Social Isolation and Cognitive Function in Appalachian Older Adults

Abstract

Objective:

Investigating the relation between social isolation and cognitive function will allow us to identify components to incorporate into cognitive interventions.

Method:

Data were collected from 267 Appalachian older adults (M = 78.5, range 70–94 years). Overall cognitive functioning and specific cognitive domains were assessed from data of a self-assembled neuropsychological battery of frequently used tasks. Social isolation, social disconnectedness, and perceived isolation were measured from the Lubben Social Network scale-6.

Results:

Results indicated a significant positive association between all predictor variables (e.g., social isolation, social disconnectedness, and perceived isolation) and outcome variables (e.g., overall cognitive function, memory, executive functioning, attention, and language abilities). Perceived isolation accounted for nearly double the amount of variance in overall cognitive functioning than social disconnectedness (10.2% vs. 5.7%).

Discussion:

Findings suggest that social isolation is associated with poorer overall cognitive functioning and this remains true across varied cognitive domains.

Keywords

older adults social isolation social disconnectedness perceived isolation cognitive function rural

Social isolation can have devastating effects on older adults, such as causing poor health status and lessened health-related quality of life (Hawton et al., 2011). Some research suggests that social isolation may become more prevalent with increasing age (Routasalo, Savikko, Tilvis, Strandberg, & Pitkala, 2006). In fact, 7.9% of older adults report some social isolation, with 4.8% being isolated or very isolated (Hawthorne, 2008). These rates may be dramatically exacerbated for older adults living in remote or rural settings. For example, 49% of a subsample of rural older adults reported being both socially isolated and lonely (Havens, Hall, Sylvestre, & Jivan, 2004). This burden is expected to increase substantially because the older adult segment of the U.S. population is the fastest growing demographic (United States Census Bureau, 2009). Consequently, there is increased need to understand the potential health consequences for older adults that may be associated with social isolation. This article focuses on investigating potential relations between social isolation and cognitive function in Appalachian older adults. It is important to know the association between social isolation and cognitive function because it will allow us to identify components to incorporate into cognitive interventions.

A variety of indicators and conceptual definitions have been used in social isolation research. For example, the diversity of indicators used by prior research to define social isolation includes: having a small social network (McPherson, Smith-Lovin, & Brashears, 2006), living alone or being unmarried (Schmaltz et al., 2007), a perceived lack of social network (Giuli et al., 2012), and feelings of loneliness (Tomaka, Thompson, & Palacios, 2006). Cornwell and Waite (2009) suggest that these indicators of social isolation can be combined into two distinct concepts: social disconnectedness and perceived isolation. Confirmatory factor analyses of a wide range of indicators of social isolation among older adults from the National Social Life, Health, and Aging Project revealed that social disconnectedness is comprised of restricted social network (e.g., social network size) and social inactivity (e.g., socializing with family and friends). Likewise, perceived isolation appeared to be comprised of two dimensions: lack of support (e.g., relying on and opening up to friends, family, and spouse) and loneliness (e.g., feeling isolated). These two concepts (perceived isolation and social disconnectedness) have a weak to moderate positive correlation (r = .25), such that individuals who are more socially disconnected tend to feel lonelier (Cornwell & Waite, 2009).

Social isolation has been found to be associated with negative mental and physical health outcomes. For example, perceived social isolation predicts cardiovascular disease (O’Luanaigh et al., 2012), sleep fragmentation (Kurina et al., 2011), and systematic inflammation (Cole, Hawkley, Arevalo, & Cacioppo, 2011). Social isolation has even been found to be a risk factor for cognitive impairment, which is prevalent in 3–42% of older adults (Ward, Arrighi, Michels, & Cedarbaum, 2012). Specifically, social isolation has been found to be associated with increased risk of developing dementia and a more rapid decline in global cognition, semantic memory, perceptual speed, and visuospatial ability (Berkman, Glass, Brissette, & Seeman, 2000; O’Luanaigh et al., 2012). Wilson et al. (2007) conducted a longitudinal study and found that loneliness did not significantly change, but loneliness increased decline and development of Alzheimer’s disease (AD) at a 4-year follow-up. In addition to declines in memory, social isolation has been found to be associated with declines in executive functioning and attentional self-regulatory processes (Cacioppo et al., 2000).

Unfortunately, the theoretical mechanism underlying the association between social isolation and cognitive functioning is largely unknown. Cacioppo and colleagues (2011) hypothesize that social isolation may influence cognitive functioning by activating neurobiological mechanisms that stimulate the hypothalamic–pituitary–adrenal axis and diminish sleep quality. The researchers propose a model by which perceived social isolation triggers implicit hypervigilance for social threat, which in turn produces attentional confirmatory and memory biases (i.e., viewing one’s social environment as threatening). Consequently, lonely individuals then willfully recede further from society; a behavior perceived as self-preserving but is ultimately self-defeating. Further withdrawal serves to confirm their initial negative social biases, which then activate neurobiological mechanisms. Therefore, Cacioppo and colleagues (2011) suggest that an interaction of chronic social isolation, hypervigilance of social threat, and activation of neurobiological mechanisms may produce heightened cognitive load and diminished cognitive abilities.

