In general, disabilities are considered a consequence of frailty rather than a cause of frailty, whereas in people with intellectual disabilities (ID), disabilities are often lifelong, which could have consequences for the feasibility and validity of frailty instruments. To better understand frailty in people with ID, we compared two broadly used concepts: the frailty phenotype (FP) and the frailty index (FI) taking into account their feasibility (e.g., percentage of participants able to complete the frailty assessments), agreement, validity (based on 5-year mortality risk), influence of motor disability, and the relation between single frailty variables and mortality. The FI and an adapted version of the FP were applied to a representative dataset of 1050 people with ID, aged 50 years and over. The FI was feasible in a larger part of the dataset (94 %) than the adapted FP: 29 % for all five items, and 81 % for at least three items. There was a slight agreement between the approaches (κ = 0.3). However defined, frailty was related with mortality, but the FI showed higher discriminative ability and a stronger relation with mortality, especially when adjusted for motor disabilities. Concluding, these results imply that the used FI is a stronger predictor for mortality and has higher feasibility than our adaptation of the FP, in older people with ID. Possible explanations of our findings are that we did not use the exact FP variables or that the FI includes multiple health domains, and the variables of the FI have lower sensitivity to lifelong disabilities and are less determined by mobility.
Keywords: Frailty; Frailty index; Frailty phenotype; Intellectual disability; Survival.