Social Isolation in The Elderly: Causes and Consequences

Extensive research in recent decades shows the extreme harm that social isolation wreaks on individuals of all ages; it has been found that the association between social isolation and health is as strong as the evidence that has linked smoking and health. Social isolation can be defined as the “absence of social interactions, contacts, and relationships” with family, friends, neighbors, and with society as a whole. The impact of social isolation in communities is profound: study after study shows that social isolation has a significant association with lower health outcomes and a host of other problems. Isolation is more common now than it used to be, in part due to the lower birthrate and higher rate of divorce of the Baby Boomer generation.

Social isolation in the elderly is particularly challenging because of the specialized needs of older adults; social isolation can either exacerbate or be exacerbated by mental illness, physical ailments, and decreased cognitive function. Holt-Lunstad, et al. also found that those who are socially connected have a 50% reduced chance of early death. One in six Baby Boomers lives alone and about one in 11 Americans age 50 or older has no spouse, romantic relationship, or child. A socially isolated older adult is more likely to be depressed or die by suicide. Death by suicide among older adults is considered to be a significant public health problem, and the aging baby boomer generation has consistently had higher suicide rates than any other cohort. The Centers for Disease Control has found that a key strategy in preventing suicide is “the promotion and strengthening of connectedness at personal, family, and community levels”. Addressing social isolation – and depression – in older adults will not only increase overall well-being but will impact the suicide rate in this population.

An examination of the current environment in Ventura County regarding its homebound elderly provides additional insight into why the proposed project is both important and innovative. The number of County residents aged 75 and older is expected to double to more than 90,000 within the next decade (250,000 age 60 and above). Programming to address the needs of these residents falls far short of the need. Wider acceptance of the need for addressing social isolation as well as a greater commitment to helping our homebound elderly is needed for true broad-based change. Thus, a comprehensive strategy that goes beyond program implementation is imperative. The Ventura County Elderly Social Isolation Project (VCESIP) is a comprehensive strategy for improving services countywide to low-income, homebound elderly. In addition to forming a coalition, VCESIP will work to develop and/or implement in-home therapeutic programs throughout the county. There are existing evidence-based programs that have the potential to be replicated in Ventura County. For example, the Community Aging in Place, Advancing Better Living for Elders (CAPABLE) program developed at Johns Hopkins University combines occupational therapy, nursing, and handyman support and offers up to 10 home visits over 6 months. The Positive Solutions program, in place in San Diego County for ten years now, has been successful in the identification and treatment of socially isolated seniors for depression. Positive Solutions program utilizes the Program to Encourage Active Rewarding Lives (PEARLS), a nationwide evidence-based program treating late-life depression to seniors in their homes. Given the challenges that frail homebound seniors have in accessing programs that can help them, in-home supportive/therapeutic services are an optimal solution for many.

Meta-analytic Review

This meta-analytic review focuses on determining the extent social relationships influence risk for mortality as well as which factors may moderate the risk. This analysis ultimately included the data from 148 studies, which encompassed 308,849 participants.

Sampling. 

Several techniques were utilized to find studies for inclusion in this analysis. Searches were done of studies from January 1900 to January 2007 using several databases, including HealthSTAR, Medline, Mental Health Abstracts, PsycINFO, Social Sciences Abstracts, Sociological Abstracts, Academic Search Premier, ERIC, and Family & Society Studies Worldwide. Multiple search terms were cross-referenced with words related to social relationships. To reduce inadvertent omissions, databases with the most citations were searched two more times. Researchers also manually examined reference sections of past reviews of studies to find articles that had not been identified in the database searches. Finally, letters were sent to authors who had published three or more articles in this area to ensure that the appropriate studies were all included.

Coding and data entry. 

The researchers laid out numerous details and requirements in order to capture the data on which they wanted to focus. For example, mortality from suicide or injury was not included, while baseline health status and pre-existing health conditions were included. OR (odds ratio) data cannot be aggregated in a meaningful way, thus it was put into natural log OR form for analysis and then transformed back to OR for interpretation. Random effects modeling was undertaken to calculate the average effect size of the studies (average OR was 1.5)

Results. 

