Mental Illness in Hispanics: A Review of the Literature

Compared with the overall population, Hispanics are younger, poorer, and less educated; they are more likely to be foreign born and to live with family; and they are less likely to speak English and to have health insurance. Their employment rate is close to the average U.S. rate (Table 1). Hispanics make up over 80% of the undocumented U.S. population, with an estimated 8.4 million in 2004 (5).

Despite a common ethnic identification, Hispanics are heterogeneous in some aspects that may influence the incidence, presentation, course, and treatment of mental illness, such as birthplace and acculturation, genetics and race, health care access and utilization, and language. For example, a recently published study found differences in alcohol, marijuana, and cocaine use among Hispanic students according to ethnic subgroup, language first spoken, parental education, urbanicity, and region (7).

The question of whether this heterogeneity detracts from the validity of studies on Hispanics does not have a simple answer. In many cases the report specifies which Hispanic population or subgroup was studied, or this information can be inferred from the location of the study. Other studies address subjects that traverse all or several of the Hispanic subgroups, such as monolingualism versus bilingualism, generational differences, or disadvantaged socioeconomic status. We simply want to emphasize that the generalizability of study results must be judged on a case-by-case basis.

Hispanics are more likely to be uninsured, and even when insured, they have less access to medical care than other Americans. For example, in 2002, 35% of Hispanics reported being uninsured (vs. 14% of whites and 21% of African Americans) (4). Worse, the number of insured Hispanics, especially privately insured, seems to be diminishing faster than in other groups (Table 3).

In the 2004 National Healthcare Disparities Report, Hispanics scored lower than non-Hispanic whites on about 90% of access measures (25). The recent National Comorbidity Survey Replication showed that, for Hispanics, the 12-month odds ratio of receiving treatment for mental disorders was 0.6 (95% CI, 0.5–0.8; p<0.05) compared with non-Hispanic whites (26). Even insured Hispanics seem to use services less. A recent study at the Department of Veterans Affairs (VA) found that the odds ratio of having a health visit in the previous year was 0.4 for Hispanics, compared with 2.2 for non-Hispanic whites (27). Besides this disadvantaged position for Hispanics as a whole, significant differences may exist in health care access among Hispanic subgroups, with a negative effect on Mexicans and Central Americans in particular (Table 4). Undocumented and/or immigrant Hispanics also have less access and use services less (28, 29).

Because Hispanics are younger overall than other ethnic groups, any issue affecting them has a disproportionate impact on the young. More than a quarter (25.7%) of Hispanic children lack health insurance, with Mexican children faring the worst (30.4%), followed by Central and South Americans (23.8%), other Hispanics (18.6%), Puerto Ricans (11.4%), and Cubans (9.3%), compared with 7.8% for non-Hispanic whites (31).

Different types of studies have shown that the quality of health care received by Hispanics is below average. Depressed Hispanics seem to be less likely than Caucasians to receive a prescription for an antidepressant, to actually take the medication, and to receive specialty care (32–35). Some studies suggest that, compared with Caucasians, Hispanics with psychosis are less likely to receive medication or to receive an atypical antipsychotic, more likely to receive a depot antipsychotic, and less likely to have appropriate trials of antipsychotic medications (36–39). The 2004 National Healthcare Disparities Report (25) indicated that Hispanics received a poorer quality of care than non-Hispanic whites on about 40% of quality measures. The reasons for this disparity seem to be complex. For example, Hispanic Medicare beneficiaries have an odds ratio of 0.36 for receiving hip replacement surgery compared with non-Hispanics (40). Given that the procedure is fully covered by Medicare, this finding suggests that underutilization of health services by Hispanics cannot be attributed solely to lack of health insurance. In a recent study of an insured adult population at the VA, researchers found a trend toward greater inequality in the delivery of outpatient mental health care for Hispanics (41). These racial/ethnic differences in the treatment of mental illness have persisted under Medicare managed care (42).

