This Article Abstract Full Text (PDF) Services Download to citation manager PubMed PubMed Citation

Ethnic Variation in the Impact of Emotion and Emotion Regulation on Health: A Replication and Extension

¹ Department of Psychology
² Intercultural Institute on Human Development & Aging, Long Island University, Brooklyn, New York.

Address correspondence to Nathan Consedine, Department of Psychology, Long Island University, 191 Willoughby Street, Suite 1A, Brooklyn, NY, 11201. E-mail: nconsedi{at}liu.edu

	Abstract

TOP Abstract Methods Results Discussion References

 
Although emotions and patterns of emotion regulation are central to models linking personality and health, the generalizability of these models to diverse populations of older adults remains untested. In this study, 1,364 community-dwelling women (aged 50–70 years) from six ethnic groups completed self-report measures of trait anger, inhibition, defensiveness, and health. As expected, reports of trait anger and emotion inhibition predicted poorer health (and defensiveness better health), even when demographics and health behaviors were controlled. However, these characteristics related to outcome differently across ethnic groups; greater anger was related to better health in all groups other than U.S-born European Americans, and increased emotion inhibition was associated with better health among immigrant Eastern European women. Results are discussed within a contextualistic model of emotions and health, and directions for future research are given.

Research has demonstrated complex relations between patterns of negative emotions and emotion regulation and physical health outcomes (Consedine, Magai, & Bonanno, 2002; Mayne, 2001). Previous research has demonstrated that emotion variables are related to several health outcomes (Mayne, 1999, 2001), including cardiovascular diseases (Krantz & McCeney, 2002; Sirois & Burg, 2003), immunological functioning (Baum & Pozluszny, 1999; Kiecolt-Glaser, McGuire, Robles, & Glaser, 2002), and early mortality (Harburg, Julius, Kaciroti, Gleiberman, & Schork, 2003). In this article we examine the relations between health and trait anger, repression, and defensiveness in a large sample of older, community-dwelling women from six ethnic groups.

Trait anger has been associated with poor health in a number of areas. It has been linked to somatic complaints in samples of older individuals (Mora, Robitaille, Leventhal, Swigar, & Leventhal, 2002). Equally, trait anger has emerged as the central affective component of the coronary-prone, Type A personality (Sirois & Burg, 2003), with links to stress reactivity and ambulatory blood pressure (Schum, Jorgensen, Verhaeghen, Sauro, & Thibodeau, 2003).

Patterns of regulating emotions and stress are also thought to have health implications. Defensiveness, alone or as part of the repressive personality (Mendolia, 2002; Weinberger, Schwartz, & Davidson, 1979), has been linked to elevated physiological responding (Niaura, Herbert, McMahon, & Sommerville, 1992) and impaired immune functioning (Jamner & Leigh, 1999). Conversely, however, repressive constructs have been linked to reports of better health (Brosschot & Janssen, 1998). Defensiveness has been related to biased appraisals of health (Myers & Reynolds, 2000). It has been suggested that self-deception (Werhun & Cox, 1999) may produce distortedly positive self-reports (Furnham, Petrides, Sisterson, & Baluch, 2003).

Finally, the inhibition of emotion expression has also been related to poorer health, for example, in the case of cancer (Cooper & Faragher, 1992). The notion that the free expression of emotion is healthy has long been maintained (Leventhal & Patrick-Miller, 2000; Pennebaker & Seagal, 1999), whereas inhibition has been linked to poorer physical health (Consedine, Magai, Cohen, & Gillespie, 2002). Inhibition heightens physiological responding (Gross, 1998), and it may be that the autonomic responding associated with inhibition (Consedine, Magai, & Bonanno, 2002) impairs the immune system (Pennebaker, Kiecolt-Glaser, & Glaser, 1988) and thus damages health.

However, although all three variables are thought to affect health by means of similar direct (Krantz & McCeney, 2002; Suinn, 2001) and indirect pathways (Consedine, Magai, & Neugut, 2004), the ethnic generalizability of these models remains in question. Theories relating personality to health have been based on data from samples of predominantly affluent European individuals (Consedine, Magai, & Bonanno, 2002; T. Q. Miller, Smith, Turner, Guijarro, & Hallet, 1996), and findings are mixed (Smith, 1992). A recent study of 1,118 lower income, community-dwelling older adults (M = 74 years) from four ethnic groups, for example, found that ethnic group membership interacted with inhibition and negative emotion in the prediction of health (Consedine, Magai, Cohen, et al., 2002). Specifically, although emotion inhibition predicted sleep disturbance in all groups, this effect was greater among African Americans and English-speaking Caribbeans. Similarly, emotion inhibition interacted with being Eastern European (vs. European American) in predicting arthritis, hypertension, and respiratory distress; emotion inhibition had a reduced effect among Eastern Europeans. It has been argued that the impact of trait emotion and emotion inhibition on health may vary depending on the meaning associated with emotion experiences and the regulation of emotion expressions (Roloff & Ifert, 2000).

Although there are few studies that have directly assessed the values that different ethnic groups hold regarding emotions and emotion regulation, it is known that they vary in their characteristic patterns of emotion, defensiveness, and emotion expression or inhibition. African Americans, for example, typically report lower negative emotion (Consedine & Magai, 2002), less depression (Baker, 2001), lower emotional expression (Brantley, O'Hea, Jones, & Mehan, 2002), and greater defensiveness (Marin, Gamba, & Marin, 1992) than Whites.

