Midlife Risk Factors for Subtypes of Dementia: A Nested Case-Control Study in Taiwan

Objective: To identify the midlife risk factors for subtypes of dementia newly developed later in life. Methods: A nested case-control study was conducted on 157 demented cases and 628 comparison cases selected from 40, 636 men and women who were enrolled from 1982 to 1992. Four comparison cases were frequency-matched on age, time at enrollment ( within 6 months) , gender, and residential township. Midlife risk factors included vascular risk factors ( body mass index [ BMI], total cholesterol, total triglycerides, blood glucose, cerebrovascular accident [ CVA] history, diabetes mellitus history, and hypertension history), cigarette smoking, and alcohol consumption. Dementia assessments were ascertained through the computerized data linkage from National Health Insurance Database from 2000 to 2002 and clinically confirmed by neurologists or psychiatrists. Conditional logistic regression analysis was used to estimate the matched odds ratio ( OR) and its 95% confidence intervals ( CI) for each risk factor. Results: A J-shaped relationship was observed between BMI ( kg/m(2 )) and dementia. The multivariate-adjusted ORs ( 95% CI) of developing dementia were 1.84 ( 1.02-3.33), 1.87 ( 1.08-3.23 ) and 2.44 ( 1.39-4.28), respectively, for BMIs of <20.5, 23 .0-25.4, >= 25.5 compared with a BMI of 20.5-22.9 as the referent group ( OR = 1.0). Similar findings were observed for Alzheimer disease ( AD) and vascular dementia ( VaD). The association between obesity ( BMI >= 25.5) and both AD and VaD was statistically significant among cigarette smokers but not among nonsmokers . Additionally, history of CVA was a significant risk factor for VaD, but not for AD. Conclusion: Being underweight, being overweight, and a cerebrovascular accident in midlife may increase the risk of dementia in late life.