Findings indicate that there is a 50% increased likelihood of survival for participants with stronger social relationships, a finding that remained consistent across age, sex, baseline health, cause of death, and follow-up period. This particular finding from this study has been used prolifically in academic and other publications that are focused on the impact of social relationships (and/or social isolation) on health. The research literature on social isolation provides strong proof that social isolation truly is a health epidemic in this country, and well-supported statistics such as this link of social relationships to mortality are useful in understanding the seriousness of the issue. For the purposes of this capstone project, this finding is helpful in bolstering the argument for the existence of the problem and the need for intervention.

Bias. 

While this meta-analysis is well-structured and well-respected in the research literature, there are some areas of potential bias to explore. The editors of this article report that this meta-analysis may underestimate the effect of social relationships on mortality. This is due to the fact that most of the studies had simple single-item measures of social isolation rather than complex ones (research indicates that effect sizes tend to be bigger when multiple measures of social relationships are used in a study). Additionally, there is the fact that proving causation between a lack of social relationships and bad health/mortality has some inherent challenges to overcome. The definitions of social relationships are heterogeneous by nature, and Howick, et al argue that the term ‘social relationship’ is a “synthesized construct relative to specific contexts that change over time.” Finally, an aspect not covered in this or other meta-analytic reviews is that some social relationships can shorten lifespan (gang affiliation/membership is one example).

Pantell article

This study is a quantitative research study exploring the relationship between social isolation and mortality using a nationally representative U.S. sample. Social isolation is compared with traditional clinical factors in terms of predicting mortality.

Sampling. 

The sample was made up of 16,849 adults using data from the National Health and Nutrition Examination Survey and the National Death Index. Sampling weights were already included so that the sample was representative of the non-institutionalized civilian population. Some participants from the original data set were excluded due to their low mortality rate/young age and incomplete mortality follow-up data. There were 8,974 women and 7,875 men, the average age being 48.4 years for women and 46.5 years for men. The study population was mostly white non-Hispanic, had 12 or more years of education, middle-level income, and good baseline health.

Data collection and analysis. 

Four predictor variables for social isolation (marital status, frequency of contact with other people, participation in religious activities, and participation in another club/organization) were compared against four comparison predictors (smoking, obesity, elevated blood pressure, and high cholesterol). The Berkman-Syme Social Network Index (SNI) was used to measure the level of social isolation. Kaplan-Meier survival tables and Cox proportional hazards models were utilized to predict mortality according to social isolation, the clinical risk factors, and covariates (age, race/ethnicity, education, income, and baseline health). All models were stratified by gender because previous studies have shown gender differences in the influence of social isolation on mortality.

Results. 

The study finds an increased risk of death among socially isolated men and women. Social isolation also predicted mortality at a higher level than some of the standard clinical risk factors (smoking and high blood pressure). These findings are consistent with the many other studies that support for the need for innovative solutions addressing social isolation, such as my capstone project.

Bias. 

The authors of this study acknowledge several limitations, which could indicate potential areas for bias. Most of the data included in this study were self-reported data from participants. Thus, true levels of social activity may not have been captured accurately. The researchers controlled for a number of potential confounders, but they acknowledge that there were possibly unmeasured confounders that would have then affected the relationship between social relationships and mortality. As with many of the studies looking at the relationship between social isolation and health, it is challenging to control for reverse causality in all instances. There are likely many cases where health conditions led to social isolation rather than the other way around.

Capable Trial

An article written by John Hopkins researchers (from various schools within John Hopkins) outlines the rationale and design of the first clinical trial of a program called CAPABLE (Community Aging in Place, Advancing Better Living for Elders) (Szanton, et. al., 2014). This article does not discuss findings but instead provides an in-depth analysis on the study design. The trial study’s purpose was to evaluate the effectiveness of CAPABLE in reducing disability in participants. This is a randomized controlled trial, two-group and single-blind.

Sampling and recruitment. 

The target population consists of low-income community-dwelling older adults with a self-care disability. Participants needed to be functionally limited but medically stable and with a high enough level of cognition to participate actively in the intervention. The participants were divided into an experimental group (n=150), whose members received ten in-home sessions (6 with the occupational therapist, up to 4 with an RN, and up to $1200 worth of safety and functional repairs to the home by a handyman). These sessions took place over a four-month period. The control group (n=150) also received 10 in-home sessions over a four-month period, but the intervention was different than what the experimental group received.