Developmental data circumscribe differences in emotion socialization that again suggest different values surrounding emotion and its expression (Deater-Deckard, Dodge, Bates, & Pettit, 1996; Payne, 1989). African American (Brody & Flor, 1998; Deater-Deckard et al.; Pinderhughes, Dodge, Bates, Pettit, & Zelli, 2000) and African Caribbean (Gopaul-McNicol, 1999) parents use greater physical means to socialize children (Brody & Flor, 1998; McGroder, 2000), and they encourage expressive control, perhaps because of discrimination (Dilworth-Anderson, 1998) or concerns regarding the consequences of free expression in unsafe environments (Brody & Flor; Payne). Conversely, however, ethnic differences in the types of situations that elicit anger (Stevenson, Herrero-Taylor, Cameron, & Davis, 2002) create the possibility that trait anger may have some adaptive value among urban, low socioeconomic status (SES), minority individuals. Finally, it may be that reports of lower trait anger or greater inhibition are indexing the presence of a global, repressive personality style (we are grateful to an anonymous reviewer for this intriguing suggestion). In either case, there is reason to suspect that self-reported emotion inhibition and trait anger may relate to health outcomes differently in these groups.

A small body of research suggests that Russians place a premium on emotional honesty (Wierzbicka, 1999), group solidarity, and the consideration of others (Ispa, 1994; Williams & Ispa, 1999), and on emotional altruism (Goldberg & Shmelev, 1993). We note here that while this characterization may be changing as the Russian Federation adapts to democracy and new economic climates (Goodwin, Nizharadze, Nguyen Luu, & Emelyanova, 1999; Ispa, 1995), it likely represents an accurate portrayal of socialization values during the time the older women in our sample (M_age = 60.8 years) were being raised.

As values that clearly require some strategic emotion regulation—the expression of negative emotions is frequently unpleasant for others (Consedine, Magai, & Bonanno, 2002)—these data may suggest underlying differences in the reasons for emotion inhibition. Western groups may experience emotion inhibition as inconsistent with notions of freedom and individuality and as externally imposed (Markus & Kitayama, 2003), whereas Russians may view inhibition as normative given their interpersonal values. Russians view a lack of self-control as highly problematic and maladaptive (Peabody, Shmelev, Andreeva, & Gramenitsky, 1993).

In this study we sought to explicitly test whether trait anger, defensiveness, and emotion inhibition related equally to health in women from six ethnic groups, and we also sought to contrast their affective profiles. Specifically, we expected that differences in cultural values and socialization among groups of U.S.-born European American, U.S.-born African American, English-speaking Caribbean, Haitian, Dominican, and Eastern European women would mean the following:

1. Anger would be lower among U.S.-born African American and Caribbean women, relative to U.S.-born European American and Eastern European women; in the absence of prior research, we made no specific predictions about either Haitian or Dominican women.
   
2. Defensiveness and inhibition would be higher among U.S.-born African American and Caribbean women, although, again, no predictions were made regarding Haitian or Dominican women.
   
3. We also expected that greater self-reported trait anger and inhibition would predict poorer self-reported health, and increased defensiveness better health, across groups, even when demographics and health behaviors were controlled. However, because evidence suggests that emotions and their regulation may (a) occur in different situations and (b) have different meanings among diverse ethnic groups, we also expected to find the following.
   
4. Self-reported emotion inhibition would predict better self-reported health for Eastern European women (Consedine, Magai, Cohen, et al., 2002).
   
5. Self-reported trait anger would predict poorer self-reported health for U.S-born European Americans, but not for other ethnic groups.
   

	METHODS

TOP Abstract Methods Results Discussion References

 
Participants
Participants in the study were 1,364 women, ranging in age from 50 to 70 years, living in Brooklyn, New York, and recruited for a study on emotions and women's health. On the basis of prior findings, we distinguished between U.S.-born African Americans and Caribbean immigrant women who were from either (a) the English-speaking territories (e.g., Barbados, Trinidad, and Tobago, Jamaica), (b) the Spanish-speaking part of Hispaniola (the Dominican Republic), or (c) the Creole-speaking part of Hispaniola (Haiti). U.S.-born European Americans served as a comparison group, with immigrants from Eastern Europe serving as a White immigration control group. This latter sample was made up of women from the former United Soviet Socialist Republic countries of Russia, Ukraine, and Belarus—they were all Russian-speaking Eastern Slavs. As Eastern Slavs, these women are ethnically similar (Margonoff & Folwarski, 1996) and so we combined their numbers to yield a sample size comparable with the other ethnic groups.

Procedures
We obtained permission to conduct the study from the Long Island University Institutional Review Board; we collected data over the course of approximately 2 years during 2000–2002. Interviewers collected data during face-to-face interviews that lasted about 90 min and that were conducted in the respondent's home or another location of their choice, such as a senior center or church. A large number of community interviewers were trained specifically for the project. Female interviewers recruited women fitting the sampling criteria from their community and interviewed them in their native language; Dominicans were interviewed in Spanish, Haitians in Creole, and Eastern Europeans in Russian. For the non-English-speaking individuals in the sample, translators translated instruments into Spanish, Creole, and Russian and then, consistent with standard ethnographic practice, backtranslated them to ensure comparability; interviewers administered measures in a standard order for all respondents.

Measures
Demographics questionnaire
We used a demographic questionnaire to elicit information regarding ethnicity, age, education, marital status, and household income.

