作者
Fei Tian,Jian Zhao,Lan Chen,Shengtao Wei,Hualiang Lin
摘要
Objective Various modifiable risk factors have been identified for dementia, but it remains unclear whether these associations and importance vary according to genetic predisposition. Methods We included 182,473 dementia‐free individuals from United Kingdom (UK) Biobank cohort. We calculated the hazard ratios (HRs) and population‐attributable fractions (PAF) of 11 modifiable risk factors (smoking status, drinking status, diet, physical activity, hypertension, diabetes, waist‐to‐hip ratio, education, depression, social contact, and air pollution) for dementia in general population and across various genetic groups based on APOE genotypes and a weighted polygenic‐risk score (PRS). Results During 12.42‐year follow‐up, a total of 2,065 incident all‐cause dementia (ACD) cases were documented. Low social contact, diabetes, and abdominal obesity were top 3 modifiable risk factors associated with an increased risk of dementia, especially in low APOE risk (ε2ε2 or ε2ε3) group. The HRs associated with low social contact, diabetes, and abdominal obesity were 3.86 (95% CI: 2.08–7.17), 2.18 (95% CI: 1.43–3.32) and 1.29 (95% CI: 0.88–1.89) for dementia in the low APOE risk. The same patterns were also observed in PRS groups. In terms of PAFs, abdominal obesity contributed the most in low genetic risk groups (PAF: 13.68 to 15.73%). In contrast, less education was the largest contributor in high APOE and high PRS groups (PAF: 11.35% [95% CI: 5.32–17.38%] and 11.24% [95% CI: 4.40–18.07%], respectively), followed by hypertension (PAF: 9.09% [95% CI: 2.01–16.17%] and 5.29% [95% CI: −2.72 to 13.29%], respectively). Overall, the total PAF decreased with increasing genetic risk, particularly for metabolic risk factors. Interpretation Our findings highlight the importance of personalized risk assessments and the development of tailored interventions based on genetic background to prevent dementia more effectively. ANN NEUROL 2025;98:1210–1221