Clinical modifiers of the association between type 1 diabetes and dementia incidence.

Background: Type 1 diabetes mellitus (T1DM) is associated with elevated dementia risk, but the mechanisms are not well understood. Prior studies suggest that co-occurring diabetes-related complications and other comorbidities may further increase dementia risk, but these studies are few and typically have small samples. Whether diabetes-related complications and other comorbidities modify the effect of T1DM on dementia risk remains unclear. Methods: Data are from participants of the All of Us (AoU) cohort ages ≥ 50 years, with complete baseline surveys, linked electronic health records (EHRs), and either T1DM or no DM. Enrollment began in 2017, with data available through October 2023, including information prior to enrollment in AoU. Incident dementia was identified based on ICD-9, ICD-10, and SNOMED codes in participants' EHRs. Baseline clinical comorbidities (diabetes complications and eye diseases, other vascular and metabolic comorbidities, and mental health conditions) were also identified using participants' EHRs, classifying each as present if at least one diagnostic code occurred on or before the baseline survey. Results: Among 232,429 participants (mean [SD] age 64.5 [9.0] years; 57.3% women), 2.3% had a T1DM diagnosis. Participants averaged 1.67 total comorbidities (SD = 2.08). T1DM and each comorbidity was associated with higher dementia incidence. T1DM was associated with higher dementia incidence among individuals with no comorbidities (HR = 1.77; 95% CI: 0.95-3.30), though the CI included 1. Each additional comorbidity increased risk (HR = 1.22; 95% CI: 1.19-1.26), with some evidence that the effect of T1DM differed by the number of comorbidities (HR = 0.94; 95% CI:0.86-1.01). The combined estimated effect of T1DM and most comorbidities was less than multiplicative. Depression was an exception; the dementia HR for individuals with both T1DM and depression (HR = 5.47; 95% CI: 4.23, 7.08) roughly reflected what would have been expected based on the HR for T1DM (HR = 2.01; 95% CI: 1.38, 2.92) times the HR for depression among those without T1DM (HR = 2.55; 95% CI: 2.22, 2.95). Conclusion: Our findings suggest that T1DM and common comorbidities independently increase dementia risk, though their combined effects are generally less than multiplicative. However, depression in the context of T1DM is associated with major elevations in dementia risk.