Health and Health Service Use Among Assisted Living Residents With Chronic or Transient Loneliness: A Retrospective Cohort Study

OBJECTIVES: To identify patterns of loneliness among assisted living (AL) residents using routinely collected resident assessment data, and to examine differences in sociodemographic characteristics, health, and health care use across loneliness patterns. DESIGN: Retrospective cohort study with 1-year follow-up. SETTINGS AND PARTICIPANTS: Residents living in publicly funded AL homes in Alberta, Canada, on December 31, 2018, with at least 2 Resident Assessment Instrument-Home Care assessments between December 1, 2015 and December 31, 2018 (N = 9053). METHODS: Using linked administrative health data, residents were classified as not lonely, transiently lonely (at least 1 lonely assessment not meeting criteria for chronic loneliness), or chronically lonely (2 or more consecutive lonely assessments more than 30 days apart). Demographic characteristics, diagnoses, and functional status were compared across groups. Health care utilization (emergency department visits and hospital admissions) and medication administration were examined during follow-up using descriptive statistics and negative binomial regression models adjusting for age, sex, and cognitive performance. RESULTS: Overall, 64.7% of residents were not lonely, 20.5% were transiently lonely, and 14.9% were chronically lonely. Chronically lonely residents had a higher prevalence of depression (32.9% vs 22.8%), pain (43.5% vs 38%), and psychiatric diagnoses (48.2% vs 44.5%) compared with transiently lonely residents. In adjusted models, both transient (incidence rate ratio, 1.09; 1.01-1.18) and chronic loneliness (incidence rate ratio, 1.21; 1.12-1.32) were associated with increased emergency department use, with chronic loneliness showing a stronger association. CONCLUSIONS AND IMPLICATIONS: Chronic loneliness was common among AL residents and was associated with poorer mental health, pain, and greater health care use. Distinguishing loneliness patterns using routine assessment data may help identify residents at heightened risk for adverse outcomes and inform targeted interventions.