Culture and Health: Recent Developments and Future Directions

Abstract

Evidence of cultural differences in relationships and emotions has accumulated over the past few decades. As findings on cultural differences in psychological processes have accumulated, there has been growing interest in investigating whether they have implications for other phenomena such as health. Using scientific advances from the MIDUS and MIDJA studies, both publicly available, we examine links between culture and health. We first briefly review the accumulated evidence on cultural influences on health correlates of psychosocial factors. We then feature two recent developments – a more micro-level perspective on biological factors that may be involved in the culture and health linkage, and a more macro-level view of socioeconomic inequality, which also matters for health. Both perspectives inform the pathways through which health effects occur. Finally, we conclude our review by highlighting the changing historical contexts surrounding these cross-cultural investigations. Specifically, we draw attention to widening of economic inequality across cultures and the world-wide COVID-19 pandemic. These happenings bring notable implications for future research on health across cultural contexts.

Evidence on cultural differences in psychological processes has accumulated over the past three decades (cf., Handbook of Cultural Psychology, edited by Cohen & Kitayama, 2019). Among many patterns and dimensions along which cultures differ from each other, independence and interdependence are some of the most widely documented ways to characterize different cultural meaning systems (Markus & Kitayama, 1991; Triandis, 1989). According to an independent way of viewing the self and relationships prevalent in Western cultural contexts, the self is defined by its internal attributes that are separate from social relationships and contexts. Alternatively, according to an interdependent way of viewing the self and relationships dominant in non-Western cultural contexts, such as East Asia, the self is viewed as fundamentally embedded in social relationships and contexts. Such divergent views of the self and relationships have been suggested to lead to cultural differences in a wide range of psychological processes, including how people relate to each other (e.g., Kim, Sherman, & Taylor, 2008) and experience emotions (Mesquita, De Leersnyder, & Albert, 2014).

As the evidence on cultural differences in psychological processes has accumulated, there is growing interest in investigating the implications of such cultural differences. That is, beyond explicating the nature of cultural differences in psychological processes, there are critical issues of whether they matter for other phenomena such as one’s health or functional capacities. Investigating health implications of cultural differences in psychological processes is important not only because of its practical significance, but because it may inform understanding of how culturally divergent psychological processes are sustained over time, or are modified. Exploring health and functional capacities, however, calls for a different approach to data collection than has been used in much of the preceding cultural psychology literature, or psychological research in general, where samples were often based on available college students, or local community participants. Guided by thinking in population-based disciplines (demography, epidemiology, sociology), health inquiries make it imperative to recruit samples that reflect the heterogeneity (e.g., variation by gender, age, race, educational status/income, marital status) of the societies from which they are drawn. Extensive prior science had documented the import of these sociodemographic factors on health.

These observations were center stage in the inception of the parallel Midlife in the U.S. (MIDUS) and Midlife in Japan (MIDJA) studies. MIDUS was conceived first in 1995 as a multidisciplinary endeavor seeking to investigate the interplay of sociodemographic, psychosocial, and behavioral factors on health, broadly defined, in a representative sample of approximately 3,500 U.S. adults, aged 25 to 74. MIDUS expanded scientifically over time to incorporate biological factors and neuroscience assessments that were obtained on subsamples of participants. In 2008, MIDJA was launched with a probability sample of approximately 1,000 midlife adults from Tokyo, Japan. The overall objective of the two endeavors was to bring cultural factors into the formulation of what matters for health and well-being as individuals age across the decades of adult life. To the extent possible, there was commitment to obtaining parallel assessments in both studies, including the collection of biomarkers on subsamples within both. There was also a commitment to longitudinal tracking of MIDUS and MIDJA participants. The overarching idea motivating these parallel studies was the belief (hypothesis) that psychosocial and behavioral factors matter for health across all societal contexts, but how they unfold may be importantly shaped by distinct norms and values across cultures.

What follows is an overview of findings that have emerged from MIDUS and MIDJA comparative studies of culture and health (cf. Figure 1). In the opening section below, we briefly distill some of the initial findings that emerged using self-reported survey data. Many of these extended prior findings in cultural psychology. We then move to two areas that are more recent developments growing up around these two studies. Prominent among them are findings that have utilized the biological data available in MIDUS and MIDJA, often while linking them to different psychosocial variables. We see the addition of such objectively measured biological factors as an important advance in cross-cultural research. Our second focus on new developments gives attention to the critical need to bring socioeconomic status variables into cross-cultural studies. A central point is that people are simultaneously embedded not only in cultural contexts, but they also occupy differing positions in socioeconomic hierarchies that are increasingly recognized as important influences on health. Finally, we conclude our review with consideration of the changing historical context within which this broad-ranging comparative research is unfolding. The global pandemic provides a dramatic example of such historical change that brings with it notable implications for future research on health across cultural contexts.

Figure 1.

An overview of the MIDUS and MIDJA content. Boxes that include two recent developments featured in this paper are highlighted in gray.