On the other hand, Fratiglioni, Paillard-Borg, and Winblad (2004) suggests that social factors combine with mental and physical activity to explain lifestyle components that have beneficial effects on cognition. They propose that these lifestyle components share a common pathway that converges within three major hypotheses for dementia and AD: the cognitive reserve hypothesis, the vascular hypothesis, and the stress hypothesis. The cognitive reserve hypothesis posits that individuals possess varying capacity to resist AD symptomology. Although the cognitive reserve hypothesis does not hold true for everyone, it could explain the lack of relation between the severity of brain pathology and the clinical manifestation in some individuals (Katzman, 1993). Specifically, greater positive lifestyle components (i.e., social, physical, and mental factors) increase cognitive reserve by making the individual more resilient to neuropathological damage. On the other hand, the vascular hypothesis suggests that the lifestyle components may provide beneficial effects on cardiovascular disease and stroke (Rosengren, Wilhelmsen, & Orth-Gomer, 2004), which in turn are risk factors for cognitive decline. Finally, the stress hypothesis suggests that active lifestyles will inherently provide more opportunities for social engagement, which may lead to lower stress. This idea is derived, in part, from the observation that long-term exposure to stress increases the risk of dementia (Wilson et al., 2003).

It is not clear which aspects of social isolation are associated with overall as well as specific cognitive domains. The uniqueness of the study is as follows: First, this study uses the abbreviated Lubben Social Network scale-6 (LSNS-6; Lubben, 1988), which is considered to be a validated screening measure for social isolation (Gottlieb & Bergen, 2010). Second, this data set includes a comprehensive assessment of major domains of cognitive function, such as memory, executive functioning, attention, and language abilities. Third, this subject pool constitutes a doubly underrepresented population in research: rural and older.

This study used data collected at West Virginia University Center on Aging (Wu et al., 2010). Consistent with prior research (Wilson et al., 2007), the hypothesis for this study is that social isolation will have a significant positive association with cognitive functioning. Therefore, older rural adults with greater social isolation will have poorer overall cognitive functioning. Similarly, it is hypothesized that social isolation will be associated with poorer memory, executive functioning (Cacioppo & Hawkley, 2009), attention (Cacioppo et al., 2000), and language abilities. We also plan to investigate the following secondary hypotheses: Social disconnectedness and perceived isolation will have a significant positive association with cognitive functioning as well as with the specific cognitive domains.

Method

Study Sample

Data were collected from 267 community-dwelling older adults. Individuals in West Virginia who were dentate (i.e., at least four natural teeth) and aged 70 and above were eligible to participate. The participants ranged in age from 70 to 94, with a mean age of 78.5. The sample consisted of 173 women and 94 men. Study participants were heterogeneous with respect to geographic location and sociodemographic characteristics (Table 1). With respect to health, the participant identified an informant (e.g., spouse, child, etc.) who reported the following percentages of vascular risk factors prior to testing: 10.5% had a stroke, 64.6% had hypertension, 17.4% had diabetes, 27.4% had cancer, 14.4% had thyroid disease, and 51.3% had hypercholesterol. The majority of the sample self-reported their general health as good, very good, or excellent (82%). Additionally, 23% of the participants in the study were cognitively impaired as determined by a neuropsychologist’s interpretation of neuropsychological scores.

Table 1.

Participant Demographics (N = 267).

Variable name	Mean (range)	Percentage
Age	78.5 (70–94)
Female		64.8
White		95.9
Marital status
Never married		1.5
Married		42.1
Divorced/separated		9.8
Widowed		46.2
Highest level of education
Elementary school or less		5.7
Some high school		11.0
High school		31.2
Some college		21.7
College		17.9
Graduate degree or above		12.5
Household income
Under $10,000		8.2
$10,000–$19,999		30.0
$20,000–$29,999		18.5
$30,000–$39,999		12.8
$40,000 or above		30.5
Vascular risk
0–2		71.7
3 or above		28.2
GDS score
0–2		83.9
3 or above		16.0
Cognitive task scaled scores
CVLT-II Word List Memory	−.10 (−3 to 3.5)
CVLT-II Recall	.34 (−2.5 to 3)
CVLT-II Recognition	−.42 (−5 to 1)
Rey-O Copy	8.69 (1–14)
Rey-O Immediate Recall	9.29 (3–17)
Rey-O Delay Recall	9.52 (4–17)
Trail Making A	9.93 (2–18)
Trail Making B	8.56 (2–18)
COWAT	−.26 (−2.67 to 2.67)
BNT	−.56 (−7.98 to .71)

Note. N = total sample size; BNT = Boston Naming Test-2nd Edition; COWAT = Controlled Oral Word Association Test; CVLT-II = California Verbal Learning Test–Second Edition–Short From; GDS = Geriatric Depression scale; Ret-O = Rey–Osterrieth Complex Figure.

