Increased prevalence of loneliness and associated risk factors during the COVID-19 pandemic: findings from the Canadian Longitudinal Study on Aging (CLSA).

backgroundOlder adults have been disproportionately impacted by COVID-19 and related preventative measures undertaken during the pandemic. Given clear evidence of the relationship between loneliness and health outcomes, it is imperative to better understand if, and how, loneliness has changed for older adults during the COVID-19 pandemic, and whom it has impacted most.

methodWe used "pre-pandemic" data collected between 2015-2018 (n = 44,817) and "during pandemic" data collected between Sept 29-Dec 29, 2020 (n = 24,114) from community-living older adults participating in the Canadian Longitudinal Study on Aging. Loneliness was measured using the 3-item UCLA Loneliness Scale. Weighted generalized estimating equations estimated the prevalence of loneliness pre-pandemic and during the pandemic. Lagged logistic regression models examined individual-level factors associated with loneliness during the pandemic.

resultsWe found the adjusted prevalence of loneliness increased to 50.5% (95% CI: 48.0%-53.1%) during the pandemic compared to 30.75% (95% CI: 28.72%-32.85%) pre-pandemic. Loneliness increased more for women (22.3% vs. 17.0%), those in urban areas (20.8% vs. 14.6%), and less for those 75 years and older (16.1% vs. 19.8% or more in all other age groups). Loneliness during the pandemic was strongly associated with pre-pandemic loneliness (aOR 4.87; 95% CI 4.49-5.28) and individual level sociodemographic factors [age < 55 vs. 75 + (aOR 1.41; CI 1.23-1.63), women (aOR 1.34; CI 1.25-1.43), and no post-secondary education vs. post-secondary education (aOR 0.73; CI 0.61-0.86)], living conditions [living alone (aOR 1.39; CI 1.27-1.52) and urban living (aOR 1.18; CI 1.07-1.30)], health status [depression (aOR 2.08; CI 1.88-2.30) and having two, or ≥ three chronic conditions (aOR 1.16; CI 1.03-1.31 and aOR 1.34; CI 1.20-1.50)], health behaviours [regular drinker vs. non-drinker (aOR 1.15; CI 1.04-1.28)], and pandemic-related factors [essential worker (aOR 0.77; CI 0.69-0.87), and spending less time alone than usual on weekdays (aOR 1.32; CI 1.19-1.46) and weekends (aOR 1.27; CI 1.14-1.41) compared to spending the same amount of time alone].

conclusionsAs has been noted for various other outcomes, the pandemic did not impact all subgroups of the population in the same way with respect to loneliness. Our results suggest that public health measures aimed at reducing loneliness during a pandemic should incorporate multifactor interventions fostering positive health behaviours and consider targeting those at high risk for loneliness.