Longitudinal patterns of smoking behaviours in adolescence and early adulthood and their association with modifiable and sociodemographic risk factors

Introduction Smoking contributes a huge burden on public health; thus, identifying risk factors for smoking remains an important area of research. This study adds to the wealth of existing literature by utilising repeated smoking measures collected in a UK sample of young adults to (a) examine differences between longitudinal smoking behaviours, (b) investigate their association with many risk factors, and c) consider how these associations may change over time. Methods This study uses longitudinal latent class analysis and 12 repeated measures to derive patterns of smoking in the Avon Longitudinal Study of Parents and Children. The association of these patterns with 402 risk factor measures collected from 0-28 years is then investigated. The selected risk factors include familial and peer factors, lifestyle and sociodemographic factors, mental health, parenthood, adverse childhood experiences and trauma. Results Five different latent classes of smoking were derived and referred to as non-smoking, short-term smoking, occasional smoking, early-onset smoking, and late-onset smoking. These showed differences in age of onset, frequency, and cessation. Other substance use, and parental and peer substance use, showed the strongest association with smoking patterns. More risk factors were associated with early-onset than late-onset smoking. Many risk factors of regular smoking did not show the same associations with occasional smoking. Fewer measures differentiated late-onset from short-term smoking. Some associations varied depending on the time of measurement or smoking pattern in question. Conclusions Findings from this study may be used to identify groups of people most vulnerable to more harmful smoking patterns despite being exposed to strong tobacco prevention efforts. This could also help better tailor smoking interventions and improve tobacco control policies.