Culture and Self-Reported Health: Initial Findings

Initial studies from MIDJA and MIDUS examined how independent and interdependent psychosocial factors are associated with physical health in U.S. and Japanese adults. Kitayama, Karasawa, Curhan, Ryff, and Markus (2010) used pilot data based on community samples from Tokyo and Sapporo that were compared to U.S. adults in MIDUS. Using a combined measure of health (including subjective health, number of chronic conditions, somatic symptoms, and sensitivity to unpleasant sensations) and well-being, the results showed that independence (operationalized as high personal control) was more strongly associated with better well-being/health in the U.S. than in Japan, whereas interdependence (high relational harmony and low relational strain) was associated more strongly associated with better well-being/health in Japan than in the U.S. Using the MIDJA and MIDUS samples, Park, Kitayama, Karasawa et al. (2013) further highlighted specific psychosocial contexts in which perceived social support was associated with better health in Japan, such as when one is experiencing high stress or is low on neuroticism. Levine, Miyamoto, Markus et al. (2016), in turn, focused on healthy eating habits (e.g., consumption of more vegetables) and found that, in the U.S., independence was associated with healthier eating habits, whereas in Japan, interdependence was associated with healthier eating habits. These findings suggested that the type of psychosocial factors that are associated with better health often vary across contexts, depending on the feelings, beliefs, or practices that are endorsed (sanctioned) within specific cultures.

Other findings examined how emotions are linked to health across cultures, beginning with a focus on dialectical beliefs, which had been suggested to characterize East Asian cultural contexts (Peng & Nisbett, 1999; Spencer-Rodgers, Williams, & Peng, 2010). According to dialectical beliefs, opposites, such as positive emotions and negative emotions, coexist and constantly change into each other. In line with such dialectical beliefs, Miyamoto and Ryff (2011) found, using MIDUS and pilot data from MIDJA, that Japanese adults were more likely than American adults to report experiencing both positive and negative emotions with moderate frequency. Furthermore, this moderate dialectical emotion type was associated with fewer physical symptoms in Japan than in the U.S., although no cultural difference was found for the subjective health rating. Reflecting both interdependence and dialectical beliefs, other studies suggested that East Asians are more likely than Americans to perceive negative sides of positive emotions and positive sides of negative emotions (Miyamoto, Ma, & Wilken, 2017; Uchida & Kitayama, 2009). Using MIDUS and MIDJA, Curhan, Sims, Markus et al. (2014) found that the links between negative emotions and measures of health (i.e., chronic health conditions and physical functioning) were weaker in Japan than in the U.S. Employing a structural equation model analysis, Ishii (2019) showed that the link between negative emotions and somatosensory amplification (i.e., sensitivity to unpleasant bodily sensation, a component of physical health in Kitayama et al., 2010) was weaker among Japanese than among Americans.

Some studies have examined longitudinal relationships between well-being or emotions and health. Focusing on MIDJA, Yoo and Ryff (2019) found that those with persistently high psychological well-being across four years showed better subjective health, fewer chronic health conditions (e.g., hypertension, arthritis), fewer health symptoms (e.g., headaches), and fewer functional limitations over time than those with persistently low well-being. Similar findings had previously been documented with data from MIDUS (Ryff, Radler, & Friedman, 2015). Willroth, Ong, Graham, and Mroczek (2020) examined both MIDUS and MIDJA and found that a lower level of negative emotions and a higher level of positive emotions across time were associated with better health (i.e., subjective health and chronic health conditions).

Taken together, the above findings point to evidence that the types of emotion that are more accepted (sanctioned) in Japan than in the U.S. (e.g., dialectical emotion, negative emotions) are less predictive of poor physical health in Japan compared to the U.S. At the same time, the results sometimes varied across different measures of health and emotions – for example, cultural differences tended to be smaller for subjective health ratings than for concrete measures of health (e.g., Curhan, Sims, Markus et al., 2014; Miyamoto & Ryff, 2011) and for the longitudinal measures (e.g., Willroth et al., 2020) than for single-time measure of emotions (e.g., Curhan, Sims, Markus et al., 2014). Such mixed findings highlight the need for further investigations to clarify the contextual aspects of culture that are more (or less) conducive for emotions to matter for health outcomes.

Importantly, these initial MIDUS and MIDJA inquiries provided researchers with opportunities to do more than demonstrate cultural differences in specific psychological processes. In addition, they examined whether and how such psychological processes mattered for individual health and functioning across different sociocultural contexts. The emerging evidence underscored the need to augment self-reported measures of health with objectively measured biological factors to elucidate possible pathways through which psychosocial factors may matter for physical health, which we cover in the next topic.

Recent Developments

Culture and Health: Probing Biological Pathways

A key innovation in the MIDUS and MIDJA studies has been to augment self-reported health assessments (subjective health, chronic conditions, functional capacities) with objective measurement of biological risk factors. These biomarkers, particularly when combined with psychosocial assessments, have offered new windows on the interplay of factors that may matter for health, and for how they may vary by cultural context. What follows summarizes some of the key findings, organized by distinct physiological systems.

Neuroendocrine and inflammatory markers.

As described above, multiple studies have documented cultural differences in experiences of positive and negative emotion. Extending such work, Park, Kitayama, Miyamoto, and Coe (2020) examined whether culture moderates the relationship between negative affect and biological stress of the neuroendocrine system, measured with diurnal cortisol. The healthy pattern of cortisol diurnal is characterized by rapid increase after awakening in the morning, followed by a steady decline over the day (i.e., declining slope). After controlling for demographic factors, personality traits, and health behaviors, findings showed that negative affect was associated with a flattening of the diurnal cortisol slope among Americans, but not among Japanese. Additional analyses showed that diurnal cortisol mediated the link between negative affect and other aspects of biological risk (inflammatory, cardiovascular) among Americans. These findings suggested cultural differences in the experience of negative affect that make it more (or less) stressful and thus differentially consequential for health.

