Abstract
Introduction
Rising concerns about social isolation and loneliness (SIL) began well before the pandemic, with many suggesting that we were facing a loneliness epidemic. While these concerns were shared among researchers and those who serve older adults and other vulnerable populations, these issues were largely underappreciated by the broader public, 1 and even less so by the health care community. Fast forward to the emergence of the Covid-19 pandemic, and suddenly “those at risk” for SIL presumably included everyone. Immediate and widespread “social distancing” policies and practices were implemented nearly globally to reduce social contact to hasten the spread of the virus, but without a critical look at whether these policies aimed to save lives may also increase the risk of earlier death. Concerns about SIL grew exponentially across the globe, with crucial questions raised about whether the pandemic-related restrictions, including forced isolation, would increase loneliness across the population; if so, whether some groups were at greater risk, and what kinds of secondary health effects might occur as a result.
The Costs of Isolation: Lives Influenced
While there continue to be many unanswered questions about the long-term effects of the pandemic, we do have some evidence pointing to the impact on loneliness and who may be at the greatest risk. For example, a meta-analysis of 32 longitudinal studies that collected data at some point before and during the pandemic demonstrates that there were increases in loneliness severity within individuals as well as increases in global prevalence rates of loneliness. 2 This experience of SIL is location agnostic and not unique to the United States. A multi-national study, including over 20 000 participants from 101 countries, further examined who might be most impacted. This data demonstrates that those with significant financial needs, mental health concerns, and those living alone (particularly those not by choice) were more likely to report severe loneliness both before and during the pandemic relative to those without those characteristics. 3 This evidence suggests that loneliness has increased both in prevalence and severity and helps us understand who may be at greatest risk for loneliness and consequently worse health outcomes. Yet, sadly, we cannot adequately answer how this relates to secondary health effects without nationally standardized sources or mechanisms of data collection. Furthermore, in the United States, we still do not have consensus on collecting standardized assessments of SIL in health encounters (despite recommendations to do so4,5), and without this, we will not know the full magnitude of risk and health consequences.
The Costs of Isolation: Health Consequences
Both authors served as committee members on a National Academy of Science Engineering and Medicine expert consensus report focused on the medical and health care relevance of SIL among older adults. 5 Ironically, and perhaps thankfully, this report was published just 2 weeks prior to the World Health Organization declaring a pandemic and immediately prior to shelter in place orders. This report summarizes decades of evidence on the health effects of SIL by researchers spanning continents and fields of study. These data, including systematic reviews and meta-analyses, find substantial evidence that SIL increases the risk for depression, 6 dementia, 7 cardiovascular disease and stroke, 8 and risk for premature all-cause mortality. 9 This report concludes that the evidence points to significant mental, cognitive and physical health morbidities with the strongest evidence associated with risk for premature mortality. 5 Furthermore, a growing body of evidence has documented the influence of SIL on various biomarkers, such as markers of chronic inflammation, 10 that may serve as the underlying mechanisms to explain the associations with poorer outcomes. Thus, prior to the Covid-19 pandemic, we had evidence of the health risks associated with SIL, and yet we were not prepared to appropriately respond to the potential harm that was unfolding before our eyes by enforcing mass isolation. We already knew SIL was a problem, but now we are faced with determining the full scope and ramifications of this problem.
We Cannot Adequately Solve a Problem Without Adequately Measuring the Problem
Measurement in Clinical Settings
What kinds of effects of SIL may be occurring at a population level that we are unaware of because we are not adequately measuring it? Evidence of the secondary mental health effects during the pandemic has become apparent because these have been measured and included in the EHR. For example, early in the pandemic, a study using the data from 61 million adults, obtained from data from the electronic health record, demonstrated that patients with a recent mental health disorder were significantly at greater risk for infection, hospitalization, and death from Covid-19 relative to those with no mental health disorder. 11 Another study examining the electronic health record found that 33% of Covid-19 patients developed a new neurological and psychiatric diagnosis within 6 months. 12 Importantly SIL are different from mental health (eg, depression or anxiety); however, this data has important implications for understanding the potential consequences of SIL and points to missed opportunities. First, well-established bi-directional associations between SIL and mental health outcomes suggest similar associations with the Covid-19 infections may exist with SIL. Second, the scope of secondary health effects due to mass isolation practices and policies are unknown because loneliness or other social indicators are not routinely collected in health assessments. If SIL were in the EHR, direct and indirect associations with COVID-19 infections and other health outcomes could be known.
Measurement of SIL within health care settings are also needed to identify those at risk and those already experiencing SIL, to understand associated health risks and track changes over time. While there is robust evidence of associations between SIL and short and long-term health outcomes, it is not clear the exactly how long 1 can remain isolated or lonely before such health effects emerge. Indeed, based on an evaluation of the evidence an IOM Report concluded that it is critical to include measurement of social connection and isolation in the electronic health records. 4 Alongside this report is ongoing work by the Gravity Project, which provides further guidance on implementing this task. Without a baseline understanding, it is nearly impossible to begin to understand how to intervene and how to prevent morbidity and mortality. Without systematically identifying these social risks, we are not only missing opportunities for affecting people’s lives, but importantly impacting the costs of health care. 13
Population Health Measurement
To better gauge impact more broadly and guide public health efforts, we also need systematic measurement of SIL nationally. For example, the UK National Office of Statistics established a standardized measure of loneliness in 2017, and subsequently were able to map loneliness during the Covid-19 pandemic and where it tended to be worse. These maps can then be used to inform solutions, guide where resources and are best prioritized, and subsequently remeasured to evaluate effectiveness. We need something similar in the US and other nations. For example, national measurement of SIL would facilitate the inclusion of these factors in the Health People 2030 objectives to establish national goals and track progress over time. Standardized and routinely collected national assessments will help us accurately estimate the prevalence rates of isolation and loneliness within the population and whether national and local efforts, societal trends, and policies change these prevalence rates over time.
National assessments of SIL in the US and globally would have been incredibly helpful to have before the Covid-19 pandemic to capture the consequences of reduced social contact, but it is not too late. The world has experienced disruptions across various sectors of society, but perhaps none felt more than education, workplace, and health. Shifts to online education, remote working, and telemedicine are likely to continue in some form and unlikely to return to pre-pandemic levels. To fully understand the effects of these shifts in societal trends, we need to look beyond the evidence of convenience, access, and productivity levels and include adequate measurement of SIL. Measurement allows us to identify features of practices and policies that may be more detrimental to SIL than others.
Conclusion
The full scope and ramifications of the broader effects on population health associated with the pandemic are likely to extend far beyond the official death toll from Covid-19. There are likely deaths that ensued from SIL, yet these were never adequately assessed, nor is it possible to indicate isolation or loneliness as an official cause of death. Many clinicians and non-clinicians saw patients, family members, and friends suffer and die and recognized that it was due to isolation—but cannot directly say so. We need to measure SIL both within health care settings and in population health to establish direct and indirect influences on health, and guide and evaluate efforts to reduce risk. Because restrictions in social contact have been experienced globally to some extent, we must take this opportunity to collect reliable data to understand the reaching long-term effects of isolation and loneliness.