Multiple strategies were used to recruit participants, including educational presentations, senior center sign-up, and regional data collection sessions. Caregivers and senior center directors and members were presented with a description of the study and were then urged to discuss participation with anyone whom they felt might fit the desired participant profile. Those who expressed interest, met with or provided contact information to project staff, so they could be given further detailed information about study participation and screened for eligibility. Study participants were recruited from 14 counties within West Virginia. Data collection sessions were conducted at 18 sites, including 12 senior or community activity centers, 3 dental or health clinics, 2 assisted living facilities, and 1 retirement housing community.

Measures

The participant survey and battery of neuropsychological instruments were administered in person by a trained psychometrician. The dependent measures of cognitive functioning were assessed using results from the neuropsychological battery. A neuropsychologist constructed the neuropsychological battery. It consisted of the following tests, administered to each participant in the subsequent order: Rey-Osterrieth Complex Figure (Rey-O; Osterrieth 1994; Rey, 1994), California Verbal Learning Test-2nd edition Short From (CVLT-II; Delis, Kaplan, Kramer, & Ober, 2000), Trail Making Test A and B (Reitan, 1958), Boston Naming Test-2nd edition (BNT; Kaplan, Goodglass, & Weintraub, 1983), North American Adult Reading Test (NAART; Blair & Spreen, 1989), Controlled Oral Word Association Test (COWAT; Benton & Hamsher, 1989), and Animal Naming Test (Barr & Brandt, 1996). The independent measure of social isolation was measured from structured questions asked as a part of the participant survey using the LSNS-6 (Lubben, 1988).

Assessment of cognitive function

The assessment of cognitive function was acquired from six tasks in the neuropsychological battery: two measures of memory including Word List Memory, Recall and Recognition as a portion of the CVLT-II as well as the Rey-O; two measures of executive functioning including Trail Making B and COWAT; one measure of attention including Trail Making A; and one measure of language abilities including the BNT. Cognitive function was assessed by two means: a total score of all six tasks and a set of scores for each of the specified cognitive domains. Studies that assess multiple domains are important because dementia criteria consist of a loss of memory and cognitive impairment in at least two cognitive domains that cause impaired functioning in daily living (Roman et al., 1993). As such, it would be beneficial to examine both overall cognition and which specific domains of cognitive function are associated with social isolation. Overall cognitive function was based upon results of all six tasks by acquiring the raw scores on each test. Raw scores were then converted to scaled scores, using the baseline mean and standard deviation (SD) in the population. The scaled scores were then averaged and standardized (Wilson et al., 2005). In addition, the same procedure was used to construct a set of scores for memory (two tasks), executive functioning (two tasks), attention (one task), and language abilities (one task).

Assessment of social isolation

Social isolation was measured using the abbreviated LSNS-6, which can be dividied into dimensions of social disconnectedness and perceived isolation. Social disconnectedness can be assessed by the size of the participant’s active social network (e.g., Items 1 & 4: How many relatives/friends do you see or hear from at least once a month?). Perceived isolation can be assessed by perceived support network (e.g., Items 2 & 5: How many relatives/friends do you feel close to such that you could call on them for help?) and perceived confidence in network (e.g., Items 3 & 6: How many relatives/friends do you feel at ease with whom you can talk about private matters?). Each question is scored on a 0–5 scale, with responses as none (0), one (1), two (2), three or four (3), five through eight (4), or nine or more (5). The social isolation score is the sum of these six questions. Therefore, scores range from 0 to 30 with greater scores indicating less social isolation. A cutoff score of 12 on the LSNS-6 is indicative of individuals that are socially isolated (Lubben, Blozik, & Gillmann, 2006; Rubinstein, Lubben, & Mintzer, 1994). Approximately 12% of our sample was at or below this cutoff score.

Social disconnectedness score is the sum of Items 1 and 4, with scores ranging from 0 to 10 with higher scores indicating greater social connectedness. Perceived isolation score is the sum of Items 2, 3, 5, and 6, with scores ranging from 0 to 20 with higher scores indicating less perceived isolation. To evaluate internal consistency, a Cronbach’s α coefficient was calculated for the LSNS-6. There was an overall Cronbach’s α of .78, with a .73 for the perceived isolation subscale (4 items) and .37 for the social disconnectedness subscale (2 items). Such results are similar to previous studies which found an overall Cronbach’s α of .83 (Lubben et al., 2006). The Cronbach’s α are likely smaller for the subscales because of the limited number of items available for each concept. Because Cronbach’s α values are sensitive to the number of items in the scale, it is suspected that social disconnectedness is low because it is compromised of only 2 items.