Another study (Miyamoto, Boylan, Coe et al., 2013) investigated negative emotion but linked it to a specific pro-inflammatory marker, interleukin-6 (IL-6), which is implicated in numerous disease outcomes. Although previous studies conducted in Western cultures had shown that negative emotion predicted elevated levels of IL-6, this investigation replicated the finding with the U.S. biomarker sample, but showed no association between negative emotion and IL-6 in the Japanese biomarker sample. The obtained interaction between culture and negative emotion in predicting a major pro-inflammatory biomarker remained after adjusting for numerous demographic, psychological, and health factors.

Several studies have utilized a composite index of biological health risk (BHR) including inflammatory markers and cardiovascular factors. Using the BHR, Kitayama and Park (2017) showed that negative emotions (negative affect, anger expression) were inversely linked with biological health among Americans. The relationships were not present (negative affect), or were reversed (anger expression) for Japanese adults (Kitayama, Park, & Boylan et al., 2015). In addition, happiness measured with eudaimonic assessments was linked with better biological health in both cultural contexts but happiness measured with hedonic assessments was unclearly linked with biological health (Kitayama, Akutsu, Uchida, & Cole, 2016).

Another study (Kitayama, Park, Miyamoto et al., 2018) investigated links between neuroticism and the above BHR composite. Behavioral adjustment (a propensity to flexibly adjust behaviors to environmental contingencies) was examined as a moderator. Japanese participants were higher in behavioral adjustment than U.S. adults, and as predicted, neuroticism was linked to lower BHR among Japanese but not Americans, which was partly explained by cultural differences in behavioral adjustment. That is, among those who were high in behavioral adjustment, neuroticism was linked to lower BHR. At the same time, neuroticism was associated with worse self-reported health regardless of behavioral adjustment or cultural context.

Finally, Kitayama and Park (2021) examined links between conscientiousness and the BHR composite, with various mediating factors considered. Among Americans, conscientiousness was associated with lower BHR, and the relationship was mediated by healthy lifestyle. In contrast, the relationship between conscientiousness and BHR was not significant among Japanese, although further analyses showed that conscientiousness was associated with greater commitment to social obligation, which in turn predicted higher BHR. Thus, the findings showed that conscientiousness may or may not be salubrious for health depending on normatively sanctioned behaviors in different cultures.

Underscoring that much of what is known about the normal functioning of IL-6 and other inflammatory markers has been generated with European samples and Americans of European descent, Coe, Love, Karasawa et al. (2011) examined levels of these biomarkers among Japanese adults, and Caucasian and African American adults from the U.S. Across adults aged 30–80, IL-6 levels were strikingly lower in Japanese individuals and also significantly lower for C-reactive protein (CRP) and fibrinogen (FBG) compared to U.S. adults. Differences in body mass index (BMI) were taken into consideration, although differences in IL-6 persisted after including this covariate. The soluble receptor for IL-6 (sIL-6r) was also evaluated. This marker was higher in Japanese than American adults, with notably low levels for African Americans. These findings underscored the importance of population diversity in studies of health and aging.

Reflecting on the fact that Japan is an exceptionally healthy East Asian country with notable longevity, Coe, Miyamoto, Love et al. (2020) examined four cultural practices (drinking tea, eating seafood, consuming vegetables, taking relaxing baths) to see if they mattered for levels of CRP and IL-6 among participants in the MIDJA biomarker sample. After controlling for demographic factors, health status, and health behaviors, findings showed that consuming a Japanese diet was associated with significantly lower levels of both biomarkers, while more frequent bathing was associated with lower IL-6 but not CRP.

The social relational realm has been interest in investigating cultural differences in biological health risk. Hartanto, Yee-Man, and Yong (2020) examined whether culture moderates the link between perceived obligation to others and both inflammatory and cardiovascular risk. Findings showed that a higher tendency to disengage from stressful social obligations is associated with better health for Americans, but poorer health outcomes among Japanese participants. These results thus underscored culturally distinct pathways between perceived obligations to others and objectively measured physical health.

Metabolic and cardiovascular risk factors.

Links between psychological resources and glucoregulation in the MIDJA sample were investigated (Boylan, Tsenkova, Miyamoto, & Ryff, 2017), to augment prior U.S. findings. Such inquiry is important, given increasing rates of type 2 diabetes in Japan. Three categories of resources were considered: hedonic well-being (life satisfaction, positive affect), eudaimonic well-being (personal growth, purpose in life, ikigai [the sense of “life worth living”; Sone et al., 2008]), and interdependent well-being (gratitude, peaceful disengagement, adjustment). After adjusting for demographic, health status, health behaviors, antidiabetic medication and negative affect, findings showed that purpose in life was associated with significantly lower HbA1c (glycosylated hemoglobin), while hedonic well-being was not associated with HbA1c and peaceful disengagement was associated with significantly higher HbA1c. Although ikigai was not linked to HbA1c, it was related to lower systolic blood pressure. Thus, while the findings on the associations between psychological resources and metabolic risk are mixed, eudaimonic well-being tends to be associated with reduced metabolic risk in Japan, which is in line with the findings in MIDUS (Boylan & Ryff, 2015). These mixed results underscore the need for continuing inquiry into which psychological resources are, or are not, associated with reduced metabolic risk.

Two studies have examined links between psychosocial factors and lipids. Yoo, Miyamoto, Rigotti, and Ryff (2017) linked positive affect to a wider array of serum lipid profiles known to be strongly predictive of risk for cardiovascular disease. Findings showed that healthier lipid profiles were evident among Americans with higher levels of positive affect, but no significant links were evident among Japanese adults. Further analyses showed that this cultural moderation was mediated by body mass index (BMI); positive affect was linked to lower BMI among Americans, but not among Japanese, and lower BMI predicted healthier lipid profiles across cultures. Experiencing positive affect, a culturally normative emotions in the U.S., seems to be associated with lower BMI (possibly through engagement in healthy behaviors, such as exercises) and in turn lead to healthier lipid profiles, but less so in a cultural context where different norms about emotions exist.

