At this point, the limited amount of empirical evidence is mixed; some evidence supports cross-cultural continuity (eg, Pressman et al., 2013), whereas other evidence is consistent with cross-cultural variation in the association between negative affect and health (eg, Miyamoto et al., 2013; Miyamoto & Ryff, 2011). One reason for these conflicting findings may be the lack of consensus in how emotion and health are measured. Some studies have measured state affect (ie, how people feel in a given moment or on a given day), and others have measured trait affect (ie, how people typically feel). Additionally, the measures of health outcomes used in these studies varied widely in terms of relative subjectivity/objectivity as well as in their clinical relevance. Finally, conclusions based on significance testing increase the possibility of inferring cross-cultural similarity when examining large samples. Thus, we focus here on comparing effect sizes. Addressing this issue, we compared the magnitude of the effect of negative affect on health in the United States and Japan using a stable index of negative affectivity and six clinically relevant, well-known self-report health metrics. The United States/Japan comparison is a relatively ideal one because both nations are modernized, democratized, industrialized societies with well-developed systems of health care. Yet these two societies are markedly different in their historically derived ideas about negative affect and in the everyday social practices that lend form and organization to affective experience (Markus & Kitayama, 1994; Mesquita & Leu, 2007). To examine this possibility, we compared survey data from two large samples of Japanese (n= 988) and American adults (n= 1,741) participating in the Midlife in the United States (MIDUS) and Survey of Midlife Development in Japan (MIDJA) survey studies. To measure negative affect, participants reported how often (1= none of the time, 5= all of the time) they had experienced negative emotions (ie, how often they had felt nervous, hopeless, lonely, afraid, jittery, irritable, ashamed, upset, angry, and frustrated) over the previous 30 days. We indexed physical health using two relatively objective measures—number of chronic conditions and degree of functional limitations—and we administered a single-item measure of subjective global health. We indexed mental health using two multiitem measures of psychological well-being and selfesteem, and we administered a single-item measure of life satisfaction. We included positive affect1 and demographic variables as covariates in our analyses (for details, see Methodological Details in the Supplemental Material available online). Japanese participants reported higher mean levels of and variance in negative affect (M= 1.80, SD= 0.62) than did Americans (M= 1.57, SD= 0.53), t (1806.31)= 9.52, p<. 001, Levene’s F (1, 2727)= 65.53, p<. 001.

Overall, we found that for each measure, negative affect significantly predicted poor health in both the United States and Japan. However, a comparison of the magnitude of the effect revealed that negative affect was indeed worse for one’s health in the United States than in Japan (see Fig. 1). Differences in negative affect–health associations (calculated as critical ratios of the differences) indicated that in the United States, compared with Japan, negative affect more strongly predicted more chronic conditions, z= 6.47; worse physical function, z= 2.45; worse psychological well-being, z= 6.59; and lower self-esteem, z= 5.65. Across cultures, negative affect similarly predicted poor global health, z= 0.62, and lower life satisfaction, z=− 0.62. Multigroup structural equation modeling confirmed these�…