Preventive Effect of Smoking Cessation on Diabetes Onset: A Target Trial Emulation Approach

Smoking cessation may transiently increase diabetes risk through weight gain, but its temporal trajectory remains unclear. We evaluated the long-term effect of smoking cessation on incident diabetes.

We conducted a target trial emulation using a sequential-cohort design with repeated annual enrollment. Using annual health checkup data of Japanese employees from 2008 to 2022, we newly enrolled, at each annual assessment, and started follow-up individuals aged 40 years or older who were currently smoking and free of diabetes. Inverse probability weighting was used to adjust for baseline confounding in every sequential cohort. We estimated cumulative risks for quit-smoking and persistent-smoking groups, and their differences and ratios.

From 52 880 unique participants, a total of 331 701 participants were followed-up in the sequential cohorts, with a mean (standard deviation) age of 49.1 (5.7) years, and 25 749 developed diabetes. The risk ratio (95% confidence interval) for diabetes in the quit-smoking group relative to the persistent-smoking group was 1.15 (1.07-1.24) at 3 years and 0.95 (0.85-1.06) at 15 years. Per-protocol analyses showed a similar pattern.

Smoking cessation was associated with a short-term increase in diabetes risk, but this excess risk diminished over time, and no long-term increase in risk was observed. These findings support sustained smoking cessation as metabolically beneficial in the long term.