Although Yoo et al. (2017) found no link between positive affect and lipids in Japan, this may not necessarily mean that positive affect is irrelevant for health in Japan. Yoo, Miyamoto, and Ryff (2016) examined positive affect in the context of different levels of social connectedness. They focused on two indicators of well-functioning physiology: HDL (high-density lipoprotein) cholesterol, and DHEA-S (dehydroepiandrosterone-sulfate). Lower levels of both biomarkers (i.e., less healthy profiles) were found among Japanese adults who reporter high positive affect in combination with low social connectedness. Such findings suggest that experiencing positive emotions without maintaining high social connectedness is associated with poorer health in Japan. In contrast, the general pattern in the U.S. was that those with higher positive affect showed healthier HDL levels regardless of social connectedness. The findings thus highlighted cultural variation in the health implications of how positive affect and social connectedness come together.

Other studies have examined biological health correlates of personality across cultures. A study using six large community samples, including MIDUS and MIDJA, examined links between personality traits and risk of diabetes (Stephan, Sutin, Luchetti, Canada, & Terracciano, 2020). After controlling for demographic factors and health status, findings showed that higher conscientiousness was consistently linked with lower HbA1c levels across most samples and in the meta-analysis as well as to lower risk of the threshold score for diabetes. This finding, based on more than 26,000 participants, underscored the potential health protective features of conscientiousness across cultural contexts.

While Stephan et al. (2020) found the link between conscientiousness and lower HbA1c levels across cultures, other studies have also shown cultural variation in the links between conscientiousness and metabolic and cardiovascular risk factors. Sutin, Stephan, Wang et al. (2015) examined the associations between personality traits and BMI among three East Asian samples, including MIDJA. Unlike previous findings in Western cultures that found associations between low consciousness and high BMI (e.g., Terracciano et al., 2009), consciousness was not associated with BMI in East Asian samples. Interestingly, extraversion and agreeableness were, however, associated with higher BMI among Asian men. Although the reason is unclear, it may be that men who are socially skilled in Asia tend to eat in a group setting, which could lead to more eating and drinking. Furthermore, as discussed in the preceding section, using a composite index of BHR including both inflammatory markers and cardiovascular factors, Kitayama and Park (2021) found that conscientiousness was associated with lower BHR among Americans, but the association was not significant among Japanese.

Researchers have also examined how personality traits predict physical inactivity (Sutin, Stephan, Luchetti et al., 2016), which is a cardiovascular risk factor and is associated with increased mortality (Schmid, Ricci, & Leitzmann, 2015). They examined the links among 16 samples, including MIDUS and MIDJA as well as other large datasets (e.g., Health and Retirement Study, National Longitudinal Study of Adolescent to Adult Health). Although samples differed in the specific measure of physical (in)activity (e.g., participation in therapies for MIDJA, engagement in vigorous physical exercise for MIDUS), the researchers found that, across all samples, respondents who scored high on neuroticism were more likely to be physically inactive, whereas those who scored high on conscientiousness were less likely to be physically inactive. Extraversion, openness, and agreeableness were also negatively associated with physical inactivity, but their associations tend to be weaker for specific activities.

It is important to underscore that the links between conscientiousness and health behaviors may vary for specific health behaviors due to different cultural norms and beliefs associated with a particular behavior. An example is the work of Lee, Gao, and Ryff (2020) focused on smoking, wherein norms vary widely across societies and by gender. They found that conscientiousness was associated with a lower likelihood of smoking (former or current) across both genders in the U.S. However, there were gender differences in Japan; conscientiousness was associated with a lower likelihood of being a former smoker among Japanese women, but was not associated with smoking status among Japanese men. Consciousness is likely effective in preventing smoking in sociocultural contexts where smoking is stigmatized (e.g., Western cultures, or for women in Japan), but less so in a sociocultural context where smoking is less stigmatized (e.g., men in Japan).

The emerging picture and remaining issues.

Numerous scientific advances have followed from the MIDUS and MIDJA biomarker samples. Several have extended prior findings on cultural differences in emotional and social experiences, showing that linkages between psychosocial factors and different aspects of physiological risk, such as BMI, inflammation, and cardiovascular malfunction, also vary by cultural context. The emerging picture is that psychosocial factors that are relatively more accepted in Japan than in the U.S. (e.g., negative emotions and neuroticism) are less likely to be associated with poor biological health in in Japan than in the U.S., whereas psychosocial factors related to positive emotions and consciousness tend to be associated with better biological health in the U.S. than in Japan, or are linked to health only in specific contexts in Japan. Social relations provided specific contexts where, or specific factors through which, psychosocial factors matter for health in Japan.

