Abstract
OBJECTIVES: To evaluate whether loneliness was associated with transition rates to and between healthcare settings and death in older adults and to examine whether associations were modified by sex. DESIGN: Retrospective cohort study. SETTING: Ontario, Canada, from December 2008 to February 2020. PARTICIPANTS: Community-dwelling, Ontario respondents (≥65 years) to the 2008/2009 Canadian Community Health Survey-Healthy Aging (CCHS-HA). EXPOSURE: Baseline loneliness was measured with the Three-Item Loneliness Scale in the CCHS-HA. Respondents were classified as not lonely, moderately lonely or severely lonely. PRIMARY OUTCOME MEASURES: Healthcare transitions were assessed over a 12-year period through linkage to health administrative records. Relative rates of transition to and between community, inpatient hospitalisation, home care and long-term care (LTC) settings, as well as death, were estimated using a continuous-time multistate transition framework. Adjusted models were weighted and tested for sex interactions. RESULTS: Of 2671 respondents (weighted n=1 398 180), 20.9% were moderately lonely and 12.0% were severely lonely. Compared with those who were not lonely, moderately lonely respondents transitioned at a faster rate from the community to home care, and severely lonely respondents transitioned at a faster rate from the community to LTC; although, both associations were attenuated with adjustment (RR(ML) 1.37; 95% CI 1.00 to 1.85 and RR(SL) 1.96; 95% CI 0.99 to 4.12, respectively). Female and male respondents had mostly similar transition patterns. CONCLUSIONS: Our findings suggest that loneliness may hasten transitions from the community to home care and LTC settings in older adults, although these transitions are mostly driven by related health and social factors.