BackgroundA systematic evaluation of the relationship between frailty and large artery atherosclerosis (LAA) burden has not been carried out. Here, we aimed to assess the association between these variables in community-dwelling older adults living in Atahualpa (rural Ecuador). MethodsParticipants underwent frailty assessment and determinations of LAA in several vascular beds. Frailty was estimated by the Edmonton Frailty Scale (EFS). LAA was investigated in the peripheral vascular bed by means of ankle-brachial index determinations, in the extracranial carotid bed by B-mode ultrasounds, and in the intracranial bed by high-resolution CT and time-of-flight MRA. Ordinal logistic regression with interaction models were fitted to assess the independent association between levels of cognitive frailty and the LAA burden. Casual mediation and sensitivity analysis, and the E value, evaluated the effect of age in this association. ResultsOut of 331 included individuals, 176 (53%) were robust and the remaining 47% were either pre-frail (n = 78) or frail (n = 77). Atherosclerosis affected only one (any) vascular bed in 111 (34%) individuals, two beds in 75 (23%), and three beds in 22 (7%); the remaining 123 (37%) had no evidence of atherosclerosis. Univariate analysis showed a significant inverse association between the robust status of cognitive frailty and LAA burden (p = 0.006). This association vanishes after considering the effect of covariates. Causal mediation analysis confirms that age captures 51.8% (95% C.I.: 34.6 to 97.2%) of the effect of the association. Sensitivity analysis and E-value computation find that the amount of bias provided by age is enough to explain away the effect estimate. ConclusionsThis study found no independent relationship between cognitive frailty and LAA burden.