Demographic variables

Demographic variables included age (in years), sex (male = 1, female = 2), education (in years), marital status (never married = 1, married = 2, divorced/separated = 3, and widowed = 4), total annual income (under $10,000 = 1, $10,000–$19,9999 = 2, $20,000–$29,999 = 3, $30,000–39,999 = 4, $40,000–$49,999 = 5, $50,000 or above = 6), and race (White = 1, non-White = 2). Vascular risk factors consisted of diabetes, hypertension, hypercholesterolemia, stroke, cardiac surgery, cancer, and thyroid disease (Black, 1992). If the participant had the condition they received a score of one and a score of zero if they did not. A composite score of vascular risk was acquired by summing of scores from the seven conditions. Therefore, the range of scores was 0–7, with high scores suggesting increased vascular risk. Depressive symptoms were assessed by the Geriatric Depression scale (GDS-15; Yesavage et al., 1983). Scores ranged from 0 to 15, with higher scores indicating greater levels of depression.

Data Analysis

To test the main and secondary hypotheses, a series of linear regression analyses were used to examine the association of the predictor variables (i.e., independent variable) with the outcome variables (i.e., dependent variables). Each of these linear regression analyses were repeated controlling for self-reported variables that were potential confounds because they had associations with cognitive function. Therefore, the relation between the demographic variables and outcome variables were investigated using Pearson product–moment correlation coefficients. Covariates were chosen based on the number of significant correlations with the outcome variables. Hierarchical linear regression analyses were run with chosen covariates entered together on Step 1 and social isolation entered on Step 2. In addition, to determine whether there was an interaction between covariates and independent variables (i.e., social isolation, social disconnectedness, and perceived isolation) in predicting cognitive functioning, these variables were combined into a new variable, and a multiple regression analysis was run using all three variables. The significance of the association between the variables were tested with an α = .05.

Results

Social isolation ranged from 4 to 30 (M = 19.45, SD = 5.82), social disconnectedness ranged from 2 to 10 (M = 7.77, SD = 1.95), and perceived isolation ranged from 2 to 20 (M = 11.65, SD = 4.39). Overall cognitive functioning ranged from −3.35 to 2.13 (M = .000, SD = 1.00; see Table 1 for mean scaled scores of cognitive tasks ). Given the significant correlations with outcome variables, participants’ age, education level, income, vascular risk, and GDS scores were used as covariates in the following analyses. The correlations of demographic variables with outcome variables are presented in Table 2.

Table 2.

Correlations of Demographic Variables With Outcome Variables.

	Age	Gender	Race	Marital status	Education level	Income	Vascular risk	GDS
Overall cognitive function	−.24^*	.12	−.06	−.11	.28^*	.26^*	−.19^*	−.21^*
Memory	−.14^*	.25^*	−.15^*	−.10	.26^*	.24^*	−.12	−.20^*
Attention	−.20^*	−.00	−.01	−.08	.15^*	.14^*	−.09	−.21^*
Language abilities	−.26^*	.09	−.08	−.14^*	.14^*	.19^*	−.21^*	−.07
Executive functioning	−.12	.08	.03	−.09	.34^*	.26^*	−.18^*	−.18^*

Note. GDS = Geriatric Depression scale.

*p < .05.

Main Hypothesis

There was a significant positive association between social isolation and overall cognitive functioning (β = .32, SE = .06). In addition, social isolation accounted for 10.1% of the variance in overall cognitive functioning. After controlling for covariates, a significant positive association was maintained between social isolation and overall cognitive functioning (β = .25, SE = .07). However, in this analysis, the association of social isolation with overall cognitive functioning was reduced by about 21%. Social isolation also had a significant positive association with all four cognitive domains: memory (β = .29, SE = .06), executive functioning (β = .24, SE = .06), attention (β = .30, SE = .06), and language abilities (β = .18, SE = .06). After controlling for covariates, results indicate that social isolation maintained a significant positive association with three of the cognitive domains: memory (β = .25, SE = .07), executive function (β = .21, SE = .07), and attention (β = .25, SE = .08). However, the positive association (β = .12, SE = .08) between social isolation and language abilities became nonsignificant (p = .11). In addition, the associations of social isolation with memory were reduced by about 15%, executive functioning was reduced by 16%, and attention was reduced by about 17%. None of the social isolation by covariate interactions were significant for the above analyses. The results of these regression analyses are presented in Table 3.

Table 3.

Regression Analyses of the Association Between Social Isolation and Outcome Variables.

