Health System Responsiveness and Falls Burden in Older Adults: A Multi-Database Global Analysis

Abstract

Background: Falls among older adults impose a growing global burden, yet whether health system responsiveness can causally reduce this burden remains unclear. This study integrates the Global Burden of Disease Study 2021 (GBD 2021), OECD Health Statistics, and World Bank Open Data to evaluate how health system capacity affects falls mortality among adults aged ≥60 years. Methods: We extracted falls-related age-standardized death rates (ASDR) from GBD 2021 (1990–2021, 204 countries), long-term care (LTC) beds and expenditure from OECD Health Statistics (2000–2023, 49 countries), and health expenditure from World Bank Open Data. Multilevel mixed-effects models examined the moderating effect of LTC capacity on falls ASDR and its modification by socio-demographic index (SDI). Panel fixed-effects models identified causal effects, mediation analysis quantified the role of health system quality index (QCI), and SDI-stratified data envelopment analysis (DEA) assessed efficiency frontiers. Individual-level validation used WHO SAGE Wave 1 data from six low- and middle-income countries (n=34,138). Results: Globally, falls incidence among older adults increased from 1990 to 2021 (EAPC=+0.49%, 95% CI: 0.07–0.91), with population aging contributing 59.36% of the mortality increase. Every 10 additional LTC beds per 1,000 population aged 65+ reduced falls ASDR by 9.8% (β=−0.098, 95% CI: −0.152 to −0.044), but only in countries with SDI>0.6 (interaction p=0.008). Panel fixed-effects models showed each 1 log-unit increase in per capita health expenditure reduced ASDR by 7.2% (β=−0.072, 95% CI: −0.118 to −0.026). QCI mediated 34.8% of the SDI–ASDR association. SDI-stratified DEA identified Switzerland, the Netherlands, and Japan as high-SDI efficiency frontiers (community-based care >60%); Chile and Costa Rica as middle-SDI frontiers; and Rwanda and Vietnam as low-SDI frontiers (community health worker models). SAGE data confirmed grip strength decline increased fall risk (OR=1.22 per 5 kg decrease), but did not support the hypothesized buffering effect of health expenditure (interaction p=0.178). Conclusion: Health system responsiveness reduces falls mortality, but its effect is strongly moderated by national development level. High-SDI countries should optimize LTC resource allocation toward community-based care; middle-SDI countries should establish LTC insurance frameworks; low-SDI countries should prioritize primary emergency and rehabilitation capacity rather than replicating high-SDI institutional LTC models.