Health risks questionnaire
We used a questionnaire to ask participants to record smoking and drinking history, as well as height and weight, which we then combined (Deurenberg, Yap, & van Staveren, 1998) to calculate body mass index, or BMI (kg/m²). We measured alcohol consumption with the Frequency-Quantity Index (Knupper, Fink, Clark, & Goffman, 1963), which asks respondents to indicate their frequency of drinking beer, wine, and spirits. We summed items from the three types of drinking to create an aggregate drinking measure ( =.73). This variable was, however, positively skewed and did not respond to several transformations; we converted the scores to a dichotomous form (present–absent) for analysis. We assessed smoking history by having interviewers ask respondents to indicate the number of cigarettes or packs of cigarettes they smoked daily and the number of years they smoked. Again, however, the variable was quite skewed and did not respond to transformation; we dichotomized it on a present–absent basis.

Self-reported physical health
We assessed physical health by using the physical health scales of the Comprehensive Assessment and Referral Evaluation (CARE; Golden, Teresi, & Gurland, 1984; Teresi, Golden, & Gurland, 1984). The CARE is a semistructured instrument that was developed to assess health among community-dwelling older adults in a national study and has been used extensively with geriatric populations, including indigent and minority populations (Consedine, Magai, Cohen et al., 2002). It has demonstrated good construct validity (Teresi, Golden, Gurland, Wilder, & Bennett, 1984) as well as concurrent and predictive validity (Teresi, Golden, & Gurland, 1984). For the current study, we combined 117 items from the somatic symptoms, heart disorder, stroke, respiratory symptoms, arthritis, leg problems, sleep disorder, hearing disorder, vision disorder, and hypertension scales to form an aggregate health complaints scale ( =.94). Questions describe objective health symptoms and are answered in a yes–no format. Items include questions such as "Do you have a blood pressure problem?" and "Are you taking a drug prescribed for hypertension?" (hypertension), "Do you have breathlessness?" and "Do you presently have a cough?" (somatic), "Do you experience backaches?" and "Do you have pain in your joints?" (arthritis), and so on. The aggregate variable was positively skewed and we square-root transformed it.

Self-reported trait anger and emotion regulation
As noted, theory suggests that characterological anger and emotion inhibition affect health by cumulatively damaging physical systems over time. We assessed self-reports of trait anger and the tendency for a person to inhibit or withdraw when emotionally aroused. We measured trait anger with the 11 items from the State-Trait Personality Inventory (Spielberger, 1986), which we summed to provide an aggregate trait anger score ( =.80). We assessed trait inhibition or withdrawal with 12 items from the Present Personality Questionnaire, a 24-item measure assessing trait expressive and inhibitory tendencies (Consedine, Magai, Cohen et al., 2002). The scale includes items such as "I try not to let my anxieties show," "I tend to suffer in silence when I am worried," and "I have difficulty expressing my anger." We combined the 12 inhibition items to form an aggregate measure ( =.83).

Defensiveness
Because socially desirable or defensive responding may obscure the relations between negative emotions and outcomes (Bardwell & Dimsdale, 2001), we used the Marlowe–Crowne Social Desirability Scale (Crowne & Marlowe, 1960). Although originally designed as a measure of social desirability, scores on this 33-item instrument are widely interpreted as accessing a defensive personality style (Jamner & Leigh, 1999; Kline, Bell, Schwartz, Hau, & Davis, 1998; Mente & Helmers, 1999). Internal consistencies for the aggregate measure are generally between.73 and.88 (Bardwell & Dimsdale), and the alpha was =.73 in the current study.

	RESULTS

TOP Abstract Methods Results Discussion References

 
Overview
We conducted our analyses to facilitate our primary interest in ethnic interactions. To identify significant demographic and health behavior covariates of health to be controlled in subsequent regressions, we first examined ethnic differences in demographic and health risk behaviors. Second, we considered ethnic differences in trait anger, inhibition, and defensiveness. Finally, we used multiple regression to examine whether these psychological variables predicted health equally in our six groups.

Ethnic Differences in Demographics and Health Behaviors
Because demographic factors and health behaviors are closely linked to health outcomes, we began our analyses by conducting descriptive multivariate analyses of covariance (MANCOVAs) and follow-up MANOVAs to identify ethnic differences in demographics and lifestyle risk factors. There were significant group differences for age, income, education, and marital status hence (see Consedine et al., 2004 for a fuller description of the sample), as well as ethnic differences in lifestyle risk factors (smoking, drinking, and BMI). We treated both demographic and lifestyle risk factors as covariates in our subsequent analyses of health. (In the interests of brevity, we have not presented the analyses of ethnic differences in either demographics or lifestyle risk behaviors. A demographic characterization of the sample can be found in Consedine et al., 2004, and readers interested in the lifestyle risk factors should contact N. Consedine directly.)

Ethnic Differences in Trait Anger, Trait Inhibition, Defensiveness, and Health
Next, we tested our hypotheses regarding ethnic differences in our three emotion variables, specifically examining whether anger was lower, and defensiveness and inhibition greater, among U.S.-born African American and Caribbean women. We began by conducting a multivariate analysis of covariance (MANCOVA), with ethnicity as the independent variable; trait anger, trait inhibition or withdrawal, defensiveness, and total impaired health as dependent variables; and age, income, education, and marital status as covariates. The model was significant for ethnicity, with a Wilks lambda value of = 28.07 and p <.01. However, age, income, and education were also significant for at least one dependent measure. We next ran a MANOVA with ethnicity, education, income, and age as factors and the three emotion measures as dependent variables. Again, we constrained the model to permit only main effects and two-way interactions; the overall model remained significant for ethnicity ( = 12.68, p <.01).

