Treating Minority Patients With Depression and Anxiety: What Does the Evidence Tell Us?

In this article, we examine the current state of knowledge regarding treating racial and ethnic minority patients with mood and anxiety disorders. Past literature has amply documented significant disparities in access to, and quality of, mental health care received by many ethnic/racial minorities in the United States. An emerging literature examines outcomes of care for ethnic/racial minorities. In this article, we briefly review the evidence on poorer access and quality of care and provide a rationale for examining potentially different outcomes of care for ethnic/racial minorities, prior to presenting new evidence about response to and outcomes of care for ethnic/racial minority patients.

The term “minority” encompasses the overlapping, evolving and still elusive terms of “race,” “ethnicity” and “culture.” A simple approach to these concepts is nicely outlined in the minority supplement to the Surgeon General's Report on Mental Health (1). The term “race” does not denote a biological category, but rather a social construct. This construct serves to group people together according to characteristics that are socially significant (e.g., visible physical traits such as skin color) and is often the basis for discrimination. The term “ethnicity” is used to group people together on the basis of a shared heritage (e.g., history, language, rituals, food preferences). Race and ethnicity are only partially overlapping concepts because individuals from the same race may have quite different heritages (e.g., Caribbean and American blacks; Native Hawaiians and Vietnamese Americans). Finally, the term “culture” denotes a shared set of beliefs, norms or values that will influence the meaning given to life events and experiences. People from the same ethnic or racial group may well have significant cultural differences, quite evident in the differences among Mexican American and Cuban American Latino groups, which contribute to the controversy about for whom to use the term “Latino” vs. the term “Hispanic” (2). (For simplicity, we have adopted whichever term was used by the respective cited study.) Cultural differences are common among Caucasian groups as well. These differences may be a product of ethnicity (e.g., Italian Americans), religion (e.g., Catholicism, Judaism) or geographic region (e.g., Southern United States).

In this article, we attempt to identify studies in which various ethnic/racial minority groups have been contrasted with Caucasian groups that are presumed to represent the dominant “Western” culture of Euro-America with its white Judeo Christian history and cultural landscape. It must be remembered, however, that all of these constructs are inherently imprecise, that various identifications are not always fixed or easily determined, that concepts of race/ethnicity should not be used as a proxy for genetic variation and that interpretation of any differences needs to consider the role of a host of conceptually relevant factors such as discrimination, social class, SES, environmental exposures, educational level, spoken language, religion, country of origin and time in country of residence (3).

The recent report of the Surgeon General on mental health documented significant disparities in access to and the quality of mental health care received by members of racial and ethnic minority groups in the United States (4). Ethnic/racial minorities lack access in part because they are more likely to be uninsured than are Caucasians, and Caucasian service providers are proportionately more available than are ethnic/racial minority service providers (1). Not surprisingly, members of African American and Hispanic American minority groups also are less likely to obtain treatment for either depression or anxiety than are their Caucasian counterparts (5). Among those who do receive care, two studies with nationally representative samples find that African American and Latino minorities are less likely to receive quality care than are Caucasians (6, 7). Even among insured population of U.S. federal employees, Caucasians are 1.7 times as likely to visit an outpatient mental health provider, and make 2.64 more mental health visits, per year, compared to both African Americans and Hispanic Americans (8). Unfortunately, most large epidemiological studies of access and care have not included sufficient samples of Asian American/Pacific Islanders or Native Americans to provide comparative rates of care and quality of care. However, data from the Chinese American Psychiatric Epidemiological Study conducted in 1993 and 1994 indicate low rates of insurance among Chinese Americans living in Los Angeles County, with those living in areas with the highest proportion of Chinese Americans having the lowest likelihood of having medical insurance (9). An earlier study, which sampled patients based on the first wave of the Epidemiologic Catchment Area study, showed that Asian Americans had a lower rate of utilization of mental health services compared to Caucasians (10), and a summary of older studies suggests that, among Asian Americans who use services, severity of illness is high, suggesting a delay in seeking treatment (1). The geographic, linguistic, cultural and economic heterogeneity of Asian American/Pacific Islander groups, including the broad range of acculturation that makes some subgroups less disadvantaged than others, makes it inadvisable to characterize them as a single entity and has limited their inclusion in some larger surveys (11). As the clustering of this diverse group has, until recently, been the norm, the interpretation of data collected from “Asian American” subjects must be done with the understanding that the results may be difficult to generalize to a specific ethnic group (12). Although data on native American populations are sparse, only 20% utilize the Indian Health Service that is federally mandated to provide care for members of recognized tribes, only half have employer-based insurance (in contrast to 72% for Caucasians) and a quarter have no health insurance (13). Thus, it appears that, overall, ethnic/racial minorities seem to have poorer access to care. Even when access appears to be equalized by insurance status, barriers other than cost, such as systems issues, language, stigma, cultural beliefs about illness and treatment, and personal experience can contribute to continued poor access, and for those who do follow through and make a visit, poorer quality of care.

