experienced aggressive harm (OR 1.92 95%CI 1.46–2.53) compared with those who did not.

Conclusion Alcohol and tobacco use are closely linked. This exploratory analysis has allowed us to show that the link extends to people who have experienced harm from another person’s drinking. People who have experienced harm are more likely to smoke, particularly for those who experience aggressive harm. The reasons for this correlation between alcohol harm and tobacco use have not been investigated before and this study shows that the relationship does not appear solely to be due to sociodemographic characteristics. This warrants further study. Smoking cessation services should take into consideration the impacts of harm from other’s alcohol consumption on the smokers attempts to quit.