Abstract
BACKGROUND: Early-onset dementia (EOD) and late-onset dementia (LOD) may have distinct modifiable risk-factor profiles. OBJECTIVE: To identify and compare factors associated with EOD and LOD using a nationwide cohort database. DESIGN: Nationwide two nested case-control studies. SETTING: We used the National Health Insurance Service-National Sample Cohort database (2004-2019). PARTICIPANTS: The initial sample size was 514,866; 5157 EOD and 39,326 LOD cases were matched 1:1 with controls based on age, sex, and the Charlson Comorbidity Index. MEASUREMENTS: Socioeconomic status, residential area, body mass index, alcohol consumption, smoking status, physical activity, blood pressure, and laboratory findings were analyzed. Multivariable logistic regression models were used to identify the risk factors. RESULTS: Higher socioeconomic status and increased frequency of physical activity were associated with a lower risk of both EOD and LOD. Rural residence, heavy alcohol consumption, and higher fasting blood sugar levels were associated with an increased risk of LOD, although there was no significant association with EOD. Overall, these factors impacted LOD more strongly than EOD. Demographic and lifestyle factors had a greater effect on LOD than blood pressure and relevant laboratory findings. CONCLUSION: Modifiable risk factors were associated with LOD and EOD. The influence of some modifiable risk factors was more pronounced in the LOD group than in the EOD group. Identifying modifiable risk factors associated with dementia can aid in the development of preventive strategies, underscoring the clinical importance of our findings.