Our data indicate that self-reported discrimination was associated with 12-month mental disorders and co-morbidity among AAs. We challenged the association between discrimination and mental disorders with several potential confounders, including other stressors, health conditions, and social desirability. Discrimination was a robust predictor of mental disorders when these factors and other sociodemographic characteristics were controlled.
One challenge came from other stressors as measured by low family cohesion, poverty, and acculturative strains. Most studies of discrimination do not consider other potential stressors, leaving open whether discrimination constitutes a unique stressor or simply captures these unmeasured stressors. Although family cohesion, poverty and acculturative stress were associated with mental disorders in unadjusted models, they were not significantly associated with disorders in fully adjusted models. Discrimination, however, remained an important correlate of mental disorders. We did not measure all possible stressors (e.g. job strain), and focused on specific types of stressors rather than more global inventories (e.g. life events). However, our findings indicate that discrimination is associated with mental disorders independent of poverty, family cohesion and acculturative stress.
Remarkably, even among immigrants, discrimination appeared to be a more important predictor of mental disorders than acculturative stress or years in the United States. It is often reported that immigrants’ mental health worsens due to the stressors associated with adjustment in a new culture (Vega, Sribney, Aguilar-Gaxiola, & Kolody, 2004
). Although acculturative stress was associated with mental disorders, acculturative stress was no longer significant once we included discrimination. Finch and colleagues (2001)
also found that acculturative stress (measured by language conflict and legal status) was not associated with chronic health conditions among Mexican Americans after including discrimination. In contrast, Noh and Kasper (2003) reported that acculturative stress was associated with depression among Koreans in Toronto, but that this association varied by level of discrimination and type of coping response. Gee, Ryan, Holt, and Laflamme (2006)
reported that the association between discrimination and mental health strengthened with increasing years in the U.S. among Latino and Black immigrants. However, the current data did not find a main effect for years in the U.S. or an interaction between years and discrimination. Further research is needed to clarify the potential relationships between discrimination and acculturation. That said, our findings and the literature indicate that discrimination is associated with health independent of acculturation. Hence, our findings do not invalidate the claim that adjustment is a stressful process, but do buttress arguments that an important issue in cultural adjustment involves dealing with discrimination (Berry, 2003
; Finch, Hummer, Kolody & Vega, 2001
The second challenge involved physical health conditions. Physical health attenuated, but did not eliminate the association between discrimination and mental disorders. This finding is consistent with arguments that discrimination, mental disorders and physical conditions lie on a common causal pathway. That is, discrimination may directly cause mental and physical disorders and these disorders may cause one another. For example, discrimination may lead to heart disease which leads to anxiety. Thus, part of the relationship between discrimination and mental health may be mediated by physical health conditions (conversely, physical health may be mediated by mental health). We are unable to establish these pathways with cross-sectional data, but these ideas should be investigated in future research.
The third challenge involved social desirability, a reporting tendency associated with the seeking of approval or the avoidance of disapproval. These ideas resonate with “loss of face,” the avoidance of shaming oneself or one’s family, which may be especially important for AAs (Gong, et al., 2003
; Zane & Yeh, 2002
). In our study, social desirability did not influence the relationship between discrimination and mental disorders. This finding is important because response factors are always a threat to the validity of self-reported data. Other studies have found that the association between self-reported ethnic harassment and mental health is not confounded by affective disposition (tendency to complain about innocuous events) (Schneider et al., 2000
). Although this gives us more confidence, future work should examine other potential response (e.g. memory) and personality factors (e.g. optimism). We discuss some other response factors below.