What theoretical framework can be invoked to understand such findings on cultural differences in health correlates of psychosocial factors? Social and cultural psychologists have suggested the idea that cultural fit – the extent to which one’s psychological experiences are in accordance with the expectations sanctioned in the given cultural context – leads to beneficial outcomes, such as higher motivation, better relationships, and greater well-being (De Leersnyder et al., 2014; Fulmer et al., 2010; Miyamoto, Yoo, & Wilken, 2019; Stephens, Fryberg, Markus, Johnson, & Covarrubias, 2012). According to a psychosocial resources model (Yoo & Miyamoto, 2018), cultural fit is further thought to lead to better health outcomes by increasing both psychological (e.g., meaning, intrinsic motivation) and social resources (e.g., social integration), which in turn lead to better health via behavioral (e.g., exercise, diet) and biological pathways (e.g., inflammation). That is, such behavioral and biological factors themselves are formulated as pathways (mechanisms) to disease outcomes (morbidity) and mortality. These proposed causal pathways need to be empirically assessed in future research. Critical to such endeavors will be longitudinal designs, such as the multiple waves of MIDUS and MIDJA. With such data, investigators can test whether the putative antecedent temporally preceded change in the putative consequent, thereby increasing confidence in the hypothesized causal directionality. Evidence of bi-directional influences can also be documented, as illustrated by recent findings from MIDUS showing that physical activity and purpose in life each influenced each other over time (Yemiscigil & Vlaev, 2021).

Although cultural fit is an important theoretical concept in advancing research on culture and health, we note that it may not always lead to better health. Indeed, some culturally sanctioned practices may be maladaptive for individuals. For example, karoshi (death from overwork) has been identified as a societal issue of concern in Japan (Iwasaki, Takahashi, & Nakata, 2006). Such practices may partly arise from cultural emphasis on fulfilling one’s duties and obligations, even at the expense of individual health. Other norms within societies that sanction unfair or discriminative treatment of certain subgroups may also have negative effects on members of those subgroups (e.g., Fuller-Rowell, Curtis, Chae, & Ryff, 2018; Fuller-Rowell, Doan, & Eccles, 2012), an issue to which we will return in discussing social inequality in the next section. These examples underscore the importance of being mindful of and critically evaluating the benefits as well as possible harms of culturally sanctioned norms and practices. It is important to understand which norms and practices are beneficial or harmful, including when, where, and for whom.

Reviewing the above findings as a whole, we also want to comment on the apparent inconsistencies evident in some of them – sometimes the evidence supports cultural differences in how psychosocial factors and health are linked, and other times similarities in such linkages are evident across cultural contexts. These patterns need further examination in the future research and will likely be illuminated by paying attention to multiple factors. First, the nature of psychosocial factors under consideration and the associated cultural norms need continuing attention. Most evidence for cultural differences has been accumulated for emotions, for which norms are known to vary across cultures (e.g., Mesquita et al., 2014; Miyamoto et al., 2017). On the other hand, other examples of findings point to cultural similarities include psychological factors such as purpose in life and ikigai, which are sanctioned in both the U.S. and Japanese cultural contexts. Second, the nature of the health outcome may also play a role in discrepant findings. Cultural differences have been more often evident when examining objective measures of health, such as inflammation and cardiovascular malfunction, whereas cultural similarities are more common in studies using self-reported health. These varying patterns need to be supported or refuted by additional science.

Finally, we draw attention to another potentially important factor in making sense of mixed findings: namely, the guiding motivations of researchers studying health across diverse countries. Culturally-oriented investigators tend to focus on cultural meanings and norms surrounding particular psychosocial factors and their health sequelae (Ryff, Love, Miyamoto, et al., 2014; Miyamoto et al., 2019). This approach typically involves comparison of a few cultures (or a single culture) due to the in-depth exploration of the meanings and norms associated with various psychosocial factors. Other investigators seek to test the scope of replicative consistency in factors that influence health across diverse countries. These inquiries more often involve multiple international comparisons with limited attention given to underlying meanings and norms associated with particular psychosocial factors within differing cultural context. This approach is more conducive to finding similarities in links between psychosocial factors and health at the general level. We draw attention to these differences here not to argue that one approach is better than the other, but rather to underscore that comparative studies of health across countries are not always explicitly oriented toward the understanding of cultural contexts, which is the central theme of our review.

Bringing Socioeconomic Inequalities into Work on Culture and Health

Specific socioeconomic contexts in which individuals are located provide certain meanings that shape how psychological processes may be linked to health. Thus, examining psychosocial processes in relation to specific socioeconomic contexts as well as cultural contexts is, in our view, essential. Perhaps due to the reliance on college students as the main source of samples in much contemporary social psychology (Arnett, 2008; Henrich, Heine, & Norenzayan, 2010), psychologists have traditionally not paid much attention to issues of inequality that exist within the societies in which they are embedded. However, there is a growing recognition that socioeconomic inequality is powerfully implicated in psychological processes and health. Furthermore, such socioeconomic inequality is widening over time. Because of the recruitment of representative samples from the U.S. and Japan, MIDUS and MIDJA have fostered research on socioeconomic inequality across cultures. Here, we explain how the intersection of culture and SES contexts is a critical and rich topic for further scientific inquiries.

Health disparities due to socioeconomic status (SES) have been widely documented (Adler et al., 1994; Marmot, 2015; Matthews & Gallo, 2011). For example, compared to high SES people (e.g., those with higher education, occupational status, or income), low SES people tend to experience not only more psychological distress, such as anxiety and depression (Gallo & Matthews, 2003), but also poorer physical health, including increased mortality (Adler et al., 1994). Using the MIDUS data, many studies have elucidated behavioral (e.g., smoking; Glei, Lee & Weinstein, 2020) and biological pathways (e.g., inflammation and hormones; Friedman & Herd, 2010; Zilioli, Imami & Slatcher, 2017) through which SES leads to mental and physical health (see also Kirsch, Love, Radler & Ryff, 2019). Experiences of discrimination have also been documented as adverse influences on health among disadvantaged minorities (Fuller-Rowell et al., 2012, 2018; Surachman, Jenkins, Santos & Almeida, 2021).

