1 Department of Psychology, Concordia University, Montreal, Quebec H4B 1R6, Canada.
2 Roger and Flo Lipitz Center to Advance Policy in Aging and Disability, Department of Health Policy and Management, Johns Hopkins Bloomberg School of Public Health, Baltimore, MD21205, USA.
3 Center for Equity in Aging, School of Nursing, Johns Hopkins University, Baltimore, MD21205, USA.
4 Department of Medicine, Division of Geriatric Medicine and Gerontology, Johns Hopkins School of Medicine, Baltimore, MD21205, USA.
5 Caring Futures Institute, College of Nursing and Allied Health, Flinders University, Adelaide, Australia.
6 Department of Epidemiology & Data Science, Amsterdam Public Health Research Institute, Amsterdam UMC - location VU University Medical Center, Amsterdam, the Netherlands.

Background: It is unclear whether social isolation and loneliness may precede frailty status or whether frailty may precipitate social isolation and loneliness. We investigated the reciprocal and temporal sequence of social isolation, loneliness, and frailty among older adults across 21 years.

Methods: We used seven waves of the Longitudinal Aging Study Amsterdam from 2302 Dutch older adults (M = 72.6 years, SD = 8.6, 52.1% female) ages 55 or older. Using random intercept cross-lagged panel models, we investigated between- and within-person associations of social isolation and loneliness with frailty. Frailty was measured using the Frailty Index. Loneliness was measured using the 11-item De Jong Gierveld Loneliness Scale. Social isolation was measured using a multi-domain 6-item scale.

Results: Social isolation and loneliness were weakly correlated across waves. At the between-person level, individuals with higher levels of frailty tended to have higher levels of social isolation but not loneliness. At the within-person level, the cross-lagged paths indicated that earlier frailty status predicted future social isolation and loneliness over time. However, prior social isolation was not associated with subsequent frailty except at time point 5 (T5). Loneliness at specific time points (T1, T4 and T6) predicted greater frailty at later time points (T2, T5 and T7). The results also supported reciprocal and contemporaneous relations between social isolation, loneliness and frailty.

Conclusions: Social isolation and loneliness are potential outcomes of frailty. Public health policies and health practitioners should prioritise interventions targeting social connection among older adults with pre-frailty or frailty.