There is an extensive literature demonstrating that multiple dimensions of social isolation are associated with poorer cognitive functioning in older adults (Evans et al., Reference Evans, Martyr, Collins, Brayne and Clare2019). Having social relationships and interactions has been associated with less cognitive decline and lower rates of dementia (Penninkilampi et al., Reference Penninkilampi, Casey, Singh and Brodaty2018), increasing cognitive reserve (Stern, Reference Stern2009) and promoting cognitively enriching experiences (Hertzog et al., Reference Hertzog, Kramer, Wilson and Lindenberger2008). A longitudinal study by Liu et al. (Reference Liu, Zhang, Choi and Langa2020) found spousal older adults had significantly lower odds of developing dementia compared to unmarried healthy older adults. Another longitudinal study found years of spousal education were associated with less cognitive decline in healthy older adult spousal partners (Xu, Reference Xu2020).

As adults grow older, their social relationships may become more selective, and their social networks become smaller. Older adults may choose to spend more time with close family members and friends rather than meeting new friends and socializing with acquaintances (Carstensen et al., Reference Carstensen, Fung and Charles2003). A narrower focus on spousal and other close relationships may be especially advantageous in preserving or enhancing cognitive health of older adults. With the growing literature on the associations of social relationships and psychological, mental, and physical health outcomes (Holt-Lunstad et al., Reference Holt-Lunstad, Smith, Baker, Harris and Stephenson2015), marriage has become a central focus of studies on relationships and health, primarily because couples have a unique influence on one another’s health (Smith et al., Reference Smith, Baron, Caska, Agnew and South2014).

Research has shown a significant bidirectional association between older adult cognitive functioning and depressive symptoms such that older adults with depressive symptoms have poor cognitive performance (Zhu et al., Reference Zhu, Li, Xie, Zhong, Wu and Blumenthal2022) and poor cognitive functioning is significantly associated with greater depressive symptoms in healthy older adults (Perrino et al., Reference Perrino, Mason, Brown, Spokane and Szapocznik2008). Studies have examined the reciprocal relationship between cognitive functioning and depressive symptoms over time in spousal older adult couples (Lee et al., Reference Lee, Paddock and Feeney2012), but few have examined these associations within a specific cultural context.

Social relationships can be influenced by cultural contexts, and the extent to which social relationships are associated with health may differ cross-culturally as well. Cultural variations in the link of self in the relationship and the regard to emotions can impact relationship quality (Campos, Reference Campos2015). With respect to link of self, people in cultures that favor independent ideals may require less social support in relationships and practice acceptance to cope with life changes; whereas individuals in cultures that value interdependent ideals may seek social assurances and support to cope with changes. The extent to which cultural ideals of independence and interdependence are attained in one’s social relationship may also contribute to one’s psychological health (Campos, Reference Campos2015). These cultural ideals may be especially important within spousal relationships and the unique influence of spouses on one another’s cognition and depressive symptoms. Another important cultural variation when examining social relationships and its impact on psychological health is the role of emotion expression. In cultures that value emotion suppression to maintain social relationships, emotion suppression is not associated with negative psychological health (Campos, Reference Campos2015). According to the review by Campos (Reference Campos2015), people who tend to suppress their emotions in their social relationships and live in cultures that encourage and value the expression of emotions will most likely face psychological challenges. This aspect of emotion suppression within the spousal relationships may be specifically important in addressing negative psychological outcomes in Hispanic cultures which value positive emotion expression to generate and maintain social interaction (Campos, Reference Campos2015).

The study by Monin et al. (Reference Monin, McAvay, Newkirk and Samper-Ternentin press) in International Psychogeriatrics provides important insights into the bidirectional associations between older adult spouses’ cognitive functioning and depressive symptoms over time using a longitudinal, dyadic path analysis with the actor–partner interdependence model. This study makes a unique research contribution by examining whether previous findings using the actor–partner independence model previously reported by Monin et al. (Reference Monin2018) in a US sample also occurred in older Mexican couples, thus highlighting the importance of examining whether findings in one society are replicable or differ across cultures. The study used a representative sample from the Mexican Health and Aging Study (MHAS), a unique dataset that links spousal data and provides extensive information on the impact of disease on the health, function, and life span of adults ages 50 and older. Depressive symptoms and cognitive function were assessed over time from 2012 to 2018, and extensive sociodemographic covariates included age, sex, language (Spanish vs. English), number of children, employment status, education, ADLs, IADLs, comorbidities, pain, and smoking status. Statistical analyses involved a generalized estimating equations approach that included both husbands and wives simultaneously and each interview (time) for cognitive functioning and depressive symptoms. Two separate crossed-lag actor–partner interdependence structural equation models were used to examine the directionality of the association between cognitive functioning and depressive symptoms. The first model assessed cognitive functioning when depressive symptoms were the outcome, while the second model assessed depressive symptoms when cognitive functioning was the outcome. Possible gender effects were also examined for both models. The results showed that one’s own depressive symptoms and cognitive functioning each prospectively predicted changes in one’s own follow-up cognitive functioning and depressive symptoms. Thus, within individuals, depressive symptoms and cognitive functioning are intertwined and the effects are not specific to outcome. However, the only partner effect found in the study for both husbands and wives was that one’s depressive symptoms predicted more depressive symptoms over time in the partner. Cognitive functioning did not predict subsequent changes in depressive symptoms or cognitive functioning in partners over time. These associations did not differ between husbands and wives. These results did differ from Monin et al.’s (Reference Monin2018) findings with US couples, which found that there were partner effects for spouse’s depressive symptoms predicting subsequent cognitive functioning over time.

