Accelerated biological aging in the association between early-life tobacco smoke exposure and dementia risk: An exploratory mediation analysis
Linling Yu, Huaxin Deng, Meng Yang, Ding-Sheng Jiang, Chenqi Liao, Danyang Sun, Jiahao Song, Hongbin Luo, Wei Liu, Xiong WangBackground
Early-life tobacco smoke exposure, including maternal smoking during gestation and smoking during childhood or adolescence, may have long-term health consequences, but its association with dementia remains uncertain.
Objective
To examine the associations of prenatal tobacco smoke exposure and childhood/adolescence smoking with incident dementia and to explore whether biological-age acceleration represents a potential intermediate component of these associations.
Methods
We conducted a prospective cohort study of 285,009 dementia-free participants from the UK Biobank. Biological aging was assessed using Klemera–Doubal method biological age acceleration (KDM-BA) and phenotypic age acceleration (PhenoAge). Cox proportional hazards models estimated hazard ratios (HRs) for incident dementia, and exploratory mediation analyses evaluated model-based indirect components through biological-age acceleration.
Results
During a mean follow-up of 13.67 years, 4621 incident dementia cases were documented. Prenatal tobacco smoke exposure (HR = 1.094, 95% CI 1.025–1.168) and childhood/adolescence smoking (HR = 1.187, 95% CI 1.106–1.275) were associated with higher dementia hazards. The corresponding standardized 10-year absolute risk differences were 0.063 and 0.124 percentage points. For childhood/adolescence smoking, the combined indirect effect through KDM-BA and PhenoAge acceleration was small (HR = 1.0062, 95% CI 1.0002–1.0125), with an estimated mediated proportion of 3.63% (95% CI 0.11%–8.77%).
Conclusions
Early-life tobacco smoke exposure was associated with modestly higher hazards of incident dementia. Exploratory mediation analyses identified a small estimated indirect component through biological-age acceleration for childhood/adolescence smoking.