Measuring bio-psycho-social frailty in older adults across multiple surveys and indicators: evidence from Italy for public health policy.

Abstract

Detecting and measuring frailty is essential to inform health policy. Numerous frailty tools exist, but their application to population data remains limited. This study focuses on Italy and adopts a policymaker's perspective to identify practical, nationally applicable tools for bio-psycho-social frailty assessment using existing survey data. Three validated bio-psycho-social indicators-the Tilburg Frailty Indicator (TFI), the Groningen Frailty Indicator (GFI), and the Edmonton Frail Scale (EFS)-were operationalized across four population surveys (SHARE, EU-SILC, EHIS, and AVQ). The study proceeded in three steps: (i) reconstruction of each indicator within each dataset to assess items' replicability; (ii) validation of reproducible indicators by examining their psychometric properties; (iii) comparison of prevalence and predictive validity across indicators to illustrate their relative strengths and limitations. The GFI was the most replicable across surveys, while the EFS showed the strongest predictive ability. Among datasets, EHIS emerged as the most suitable platform, offering higher indicators' reproducibility and strong validity. Regional analyses revealed a clear North-South gradient, and regression models confirmed that frailty adds explanatory power for health outcomes beyond chronological age and other sociodemographic characteristics. The study demonstrates that bio-psycho-social frailty can be measured using existing national surveys, providing a low-cost and reproducible framework for national monitoring. By highlighting the added value of frailty over age in predicting health needs, it lays the groundwork for integrating frailty assessment into public health planning and supporting more equitable and effective policy responses to population ageing.