The minority-focused supplement to the Surgeon General's Report on Mental Health (1) reviewed in detail the limited science base on racial and ethnic minority mental health, including data on treatment. Analyses of participants in psychiatric clinical trials conducted between 1986 and 1994 documented a surprising and shocking absence of ethnic/racial minority participants. No ethno-specific analyses were reported for 25 trials involving 2813 schizophrenic patients, for 16 trials involving 921 bipolar patients and for 27 trials involving 3860 depressed patients. Adequate samples of clearly identified ethnic/racial minorities are missing from most of these trials. For example, only 7% of those participating in depression trials were clearly identified as an ethnic or racial minority. Although this situation has recently improved under impetus from the NIMH, there still remains a dearth of evidence-based information about the treatment of depression and anxiety in ethnic/racial minority patients, and reviews of this area have been more anecdotal, impressionistic or, at best, observational and descriptive, rather than evidence based.

Now that the rationale for examining this area is clear, this article will review the biological and cultural bases for possible differences in treatment participation, response and outcome, and the available studies that examine either alone or, in comparison with Caucasian populations, how members of racial and ethnic minority groups with depression and anxiety respond to evidence-based treatments. The review concludes with some practical approaches to improve treatment and care for this underserved population that might be utilized by clinicians in their day-to-day practice.

Numerous factors affect the way an individual will respond to medication, including biological factors, cultural factors, environmental factors and factors related to the medication itself (formulation, dose and route of administration) (14). Medication effects are determined in large part by pharmacokinetics and pharmacodynamics. These processes are, to a degree, genetically determined, because genes control the expression and function of enzymes responsible for drug metabolism, as well as proteins that modulate the action of neurotransmitter drug targets (e.g., transporters, receptors) (15). Because the distribution of genetic polymorphisms varies across some ethnic and racial groups, this distribution will create pharmacokinetic and pharmacodynamic differences that could result in differential medication sensitivity or efficacy for different racial and ethnic groups (15). These polymorphisms can also interact with lifestyle variations (e.g., diet and use of alternative medicine) to produce gene-environment interactions and effects that would not occur on a genetic basis alone (15).

The drug metabolism polymorphisms of major interest involve alleles that code for essentially inactive cytochrome P-450 proteins, with the probability of an allele in a population determining the frequency of intermediate (heterozygous) and poor (homozygous) metabolizers. Rates of poor metabolizers are largest for Asian, rather than Latino or African American populations, although there are also differences between European and other ethnic and racial groups, and it is likely there are many unknown and undiscovered polymorphisms that could be expressed differently in these various groups. Cytochrome P-450 2D6, which metabolizes several antidepressants, is less active in East Asians, which poses the risk for higher blood levels and more side effects within this population. In contrast, some individuals may have multiple copies of functional alleles, causing them to be “ultra-rapid metabolizers.” Such individuals appear to be overrepresented in some Arab and Ethiopian populations and would be less likely to attain the adequate blood levels necessary for optimal clinical response (15, 16). Cytochrome P-450 2C19, which is involved in the metabolism of some benzodiazepines, can also have different rates of activity according to population. Individuals of East Asian decent, for example, are more likely to be “poor metabolizers” of drugs that are substrates of cytochrome P-450 2C19, and are, thus, more susceptible to higher blood levels compared to other populations with the same dosages.

In addition to genetic factors, certain cultural practices that impact biology can interact with drug metabolism. For example, CYP1A2, which metabolizes several antipsychotic and antidepressant drugs, is “highly inducible,” meaning that its activity can be increased by multiple external stimuli that may vary according to cultural practices. In particular, diet, caffeine and tobacco use, and hydrocarbon cooking byproducts all impact the activity of 1A2 (15). As another example, the metabolic action of enzyme CYP3A4, which is involved in the metabolism of more medications than any other enzyme, including many psychotropics such as benzodiazepines and some antidepressants, can be readily blocked by grapefruit juice but increased by St. John's Wort (15, 17). Furthermore, because of the paucity of pharmacokinetic studies of herbal remedies, it is likely that other herbal remedies have either enhancing or dampening effects on drug metabolism, which have not yet been identified. Because use of herbal products varies greatly across cultures (18), this is another potential source of differences in pharmacokinetics among ethnic/racial minority populations. Finally, changes in diet can have a powerful pharmacokinetic effect (e.g., the metabolism of African people living in their native land, eating their traditional foods, is known to change significantly when they immigrate to Europe and change their diet) (19). Therefore, recent immigration can be an important consideration in drug metabolism.