	Model A				Model B
Outcome	R ²	ΔR ²	F	β	R ²	ΔR ²	F	β
Overall cognitive function	.10		26.81^*	.32^*	.26	.06	6.99^*	.25^*
Attention	.09		25.19^*	.30^*	.16	.06	5.08^*	.25^*
Memory	.09		23.11^*	.29^*	.19	.06	5.35^*	.25^*
Executive function	.06		15.59^*	.24^*	.21	.04	6.98^*	.21^*
Language abilities	.03		8.59^*	.18^*	.15	.01	4.86^*	.12

Note. Model A = Regression analyses without covariates; Model B = Regression analyses controlling for covariates.

*p < .05

Secondary Hypotheses

Social disconnectedness

Given the low Cronbach’s α on social disconnectedness, the following results should be interpreted with caution. There was a significant positive association between social disconnectedness and overall cognitive functioning (β = .24, SE = .06). In addition, social disconnectedness accounted for 5.7% of the variance in overall cognitive functioning. After controlling for covariates, a significant positive association was maintained between social disconnectedness and overall cognitive functioning (β = .21, SE = .07). However, this association was reduced by about 14%. Social disconnectedness also had a significant positive association with all four cognitive domains: memory (β = .26, SE = .06), executive functioning (β = .15, SE = .06), attention (β = .21, SE = .06), and language abilities (β = .16, SE = .06). After controlling for covariates, results indicate that social disconnectedness maintained a significant positive association with two of the cognitive domains: memory (β = .25, SE = .07) and attention (β = .18, SE = .08). However, social disconnectedness no longer had a significant positive association with executive functioning (β = .14, SE = .07, p = .06) and language abilities (β = .13, SE = .07, p = .09). In addition, the association of social disconnectedness with memory was reduced by about 7% and attention was reduced by 15%. None of the social disconnectedness by covariate interactions were significant for the above analyses.

Perceived isolation

There was a significant positive association between perceived isolation and overall cognitive functioning (β = .32, SE = .06). In addition, perceived isolation accounted for 10.2% of the variance in overall cognitive functioning. After controlling for covariates, a significant positive association was maintained between perceived isolation and overall cognitive functioning (β = .24, SE = .07). However, this association was reduced by about 24%. Perceived isolation also had a significant positive association with all four cognitive domains: memory (β = .28, SE = .06), executive functioning (β = .26, SE = .06), attention (β = .31, SE = .06), and language abilities (β = .18, SE = .06). After controlling for covariates, results indicate that perceived isolation maintained a significant positive association with three of the cognitive domains: memory (β = .22, SE = .07), executive functioning (β = .22, SE = .07), and attention (β = .25, SE = .08). However, perceived isolation no longer had a significant positive association with language abilities (β = .11, SE = .08, p = .15). In addition, the associations of perceived isolation with memory were reduced by about 18%, executive function was reduced by about 17%, and attention was reduced by 18%. None of the perceived isolation by covariate interactions were significant for the above analyses. The results of the secondary analyses are presented in Table 4.

Table 4.

Regression Analyses of the Association Between Social Disconnectedness and Perceived Isolation With Outcome Variables.

	Perceived isolation
	Model A				Model B				Model A				Model B
Outcome	R ²	ΔR ²	F	β	R ²	ΔR ²	F	β	R ²	ΔR ²	F	β	R ²	ΔR ²	F	β
Overall cognitive functioning	.06		14. 27 ^*	.24^*	.24	.04	8.26^*	.21^*	.10		26.67^*	.32^*	.26	.06	8.92^*	.24^*
Executive functioning	.02		5.79^*	.15^*	.19	.02	6.08^*	.14	.07		18.07^*	.26^*	.21	.04	7.16^*	.22^*
Attention	.05		12.01^*	.21^*	.13	.03	4.10^*	.18^*	.10		26.59^*	.31^*	.16	.06	5.13^*	.25^*
Memory	.07		18.30^*	.26^*	.19	.06	4.99^*	.25^*	.08		20.06^*	.28^*	.18	.05	5.89^*	.22^*
Language abilities	.03		6.63^*	.16^*	.16	.02	4.95^*	.13	.03		8.54^*	.18^*	.15	.01	4.77^*	.11

Note. Model A = Regression analyses without covariates; Model B = Regression analyses controlling for covariates.

*p < .05.

Discussion

Overall, the results of this study demonstrate that social isolation has a significant positive association with overall cognitive functioning, memory, executive functioning, attention, and language abilities. These findings are in agreement with prior studies (Berkman et al., 2000; O’Luanaigh et al., 2012) and our hypothesis that greater social isolation will be associated with poorer overall cognitive functioning. Controlling for covariates decreased, but did not render nonsignificant, the association of social isolation with overall cognitive functioning and with each of the specified cognitive domains.

