Social isolation, loneliness, and positive affect before and during the COVID-19 pandemic in very old adults living in Germany: a quasi-experimental multi-sample study.

backgroundSocial relationships and the absence of loneliness are vital protective factors for the mental health of ageing individuals. The COVID-19 pandemic poses a dual threat for older adults, exposing them to health risks from SARS-CoV-2 while also restricting access to social resources through public health measures like social distancing. Positive affect, as an ability to experience positive emotions, may help to adapt to stressful situations. However, few studies have explored the impact of the pandemic on the relationship between social resources and positive affect, especially in the very old population, often due to data limitations. This study aims to investigate the cumulative risks of insufficient social resources on positive affect in very old adults before and during the COVID-19 pandemic.

methodsFor a quantitative examination of very old adults (≥ 80 years), we used a pooled multi-sample design incorporating German Ageing Survey (DEAS) waves 3 to 7 and Old Age in Germany (D80+) study. Stratifying the sample into the pre-pandemic (DEAS waves 3-6) and pandemic (DEAS wave 7 and D80+) periods, we identified 8,751 individuals for analysis, with 1,816 measured during and 6,935 individuals before the pandemic. We applied inverse probability of treatment weighting (IPTW) for the likelihood of being part of the DEAS sample and participating during the pandemic. The product of both IPTWs were used in the mediation analysis to assess the association of isolation on positive affect directly and indirectly through loneliness before and during the pandemic.

resultsThe highest total effect of social isolation on positive affect occurred during the pandemic (0.283***), with loneliness mediating 34%. Before the pandemic, only the indirect effect of loneliness (-0.097***) remained, with an approximate mediation of 70%. IPTWs effectively balanced and reduced selectivity between the time periods, and results remained robust in sensitivity tests including missing value imputation and error-term correlation.

conclusionsCumulative risks during the COVID-19 pandemic escalated, with pandemic responses such as lockdowns increasing the adverse effects of social isolation and loneliness on positive affect of very old adults. Our pooled-multi-sample design allows a quantitative investigation of an understudied population, offering researchers new insights and an empirical strategy to utilize.