Social isolation change associated with increased risk of mortality, disability, and dementia

By Pooja Toshniwal PahariaReviewed by Danielle Ellis, B.Sc.Jul 29 2024

In a recent study published in JAMA Network Open, researchers evaluated the relationship between social isolation and health outcomes such as cognitive and physical function decline, cardiovascular disease (CVD), stroke, and mortality in older adult residents of the United States (US).

Background

Social isolation is a serious concern among older persons, resulting in a smaller social network and fewer social engagements. This susceptibility is a primary issue in health and social policy debates. Social isolation is associated with poor health outcomes such as higher mortality, cognitive decline, heart disease, and lower physical activity. However, the relationship between social isolation changes and health effects is poorly known.

Most social isolation research includes cross-sectional assessments, which ignore the link between social isolation changes and future health effects. While some studies have shown links, such as functional limits and memory deterioration, this field is still under-researched, making it more difficult to evaluate treatments targeting social isolation prevention or encouraging social interactions to enhance health outcomes.

About the study

In the present study, researchers investigated whether social isolation changes impact long-term health outcomes among US adults aged ≥ 50 years.

The researchers assessed social isolation alterations in four years and their impact on health among 13,649 Health and Retirement Study (HRS) participants surveyed between 2006 and 2020. They analyzed data between 11 October 2023 and 26 April 2024. They stratified the study participants into stable, decreased, or increased social isolation groups by their isolation status at baseline.

The primary exposure included social isolation changes measured using the five-item Steptoe Social Isolation Index (SII) of the Leave-Behind Questionnaires (LBQ) from the baseline evaluation to that conducted four years later. Primary outcomes included mortality, disability, Alzheimer’s disease and related dementia, cardiovascular disease, and stroke.

The researchers assessed dementia, cardiovascular disease, and stroke using Medicare records linked to HRS, classified by the Medicare Chronic Conditions Data Warehouse (CCW) and the International Classification of Diseases, tenth revision (ICD-10) codes. They confirmed mortality and the death year using the Social Security Death Index (SSDI) and National Death Index (NDI). They assessed disability through self-reported daily living activities and dependencies related to walking, eating, bathing, dressing, getting in or out of bed, and toilet use.

The researchers estimated incidence rates (IR) per 1,000 individual years and used Cox proportional hazard regressions and inverse probability-type treatment weighting (IPTW) to determine adjusted hazard ratios (AHR) for analysis. Study covariates included gender, age, race, ethnicity, comorbidities, education, body mass index, marital status, smoking status, baseline year, HRS cohort, self-respondent versus proxy, and total assets. The models also adjusted for Center for Epidemiological Studies-Depression (CES-D) scores, Activities of Daily Living (ADL) scores, and 27-point cognition scores.

The researchers performed sensitivity analyses by defining social isolation changes based on binary social isolation status transitions. Particularly for initially isolated individuals, they categorized change groups as converting from isolation to non-isolation versus remaining isolated. In addition, they excluded people who were deceased within two years of the second evaluation and incorporated HRS weights at study initiation.

Results

The mean age of 13,649 study participants was 65; 59% (n = 8,011) were female, 9,093 (67%) did not have baseline social isolation, and 4,556 (33%) did. Among non-isolated participants, 1,055 (12%) experienced reduced isolation, 4,553 (50%) were stable, and 3,485 (38%) experienced increased social isolation at the subsequent evaluation.

Among baseline-isolated adults, 2,067 (45%) experienced decreased social isolation, 1,796 (39%) were stable, and 693 (15%) were more socially isolated at the subsequent assessment. Compared to individuals with decreased social isolation, the increased socially isolated ones showed an increased likelihood of being female, older, less educated, married, and white of non-Hispanic ethnicity. At study initiation, they were also likely to have more assets, poorer cognitive and physical functions, and increased comorbidities. Over four years, they showed higher incidence rates of stroke, depression, psychiatric disorders, and worsened cognition and physical function.

Among individuals with baseline social isolation, those experiencing increased social isolation had more deaths (IR, 68) than stable individuals (IR, 44) or those with decreased isolation (IR, 38). Increased social isolation was related to higher mortality (AHR, 1.3), dementia (AHR, 1.4), and disability (AHR, 1.4) risks compared to stable isolation. Decreased social isolation was related to a lower mortality risk for individuals without baseline social isolation. Sensitivity analyses yielded similar results.

Conclusion

Overall, the study found that increased social isolation among older adults increased the risk of disability, dementia, and death, independent of baseline social isolation. The findings highlight the necessity of treatments aimed at preventing growing social isolation in older individuals to reduce its negative consequences on mortality and cognitive and physical decline.