J Am Med Dir Assoc. 2026 Jul 7;27(8):106342. doi: 10.1016/j.jamda.2026.106342. Online ahead of print.
ABSTRACT
OBJECTIVES: To examine the short-term within-individual changes in residents’ depressive and behavioral symptoms following the March 2020 lockdown in US nursing homes.
DESIGN: Retrospective cohort study.
SETTING AND PARTICIPANTS: Nursing home residents in US nursing homes operating from 2019 to 2020.
METHODS: Residents in 2020 were considered exposed to the lockdown, and residents in 2019 were considered not exposed to the lockdown. Using the national Minimum Data Set 3.0, outcomes included changes in depressive symptoms (Patient Health Questionnaire [PHQ]-9 or PHQ-10-OV) and behavioral symptoms (Agitated and Reactive Behavior Scale), examined as continuous score changes and binary adverse severity changes between baseline and follow-up (PHQ-9: n2020 = 499,984, n2019 = 514,378; PHQ-10-OV: n2020 = 47,371, n2019 = 51,398; Agitated and Reactive Behavior Scale: n2020 = 628,332, n2019 = 641,403). Mixed-effects models with facility random intercepts estimated changes in symptom scores and generalized estimating equations with a log link to estimate the risk ratios of adverse severity changes. Statistical interactions between the lockdown and resident cognitive impairment, facility ownership, size, overall ratings, and staffing levels were examined. Sensitivity analyses were conducted among residents in states with state-level visitation bans.
RESULTS: No clinically meaningful within-individual, short-term changes in depressive or behavioral symptom scores were observed. Associations did not meaningfully differ across resident cognitive impairment levels, facility characteristics, or states with visitation bans. For residents with minimal or mild depressive symptoms at baseline, those exposed to the lockdown were 22% and 39% more likely to experience worsening symptoms than those unexposed, whereas those with very severe behavioral symptoms at baseline were 7% less likely to show no improvement. These associations were slightly stronger in states with visitation bans.
CONCLUSIONS AND IMPLICATIONS: Findings should be interpreted with measurement sensitivity, residents’ lived experiences, and nursing home efforts to address social isolation in mind, underscoring the need for more responsive and timely assessments and for balanced approaches that consider both infection control and residents’ psychosocial well-being.
PMID:42413137 | DOI:10.1016/j.jamda.2026.106342