Genetic variation also can have a major effect on the pharmacodynamics of medications (their intrinsic effects at sites of action) by impacting the encoding of effector proteins, which transduce the effects of molecules after they bind to receptor sites in the brain. Many antidepressants act through the major monoamine systems, especially the serotonergic system. One allele of emerging interest in pharmacogenomic studies is the promoter region of the gene coding for the serotonin transporter, which mediates the transport and reuptake of serotonin. The prevalence of the “long” allele for this promoter region (which increases the transcription and activity of the transporter resulting in more rapid and complete reuptake of serotonin back into the nerve terminal) is greatest among people with African lineage (70%), whereas Caucasians have a rate of roughly 50% and East Asians only 17% (20).

In addition, evidence suggests that there are racial and ethnic differences not just in the distribution of this long allele, but also in its behavioral consequences. For example, two studies conducted with Caucasian populations found a positive correlation between having the long allele and response to antidepressants, whereas the opposite was found in a study of Asian participants. In this study, response to antidepressants was greatest among subjects whose genotypes were homozygous for the short allele (21–23). A more recent study examined potential intervening mechanisms between genetic factors and behavioral responses to medications, comparing Caucasian and African American groups. In this study, being homozygous for the short allele resulted in opposite effects in African Americans than in Caucasians. African Americans homozygous for the short allele had higher levels of the metabolite 5-HIAA in their CSF (compared to African Americans with at least one long allele), whereas Caucasians of the same genotype had significantly lower levels of 5-HIAA (compared to Caucasians with at least one copy of the long allele) (24).

These studies suggest that biological bases of potential differences in responses to pharmacotherapy between ethnic/racial minority populations are complex. Although there is some evidence of differences in specific genes known to affect medication effectiveness, it also appears that the same allele may function differently in different racial or ethnic groups, both at a behavioral and at a more “upstream” biological level, perhaps as a consequence of differences in additional linked genes that produce counteracting effects. Future studies would benefit from taking these various influences into account.

Catechol-O-methyltransferase (COMT) activity, important for the metabolism of dopamine, is also influenced by a single nucleotide mutation that results in lower enzymatic activity. This mutation has been shown to vary differentially across ethnicities (26% in African Americans, 18% in Asians and 50% in Caucasians). Recent work indicating that this genotype may be related to risk factors for the development of schizophrenia suggests that it has great potential clinical importance (25). Activity of COMT is also related to the likelihood of experiencing side-effects from levodopa in the treatment of parkinsonism, and the different ethnic distributions of this mutation could help to explain the higher percentage of Asians that respond poorly to levodopa (26).

In considering the small but growing literature, suggesting that African Americans and Asian Americans may, on average, show proportional differences in certain polymorphisms affecting both drug metabolism and cellular proteins that modulate receptor effects, it should be emphasized that this research looks for variations in rates of these differences across populations, not in absolute population differences. In any individual, for example, genetic variation in the metabolism of medications cannot be deduced from racial or ethnic group membership alone as all rates of metabolism can be found in each of the major “racial” groupings that have been examined.

Early studies of African American treatment response tended to show mixed results. Preliminary studies found that African Americans and Caucasians respond similarly to both medication and behavioral treatment for PTSD (63, 64), whereas African Americans were found less responsive than Caucasians in a pilot study of behavioral treatment for agoraphobia (65). In another study of the impact of psychotherapy or antidepressant medications given to primary care medical patients, African Americans proved similar to Caucasians in symptom resolution, but African Americans showed less improvement than Caucasians in functioning (55). A 1998 study of depressed HIV patients found that African American patients were less likely to respond to treatment with fluoxetine than Caucasians (66). In other studies, African American and Caucasian children and adults (67, 68) responded equally well to anxiety interventions.

