Onset age of diabetes and incident dementia: A prospective cohort study.

BACKGROUND:Relationship between age at diagnosis of diabetes and dementia is lacking. The aim of the study was to investigate whether diabetes onset at a younger age was associated with a higher incidence of dementia. METHODS:466,207 participants free of dementia in the UK biobank (UKB) were included in the analysis. Propensity score matching (PSM) was adopted to match diabetic and non-diabetic participants in different onset age of diabetes groups to evaluate onset age of diabetes and incident dementia. RESULTS:Compared with non-diabetic participants, diabetes participants had an adjusted hazard ratio (HR) of 1.87 (95 % confidence interval [CI]: 1.73-2.03) for all-cause dementia, 1.85 (95 % CI: 1.60-2.04) for Alzheimer's disease (AD), and 2.86 (95 % CI: 2.47-3.32) for vascular dementia (VD). Among diabetic participants who reported onset age, the adjusted HRs for incident all-cause dementia, AD, and VD were 1.20 (95 % CI: 1.14-1.25), 1.19 (95 % CI: 1.10-1.29), and 1.19 (95 % CI: 1.10-1.28), respectively, per 10 years decrease in age at diabetes onset. After PSM, strength of association between diabetes and all-cause dementia increased with decreasing onset age of diabetes (≥60 years: HR = 1.47, 95 % CI: 1.25-1.74; 45-59 years: HR = 1.66, 95 % CI: 1.40-1.96; <45 years: HR = 2.92, 95 % CI: 2.13-4.01) after multivariable adjustment. Similarly, diabetic participants with onset age <45 years had greatest HRs for incident AD and VD, compared with their matched controls. LIMITATIONS:Our results only reflect the characteristics of UKB participants. CONCLUSIONS:Younger age at diabetes onset was significantly associated with a higher risk of dementia in this longitudinal cohort study.