Participant recruitment is described as a “multi-faceted community effort.” Study coordinators worked with agencies serving seniors such as senior centers and Meals on Wheels to find potential participants. A targeted direct mailing was also sent to low-income neighborhoods with high numbers of seniors. Participants were then stratified by gender and then randomized into two groups using a computer program. A pilot trial had been conducted prior to this larger trial, and in order to determine the sample size of this trial, calculations were made based on the effect size of the pilot.

Data collection. 

A prolific number of measurement tools are included in this study and are administered at various times before, during and after the intervention is implemented:

• Measurement tool
• Purpose
• Short Portable Mental Status Questionnaire
• Assesses cognition level
• Short Physical Performance Battery (SPPB)
• Three objective tests of physical function. The SPPB has been shown to have high reliability.
• Late-Life Function and Disability Instrument
• A self-report measure of disabilities and is correlated with the SPPB and self-report function questions.
• Sociodemographic Questionnaire
• Self-report assessment of basic characteristics.
• Patient Activation Scale
• Measures patient activation in relation to medical visits.
• Patient Health Questionnaire-9
• Assesses for depression. Validated for diagnosing depressing and determining level of severity.
• Brief Pain Inventory
• Measures intensity, stress, and interference with life from pain. Test-retest reliability and inter-rater reliability are strong according to article authors
• CC Home Safety Checklist
• 43 item checklist developed at the CDC.
• EuroQol questionnaire (EQ-5D)
• Health-related quality of life. 5 items.
• Control-Oriented Strategy
• Measures behavioral and cognitive processes that facilitate adaptation to life challenges.

Results. 

While this article does not include the results of the trial, this study and other studies evaluating the efficacy of the CAPABLE program have shown it to be a valuable intervention in helping older adults with some physical limitations. In the context of this capstone project, CAPABLE has been shown to be an intervention that may be very appropriate and relevant in helping Ventura County’s homebound elderly population.

Bias. 

The design of this study is dense and complex. Questions exist around the sampling procedures. For example, it is not clear from the article how potential participants were recruited through social service agencies. Was there a specific process or did agency staff simply recommend to the research team who they thought might benefit from the CAPABLE program? Agency staff bias may have come into play on who was chosen as a good candidate for the trial. Regarding mail recruitment, whoever is motivated to respond, has enough ability to read and respond to mail, or has the means to utilize mail/stamps, etc. may result in a sample that is not representative. Measurement tools used for the most part seem to be robust in terms of reliability and validity.

Conclusion

Any initiative in the area of social isolation, including both its causes and its impacts, benefits greatly from the large body of existing research. There are many well-reputed studies that have been validated by additional research over time. Nonetheless, taking a deeper look into the methods utilized by said studies is a valuable exercise to undertake to fully understand the limitations of studies using even the most robust methods.

References

1. Berkman, L.F., Glass, T., Brissette, I., Seeman, T.E. (2000). From social integration to health: Durkheim in the new millennium. Social Science & Medicine, 51, pp. 843-857.
2. Holt-Lunstad, J., Smith, T., & Layton, J. (2010). Social Relationships and Mortality Risk: A Meta-analytic Review.  PLoS Med 7(7): e1000316. https://doi.org/10.1371/journal.pmed.1000316
3. Howick, J., Kelly, P., Kelly, M. (2019). Establishing a causal link between social relationships and health using the Bradford Hill Guidelines. SSM - Population Health. Volume 8, 100402, ISSN 2352-8273. https://doi.org/10.1016/j.ssmph.2019.100402
4. Pantell, M., Rehkopf, D. Jutte, D., Syme, S.L., Balmes, J., & Adler, N. (2013). Social Isolation: A Predictor of Mortality Comparable to Traditional Clinical Risk Factors. American Journal of Public Health. 2013 Nov; 103(11): 2056–2062.  doi: 10.2105/AJPH.2013.301261
5. Szanton, S., Wolff, J., Leff, B., et al. (2014). CAPABLE trial: a randomized controlled trial of nurse, occupational therapist and handyman to reduce disability among older adults: rationale and design. Contemp Clin Trials. 2014; 38(1): 102-112. doi: 10.1016/j.cct.2014.03.005