Cutoff points for grip strength with a credibility interval as an alternative method to determine frailty assessment in older adults

Casado-Pena J.P.; Arrunategui-Salas G.; Montufar-Leon C.; Rosas V.G.S.Y.; Bayes C.L.; Valdivieso L.; Vieira L.H.P.; Casado F.L. · Human Movement (2026) · DOI: 10.5114/hm/214838

Resumen

Purpose. Grip strength cut-off points are often used to measure frailty, but they are typically reported without an assessment of uncertainty. The Bayesian inference strategy presented in this paper can help geriatricians determine a patient’s frailty status and recommend appropriate therapy to address frailty or prevention strategies to mitigate it, particularly when datasets are too small or have limited diversity to establish population-specific classifications of clinical variables. Methods. The research analysed data from a hybrid cohort from a cross-sectional observational study comprising 450 Peruvian older adults (18%, data collected in the present work) and 2,062 Mexican older adults (82%, data openly available). Peruvian participants were enrolled in a geriatric hospital in Lima. Grip strength was computed using a hand dynamometer, while walking speed was determined through the 4-metre walk test (4MWT). Data for the Mexican samples were also retrieved and re-analysed from the Multi-Country Study on Global Aging and Adult Health 2009/10. Results. With a Bayesian approach, we estimated that the estimated cutoff points for older adults were 22.20 kgF (95% CI: 16.83–30.33) and 30.84 kgF (95% CI: 30.00–35.96) for males, and 15.18 kgF (95% CI: 14.02–18.66) and 22.88 kgF (95% CI: 18.72–25.41) for females. Conclusions. The cutoff points calculated in the present study through Bayesian inference can be useful when qualifying the outcomes of grip strength to assist measure the frailty of older subjects from certain Latin American regions. The cutoff points calculated in the present study through Bayesian inference can be useful when qualifying the outcomes of grip strength to measure the frailty of older subjects from some Latin American regions (i.e., Mexico and Peru). Here, the results likely apply to the former to a greater extent than to the latter. © Wroclaw University of Health and Sport Sciences.

Instituciones: Pontificia Universidad Católica del Perú