Cultural influences on SES-Health links.

How might the surrounding cultural context play a role in the links between SES and psychological and physical health? Although some pathways through which SES is linked to health and well-being may be culturally similar – for example, Yu and Blader (2020) found that higher SES is linked to higher well-being through the perceived respect from their coworkers in both MIDUS and MIDJA, the investigations of MIDUS and MIDJA have shown at least three ways through which cultural contexts influence how SES is linked to health. First, Kan et al. (2014) examined factors that mediate the association between SES and physical health (i.e., self-rated health and chronic conditions). They found that, whereas a sense of control mediated the association across cultures, self-esteem mediated the association in the U.S. but less so in Japan. Thus, self-esteem seemed to play a more prominent role in the pathway between SES and health in a cultural context where having a positive self-view is valued (Heine, Lehman, Markus, & Kitayama, 1999).

Second, which particular facets of SES are related to psychological well-being may differ across cultures. SES has both objective (e.g., education) and subjective facets (e.g., subjective perceptions of one’s status within the society; Adler et al., 2000). Curhan, Levine, Markus et al. (2014) showed that subjective SES was more strongly associated with well-being in the U.S., where personal judgments of the self are valued more so than in Japan. In contrast, objective SES was more strongly associated with well-being in Japan, where others’ judgment of the self plays a more prominent role, than in the U.S.

Third, the strength of the link between SES and health may differ across cultures. Several studies found weaker links between SES and psychological and physical health (e.g., depressive symptoms, mortality, morbidity, and health behaviors) in East Asian cultures (e.g., Inaba et al., 2005; Kagamimori, Gaina, & Nasermoaddeli, 2009; Kim, Symons, & Popkin, 2004; Lahelma et al., 2010; Takao, Kawakami, & Ohtsu, 2003). Using MIDUS/MIDJA, Takahashi and colleagues (Takahashi, Fujiwara, Nakayama, & Kawachi, 2018) showed that such cultural differences extend to the expected trajectory of one’s health. Respondents were asked to rate their current health status, their past health status from 10 years ago, and their predicted future health status in 10 years. In the U.S., there was an SES gradient in the expected trajectory of health decline; low SES people expected their health to decline more steeply than high SES people. On the other hand, in Japan, people expected their health to decline, but no SES gradient in the expected trajectory was found. Structural and institutional factors, such as the availability of the universal healthcare system, likely underlie such differences, although the cultural meanings associated with SES may also play some role, which we turn to next.

Cultural meanings of SES: Privileges and burdens.

The importance of cultural contexts in shaping meanings and expectations associated with varying SES strata and their psychological sequelae have been suggested by cultural psychologists (Ishii & Eisen, 2020; Miyamoto, Yoo, Levine et al., 2018; Schooler, 2007; Snibbe & Markus, 2005; Stephens, Markus, & Phillips, 2014). In American cultural contexts, where the independent view of the self is dominant (Markus & Kitayama, 1991), many findings have shown that high SES people tend to focus on pursuing the self-set goals (i.e., self-orientation) that reinforce the independence of the self (Stephens et al., 2014). For example, higher SES was associated with higher self-esteem and entitlement in the U.S. (Côté et al., 2021; Kohn, 1969; Piff, 2014). On the other hand, in East Asian cultural contexts, where the interdependent view of the self is dominant (Markus & Kitayama, 1991), high SES is linked not only to higher self-orientation (Kohn, Naoi, Schoenbach, Schooler, & Slomczynski, 1990; Takemura, Hamamura, Guan, & Suzuki, 2016), but also to the maintenance of relationships and fulfillment of social responsibilities (i.e., other-orientation; Miyamoto et al., 2018; Na, McDonough, Chan, & Park , 2016; Naoi & Schooler, 1985; Rule et al., 2010) that support interdependence of the self. Based on analyses of MIDUS and MIDJA data as well as data from the World Values Survey, Miyamoto and colleagues (2018) showed that higher SES was associated with both self-orientation (e.g., self-esteem) and other-orientation (e.g., sympathy, support given to others) in Japan, whereas the association between SES and other-orientation was weaker or absent in the U.S. The fact that high social status people in Japan tend to engage in both self-orientation and other-orientation may confer privileges and burdens to them.

In terms of privilege, Park and colleagues (2013) examined how SES is linked to expression of anger using MIDUS/MIDJA. They theorized that, in Japan, where there is a normative pressure to suppress anger, only those who have high social status are conferred a privilege to express anger. The results showed that higher objective SES (education, occupation) was linked to more anger expression in Japan, and the link was mediated by the decision-making authority (i.e., the extent to which they had the authority to make decisions at work). In contrast, in the U.S., where frustration resulting from thwarted goal pursuit is the main determinant of anger expression, those with lower social status who tend to experience more life difficulties may express anger. Supporting this prediction, they found that lower subjective SES was linked to more anger expression in the U.S., which was mediated by feelings of frustration.

At the same time, higher social status, especially in terms of occupational positions, are also associated with possible burdens in Japan. Compared to Western cultures, people who occupy a managerial position in Japan tend to work longer hours, which is linked to unhealthy lifestyles (e.g., less sleep, less exercise) and higher stress (Maruyama & Morimoto, 1996). Further, the rise in the suicide rate since the 1990s in Japan was most noticeable among people in managerial positions, likely tied to their increased responsibilities and job demands during the recession (Wada et al., 2012). Similarly, one-third of victims of karoshi (death from overwork) in years following the economic crisis in Japan were from professional and administrative positions, despite the smaller share of such higher occupational positions in the workforce (Kondo & Oh, 2010). A recent survey with Japanese respondents in the workforce (n = 9,123) showed that higher managerial status was associated with higher well-being yet worse self-reported physical health (e.g., sleep problems, pain) in Japan (Miyamoto, 2021) after controlling for demographic factors (e.g., age, gender). Such a negative association between managerial status and health was partly explained by the tendency for people who occupy higher managerial status to have excessive working styles (Miyamoto, 2021). Burdens associated with higher status may partly underlie the weaker (or even reversed) link between higher social status and health in Japan.