Greater anger was predicted by ethnicity, F(5, 1,334) = 16.95, p <.01, and younger age, F(1, 1,334) = 6.38, p <.05. Post hoc tests supported our expectation that self-reported anger would be lower among African American and Caribbean women than among Eastern European and U.S.-born European American women. Dominican women reported more anger than women from all other groups (see Table 1).

Defensiveness was predicted by ethnicity, F(5, 1,334) = 18.37, p <. 01, and greater age, F(1, 1,334) = 7.15, p <.05, and there was an Income x Education interaction, F(5, 1,334) = 6.03, p <.05; more poorly educated women were more defensive at greater levels of income whereas better educated women were less. As expected, English Caribbean women were more defensive than Eastern European or U.S.-born European American women, although African Americans did not show the predicted difference with the two European-descent groups.

Finally, scores on the total impaired health measure were related to ethnicity, F(5, 1,334) = 15.84, p <.05; education, F(1, 1,334) = 21.62, p <.05; and greater age, F(1, 1,334) = 38.82, p <.05. However, the ethnic effect was qualified by an Ethnicity x Income interaction, F(5, 1,334) = 2.67, p <.05, in which poorer health was predicted by lower income for U.S.-born African American, English Caribbean, and Eastern European women, but it was not related to income among women from the other three groups.

Zero-Order Correlations and Health
As we can see in Table 2, health impairment was positively associated with age, drinking and smoking history, BMI, trait anger, and inhibition, and it was negatively related to income, education, and defensiveness. The emotion variables were also interrelated, with anger being inversely related to defensiveness but positively related to inhibition; inhibition was negatively related to defensiveness.

Consistent with our expectations, 6 of the 15 interaction terms were significant, with a further 4 having marginally significant relations with health (see Table 4). We predicted that self-reported trait anger would interact with ethnic group such that it would be associated with poorer health among U.S.-born European Americans but not among other groups. This hypothesis was partially supported. The negative interaction terms between anger and being Haitian or Dominican (together with the marginal interactions for three other groups) show that, although higher anger was related to poorer health among U.S-born European Americans, it was related to better health in several minority groups.

Second, a significant Eastern European by inhibition action term supported our expectation that although greater inhibition would predict poorer health overall, it would predict better health among Eastern European women. Finally, there were significant interactions between defensiveness and being U.S-born African American, English Caribbean, or Haitian. These terms indicated that although defensiveness was related to fewer reports of health impairment among these groups, it was not related among the other three groups.

	DISCUSSION

TOP Abstract Methods Results Discussion References

 
The current study was a replication and extension of a previous investigation (Consedine, Magai, Cohen et al., 2002) and contributes empirical evidence to the body of work linking emotion and emotion regulatory tendencies to physical health (Harburg et al., 2003; Schum et al., 2003). As expected, women from the six ethnic groups studied reported differing levels of trait anger, defensiveness, and emotion inhibition, and these variables showed the predicted relations with health—trait anger and emotion inhibition both predicted poorer health and defensiveness better health—even when background demographic variables, health risk behaviors, and ethnicity were controlled. Most importantly, however, all three emotion variables interacted with ethnicity, demonstrating that the relations between these variables and health outcome are not equivalent across ethnic groups. In the paragraphs that follow, we discuss these results more fully.

Ethnic Differences in Anger, Inhibition, and Defensiveness
As predicted, trait anger was lower among U.S.-born African American and Caribbean women than among either U.S.-born European or Eastern European women (Consedine & Magai, 2002), and it was also lower among Haitians. It should be noted, however, that the tendency of ethnic minorities to report less negative emotion did not generalize to the Dominican group, suggesting that minority status per se does not necessarily predispose either greater or lesser emotionality than majority groups. In terms of defensiveness, although individuals in a precisely operationalized U.S.-born African American sample were no more defensive than U.S.-born European Americans, a finding that contradicts prior research among less differentiated samples (Marin et al., 1992), English Caribbean, Haitian, and Dominican women were all more defensive than the two European-descent samples. Exactly why this finding obtained remains unclear at this time, although the fact that it obtains for the three racial minority immigrant groups, at least some of whose emotion expressivity has likely been restrictively socialized (Gopaul-McNicol, 1999), implies that a combination of early predispositions and minority issues may be responsible.

However, our hypotheses regarding ethnic differences in the tendency to inhibit emotion expression were not supported. Prior research among African Americans (Brantley et al., 2002) and Jamaicans (Consedine & Magai, 2002) had led us to expect that U.S.-born African American and English Caribbean women would be more inhibited than either U.S.-born European Americans or Eastern Europeans (Consedine & Magai, 2002; Wierzbicka, 1999). However, Eastern European women were more inhibited than most other groups in the current study. One possible reason for this discrepancy relates to the exclusively female composition of our sample; it may be that the characterization of Russians as highly expressive (cf. Wierzbicka) is less germane to women, although data from Western samples have generally suggested that women are more expressive than men (Kring, 2000). Second, it may be that the result reflects a cohort effect that is particular to this sample. Born and raised during WWII and under the leadership of Joseph Stalin (d. 1953), who ushered in an era of restriction and fear, these women may have acquired more inhibited styles than is typical for their culture or is reflected in their language (cf. Wierzbicka). Finally, it may be that the difference between the current results and those of Consedine, Magai, Cohen, and colleagues (2002) reflects the fact that the individuals in the present sample are inhibiting emotion in their current lives more frequently than was implied by the childhood retrospective measure used in this earlier work. Indeed, older adults are generally thought to control expressions of negative emotion more frequently than younger or middle-aged participants (Carstensen, Gottman, & Levenson, 1995), perhaps because of the negative interpersonal consequences (Consedine, Magai, & Bonanno, 2002).