Data analyses were also conducted to examine the associations between two distinct dimensions of social isolation (i.e., social disconnectedness and perceived isolation) and cognitive function. Both factors were found to have a significant positive association with overall cognitive functioning and with the specified cognitive domains, in support of our proposed hypotheses. After controlling for covariates, all significant associations between the predictor variables and outcome variables were attenuated. In four cases (social isolation and language abilities; social disconnectedness and language abilities; social disconnectedness and executive functioning; perceived isolation and language abilities), the associations were reduced to nonsignificance. Given that the associations between predictor variables and language abilities often become nonsignificant when covariates were entered into the model, these covariates likely account for language abilities beyond that of the predictor variables. Prior research, which indicates that education and age are significantly associated with older adult’s performance on language tasks (Snitz et al., 2009), supports this notion.

Results further suggest that perceived isolation accounted for nearly double the amount of variance in overall cognitive functioning than social disconnectedness (10.2% vs. 5.7%). Perceived social isolation is more closely related to the quality of social interactions, whereas social disconnectedness is more closely related to the quantity (Hawkley et al., 2008). As such, our results suggest that the quality of social interactions is more important than quantity in cognitive function outcomes of older adults. The general assumption is that the quality of social contact is more important for well-being than the quantity because more frequent contacts are not always supportive (Rook & Pietromonaco, 1987). Consistent with our results, Wilson et al. (2007) found that social network was not related to incident of AD, whereas perceived loneliness was associated with cognitive decline and development of AD.

Many explanations may be offered in attempt to define the associations between social isolation and cognitive function. For example, the dependency of our variables may be reversed, in that the cognitive capability of an individual determines their engagement in social activities. Alternatively, less social isolation may simply indicate a positive lifestyle (physical activity or engaged lifestyle) in general, which could be more highly related to overall mental health status. Furthermore, the results suggest that social isolation is related to depression level, therefore, depressive affect may precede increase social isolation. Many social factors, such as stress or life strain, negative life events, and poor social support may also influence an older adult’s perception of social isolation and ability to cope with such events. In addition, social isolation may explain decline in cognitive functioning through the proposed theoretical mechanisms, such as activating neurobiological mechanisms, the cognitive reserve hypothesis, the vascular hypothesis, or the stress hypothesis (Cacioppo, Hawkley, Norman, & Bernston, 2011; Katzman, 1993; Rosengren et al., 2004).

Study Limitations

Although the results of the current study are informative, there were several limitations. First, the study design was cross-sectional. Longitudinal studies will be needed to elucidate the direction of the association between social isolation and cognition. However, longitudinal studies require special attention to sample size because of potential dropout problems due to morbidity and mortality (Souder, 1992). Second, the analyses are based upon a convenient sample of participants who were predominantly Caucasian or White (95%) and rural (44.2%). While the findings may not be generalizable to individuals from diverse racial/ethnic background, these figures are consistent with the ethnic/racial distribution of the state of West Virginia which consisted of 96.5% of elders aged 65 and above classified as White based on the 2000 Census (Center on Aging, 2003). Additionally, caution should be made in interpretations of results because the neuropsychological battery was administered in a set order and some of the domains were based on one or two tasks.

Future Directions

Even though the present study is unable to determine causality, these data provide knowledge base for development of intervention strategies to improve cognitive health. A systematic review of interventions targeting social isolation found that those which were offered in group format, defined as being theoretically based, and required active participation were most effective (Dickens, Richards, Greaves, & Campbell, 2011). Given our findings, it would be beneficial for interventions to not only target increasing social network size but also increase perceived level of support. In addition, mixed methods should be incorporated into future research because qualitative questions allow researchers to obtain an in-depth understanding of participants (e.g., long- vs. short-term isolation, perceptions about isolation, etc.) and how their perspective contributes to the understanding of social isolation. Qualitative methods would also allow researchers to gather strategies for incorporating excluded participants into this type of research.

Future studies should (a) continue to examine social isolation with a validated measure, such as the LSNS-6, (b) build upon the research investigating the effects of social isolation on different cognitive domains using a comprehensive neuropsychological battery, (c) emulate the current study using a larger, more diverse sample of participants, and (d) investigate different interventions for decreasing social isolation.

Summary and Conclusions

This series of cross-sectional analyses demonstrated that aspects of social isolation are positively related to overall and specific domains of cognition. In combination with the imminent, substantial growth of the elderly population and the established relations between social isolation and mental and physical disorders, this investigation provides further justification for determining the nature of the relations between social isolation and cognitive function. It is our hope that future studies will generate prognostic and/or therapeutic modalities to improve social isolation in older adults.

Footnotes

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This project is funded by the National Institutes of Health/National Institute for Dental and Craniofacial Research (5R21DE016970) (Principal Investigator: Bei Wu), the West Virginia University Health Sciences Center, and the West Virginia University School of Dentistry.

References

Barr

Brandt

(1996). Word list generation deficits in dementia. Journal of Clinical and Experimental Neuropsychology, 18, 810–822.

