Tobacco Smoking and Dementia in a Kentucky Cohort: A Competing Risk Analysis

Abstract

Tobacco smoking was examined as a risk for dementia and neuropathological burden in 531 initially cognitively normal older adults followed longitudinally at the University of Kentucky's Alzheimer's Disease Center. The cohort was followed for an average of 11.5 years; 111 (20.9%) participants were diagnosed with dementia, while 242 (45.6%) died without dementia. At baseline, 49 (9.2%) participants reported current smoking (median pack-years = 47.3) and 231 (43.5%) former smoking (median pack-years = 24.5). The hazard ratio (HR) for dementia for former smokers versus never smokers based on the Cox model was 1.64 (95% CI: 1.09, 2.46), while the HR for current smokers versus never smokers was 1.20 (0.50, 2.87). However, the Fine-Gray model, which accounts for the competing risk of death without dementia, yielded a subdistribution hazard ratio (sHR) = 1.21 (0.81, 1.80) for former and 0.70 (0.30, 1.64) for current smokers. In contrast, current smoking increased incidence of death without dementia (sHR = 2.38; 1.52, 3.72). All analyses were adjusted for baseline age, education, sex, diabetes, head injury, hypertension, overweight, APOEɛ4, family history of dementia, and use of hormone replacement therapy. Once adjusted for the competing risk of death without dementia, smoking was not associated with incident dementia. This finding was supported by neuropathology on 302 of the participants.

Keywords: Competing risks; dementia; dementia free death; lifetime smoking.

Figure 1:

Cumulative incidence curves by baseline smoking status for the competing events death without dementia (Panel A) and dementia (Panel B). The predicted curves are based on hypothetical female participants who are 73 years old at baseline, with high education, no APOE-ε4, no family history of dementia, hypertension, no diabetes, no HRT use, normal weight, and no head injury.