Relations Among Self-Reported Anger, Inhibition, Defensiveness, and Health
The current data are in keeping with prior research that has linked emotion and emotion regulatory styles to health and reports of health (Mayne, 2001; Sirois & Burg, 2003). Trait anger and emotion inhibition predicted poorer health whereas increased defensiveness predicted reports of better health, even when demographic variables, health behaviors, and ethnicity were controlled. The results with respect to anger are consistent with a large body of literature detailing anger's role in disease processes (Schum et al., 2003; Sirois & Burg), as are the data regarding emotion inhibition (Consedine, Magai, Cohen, et al., 2002; Pennebaker & Seagal, 1999).

Although our reliance on self-report measures and design do not enable causal interpretation, our data do suggest that some interpretations are more likely. It might be that inhibition and anger are related to health through their relation with health behaviors, although the fact that we controlled for smoking, drinking, and BMI suggests such an interpretation would be incomplete. Similarly, inhibition and anger might be linked to health through differences in reporting tendencies. Again, however, anger and inhibition were associated with poorer health even when defensiveness was controlled. More likely, in our opinion, is that trait anger and emotion inhibition are implicated in disease processes through a hyperactivated autonomic response (Gross, 1998; Mayne, 2001), and the consequent suppression of immune functioning (Kiecolt-Glaser et al., 2002; G. E. Miller, Cohen, & Herbert, 1999). Alternately, it may be that poor health causes negative emotions (Leventhal & Patrick-Miller, 2000) and emotion inhibition, although the trait nature of our measures suggests that to be unlikely (see Consedine, Magai, Cohen, et al., 2002).

Consistent with prior research, defensiveness was related to reports of better health (Brosschot & Janssen, 1998; Myers & Reynolds, 2000), although this effect varied across ethnic groups. Prior researchers have suggested that defensive individuals are positively self-deceptive in health reporting (Furnham et al., 2003), although the absence of an objective health measure means that this thesis cannot be directly assessed in our data. Further research should continue to disentangle these relations, use both subjective and objective measures of health, and implement longitudinal designs.

Ethnic Variation in the Relations Among Self-Reported Anger, Inhibition, Defensiveness, and Health
Perhaps the most important finding in the current study is the predicted demonstration that emotion variables do not relate to health equally across ethnic groups (Consedine, Magai, Cohen, et al., 2002). Adding interaction terms to the regression changed the magnitude of the main effects and, moreover, indicated variation in the relations themselves. Ethnicity interacted with all three emotion variables, with a total of 6 significant interactions and 4 marginally significant interactions out of a possible 15. Although our data give little direct explanatory guidance, norms regarding emotion experience, emotion regulation, agency, and typical eliciting situations vary cross-culturally (Mesquita & Walker, 2003), and, presumably, across ethnic groups. Therefore, the interaction effects we report here may reflect the possibility that potentially health-damaging emotions arise in different contexts for women of different groups, or that their experiences of emotions, emotion inhibition, and their outcomes are less arousing and conflicted (and thus health damaging) than those of majority U.S.-born European American group. We subsequently address some explanatory possibilities in greater detail.

Adding interaction terms strengthened anger's negative main effect on health and indicated that anger was inversely related to health among Haitians and Dominicans, despite the fact that Dominicans reported greater trait anger than all other women; U.S-born African American, English Caribbean, and Eastern European interaction terms revealed marginal interaction terms in the same direction. Although we cannot assume that the same mechanism is responsible for each of these interactions, one possibility is that more frequent experiences of anger are adaptive among urban, low SES, minority individuals perhaps because they enable women to more effectively address access barriers to health. Alternately, it may be that reports of low trait anger are indexing a repressive personality style that is deleterious to health. Finally, it seems likely that experiences of anger carry different meanings for women from diverse ethnic backgrounds. Specifically, we suspect that anger experiences are likely less negative for our Eastern European women, because of values involving emotional honesty and openness (Wierzbicka, 1999), and less negative for Spanish-descent cultures, such as Dominicans, for whom anger is a normative response to perceived honor insults (Rodriguez-Mosquera, Manstead, & Fischer, 2002). Data describing emotion values in the other ethnic groups are not yet available.

For emotion inhibition, the interaction terms reduced the main effect such that inhibition was only marginally associated with poorer health. As we predicted, however, there was an interaction between inhibition and being Eastern European such that inhibiting was less closely linked to health for these women, again, despite their reporting the highest level of inhibition. This finding is consistent with data describing the importance of considering others' emotions among Russians (Goldberg & Shmelev, 1993; Ispa, 1994). Our suspicion here is that values of this kind are acquired developmentally and require regular emotion inhibition, particularly among older individuals who avoid negative interactions (Carstensen et al., 1995; Gross et al., 1997). However, because the inhibition is consistent with the values of the individual, it may be experienced as normative and desirable and thus not incur the health costs that appear to accompany inhibition in other age or ethnic groups. More fully, it may be that ethnic groups from the West experience inhibiting emotion as inconsistent with their values surrounding individuality and freedom (Markus & Kitayama, 2003). Alternately, it may be that the absence of inhibition among Eastern European women is indexing a lack of self-control, something that research suggests Russian culture may find culturally inappropriate and maladaptive (Peabody et al., 1993).

