Frailty as a Modifier of the Associations Between Vascular Risk Factors and Incident Dementia.

Abstract (English)

BACKGROUND AND OBJECTIVES: The contribution of late-life vascular risk factors to dementia risk remains controversial. Because low blood pressure (BP) has been associated with worse clinical outcomes in frail individuals, we hypothesized hypertension, but not diabetes or smoking, is associated with higher dementia risk in robust than in frail older adults. METHODS: We performed a prospective cohort analysis of the Atherosclerosis Risk in Communities Neurocognitive Study (ARIC-NCS) over 11 years (2011-2022). We included all community-living White and Black participants aged 67-89 years without dementia at baseline visit 5 (2011-2013) from ARIC-NCS field centers (Jackson, Mississippi; Forsyth County, North Carolina; Washington County, Maryland; Minneapolis suburbs, Minnesota). The primary vascular risk factors measured at baseline included elevated BP (systolic BP 120-129 mm Hg and diastolic BP < 80 mm Hg), hypertension (systolic BP ≥ 130 mm Hg, diastolic BP ≥ 80 mm Hg, or use of medication for BP), diabetes (fasting glucose ≥126 mg/dL, nonfasting glucose ≥200 mg/dL, self-reported physician's diagnosis, or use of diabetes medication), and former and current smoking (self-reported). We defined frailty status using the Fried criteria (5 components that include low energy, low physical activity, slowness, weakness, and weight loss). We estimated cause-specific hazard ratios (HRs) of incident dementia (ascertained from in-person neuropsychological assessments, semiannual participant or informant report, or surveillance of claims from hospitalizations and death certificates) with Cox proportional hazards models including a multiplicative interaction between vascular factors and frailty. HRs of dementia were then stratified by frailty (robust [no frailty components present] vs prefrail/frail [at least 1 frailty component present]). RESULTS: There were 377 (15.8%) dementia cases in robust participants (n = 2,383; mean age, 74.2 years; 55.4% female) and 812 (30.0%) in prefrail/frail participants (n = 2,710; mean age 76.4 years; 61.3% female). There was a significant interaction between BP and frailty on dementia risk (<i>p</i> = 0.026). For robust participants, HRs were 1.03 (95% CI 0.65-1.64) for elevated BP and 1.39 (95% CI 1.00-1.94) for hypertension relative to normal BP. For prefrail/frail participants, HRs were 0.68 (95% CI 0.49-0.95) and 0.82 (95% CI 0.66-1.01), respectively. Diabetes and current smoking were associated with higher dementia risk in both robust and prefrail/frail individuals. DISCUSSION: Late-life hypertension was associated with a lower relative risk of dementia in prefrail/frail participants, but the association was positive in robust participants. BP interpretation and management to support brain health in older adults could consider age-related functional status.