Benton

A. L.

Hamsher

K. eds.

(1989). Multilingual aphasia examination. Iowa City, IA: AJA associates.

Berkman

L. F.

Glass

Brissette

Seeman

T. E.

(2000). From social integration to health: Durkheim in the new millennium. Social Science & Medicine, 51, 843–857.

Black

(1992). Cardiovascular risk factors. In Zaret

B. L.

Cohen

L. S.

Moser

(Eds.), The Yale University school of medicine heart book (pp. 23–35). New York, NY: Hearst Book.

Blair

J. R.

Spreen

(1989). Predicting premorbid IQ: A revision of the National Adult Reading Test. The Clinical Neuropsychologist, 3, 129–136.

Cacioppo

J. T.

Ernst

J. M.

Burleson

M. H.

McClintock

M. K.

Malarkey

W. B.

Hawkley

L. C.

… Berntson

G. G.

(2000). Lonely traits and concomitant physiological processes: The MacArthur social neuroscience studies. International Journal of Psychophysiology, 35, 143–154.

Cacioppo

J. T.

Hawkley

L. C.

(2009). Perceived social isolation and cognition. Trends in Cognitive Sciences, 13, 447–454.

Cacioppo

J. T.

Hawkley

L. C.

Norman

G. J.

Bernston

(2011). Social isolation. Annals of the New York Academy of Sciences, 1231, 17–22.

Cole

S. W.

Hawkley

L. C.

Arevalo

J. M. G.

Cacioppo

J. T.

(2011). Transcript origin analysis identifies antigen-presenting cells as primary targets of socially regulated gene expression in leukocytes. Proceedings of the National Academy of Sciences, 108, 3080–3085.

10.

Cornwell

E. Y.

Waite

L. J.

(2009). Measuring social isolation among older adults using multiple indicators from the NSHAP Study. Journal of Gerontology: Series B, 64, 38–46.

11.

Delis

D. C.

Kaplan

Kramer

J. H.

Ober

B. A.

(2000). California verbal learning test-second edition (CVLT-II) manual. San Antonio, TX: Psychological Corporation.

12.

Dickens

A. P.

Richards

S. H.

Greaves

C. J.

Campbell

J. L.

(2011). Interventions targeting social isolation in older people: a systematic review. Biological Medicine Central Public Health, 11, 647.

13.

Fratiglioni

L. F.

Paillard-Borg

Winblad

(2004). An active and socially integrated lifestyle in late life might protect against dementia. Lancet Neurology, 3, 343–353.

14.

Giuli

Spazzafumo

Sirolla

Abbatecola

A. M.

Lattanzio

Postacchini

(2012). Social isolation risk factors in older hospitalized individuals. Archives of Gerontology and Geriatrics, 55, 580–585.

15.

Gottlieb

B. H.

Bergen

A. E.

(2010). Social support concepts and measures. Journal of Psychosomatic Research, 69, 511–520.

16.

Havens

Hall

Sylvestre

Jivan

(2004). Social isolation and loneliness: differences between older rural and urban Manitobans. Canadian Journal on Aging, 23, 129–140.

17.

Hawkley

L. C.

Hughes

M. E.

Waite

L. J.

Masi

C. M.

Thisted

R. A.

Cacioppo

J. T.

(2008). From social structural factors to perceptions of relationship quality and loneliness: The Chicago health, aging, and social relations study. Journal of Gerontology: Social Sciences, 63, 375–384.

18.

Hawthorne

(2008). Perceived social isolation in a community sample: its prevalence and correlates with aspects of peoples’ lives. Social Psychiatry and Psychiatric Epidemiology, 43, 140–150.

19.

Hawton

Green

Dickens

A. P.

Richards

S. H.

Taylor

R. S.

Edwards

… Campbell

J. L.

(2011). The impact of social isolation on the health status and health-related quality of life of older people. Quality of Life Research, 20, 57–67.

20.

Kaplan

Goodglass

Weintraub

(1983). Boston Naming Test. Philadelphia: Lea and Febiger.

21.

Katzman

(1993). Views and reviews: Education and the prevalence of dementia and Alzheimer’s disease. Neurology, 43, 13–20.

22.

Kurina

L. M.

Knutson

K. L.

Hawkley

L. C.

Cacioppo

J. T.

Lauderdale

D. S.

Ober

(2011). Loneliness is associated with sleep fragmentation in a communal Society. Sleep, 34, 1519–1526.

23.

Lubben

J. E.

(1988). Assessing social networks among elderly populations. Family & Community Health, 11, 42–52.

24.

Lubben

J. E.

Blozik

Gillmann

(2006). Performance of an abbreviated version of the Lubben Social Network scale among three European community-dwelling older adult population. The Gerontologist, 46, 503–513.

25.

McPherson

Smith-Lovin

Brashears

M. E.

(2006). Social isolation in America: Changes in core discussion networks over two decades. American Sociological Review, 71, 353–375.