Finally, defensiveness interacted with being U.S.-born African American, English Caribbean, and Haitian, such that although defensiveness was not associated with health reporting among U.S.-born European Americans (the absence of a main effect at Step 2), it was associated with reports of better health for these women. Although this pattern is made complex by the Dominican women (who have high defensiveness in the absence of a relation with health), the general trend is consistent with earlier work suggesting that defensive individuals report either better (Brosschot & Janssen, 1998) or positively biased appraisals of health (Myers & Reynolds, 2000). Exactly why defensiveness should be associated with reports of better health for some ethnic groups but not others remains unclear at this time, although it seems possible that health symptoms have different meanings for women from different groups and are differentially biased by defensiveness. Given the reliance that modern medicine places on patient reports of symptomatology, studies that replicate this finding, use more rigorous and multimodal operationalizations of defensiveness, and begin to understand what health reports mean to women from different ethnic groups are critically needed.

Conclusions
If anything, our data suggest that researchers need to be extremely careful in assuming that models relating psychological characteristics to health are readily transposed to new ethnic contexts. Although trait anger, emotion inhibition, and defensiveness were significantly linked to self-reported health, the strength of these relations varied across ethnic groups in every case. Although preliminary, our interpretation of these complex results suggests that researchers interested in how psychological variables affect health must consider the meaning of the psychological variables within their cultural contexts. Researchers have previously argued that the health consequences of variables such as emotion and emotion inhibition may vary depending on a number of other factors (Consedine, Magai, & Bonanno, 2002; Roloff & Ifert, 2000). In this contextualistic view, developmentally and culturally acquired variations in the values or norms associated with experiences, expression, and regulatory efforts influence how psychologically and physiologically demanding emotions and regulatory attempts are, and thus their health consequences. Culturally derived variations in meaning may comprise a critical intervening step in the pathway from basic biological processes to health outcomes.

	Acknowledgments

 
This research was supported by grants from the National Institute of General Medical Science (Grant 2SO6 GM54650), the National Institute on Aging (Grant KO7 AG00921) and the National Cancer Institute (Grant 1P20 CA 91372 & 1 U54 CA 101388) to C. Magai. We thank Yulia S. Krivoshekova for her assistance in understanding our counterintuitive finding regarding inhibition in the Eastern European sample and for sourcing and translating Russian-language publications.

	Footnotes

 
Decision Editor: Thomas M. Hess

Received for publication April 27, 2004. Accepted for publication January 10, 2005.

	References

TOP Abstract Methods Results Discussion References

 

