How have the associations between different measures of socioeconomic position and smoking changed in England between 2013 and 2025? A monthly cross-sectional population survey

Introduction We examined how the association between five measures of socio-economic position (SEP) - social grade, housing tenure, income, education, and employment - and smoking status changed between 2013 and 2019. We also identified whether the strongest independent SEP predictor of smoking in England varied over time. To provide additional context for interpreting pre-pandemic trends, we conducted exploratory extended analyses including data to May 2025 for education and social grade. Methods Data from the Smoking Toolkit Study (n=139,314 adults aged ≥16 years) were analysed using weighted Generalised Additive Models (GAMs). Each SEP measure was regressed onto smoking status with an interaction term for time (modelled linearly). Unplanned analyses extended the timeframe (2013–2025) for education and social grade. Year-stratified ridge logistic regression models including all SEP indicators simultaneously were used to assess their relative independent predictive contribution. Results For all SEP measures, individuals in the more disadvantaged categories started the study period with significantly higher odds of smoking than the reference groups. No time interactions were found for employment, social grade, or housing tenure, indicating that the significant baseline differences persisted across the series. Education and income showed significant time interactions, with widening differences primarily between higher and middle socioeconomic groups. Smoking declined by 0.3% monthly among university graduates, but only 0.0–0.1% among less-educated groups. A similar decline was seen in higher-income groups (£40,000+), with no change in the £25,000–£39,999 group. Unplanned analyses extending the series to 2025 suggested a possible narrowing of inequalities for education and social grade, with faster declines among some more disadvantaged SEP groups. While the strength of the association between housing tenure and smoking declined slightly over time, it remained the strongest independent predictor each year, followed by social grade. Conclusions From 2013–2019, income- and education-related smoking inequalities widened, while inequalities based on social grade, housing tenure, and employment status remained stable. Housing tenure remained the strongest independent predictor of smoking across all years, highlighting the importance of targeting cessation support in social housing. More recent data (for social grade and education) hint at a possible narrowing of these gaps.