In the current paper, Monin et al. (Reference Monin, McAvay, Newkirk and Samper-Ternentin press) do an admirable job of suggesting possible reasons for the differences in these findings. For example, the authors note that differences in measures across the studies, and other methodological differences, make the comparisons imperfect. They also offer suggestions on cultural mechanisms, noting that previous research suggests that older Mexicans have larger social networks than Mexican younger adults, and that social network may not narrow with age as is often found in US studies. This is intriguing in it that it suggests culture-specific variations in the operation of socioemotional selectivity theory (Carstensen et al., Reference Carstensen, Fung and Charles2003). Fung et al. (Reference Fung, Stoeber, Yeung and Lang2008) have addressed this issue of the possible cultural specificity of various aspects of socioemotional selectivity theory. They review evidence and provide results suggesting that the key motivations affecting older adults and their social relationships (viewing future time as limited and valuing close relationships) are likely universal, but that cultures may vary in which individuals within the social network (spouses, family or nonfamily, extended family) are part of one’s inner circle in late life.

Monin et al. (Reference Monin, McAvay, Newkirk and Samper-Ternentin press) study is an outstanding example of the creative use of data from diverse samples, using a longitudinal design, rigorous statistical approaches, and a strong theoretical framework. A recent themed issue of International Psychogeriatrics (August 2021) included attention to several important issues in longitudinal studies of aging that are relevant to the current study and that suggest future extensions of this work with older couples. Cano-López et al. (Reference Cano-López2021) reported on the results of a longitudinal study within individuals of the effects of depressive symptoms on cognitive functioning in older adults and found that education moderated this relationship. Depression affected subsequent cognitive functioning only among those with low educational attainment. In a commentary on this paper by Cano-López et al. (Reference Cano-López2021), Reppermund (Reference Reppermund2021) notes the potential value of studying changes in complex instrumental activities of daily living (IADL) such as the ability to manage medications and prepare meals as an important mechanism of understanding changes in the association of depressive symptoms and cognitive functioning over time. While Monin et al. (Reference Monin, McAvay, Newkirk and Samper-Ternentin press) did carefully control for education and IADL in their analyses, perhaps effects across partners are modified by risk and protective factors such as educational attainment, and variables such as changes in IADL may be considered not only as covariates but as mechanisms.

Future studies can build on this work and expand the research questions to examine other possible mechanisms that may influence cognitive and mental health of spousal older adults. The authors mention how spouses share similar environments and stressors and how a partner copes with a stressor that can influence each partner’s health outcomes through several pathways including psychological and behavioral ones. Research has found supportive and straining aspects of relationships to be integral parts in the association between social relationships and health outcomes (Campos, Reference Campos2015). A spouse’s perceived social support from own’s spouse may help a spouse better cope with stressors, and social support may mediate the relationship between cognitive and mental health spousal dyadic relationships. Having social support within a spousal relationship may also improve marital quality. Related to this, a dyadic study found that older men’s perception of negative marital quality was associated with increased cognitive decline, which was not found in older women (Liu et al., Reference Liu, Zhang and Zhang2021). Therefore, future studies may benefit from examining such relationship factors and how they may modify changes in cognition and mood in older couples.

Overall, the study by Monin et al. (Reference Monin, McAvay, Newkirk and Samper-Ternentin press) provides an exemplary approach to studying the complex relationship between psychosocial factors such as depressive symptoms, and cognitive functioning, while also addressing the challenge of understanding how these effects vary across cultures. Future researchers may need to acknowledge and incorporate aspects of sociocultural diversities within relationships in their research in order to broaden our understanding of the association of social relationships with health.