Bidirectional and dynamic relationships between social isolation and physical functioning among older adults: A cross-lagged panel model of US national survey data.

BACKGROUND Previous studies examining unidirectional relationships between social isolation and physical functioning amongst older adults may be biased due to reverse causality. This study leveraged data from a US national sample of older adults and a novel estimation method to identify bidirectional associations between these two phenomena and their associated temporal dynamics. METHODS The analyses were based on nine waves of panel data from a sample of adults aged 65+ years from the US National Health and Aging Trends Study (n=12,427 individuals) and a cross-lagged panel model. Social isolation was measured using the Social Isolation Index (SII) and physical functioning through the Short Physical Performance Battery (SPPB). RESULTS One SD increase in the SPPB was associated with a -0.013 SD decrease (95% CI: -0.023, -0.004) in the SII a year later, whereas one SD increase in the SII was associated with a larger -0.058 SD decrease (95% CI: -0.069, -0.046) in the SPPB a year later. The SII→SPPB effect lasted for ~5 years, whereas the SPPB→SII lasted for ~3 years. CONCLUSIONS This study confirmed the existence of statistically significant bidirectional associations between social isolation and physical functioning amongst older adults in the US, net of reverse causation. Since the effect of social isolation dominates, the findings indicate that public-health strategies to promote successful aging should prioritise interventions that enrich older adults' social networks. Further, the effect time horizons yielded by the model point to the optimal timing for the re-delivery of interventions.