• Baker, F. M. (2001). Diagnosing depression in African Americans. Community Mental Health Journal, 37,31-38.[Medline]
• Bardwell, W. A., Dimsdale, J. E. (2001). The impact of ethnicity and response bias on the self-report of negative affect. Journal of Applied Biobehavioral Research, 6,27-38.
• Baum, A., Pozluszny, D. M. (1999). Health psychology: Mapping biobehavioral contributions to health and illness. Annual Review of Psychology, 50,137-163.[Medline]
• Brantley, P. J., O'Hea, E. L., Jones, G., Mehan, D. J. (2002). The influence of income level and ethnicity on coping strategies. Journal of Psychopathology & Behavioral Assessment, 24,39-45.
• Brody, G. H., Flor, D. L. (1998). Maternal resources, parenting practices and child competence in rural, single-parent African American families. Child Development, 69,803-816.[Medline]
• Brosschot, J. F., Janssen, E. (1998). Continuous monitoring of affective-autonomic response dissociation in repressors during negative emotional stimulation. Personality & Individual Differences, 25,69-84.
• Carstensen, L. L., Gottman, J. M., Levenson, R. W. (1995). Emotional behavior in long-term marriage. Psychology & Aging, 10,140-149.
• Consedine, N. S., Magai, C. (2002). The uncharted waters of emotion: Ethnicity, trait emotion and emotion expression in older adults. Journal of Cross-Cultural Gerontology, 17,71-100.[Medline]
• Consedine, N. S., Magai, C., Bonanno, G. A. (2002). Moderators of the emotion inhibition-health relationship: A review and research agenda. Review of General Psychology, 6,204-228.
• Consedine, N. S., Magai, C., Cohen, C. I., Gillespie, M. (2002). Ethnic variation in the impact of negative affect and emotion inhibition on the health of older adults. Journal of Gerontology: Psychological Sciences, 57B,P396-P408.
• Consedine, N. S., Magai, C., Neugut, A. I. (2004). The contribution of emotional characteristics to breast cancer screening. Preventive Medicine, 38,64-77.[Medline]
• Cooper, C. L., Faragher, E. B. (1992). Coping strategies and breast disorders/cancer. Psychological Medicine, 22,447-455.[Medline]
• Crowne, D. P., Marlowe, D. (1960). A new scale of social desirability independent of psychopathology. Journal of Consulting Psychology, 24,349-354.[Medline]
• Deater-Deckard, K., Dodge, K. A., Bates, J. E., Pettit, G. S. (1996). Physical discipline among African American and European American mothers: Links to children's externalizing behavior. Developmental Psychology, 32,1065-1072.
• Deurenberg, P., Yap, M., van Staveren, W. A. (1998). Body mass index and percent body fat: A meta-analysis among different ethnic groups. International Journal of Obesity, 22,1164-1171.[Medline]
• Dilworth-Anderson, P. (1998). Emotional well-being in adult and later life among African Americans: A cultural and sociocultural perspective. In K. W. Schaie & M. P. Lawton (Eds.), Annual review of gerontology and geriatrics: Focus on emotion and adult development (Vol. 17, pp. 282–302). New York: Springer.
• Furnham, A., Petrides, K. V., Sisterson, G., Baluch, B. (2003). Repressive coping style and positive self-presentation. British Journal of Health Psychology, 8,223-249.[Medline]
• Goldberg, L., Shmelev, A. G. (1993). Cross-cultural study of lexicon of personality traits: "Big Five" of factors in English and Russian languages. Psikhologicheskii Zhurnal, 14,32-39.
• Golden, R. R., Teresi, J. A., Gurland, B. J. (1984). Development of indicator scales for the Comprehensive Assessment and Referral Evaluation (CARE). The Journal of Gerontology, 39,138-146.
• Goodwin, R., Nizharadze, G., Nguyen Luu, L. A., Emelyanova, T. (1999). Glasnost and the art of communication: A multi-level analysis of intimate disclosure across three former communist cultures. Journal of Cross-Cultural Psychology, 30,72-90.[Abstract/Free Full Text]
• Gopaul-McNicol, S. (1999). Ethnocultural perspectives on childrearing practices in the Caribbean. International Social Work, 42,79-86.[Abstract/Free Full Text]
• Gross, J. J. (1998). Antecedent- and response-focused emotion regulation: Divergent consequences for experience, expression, and physiology. Journal of Personality & Social Psychology, 74,224-237.[Medline]
• Gross, J. J., Carstensen, L. L., Pasupathi, M., Tsai, J., Goetestam-Skorpen, C. G., Hsu, A. Y. C. (1997). Emotion and aging: Experience, expression, and control. Psychology & Aging, 12,590-599.
• Harburg, E., Julius, M., Kaciroti, N., Gleiberman, L., Schork, M. A. (2003). Expressive/suppressive anger-coping responses, gender, and types of mortality: A 17-year follow-up (Tecumseh, Michigan, 1971–1988). Psychosomatic Medicine, 65,588-597.[Abstract/Free Full Text]
• Ispa, J. M. (1994). Child rearing ideas and feelings of Russian and American mothers and early childhood teachers: Some comparisons. In S. Reifel (Ed.), Advances in early education and day care. Vol. 6. Topics in early literacy, teacher preparation, and international perspectives on early care (pp. 235–257). Greenwich, CT: JAI Press.
• Ispa, J. M. (1995). Ideas about infant and toddler care among Russian child care teachers, mothers, and university students. Early Childhood Research Quarterly, 10,359-379.
• Jamner, L. D., Leigh, H. (1999). Repressive/defensive coping, endogenous opioids and health: How a life so perfect can make you sick. Psychiatry Research, 85,17-31.[Medline]
• Kiecolt-Glaser, J. K., McGuire, L., Robles, T. F., Glaser, R. (2002). Psychoneuroimmunology: Psychological influences on immune function and health. Journal of Consulting & Clinical Psychology, 70,537-547.[Medline]
• Kline, J. P., Bell, I., Schwartz, G., Hau, V., Davis, T. (1998). Repressive and defensive coping styles predict resting plasma endorphin levels in the elderly. Biological Psychology, 49,295-302.[Medline]
• Knupper, G., Fink, R., Clark, W. B., Goffman, A. S. (1963). Factors relating to amount of drinking in an urban community (California Drinking Practices Study Report No. 6). Berkeley, CA: State Department of Public Health, Division of Alcohol Rehabilitation.
• Krantz, D. S., McCeney, M. K. (2002). Effects of psychological and social factors on organic disease: A critical assessment of research on coronary heart disease. Annual Review of Psychology, 53,341-369.[Medline]
• Kring, A. M. (2000). Gender and anger. In A. H. Fischer (Ed.), Gender and emotion: Social psychological perspectives (pp. 211–231). New York: Cambridge University Press.
