In a recently published study, Gerontology and Geriatrics Archives, Canadian researchers investigated the individual and combined effects of social isolation (SI) and loneliness (LON) on memory in middle-aged and older adults over a six-year period.
Results showed that experiencing both SI and LON had the greatest negative impact on memory, followed by LON only, SI only, and neither condition, suggesting that addressing both factors simultaneously may improve memory in this population.
study: Exploring the differential effects of social isolation, loneliness, and their combination on memory in older adults: A 6-year longitudinal study from the CLSAImage credit: Perfect Wave/Shutterstock.com
background
Memory involves the encoding, storage, and retrieval of information. Memory decline, common with age, can affect daily activities and may be a sign of neurocognitive disorders such as Alzheimer’s disease.
More than 55 million people worldwide live with these disorders, and the number is expected to grow. Social isolation (SI) and loneliness (LON) are potentially preventable factors that exacerbate age-related memory loss. SI is associated with a lack of structural social support, while LON reflects a subjective dissatisfaction with social connections. Both factors, although distinct, lead to poorer health outcomes.
Existing studies have primarily investigated the effects on global cognition and other cognitive domains, rather than memory, and these studies often overlook the combined effects of SI and LON and focus on older adults, ignoring the middle-aged population.
The combined effects of SI and LON pose unique health risks, including cardiovascular disease and depression, especially among older, widowed, and low-income women.
Longer-term and more comprehensive studies are needed to better understand these relationships and develop effective interventions. Therefore, in the present study, we aimed to fill this gap by using a longitudinal and multidimensional approach to examine the specific and combined effects of SI and LON on memory, particularly in middle-aged and older adults.
About the Research
Data are from the Canadian Longitudinal Study on Ageing (CLSA) follow-up cohort, which initially recruited adults aged 45–85 years between 2011 and 2015, with follow-up data collected in 2015–2018 and 2018–2021. Participants were enrolled from the Canadian Community Health Surveys and additional methods such as mailing and random-digit dialing.
Exclusions included residents of Native American reservations, territories, long-term care facilities, full-time military personnel, non-permanent residents, and people determined to have a cognitive impairment. The cohort was stratified by age, sex, province, education, and distance from the research center, and data were provided via computer-assisted telephone interviews.
A five-point scale was used to measure SI and was divided into non-isolated and isolated categories. LON was assessed using questions from the CES-D-10 and was divided into lonely and not lonely categories. These scales were combined to create four groups: only socially isolated, only lonely, both isolated and lonely, and neither.
Memory was assessed using the modified Rey Auditory Verbal Learning Test (RAVLT), which assessed episodic and working memory with a focus on immediate and delayed recall.
Scores were standardized to Z-scores for the cognitively healthy subset and combined for analysis. Covariates included sociodemographic attributes, functional ability, lifestyle variables, and chronic health conditions.
The analytic sample included 14,658 CLSA participants, and three sets of analyses were conducted: The primary analysis used a modified “all available data” (AADA) approach.
Two sensitivity analyses were performed, one using unmodified AADA and one using multiple imputation. Additionally, statistical analyses included descriptive assessment, mixed-effects regression, and assessment of clinical significance using Cohen’s d.
Results and discussion
Approximately 82.90% of participants were neither socially isolated nor lonely. A minority of participants experienced SI only (7.88%), LON only (7.98%), or both (1.23%). Most participants showed average memory scores throughout the follow-up period. A significant inverse correlation was found between the SI/LON group and memory, with the “neither” group having the highest scores and the “both” group having the lowest scores.
The “neither” group had a better socioeconomic and health profile. In the SI and LON group, memory decline was observed over 6 years. The “both” and “LON only” groups were clinically significant and showed a negative association with memory, whereas the “SI only” group showed no significant effect.
This study uniquely explores the combined effects of SI and LON on memory in middle-aged and older adults, using comprehensive covariates and longitudinal data to capture change over a 6-year period.
However, due to healthier participants and potential bias in LON measurements, our sample may not be representative of the broader population and may underestimate the associations between SI, LON, and memory.
Conclusion
In conclusion, this study suggests that experiencing both SI and LON results in the most severe memory impairments, with the impact of LON being stronger than SI, highlighting the need for targeted interventions.
In the future, programs should address structural barriers for combined groups and provide tailored support for those experiencing only LON or only SI.
It is important to continually monitor transitions between these categories as individuals face challenges related to social participation and cognitive health.