The associations between discrimination and health were stronger for depressive disorders than for anxiety disorders, suggesting that the potential effects of discrimination vary by outcome. Prior studies have also found associations between discrimination and depression and anxiety among Asians in the U.S. and internationally (Bhui, et al., 2005
; Gee, 2002
; Karlsen & Nazroo, 2002
; Noh & Kaspar, 2003
). Although AAs were not examined, Kessler and colleagues (1999)
found that everyday discrimination (using a measure similar to ours) was associated with a 2.1 greater odds of depression and 3.3 greater odds of generalized anxiety disorder among the U.S. general population. It would have been informative to disaggregate our disorders into more specific categories, but small numbers of cases precluded disaggregation. Further, health associations likely vary by the type of discrimination measured. For example, the potential effects of structural discrimination may differ from those of interpersonal discrimination (Gee, 2002
). Our study only considers interpersonal discrimination, therefore it is possible that the exclusion of other types of discrimination (e.g. structural, internalized oppression) may underestimate the full association of discrimination with mental disorders.
The average levels of discrimination reported by AAs nationwide (1.8 on a 1–6 point scale) are relatively close to those reported by African Americans in Detroit (2.3, using the same scale) (Williams et al., 1997
). This suggests that discrimination may be a shared experience that may help foster bridges between diverse groups. Further, discrimination may have been underreported for several reasons. First, the everyday discrimination scale was designed for African Americans and does not measure several phenomena that may be especially relevant for Asians, such as being a “perpetual foreigner” or stereotyped as passive (Liang et al., 2004
; Young & Takeuchi, 1998
). Second, potential undesirable consequences of reporting, including invalidation by others, may cause individuals to reappraise a discriminatory incident as non-discriminatory to avoid such challenges (Harrell, 2000
; Kuo, 1995
). Third, experiences of discrimination may be simply forgotten or unrecognized. Underreporting of discrimination would tend to bias our findings towards the null.
Discrimination may also be overreported in some circumstances, as when individuals encounter paranoia. This raises questions about the causal direction between discrimination and mental disorders. That is, although we presume that discrimination causes mental disorders, it is possible that mental disorders cause one to experience and/or perceive discrimination. We cannot establish these directions given cross-sectional data. However, several prospective studies have reported that discrimination predicts illness (Jackson, et al., 1996
; Schulz, Gravlee, Williams, Israel, Mentz, & Rowe, 2006
). Pavalko, Mossakowski, and Hamilton (2003)
found that self-reported discrimination predicted illness, but that illness did not predict reports of discrimination over a period of seven years. That said, more longitudinal research is needed.
Before concluding, a few additional caveats should be mentioned. First, the NLAAS oversampled Chinese, Filipinos and Vietnamese, but we did not disaggregate our analysis by ethnic subgroup because cell sizes became sparse. In NLAAS, Filipinos reported the highest levels of discrimination (mean=1.9) and Vietnamese the lowest (mean=1.5), suggesting that future studies should investigate possibly higher risk among Filipinos. Further, our sample includes some Pacific Islanders, but their experiences are largely not represented in our data. Hence, our analyses apply to AAs in the aggregate, but future research should investigate Pacific Islanders and subgroups within Asians and Pacific Islanders. Second, we focus on psychiatric disorders as defined by the DSM-IV. While this allows for comparability with other studies, DSM-IV criteria may be biased towards western ways of expressing mental health problems and may underestimate rates of mental disorders (Takeuchi, Chun, Gong, & Shen, 2002
Despite the caveats, our study has several strengths. To our knowledge, this is the first nationally representative study of discrimination and mental disorders among AAs. Additionally, we employ standard measures that allow for comparability with other studies. Finally, we control for a number of important factors, including social desirability.
In conclusion, our study finds that self-reported discrimination is associated with greater chance of having a mental disorder and multiple mental disorders within the past year among Asian Americans. It is important to note that the analyses presented here represent a conservative test of the association between self-reported discrimination and mental disorders. That perceived discrimination still has an association with mental disorders given the other variables is impressive. There are several avenues for future research, including an assessment of the factors that may buffer against the effects of discrimination, ethnic subgroup differences, and a more detailed examination of the specific disorders. Moreover, there is a critical need for longitudinal studies. It is still premature to conclude that discrimination causes mental disorders, but these caveats do not preclude the continuation and development of civil rights legislation and multicultural training. Indeed, if discrimination is found to be a causal risk factor for mental disorders, as our results imply, then policies designed to promote civil rights may not only buttress the foundations of a civil society, but also a healthy one.