Lessons from Sweden: where can older adults shelter from COVID-19?

Hospitalisation and mortality during the COVID-19 pandemic has been concentrated in older adults,1, 2 with devastating outbreaks occurring in congregate settings. Evidence-based mitigation approaches for older adults must tailor strategies to reduce exposure. Therefore, knowing COVID-19 case numbers and mortality across different residential settings is imperative to gain insight into the characteristics of transmission.

In The Lancet Healthy Longevity, Maria Brandén and colleagues use compelling data from a population-based observational cohort to report COVID-19 deaths among older adults during the first epidemic wave in Stockholm, Sweden. The authors categorise all deaths among individuals aged 70 years and older in Stockholm from March 12 to May 8, 2020, using administrative data to link deaths to detailed household characteristics, geographic location, and household size. Older adults living in care homes had the greatest increase in risk of death among all categories of household. Older adults who lived with an adult of working age (<66 years) also had higher risk of COVID-19 mortality compared with those who lived with other older adults. Furthermore, older adults living alone had a similar risk to those living with younger adults. This could be due to a lack of social support for necessary activities, including pharmacy and food shopping, which prevent older adults living alone from physical distancing. Older adults in crowded housing and those in population-dense neighbourhoods also had increased COVID-19-related mortality compared with those living in less crowded or less densely populated settings, respectively. The authors conclude that contact with working-age adults, whether in a household, a care home, or in a neighbourhood with high population density, was associated with higher mortality from COVID-19 among older adults.

There is heightened awareness of care homes as settings with increased COVID-19 risk, but until now, the risks for older adults living in various community settings had not been elucidated. The Swedish data give a detailed and concerning picture of the risk for all older adults across multiple residential contexts. The findings also support a mathematical model of COVID-19 transmission and mortality, which predicted that without significant social distancing in other age groups, sheltering behaviours for older adults are not expected to be protective. The Swedish experience underscores the inadequacy of some current advice for older adults to shelter at home, given the ongoing risk of those in multigenerational households.

These data document the early phase of the pandemic, when there was little understanding of COVID-19 transmission, a scarcity of personal protective equipment in care homes, and no evidence-based therapeutic interventions for older people. Although advances have been made in all of these areas, which are likely to reduce mortality in future cohorts, other strategies are needed to comprehensively reduce risk for older adults. Care homes will require ongoing intense vigilance to control outbreaks, but these data suggest that we must also scrutinise multigenerational households. Despite the considerable emotional and physical support that caregivers and family can offer older adults, isolating at home with younger household members appears to still confer considerable risk for COVID-19 exposure. Older adults living alone also had high mortality, suggesting that isolation without adequate support of wellbeing is not a solution to this quandary either.

The lowest risk of COVID-19 mortality was in older adults whose household included only other older adults (presumed to be couples or family members). A research priority should be understanding the circumstances of social support that protected those households, which could help to identify a pandemic mitigation strategy.

A limitation of these data is that deaths due to COVID-19 could have been misclassified due to testing shortages and policies limiting testing to those with symptoms. The authors mitigate this problem by including all deaths and by comparing death rates to those from 2019, which confirmed their findings. Another limitation is that these data contain little information about older adults living in households that include children, which appears to be uncommon in Stockholm, but is common or the norm in other parts of the world. Living with children seems likely to confer a higher risk to older adults, but further evaluation is required.

The work by Brandén and colleagues offers insight into the devastating impact of COVID-19 among older adults living in settings with close intergenerational contact. The passive stance taken in Sweden to avoid a generalised lockdown resulted in increased deaths, not only in care homes, but also among community-dwelling older adults. As well as the death toll, the additional impacts of COVID-19 on older adults must be acknowledged: lingering health effects in those who survive the illness and the crisis levels of isolation that will lead to long-term loneliness and poorer health. COVID-19 testing is now more accessible, active contact-tracing is occurring, and face coverings are important measures in reducing infection and spread, but these practices are unlikely to prevent transmission within households. Household-level preventive measures will be needed to protect older adults as the pandemic continues. Vaccines and monoclonal antibodies are promising tools that could reduce the risk of infection and spread; older adults and their support networks should be among the earliest recipients once they are available. The Swedish experience shows that COVID-19 mortality will continue to place its heaviest burden on older adults unless a comprehensive public health strategy is implemented going forward.