AbstractBackgroundFrailty—feelings of weakness, fatigue, and lack of physiological resiliency—places older adults at increased risk for Alzheimer’s Disease (AD) and related dementias. A possible mechanism underlying these associations may be the effects of frailty on brain aging prior to old age.MethodA prospective observational study of approximately 502 community‐based men of predominantly European ancestry without dementia at baseline, from the Vietnam Era Twin Study of Aging (VETSA). Frailty index and MRI imaging were collected at three waves at average ages 56, 62, and 68. We created frailty index (FI) scores using 37 items from previously validated cumulative deficit model indices. Imaging measures included brain‐age (BA; Cole et al. 2018), as well as a validated thickness/hippocampal volume AD‐signature (CTS), and a novel gray matter mean diffusivity AD‐signature (MDS), both weighted averages of MRI‐based morphometry or diffusivity in eight AD‐related brain regions. Different longitudinal models were tested using structural equation modeling in Mplus. Models were adjusted for age, race/ethnicity, education, and smoking.ResultFI was highly correlated over time (range rs = 0.69 to 0.77), as were the brain signatures (range rs = 0.52 to 0.85). The best fitting model was the simplex/auto‐regressive model with correlated random intercepts. Figures 1‐3 show significant associations in the covariate‐adjusted models. FI is significantly associated with brain‐age and the cortical thickness/hippocampal volume AD‐signature random intercepts but not the mean diffusivity signature. Associations attenuate when smoking is added to the model. Moreover, when the cortical thickness/hippocampal volume and mean diffusivity AD‐signatures are adjusted for brain‐age then the associations further attenuate and are no longer significant. No cross‐lag associations were significant.ConclusionThe results are striking given the relatively young age of the sample. Across middle age, frailty was associated with worse brain age and a smaller cortical thickness/hippocampal volume AD‐signature. These associations do not appear to be causal, rather biological aging processes running in parallel. Combined with recent evidence showing associations between frailty and AD‐related biomarkers, the results suggest frailty is not just a geriatric disease and support the need for early intervention in biological aging.