Path model explaining the association between fear of falling and health-related quality of life in (pre-)frail older adults

Fear of falling (FoF) is estimated to be prevalent in over 50% of older adults and several studies suggest that it negatively affects health-related quality of life (HrQoL). Unlike previous studies that examined only few mediating variables, this study aimed to develop a more comprehensive path model explaining the association between FoF and HrQoL. A theoretical path model was developed based on existing evidence and expert feedback and fitted to cross-sectional baseline data on 385 community-dwelling (pre-)frail older adults from the PromeTheus randomized controlled trial using robust weighted least squares estimation. FoF and HrQoL were operationalized by the Short Falls Efficacy Scale International and EQ-5D Index, respectively. The model included potential explanatory pathways through physical activity (German Physical Activity Questionnaire for middle-aged and older adults), physical capacity (Short Physical Performance Battery), physical performance (Late-Life Function and Disability Instrument [LLFDI] function component), disability (LLFDI disability component - short form), and affect (visual analogue scales on 'happiness', 'sadness', 'calmness' and 'tension'). Age, sex, education, and previous falls were considered as covariates. The model demonstrated good fit to the data and the remaining direct effect of FoF on HrQoL was small (β=-0.05). Physical capacity and physical performance were the most important mediators (combined indirect effect of β=-0.17, accounting for > 50% of the total effect). Pathways of minor individual relevance (e.g. through disability or affect) contributed considerably to the total indirect effect when combined. Controlling for sociodemographic data and previous falls only had minor effects on model fit and path coefficients. Physical capacity and physical performance are particularly important levers for reducing the impact of FoF on HrQoL through interventions. However, the other pathways also had a considerable influence when taken together. Hence, research on the association of FoF and HrQol should acknowledge the complexity of causal pathways that may explain this association and not neglect minor pathways. The proposed model should be tested on an alternative sample, using longitudinal data, and extended to include additional explanatory factors (e.g. activity avoidance). German Clinical Trials Register, ID: DRKS00024638, https://drks.de/search/en/trial/DRKS00024638 , date of registration: March 11th 2021.