Monica E. Nelson

Sleep disorders often predict or co-occur with cognitive decline. Yet, little is known how the relationship unfolds among older adults at risk for cognitive decline. To examine the associations of sleep disorders with cognitive decline in older adults with unimpaired cognition, or impaired cognition (mild cognitive impairment [MCI] and dementia). 5,822 participants (M_age=70) of the National Alzheimer’s Coordinating Center database with unimpaired or impaired cognition were followed for three subsequent waves. Four types of clinician-diagnosed sleep disorders were reported: sleep apnea, hyposomnia/insomnia, REM sleep behavior disorder, or “other.” Cognition over time was measured by the Montreal Cognitive Assessment (MoCA) or an estimate of general cognitive ability (GCA) derived from scores based on 12 neuropsychological tests. Growth curve models were estimated adjusting for covariates. In participants with impaired cognition, baseline sleep apnea was related to better baseline MoCA performance (b=0.65, 95%CI=[0.07, 1.23]) and less decline in GCA over time (b=0.06, 95%CI=[0.001, 0.12]). Baseline insomnia was related to better baseline MoCA (b=1.54, 95%CI=[0.88, 2.21]) and less decline in MoCA over time (b=0.56, 95%CI=[0.20, 0.92]). Furthermore, having more sleep disorders (across the four types) at baseline predicted better baseline MoCA and GCA, and less decline in MoCA and GCA over time. These results were only found in those with impaired cognition and generally consistent when using self-reported symptoms of sleep apnea or insomnia. Participants with sleep disorder diagnoses may have better access to healthcare, which may help maintain cognition through improved sleep.

BACKGROUND: Stroke commonly leads to disability and depression. Social connection and engagement can be protective against functional decline and depression in the general population. We investigated the effects of social connection and engagement on trajectories of function and depressive symptoms in stroke. METHODS: This is a longitudinal study, which included 898 participants with incident stroke from the HRS study (Health and Retirement Study) between 1998 and 2012. Multilevel modeling was used to examine associations of social connection and engagement with changes in functional limitations in instrumental activities of daily living (IADLs) and depressive symptoms over time. Models controlled for age, gender, education, and race/ethnicity. Moderation analyses examined whether high social connection and engagement reduced depressive symptoms for survivors with high IADL impairment. RESULTS: Social connection and engagement were generally associated with fewer IADL limitations and depressive symptoms at the time of stroke and after stroke. For example, participants who felt lonely and did not provide help to others before stroke had more IADL limitations. Prestroke volunteering was associated with less increase in IADL limitations with stroke and increase in having friends and providing help to others compared with one’s prestroke status were associated with fewer IADL limitations after stroke. For depressive symptoms, participants who felt lonely and did not have a friend or partner before stroke had more depressive symptoms, and participants who had children residing nearby before stroke showed less increase in depressive symptoms. Moderation effects were not found for social connection and engagement on high IADL impairment and depressive symptoms. CONCLUSIONS: Findings suggest that social connection and engagement may reduce the negative physical and psychological outcomes of stroke, both at baseline and after stroke. Efforts to enhance social engagement and diminish loneliness may both enhance population well-being and enhance resilience and recovery from stroke and other illnesses.

Background Stroke commonly leads to disability and depression. Social connection and engagement can be protective against functional decline and depression in the general population. We investigated the effects of social connection and engagement on trajectories of function and depressive symptoms in stroke. Methods Participants were 898 individuals with incident stroke from the Health and Retirement Study between 1998-2012. Multilevel modeling was used to examine associations of social connection and engagement with changes in functional limitations in instrumental activities of daily living (IADLs) and depressive symptoms over time. Models controlled for age, gender, education, and race/ethnicity. Moderation analyses examined whether high social connection and engagement reduced depressive symptoms for survivors with high IADL impairment. Results Social connection and engagement were generally associated with fewer IADL limitations and depressive symptoms at the time of stroke and after stroke. For example, participants who felt lonely and did not provide help to others before stroke had more IADL limitations. Pre-stroke volunteering was associated with less increase in IADL limitations with stroke and increase in having friends and providing help to others compared to one’s pre-stroke status were associated with fewer IADL limitations after stroke. For depressive symptoms, participants who felt lonely and did not have a friend or partner before stroke had more depressive symptoms, and participants who had children residing nearby before stroke showed less increase in depressive symptoms. Moderation effects were not found for social connection and engagement on high IADL impairment and depressive symptoms. Conclusions Findings suggest that social connection and engagement may reduce the negative physical and psychological outcomes of stroke, both at baseline and after stroke. Efforts to enhance social engagement and diminish loneliness may both enhance population well-being and enhance resilience and recovery from stroke and other illnesses.

Background: Dementia syndrome is one of the most devastating conditions in older adults. As treatments to stop neurodegeneration become available, accurate and timely diagnosis will increase in importance. One issue is that cognitive performance sometimes does not match the corresponding level of neuropathology, affecting diagnostic accuracy. Cognitive reserve (CR), which can preserve cognitive function despite underlying neuropathology, explains at least some variability in cognitive performance. We examined the influence of CR proxies (education and occupational position) on the relationship between hippocampal or total gray matter volume and cognition.Methods: We used data from the Czech Brain Aging Study. Participants were clinically confirmed to be without dementia (n = 457, including subjective cognitive decline and amnestic mild cognitive impairment) or with dementia syndrome (n = 113).Results: For participants without dementia, higher education magnified the associations between (a) hippocampal volume and executive control (b = 0.09, p = 0.033), (b) total gray matter volume and language (b = 0.12, p < 0.001), and (c) total gray matter volume and memory (b = 0.08, p = 0.018). Similarly, higher occupational position magnified the association between total gray matter volume and (a) attention/working memory (b = 0.09, p = 0.009), (b) language (b = 0.13, p = 0.002), and (c) memory (b = 0.10, p = 0.013). For participants with dementia, the associations between hippocampal (b = –0.26, p = 0.024) and total gray matter (b = –0.28, p = 0.024) volume and visuospatial skills decreased in magnitude with higher education.Conclusion: We found that the association between brain volume and cognitive performance varies based on CR, with greater CR related to a stronger link between brain volume and cognition before, and a weaker link after, dementia diagnosis.