Three early studies examined treatment outcome of Latinos with depression, but without including comparison groups. Care for depression was given to unmarried Puerto Rican mothers with depressive symptoms (69), to Mexican American women (70) and to Puerto Rican adolescents (71). All found that those who were treated had favorable results. Three recent studies have also examined outcome for anxiety disorders. A 2003 study of exposure-based CBT for phobic and anxiety disorders was effective, and similarly so, for both Hispanic/Latino and European American youths (aged 6–16 years) (72). A group of pioneering researchers in the Boston, MA, area has published exciting work demonstrating the efficacy of culturally modified cognitive-behavioral therapy for two different groups of refugees from Southeast Asia. One of these studies, a pilot project by Otto et al (73, 74), found that treatment-refractory Cambodian refugees with PTSD did better after receiving a combined treatment of sertraline and CBT vs. sertraline alone. The second study, a pilot project by Hinton et al (75), found that a culturally adapted CBT regimen significantly improved symptoms and reduced the severity of panic attacks experienced by treatment-resistant Vietnamese refugees suffering from PTSD and experiencing panic attacks.

A study of multiethnic, economically disadvantaged, English-speaking and Spanish-speaking patients with major depressive disorder found no difference in response to CBT across ethnic groups (76). Case management augmentation to the therapy was effective in engaging and keeping patients in care across all ethnic groups. However, augmenting cognitive-behavioral therapy with supplemental case management was shown to improve response for Spanish-speaking depressed patients, but not for English-speaking patients. In post hoc analyses, the supplemental case management was found to be less effective for African Americans.

Recently, larger studies have examined response to care for ethnic/racial minority patients. Minority status predicted poor paroxetine response in a study of primary care panic patients. Although roughly 30% of subjects were “non-White” Americans, non-Whites comprised 45% of the nonresponders and only 22% of the responders (77). However, lower income was an even stronger predictor in the regression, suggesting that, in this sample, which included many poor Caucasian patients, poverty was a more powerful determinant than minority status.

Despite the finding in this study of poor response of impoverished patients, two recent studies of depression have found interventions effective for poor women. Female primary care depression patients seen in public sector clinics in Santiago, Chile, were found to respond significantly better to a multifaceted, structured, “stepped-care” model than to treatment-as-usual (78). This 3-month intervention included a psychoeducational group led by a nonmedical health worker, structured and systematic follow-up, and drug treatment for patients with severe depression. In this study of impoverished women, 70% of the intervention women were recovered as compared with 30% of the usual care group. A study comparing medication intervention, psychotherapy intervention and usual care for low-income, young ethnic/racial minority women in the United States found that both interventions reduced depression ratings and improved social functioning significantly more than the usual care. Medication also improved instrumental role functioning, although psychotherapy did not (79).

Ethnic/racial minority patients have also been found to respond well when the impact of quality improvement interventions for depression is evaluated in primary health care settings. In a recent study of 46 primary care practices in 6 U.S. managed care organizations, the impact of two QI programs that trained local experts to educate clinicians, nurses to educate and follow-up with patients and psychotherapists to conduct CBT were compared with care as usual. The interventions significantly improved appropriate care for depression similarly within each ethnic group. However, the interventions significantly decreased the likelihood that Latinos and African Americans would report probable depression at 6 and 12 months of follow-up, whereas the Caucasian intervention sample did not differ from controls in reported probable depression (80).

Two recent analyses further document the comparable response to treatment of ethnic/racial minority patients with depression and anxiety. A pooled analysis of the large paroxetine clinical trials database showed that response and remission rates for patients with both depression and various anxiety disorders were similar in Latino, African American and Asian American, compared with Caucasian patients (81). Interestingly, this study replicated a previous finding (57) of a higher placebo response rate in Latino patients. Finally, a secondary analysis of the large primary care IMPACT study of geriatric depression treatment, which provided stepped care with both medications and problem solving therapy, documented comparable response in minority and majority patients (82).

From our review of the literature, two consistent findings appear to be emerging. First, ethnic/racial minorities are, with some exceptions (e.g., highly acculturated citizens of Indian subcontinental heritage), significantly less likely to obtain care than their nonminority counterparts. Access to care appears limited by insurance and available providers, but even insured ethnic/racial minorities are less likely to be receiving adequate care than are Caucasians, reinforcing the importance of multiple other factors reviewed above. Focused efforts to increase the amount and quality of mental health care available to people in the service sector where they are most likely to be encountered, in this case, primary care, may be more efficient than broader efforts to restructure the delivery of specialty focused care, though this is also important. Receiving care in the primary care setting may work to reduce the stigma of mental health care, although any medical care system may not synergize well with certain cultural beliefs about illness and treatment (41).