The emerging picture and remaining issues.

There is rapidly accumulating evidence on socioeconomic inequality in health and their psychological correlates across cultures using the MIDUS and MIDJA samples. Some studies have shown different ways through which culture influences how SES is linked to health. Furthermore, other studies have shown culturally shared and depended psychological correlates of SES, and highlighted both privileges and burdens associated with higher social status in Japan. The emerging picture is that, while socioeconomic inequality exists across cultures, how it manifests in psychological processes and health depends on cultural meanings and expectations associated with varying socioeconomic status. While higher status in Japan is associated with privilege to express anger and thus may signal dominance, it is also associated with higher burden and responsibility, which can have some health costs.

The exploration of socioeconomic inequality across cultures has just begun. There are many important questions yet to be explored and addressed. For example, while most studies have focused on cultural meanings and expectations associated with higher status people in Japan, what are the implications for lower social status people in Japan? People who are at high risk for “NEET” (i.e., Not engaged in Employment, Education, or Training) in Japan have shown lower levels of interdependence compared to those who are at low risk (Norasakkunkit & Uchida, 2011). Building on such findings, it would be fruitful for future research to elucidate cultural meanings and expectations associated with lower SES strata and their psychological meanings. One promising avenue is to employ qualitative strategies to illuminate cultural meanings surrounding lower (vs. higher) SES strata. Illustrating this idea with MIDUS, Markus, Ryff, Curhan and Palmersheim (2004) conducted in-depth qualitative interviews with a subset of participants to examine meanings of well-being among high school-educated and college-educated adults in the U.S. The results revealed that among high school-educated adults, narratives about meanings of well-being centered around family relationships, financial security, and jobs, as well as adjusting to the world (e.g., high school-educated woman: “A good life is having the things you need. Having your health. You know, with the things you need. […] I’m talkin’ about havin’ a roof over your head, a job, some kind of security. That would be to me—it’s havin’ a good life. Friends and family. Without them, maybe I’d be a little lonely.” p.286; high school-educated man: “A good wife. How she takes care of me. She’s nice and she’s sweet. I just love her. And each day God blesses me to see her. It seems like I grow more in love with her. It’s just, she’s just a sweet person.” pp.301–302), whereas enjoyment, achievement, and influencing the world were more prominent among the college-educated adults’ narratives about meanings of well-being. Such a qualitative approach to well-being among those from low (vs high) SES levels in Japan could provide informative bottom-up insights into the cultural meanings and expectations surrounding life from differing social strata in Japan.

Also, the findings examining psychological correlates of SES in diverse cultures have relied on self-reported measures of psychological processes and behaviors (e.g., support provided to others, anger expression, perceived discrimination). However, how high (vs. low) status people perceive and interpret their own behavior may diverge from how such behavior is perceived and interpreted by other people from the same or different SES strata. Exploring such potential discrepancy between own and other person’s perspectives may illuminate interpersonal dynamics that underpin the observed psychological correlates of SES in Japan. Furthermore, what are the consequences of the increasing inequality across cultures? In the next (and last) section, we turn to the issue of increasing economic inequality that faces many societies around the world.

Summary and Our Changing Historical Contexts

We have focused this review on scientific advances in two key areas of research at the interface of culture and health. The first emphasized a more micro perspective on biological pathways that are part of how psychosocial and behavioral factors matter for diverse health outcomes in different cultural contexts. The second emphasized more macro-level influences involving socioeconomic inequality that are central to understanding health differences within cultures. Both of these directions represent relatively new questions in cultural psychology and thus are important directions for continuing future research, possibly using publicly available data from MIDJA and MIDUS. While such inquiries are unfolding, it is important to acknowledge change in the surrounding historical contexts within which such science is conducted. We therefore conclude this review with a call for attention to two specific historical issues: widening economic inequality and the world-wide COVID-19 pandemic.

Widening Economic Inequality

Sadly, economic inequality has been increasing over time in most OECD countries since the mid-1980s (OECD, 2011) though the increase is notably steeper in the U.S. compared to European countries (Graham, 2017; Piketty & Saez, 2014; Reeves, 2018). These problems were exacerbated by the Great Recession of 2008, which fueled dramatic increases in poverty and accompanying health costs due to job loss, unemployment, and financial strain (Burgard & Kalousova, 2015; Kirsch & Ryff, 2016). This heightened trauma among the disadvantaged has been linked with growing “deaths of despair” – i.e., suicides, deaths due to addictions, such as opioids and alcohol among midlife adults in the U.S. (Case & Deaton, 2015, 2020).