Objectives: The spousal relationship is one of the most important social contexts in old age, and the loss of a spouse/partner is associated with stress and cognitive decline. In the present study, we examined whether social relationships can buffer potential negative effects of spousal loss on cognition. We examined the role of social network, social activities, and perceived deficiencies in social relationships (loneliness). Method: We used longitudinal data between 1998 and 2012 from 2,074 participants of the Health and Retirement Study, who had experienced spousal loss during the study period. Multilevel modeling was used to examine how time-varying indicators of social network, social activities, and loneliness were related to age-related trajectories of episodic memory prior to and after spousal loss. Analyses controlled for gender, race/ethnicity, education, time-varying functional health, and being repartnered/remarried. Results: Having children living within 10 miles and providing help to others buffered negative effects of widowhood on episodic memory. In addition, within-person increase in providing help to others buffered against decline in episodic memory after spousal loss. Having friends in the neighborhood, more frequent social visits, providing help to others, volunteering, and lack of loneliness were related to higher episodic memory, while having relatives in the neighborhood was related to lower episodic memory. Discussion: Our findings suggest that social networks, social activities, and loneliness are related to levels of cognitive function at the time of spousal loss and that social relationships can buffer negative effects of spousal loss on cognitive function. Implications for future research are discussed.

Background and Objectives: Positive associations between social connection/engagement and cognitive function are well documented. However, little is known about whether social connection/engagement can buffer the impact of serious brain injury such as stroke on cognitive functioning. Research Design and Methods: Participants were 898 individuals with incident stroke from the Health and Retirement Study between 1998 and 2012. Multilevel modeling was used to examine how social connection/engagement was associated with episodic memory pre- and poststroke. Models controlled for age, gender, education, race/ethnicity, number of health conditions, and functional health. Results: Participants who were lonely prestroke recalled significantly fewer words at the time of stroke, and participants who had children residing within 10 miles prestroke showed significantly less decline in word recall over time. Participants who provided help to others prestroke showed less stroke-related decline in word recall. Within-person increase in partnered status, having friends, and helping others were related to better word recall in the poststroke period. Discussion and Implications: Higher prestroke levels of social connection/engagement predicted better episodic memory at stroke, a smaller decline in episodic memory with stroke, and less decline in episodic memory over time. Increases in social connection/engagement from pre- to poststroke also predicted better poststroke episodic memory. Beyond the widely documented benefits of social connection/engagement to well-being, they may also increase cognitive stimulation and cognitive reserve and thus contribute to stroke recovery in the cognitive domain. Social connection/engagement is an important and modifiable risk factor in older adults.

We examined associations between job strain and trajectories of change in cognitive functioning (general cognitive ability plus verbal, spatial, memory, and speed domains) before and after retirement. Data on indicators of job strain, retirement age, and cognitive factors were available from 307 members of the Swedish Adoption/Twin Study of Aging. Participants were followed up for up to 27 years (mean = 15.4, SD = 8.5). In growth curve analyses controlling for age, sex, education, depressive symptoms, cardiovascular health, and twinness, greater job strain was associated with general cognitive ability (estimate = −1.33, p = .002), worse memory (estimate = −1.22, p = .007), speed (estimate = −1.11, p = .012), and spatial ability (estimate = −0.96, p = .043) at retirement. Greater job strain was also associated with less improvement in general cognitive ability before retirement and a somewhat slower decline after retirement. The sex-stratified analyses showed that the smaller gains of general cognitive ability before retirement (estimate = −1.09, p = .005) were only observed in women. Domain-specific analyses revealed that greater job strain was associated with less improvement in spatial (estimate = −1.35, p = .010) and verbal (estimate = −0.64, p = .047) ability before retirement in women and a slower decline in memory after retirement in women (estimate = 0.85, p = .008) and men (estimate = 1.12, p = .013). Neither preretirement nor postretirement speed was affected significantly by job strain. Greater job strain may have a negative influence on overall cognitive functioning prior to and at retirement, while interrupting exposure to job strain (postretirement) may slow the rate of cognitive aging. Reducing the level of stress at work should be seen as a potential target for intervention to improve cognitive aging outcomes.

Cognitive reserve (CR) may reduce the risk of dementia. We summarized the effect of CR on progression to mild cognitive impairment (MCI) or dementia in studies accounting for Alzheimer’s disease (AD)-related structural pathology and biomarkers. Literature search was conducted in Web of Science, PubMed, Embase, and PsycINFO. Relevant articles were longitudinal, in English, and investigating MCI or dementia incidence. Meta-analysis was conducted on nine articles, four measuring CR as cognitive residual of neuropathology and five as composite psychosocial proxies (e.g., education). High CR was related to a 47% reduced relative risk of MCI or dementia (pooled-hazard ratio: 0.53 [0.35, 0.81]), with residual-based CR reducing risk by 62% and proxy-based CR by 48%. CR protects against MCI and dementia progression above and beyond the effect of AD-related structural pathology and biomarkers. The finding that proxy-based measures of CR rivaled residual-based measures in terms of effect on dementia incidence underscores the importance of early- and mid-life factors in preventing dementia later.