Despite problems in access to care, an emerging literature suggests that evidence-based interventions are likely to be effective among ethnic/racial minorities. Recent studies suggest that such care is effective even among economically disadvantaged ethnic/racial minorities. It should be noted, however, that many of these studies were designed to enhance delivery of treatment by overcoming barriers to care and did target barriers specific to the ethno-racial patients in their studies as part of their treatment delivery strategy. Thus, facilitating treatment session attendance by assisting patients with childcare, transportation and other remedies to overcome logistical barriers seems to be a beneficial strategy that clinics might pursue.

Discrepancies between the type, quality and effectiveness of treatments provided to ethnic/racial minority populations and those provided to Caucasian populations are prevalent. As there are numerous apparent causes for these discrepancies, there are also a variety of means of improving the treatment of anxiety and depression among ethnic/racial minority populations. In order to provide adequate and effective care to ethnic/racial minorities, it is imperative to understand and appreciate past experiences and current beliefs, and how these have in turn shaped attitudes, preferences and views of the “acceptability” of treatment. Beyond simply understanding a patient's opinions, it is always advisable to try to accommodate a patient's wishes. For instance, if possible, matching treatment modality with the preferences of the patient is likely to increase adherence, as is referring a patient to someone of their own ethnicity, race or background, or to an ethnicity-specific treatment program, although the latter is not always feasible given the paucity of available minority clinicians and specific programs (83). Given this fact, it is vital for practitioners to recognize “characteristic” cultures of various ethnic/racial minority groups and how these cultural influences contribute to these factors. One must be cautious, however, that, while observing and recognizing common cultural persuasions, one does not stereotype members of ethnic/racial minority groups. It is this kind of stereotyping that may, in situations where cognitive uncertainty is present, contribute to unconsciously biased clinician decision making (31).

Evidence-based procedures for increasing the cultural acceptance of specific treatment modalities are scarce. Suggestion has been made that assessing and directly addressing treatment concerns held by patients are a way to enhance treatment among ethnic/racial minority patients (49).

Psychotherapeutically, Otto et al (73, 74) and Hinton et al (75) demonstrated that CBT adapted for the specific illness, treatment beliefs and world views of Southeast Asian cultures has beneficial effects on treatment adherence and outcome in Cambodian and Vietnamese refugees with PTSD. It is likely that the overall strategies used by these authors to tailor CBT could be adapted to other cultures and populations.

Psychopharmacologically, clinicians may benefit by increasing the amount and level of communication with a patient that revolves around medication treatment. Clinicians must be cognizant of the amount of information they convey to mainstream patients and try to provide this information uniformly across patient populations. Further, it may be of value for clinicians to find out if a patient possesses less negative feelings toward herbal remedies vs. mainstream psychiatric medications. If so, this knowledge can be used by the clinician to assuage the concerns the patient has about modern psychiatric medications. As it has also been shown that the views of a patient's caregiver affects adherence to antidepressant medication, clinicians should involve a patient's spouse or other caregiver in treatment decision making, particularly among those ethnic/racial minority populations that are especially sociocentric (84).

Biologically, pharmacokinetics and pharmacodynamics must be considered by the prescribing clinician. Herbal medication, or other traditional healing practices, is often used disproportionately among certain ethnic/racial minority populations (85). Thus, must be factored in when prescribing and evaluating response, as it can impact drug metabolism. Due to these biological mechanisms, medication and dose may need to be adjusted to attain blood levels necessary for therapeutic response. Given the current state of the literature, this is particularly imperative for Asian American populations, although limited study suggests that clinicians should consider potential differences in other ethnic/racial minority groups as well. In general, careful monitoring of medication should include slow titration, enhanced sensitivity to the possibility of side effects and drug interactions, and frequent visits or telephone checks.

Although the available studies in this area are scarce, and it is paramount that more are initiated, the available evidence demonstrates fairly comparable efficacy between ethnic/racial minority populations and Caucasian populations with both psychopharmacological and psychotherapeutic treatments. Evidence seems to suggest that providing “enhanced” interventions also improves patient response. It has also been shown that case management strategies are possibly an effective tool to overcome various barriers to treatment that members of ethnic/racial minority groups often face. Of particular importance is understanding that dropout from treatment and suboptimal adherence are continual risks, and addressing these concerns must be an ongoing and evolving process.