A unique feature of the MIDUS design made it possible to examine historical change in health inequalities. Specifically, MIDUS includes two national samples of U.S. adults situated on either side of the Great Recession. The baseline sample (aged 25 to 74) was recruited in 1995, and was subsequently followed by recruitment of a new “refresher” sample (same ages) in 2012. Over this period of time, educational attainment in the U.S. improved: college-educated adults increased by 24.8% to 33.2% and those with less than a high school degree of education decreased from 15.3% to 11.3%. Despite such gains the post-Recession sample reported less household income (after adjusting for inflation) and lower financial stability than the pre-Recession sample (Kirsch et al., 2019). The post-Recession sample also had worse health (chronic conditions, body mass index, functional limitations, physical symptoms) and lower levels of many aspects of eudaimonic and hedonic well-being. Further work from MIDUS comparing these two national samples (Goldman, Glei, & Weinstein, 2018) showed that mental health (using multiple negative and positive indicators) had become more compromised among the later refresher (post-Recession) sample compared to the baseline MIDUS sample particularly among those of lower socioeconomic positions (measured with a composite of education, occupation, income, wealth).

The economic inequality has also been rising consistently in Japan since the mid-1980s (Shirahase, 2014). Parallel cross-time trends of health in Japan need to be examined as well, thereby underscoring the need for recruitment of new probability samples of sociodemographically diverse participants. That is to say, government-supported research in both the Japan and the U.S. is needed to facilitate continuing scientific inquiries that build on the kinds of research we have described above, while also remain responsive to changing historical contexts. Relevant questions are whether historic economic crises lead to comparable widening of health inequalities across all cultures, or whether some countries are less prone toward deepening discrepancies in life resources and opportunities? Prior research on the impact of the economic crisis in Japan, conducted with a nationally representative sample over two decades, found that health inequalities across income groups actually narrowed after the economic crisis in Japan. Interestingly, this outcome was mostly driven by the decline in health among middle to high-income people (Kachi, Inoue, Nishikitani, Tsurugano, & Yano, 2013; see also Kondo, Subramanian, Kawachi, Takeda, & Yamagata, 2008). The authors speculated that the health decline of higher-income people may reflect corporate restructuring that had increased the workloads of those with higher incomes (Kachi et al., 2013). Missing from these national health studies are the rich psychosocial, behavioral, and biological assessments involved in the MIDJA and MIDUS endeavors, thus underscoring the need for broadening the multidisciplinary scope of national health research, guided by a commitment to understand cultural factors of the sort covered in this review.

The COVID-19 Pandemic

Over the past year the world has been shaken by the COVID-19 pandemic that has contributed to soaring death rates across countries, while also in many places, aggravating pre-existing disparities in unemployment, education, housing, health and survival. A study done in the U.K. found that risk of COVID-19 hospitalization was higher among low SES people than among higher SES people (Patel, Paranjpe, Kathiresan, Rivas, & Khera, 2020). Data from the U.S. showed that the economic fallout of the pandemic hit lower-income Americans the hardest, measured in terms of problems paying bills, receiving help from a foodbank, job loss, and risk of eviction (Kochhar, 2020). More recent evidence has shown that that the wealthiest segments of society were more likely to stay at home during the lockdown and rely on delivery services compared to those at the low end of the income hierarchy (Serkez, 2021). The disadvantaged, in turn, were more likely to experience changes in employment and have greater problems with online schooling (availability of a computer, evidence of less progress in math coursework) compared to the advantaged. Finally, life expectancy fell for all in the U.S. over the past year, although notably more so for Black compared to White Americans (Serkez, 2021).

Data from Japan indicates that gender differences are important to tracking effects of the pandemic. Risk of unemployment and household burdens have been placed disproportionately among women (Nomura et al., 2021). Re-invoking the above theme of increased deaths of despair in the U.S., the suicide rate during the COVID-19 pandemic increased significantly among Japanese women (20 to 30% more compared to the previous year) but not among Japanese men (Nomura et al., 2021). Overlapping with the gender differences is the divide between regular workers and nonregular workers, many of whom are women. Compared to regular workers, nonregular workers have less job security, lower compensation, and less benefit (Inoue & Dooley, 2020). The gap between regular and nonregular workers became even more evident during the pandemic; while the number of regular workers increased in 2020 from a previous year, the number of nonregular workers dropped for 750,000 (Ministry of Internal Affairs and Communications, 2021). Furthermore, foreign technical trainees, who come from developing countries to work in Japan and often experience harsh working conditions (Oono, 2019), are especially vulnerable to social isolation (Nagata, 2020) and job losses (Uchihashi, 2021) during the pandemic. These early reports thus implicate different patterns of impact in differing countries: socioeconomic status and race are key indicators of disproportionate suffering in the U.S., whereas gender differences and vulnerable workers appear prominent in patterns of impact in Japan. Older people in both contexts have been at increased risk of death, but how their burdens may have been managed may vary by country and cultural context.

Missing from these initial pandemic reports are further important questions – for example, what psychological and social experiences underlie these experienced stresses of the pandemic, and how might they vary depending on sociodemographic factors as well as cultural context? How might those in positions of privilege and power (defined by educational, economic, occupational status) versus those who are disadvantaged be at risk for, or protected from, pandemic-related burdens and trauma? Are their cultural differences in conceptions of social duties and responsibilities during periods of enormous societal strain? How are these communicated and what is their impact? In thinking about constructive responses, what kinds of interventions and public policy actions are needed to assist those who are most vulnerable in coping with pandemic-related losses as well as to mobilize the more comfortable to do their part? And which interventions and policies are most effective in particular cultural contexts? These wide-ranging questions are at the heart of how countries around the globe must respond to the major historical challenges currently unfolding. Drawing of the recent scientific advances summarized in our review, we submit that culturally informed perspectives will be needed to understand the health impacts of growing inequality and the COVID-19 pandemic. What people are facing, how they can most adaptively respond, and what will facilitate long-term recovery (individually and societally) from adverse historical events demands looking through the lens of culture.