Abstract Body (Do not enter title and authors here): Background
Diabetes is associated with increased risks of cognitive dysfunction and dementia. Although genetic risk for dementia is often seen as deterministic, healthy lifestyles may attenuate its impact. We investigated how Life’s Essential 8 (LE8) and genetic risk for dementia jointly affect the risk of mild cognitive impairment, dementia, and MRI-based brain structural outcomes in individuals with diabetes.

Methods
We included 15,613 dementia-free participants with diabetes from the UK Biobank (UKB) and 20,160 from All of Us (AoU). LE8 was determined by eight metrics (nicotine exposure, physical activity, diet, sleep, BMI, blood pressure, glucose, and lipids) and categorized as low, moderate, and high cardiovascular health (CVH). A polygenic risk score (PRS) for Alzheimer’s disease (AD) was used to evaluate genetic risk for dementia, categorized by tertiles (high, moderate, and low). We performed Cox regressions for the association between LE8 and risk of MCI and all-cause or AD dementia, and linear regressions for structural brain outcomes (UKB only), including total brain volume, white or grey matter volume, left or right hippocampal volume, and white matter hyperintensity volume (WMH), adjusted for age, sex, race, Deprivation Index, diabetes duration, diabetes medications, eGFR, and PRS. The interactions between LE8 and PRS regarding the risk of MCI and dementia were tested on both multiplicative and additive scales.

Results
During 13 follow-up years in UKB, 398 MCI and 390 dementia cases were identified in participants with diabetes. In the multivariable-adjusted models, individuals with moderate or high (vs. low) CVH had a lower risk of MCI (Hazard ratio [HR]: 0.82, 95% CI: 0.65–0.99) and dementia (HR 0.93, 0.73–1.17). Increases in LE8 were positively and significantly associated with grey matter volume (adjusted β per LE8 20 increase 0.004, p=0.0002) and negatively associated with WMH (adjusted β -0.06, p=0.0008). In the high PRS group, those with moderate or high CVH (vs. low CVH) had a lower risk of MCI (HR 0.55, 0.34-0.9) and dementia (HR 0.75, 0.47-1.2). P-values for interactions range from 0.00003 to 0.049. AD dementia analysis yields similar results with attenuated statistical significance. Results from AoU (667 MCI and 538 dementia cases during 15 follow-up years) are generally consistent with UKB.

Conclusions
Maintaining optimal CVH may offset the deleterious effects of genetic disposition to dementia in people with diabetes.