26.

O’Luanaigh

O’Connell

Chin

A. V.

Hamilton

Coen

Walsh

… Lawlor

B. A.

(2012). Loneliness and vascular biomarkers: The Dubline Healthy ageing study. International Journal of Geratric Psychiatry, 27, 83–88.

27.

Osterrieth

P. A.

(1944). Le test de copie d’une figure complexe. Archives de Psychologie, 30, 206–356.

28.

Reitan

R. M.

(1958). Validity of the trail making test as an indicator of organic brain damage. Perceptual and Motor Skills, 19, 199–206.

29.

Rey

(1941). L’examen psychologique dans les cas d’encephalopahie traumatique. Archives de Psychologie, 28, 286–340.

30.

Rook

K. S.

Pietromonaco

(1987). Close relationships: Ties that heal or ties that bind? Advances in Personal Relationships, 1, 1–35.

31.

Roman

G. C.

Tatemichi

T. K.

Erkinjunntti

Cummings

J. L.

Masdeu

J. C.

Garcia

J. H.

… Scheinberg

(1993). Vascular dementia: diagnostic criteria for research studies. Report of the NINDS-AIREN International Workshop. Neurology, 43, 250–260.

32.

Rosengren

Wilhelmsen

Orth-Gomer

(2004). Coronary disease in relation to social support and social class in Swedish men: A 15 year follow-up in the study of men born in 1933. European Heart Journal, 25, 56–63.

33.

Routasalo

P. E.

Savikko

Tilvis

R. S.

Strandberg

T. E.

Pitkala

K. H.

(2006). Social contacts and their relationship to loneliness among aged people—A population-based study. Gerontology, 52, 181–187.

34.

Rubinstein

R. L.

Lubben

J. E.

Mintzer

J. E.

(1994). Social isolation and social support: An applied perspective. Journal of Applied Gerontology, 13, 58–72.

35.

Schmaltz

H. N.

Southern

Ghali

W. A.

Jelinshi

S. E.

Parsons

G. A.

King

K. M.

Maxwell

C. J.

(2007). Living alone, patient sex and mortality after acute myocardial infarction. Journal of General Internal Medicine, 22, 572–578.

36.

Snitz

B. E.

Unverzagt

F. W.

Chang

C. C.

Bilt

J. V.

Gao

Saxton

… Ganguli

(2009). Effects of age, gender, education and race on two tests of language ability in community-based older adults. International Psychogeriatric, 21, 1051–1062.

37.

Souder

J. E.

(1992). The consumer approach to recruitment of elder subjects. Nursing Research, 41, 314–316.

38.

Tomaka

Thompson

Palacios

(2006). The relation of social isolation, loneliness, and social support to disease outcomes among the elderly. Journal of Aging & Health, 18, 359–384.

39.

United States Census Bureau News. (2009, July). Unprecedented global aging examined in New Census Bureau Report commissioned by the National Institute on Aging (Issued brief No. CB09-108). Washington, DC: Bernstein, R., & Cire, B. Retrieved from http://www.census.gov/PressRelease/www/releases/archives/aging_population/013988.html

40.

Ward

Arrighi

H. M.

Michels

Cedarbaum

J. M.

(2012). Mild cognitive impairment: Disparity of incidence and prevalence estimates. Alzheimer’s & Dementia: The Journal of the Alzheimer’s Association, 8, 14–21.

41.

Wilson

R. S.

Barnes

L. L.

Krueger

K. R.

Hoganson

Bienias

J. L.

Bennett

D. A.

(2005). Early and late life cognitive activity and cognitive systems in old age. Journal of the International Neuropsychological Society, 11, 400–407.

42.

Wilson

R. S.

Evans

D. A.

Bienias

J. L.

Mendes de Leon

C. F.

Schneider

J. A.

Bennett

D. A.

(2003). Proneness to psychological distress is associated with risk of Alzheimer’s disease. Neurology, 61, 1479–1485.

43.

Wilson

Krueger

Arnold

Schneider

Kelly

Barnes

… Bennett

(2007). Loneliness and risk of Alzheimer Disease. Archives of General Psychiatry, 64, 234–240.

44.

Wu, B., Aging in West Virginia: Findings from the 2000 Census [Report]. (September, 2003), Center on Aging, West Virginia University., Morgantown, WV

45.

Goedereis

Crout

Plassman

P. L.

DiNapoli

McNeil

D. W.

… Bai

(2010). Recruitment of rural and cognitively impaired older adults for dental research. Special Care in Dentistry, 30, 193–199.

46.

Yesavage

J. A.

Brink

T. L.

Rose

T. L.

Lum

Huange

Adey

Leirer

V. O.

(1983). Development and validation of a geriatric depression screening scale: A preliminary report. Journal of Psychiatric Research, 17, 37–49.