Differential associations between smoking, e-cigarette use, and diabetes prevalence.

BACKGROUND

Cigarette smoking is a well-established risk factor for diabetes, but the relationship between e-cigarette use and diabetes remains uncertain. Evidence to date has been drawn almost entirely from North America and Asia, with little information from European populations.

METHODS

We conducted a cross-sectional study of 17,854 adults aged 16 years and older from the 2017, 2018, 2019, and 2021 waves of the nationally representative Scottish Health Survey. Diabetes status was based on self-report of doctor-diagnosed diabetes. Participants were classified into six mutually exclusive categories of smoking and e-cigarette use: never users of either cigarettes or e-cigarettes, ex-smokers (former smokers who never used e-cigarettes), current exclusive cigarette smokers, current exclusive e-cigarette users, current dual users, and former e-cigarette users. Weighted prevalence estimates and survey-weighted binary logistic regression models were used to examine associations, adjusting for age group, sex, education, deprivation quintile, ethnicity, alcohol use, physical activity, and hypertension.

RESULTS

Diabetes prevalence was highest among ex-smokers (11.3 %, 95 % CI: 10.1-12.5). Prevalence was 5.7 % (95 % CI: 5.2-6.2) among never users of either cigarettes or e-cigarettes, 6.2 % (95 % CI: 4.9-7.9) among current exclusive cigarette smokers, 4.9 % (95 % CI: 3.4-7.1) among current exclusive e-cigarette users, 8.3 % (95 % CI: 5.8-11.8) among current dual users, and 5.1 % (95 % CI: 4.1-6.3) among former e-cigarette users. In adjusted models, ex-smokers had 35 % higher odds of diabetes compared with never users of either cigarettes or e-cigarettes (OR = 1.35, 95 % CI = 1.14-1.60, p < 0.001), whereas current exclusive smokers (OR = 0.78, 95 % CI = 0.58-1.03, p = 0.084), current exclusive e-cigarette users (OR = 0.81, 95 % CI = 0.53-1.22, p = 0.309), current dual users (OR = 1.49, 95 % CI = 0.94-2.38, p = 0.091), and former e-cigarette users (OR = 1.00, 95 % CI = 0.78-1.29, p = 0.973) were not significantly different from never users. Sensitivity analyses restricting ex-smokers to those with ≥5 years since cessation and limiting the sample to adults aged ≥45 years reproduced the same pattern of results.

CONCLUSIONS

In this nationally representative study of Scottish adults, excess diabetes prevalence was observed among ex-smokers, a pattern that may reflect both reverse causation if individuals quit smoking after diagnosis and the lasting metabolic effects of cumulative smoking exposure. Neither current nor former e-cigarette use was associated with diabetes, and the observed variation in prevalence appeared linked to smoking history rather than e-cigarette use. However, because vaping is relatively recent, further longitudinal research is needed to clarify any long-term risks.