• Leventhal, H., Patrick-Miller, L. (2000). Emotions and physical illness: Causes and indicators of vulnerability. In M. Lewis & J. M. Haviland (Eds.), Handbook of emotions (2nd ed., pp. 523–537). New York: Guilford Press.
• Marganoff, P. P., Folwarski, J. (1996). Slavic families: An overview. In M. McGoldrick, J. Giordano, & J. K. Pearce (Eds.), Ethnicity and family therapy (2nd ed., pp. 658–672). New York: Guilford Press.
• Marin, G., Gamba, R. J., Marin, B. V. (1992). Extreme response style and acquiescence among Hispanics: The role of acculturation and education. Journal of Cross-Cultural Psychology, 23,498-509.[Abstract]
• Markus, H. R., Kitayama, S. (2003). Culture, self, and the reality of the social. Psychological Inquiry, 14,277-283.
• Mayne, T. J. (1999). Negative affect and health: The importance of being earnest. Cognition & Emotion, 13,601-635.
• Mayne, T. J. (2001). Emotions and health. In T. J. Mayne & G. A. Bonanno (Eds.), Emotions: Current issues and future directions (pp. 361–397). New York: Guilford Press.
• McGroder, S. M. (2000). Parenting among low-income African American single mothers with preschool-age children: Patterns, predictors and developmental correlates. Child Development, 71,752-771.[Medline]
• Mendolia, M. (2002). An index of self-regulation of emotion and the study of repression in social contexts that threaten or do not threaten self-concept. Emotion, 2,215-232.[Medline]
• Mente, A., Helmers, K. F. (1999). Defensive hostility and cardiovascular responses to stress in young men. Personality & Individual Differences, 27,683-694.
• Mesquita, B., Walker, R. (2003). Cultural differences in emotions: A context for interpreting emotional experiences. Behavior Research & Therapy, 41,777-793.
• Miller, G. E., Cohen, S., Herbert, T. B. (1999). Pathways linking major depression and immunity in ambulatory female patients. Psychosomatic Medicine, 61,850-860.[Abstract/Free Full Text]
• Miller, T. Q., Smith, T. W., Turner, C. W., Guijarro, M. L., Hallet, A. J. (1996). A meta-analytic review of research on hostility and physical health. Psychological Bulletin, 119,322-348.[Medline]
• Mora, P. A., Robitaille, C., Leventhal, H., Swigar, M., Leventhal, E. A. (2002). Trait negative affect relates to prior-week symptoms, but not to reports of illness episodes, illness symptoms, and care seeking among older persons. Psychosomatic Medicine, 64,436-449.[Abstract/Free Full Text]
• Myers, L. B., Reynolds, D. (2000). How optimistic are repressors? The relationship between repressive coping, controllability, self-esteem and comparative optimism for health-related events. Psychology & Health, 15,677-687.
• Niaura, R., Herbert, P. N., McMahon, N., Sommerville, L. (1992). Repressive coping and blood lipids in men and women. Psychosomatic Medicine, 54,698-706.[Abstract/Free Full Text]
• Payne, M. A. (1989). Use and abuse of corporal punishment: A Caribbean view. Child Abuse and Neglect, 13,389-401.[Medline]
• Peabody, D., Shmelev, A. G., Andreeva, M. K., Gramenitsky, A. E. (1993). Psikhosemanticheskiy analiz stereotipov russkogo haractera: Krosskul'turnyi aspect [Psychosemantic study of stereotypes of Russian character: Cross-cultural aspect]. Voprosy Psychologii, 3,101-109.
• Pennebaker, J. W., Kiecolt-Glaser, J. K., Glaser, R. (1988). Disclosure of traumas and immune function: Health implications for psychotherapy. Journal of Consulting & Clinical Psychology, 56,239-245.[Medline]
• Pennebaker, J. W., Seagal, J. D. (1999). Forming a story: The health benefits of narrative. Journal of Clinical Psychology, 55,1243-1254.[Medline]
• Pinderhughes, E. E., Dodge, K. A., Bates, J. E., Pettit, G. S., Zelli, A. (2000). Discipline responses: Influences of parents' socioeconomic status, ethnicity, beliefs about parenting, stress and cognitive-emotional processes. Journal of Family Psychology, 14,380-400.
• Rodriguez-Mosquera, P. M., Manstead, A. S. R., Fischer, A. H. (2002). The role of honor concerns in emotional reactions to offences. Cognition & Emotion, 16,143-163.
• Roloff, M. E., Ifert, D. E. (2000). Conflict management through avoidance: Withholding complaints, suppressing arguments, and declaring topics taboo. In S. Petronio (Ed.), Balancing the secrets of private disclosures (pp. 151–163). Mahwah, NJ: Erlbaum.
• Schum, J. L., Jorgensen, R. S., Verhaeghen, P., Sauro, M., Thibodeau, R. (2003). Trait anger, anger expression, and ambulatory blood pressure: A meta-analytic review. Journal of Behavioral Medicine, 26,395-415.[Medline]
• Sirois, B. C., Burg, M. M. (2003). Negative emotion and coronary heart disease. Behavioral Modification, 27,83-102.
• Smith, T. W. (1992). Hostility and health: Current status of a psychosomatic hypothesis. Health Psychology, 11,139-150.[Medline]
• Spielberger, L. (1986). Preliminary manual for the State-Trait Personality Inventory (STPI). Unpublished manuscript, Tampa, FL.
• Stevenson, H. C., Herrero-Taylor, T., Cameron, R., Davis, G. Y. (2002). "Mitigating instigation": Cultural phenomenological influences of anger and fighting among "big boned" and "baby faced" African American youth. Journal of Youth & Adolescence, 31,473-485.
• Suinn, R. M. (2001). The terrible twos—anger and anxiety. American Psychologist, 56,27-36.[Medline]
• Teresi, J. A., Golden, R. R., Gurland, B. J. (1984). Concurrent and predictive validity of indicator scales developed for the Comprehensive Assessment and Referral Evaluation interview schedule. Journal of Gerontology, 39,158-165.
• Teresi, J. A., Golden, R. R., Gurland, B. J., Wilder, D. E., Bennett, R. G. (1984). Construct validity of indicator scales developed from the Comprehensive Assessment and Referral Evaluation interview schedule. The Journal of Gerontology, 39,147-157.
• Weinberger, D. A., Schwartz, G. E., Davidson, R. J. (1979). Low anxious, high anxious, and repressive coping styles: Psychometric patterns and behavioral and physiological responses to stress. Journal of Abnormal Psychology, 88,369-380.[Medline]
• Werhun, C. D., Cox, B. J. (1999). Levels of anxiety sensitivity in relation to repressive and self-deceptive coping styles. Journal of Anxiety Disorders, 13,601-609.[Medline]
• Wierzbicka, A. (1999). Emotions across languages and cultures: Diversity and universals. Paris: Cambridge University Press.
• Williams, D., Ispa, J. M. (1999). A comparison of the child-rearing goals of Russian and U.S. university students. Journal of Cross-Cultural Psychology, 30,540-546.[Abstract/Free Full Text]

This Article Abstract Full Text (PDF) Services Download to citation manager PubMed PubMed Citation