The underutilization of mental health care by Hispanics, especially but not exclusively by Mexican Americans, has been noted for decades (43–46). This underutilization was recently demonstrated for Hispanics with schizophrenia or schizoaffective disorder who were receiving services in the public sector in San Diego: 19% of Latinos received case management services, compared with 30% of Caucasians and 17% of African Americans (47). Ethnic differences in utilization are not a simple issue. As shown in a study using data from the National Comorbidity Survey (NCS), to understand ethnic or racial disparities in specialty care, the effects of ethnicity or race should be analyzed in combination with variables related to socioeconomic status and environmental context (48).

Hispanics are significantly more likely to receive attention for mental disorders in primary care settings than in specialized settings, which is probably related to their underutilization of mental health care. Mexican Americans with mental disorders were found to have a 12-month utilization rate of 18.4% for general care providers and 8.8% for mental health specialists (49).

Among factors that complicate the issue of reduced access to mental health care for Hispanics, two are worth mentioning specifically: 1) knowing where to find a provider increases significantly the likelihood of using specialty mental health care (50); and 2) Latinos are more likely than Caucasians to perceive the health care system as unfair because of race or ethnicity (4, 51).

Among Hispanics, the risk of health factors that can complicate mental disorders or their treatment, such as obesity, diabetes mellitus, and lack of exercise, is growing rapidly. For some risky behaviors, such as use of alcohol, tobacco, and illegal drugs, Hispanics seem to be at lower risk than Caucasians (Table 5). Yet, again, differences exist in alcohol-related problems among Hispanic subgroups, with Mexican Americans demonstrating more problems than Cuban or Puerto Rican Americans. Furthermore, Hispanics seem to be at higher risk of some alcohol-related problems, such as driving under the influence, male-to-female partner violence, and alcohol-related mortality (cirrhosis) (52, 53).

Like the risk of mental illness, these health risk factors seem to increase with acculturation and time since entering the United States. An analysis using data from the National Longitudinal Study of Adolescent Health (57) suggested that the acculturation of overweight-related behaviors, such as diet, smoking, and inactivity, is rapid; moreover, longer residence in the United States was associated with increased weight gain among Puerto Ricans and Cubans.

However, other studies show differences between Hispanic subgroups in this regard. For example, poorer health was associated with higher socioeconomic status and acculturation among Mexicans but with lower socioeconomic status and acculturation among Latinos from the Caribbean Islands (58).

As noted earlier, Hispanics are more likely to live with family (1). They also seem to be more likely to engage in closed networks, as opposed to the open networks of most Americans of Caucasian origin. This may have both beneficial and adverse effects. Closed networks appear to provide more emotional support to their members; however, they may eventually pose barriers to problem solving because of the lack of bridges to outside networks (59).

A recently published study comparing Mexican and Anglo American patients with schizophrenia and their families (60) found that “family warmth” seemed to work in opposite ways in the two ethnic groups—as a protective factor for the Mexican patients and as a risk factor for the Anglo American patients. In another study (61), researchers found that greater self-reported family cohesion appeared to have a protective effect against emotional distress for Latino and African American relatives of patients with schizophrenia, but not for Caucasian American relatives. Researchers in California found that Mexican American patients and relatives reported lower rates of expressed emotion than Caucasians. High expressed emotion predicted relapse for Caucasians but not for Mexican Americans (62).

About half of adult Hispanics report having low English proficiency. According to the 2000 census, about 80% of Hispanics age 5 years and older speak Spanish at home, and half of those harbor doubts about their proficiency in English (63). According to the 2002 National Survey of Latinos, 47% of adult Hispanics are Spanish dominant, 28% are bilingual, and 25% are English dominant (4). A significant number of Hispanics report difficulty communicating with doctors and other health care providers (4) (Table 6). This is worrisome, as poor English fluency in itself is a risk factor for less use of health care and mental health care services among Hispanics (64) (Table 7). Hispanic children whose parents had low English proficiency had triple the odds for fair or poor health status (65).

A significant number of Hispanics also have inadequate health literacy, even in Spanish, and have mistaken beliefs about mental illness. In a four-city study, rates of inadequate health literacy in Spanish among Spanish seniors ranged from 21.2% to 60% (66). A study of beliefs about the cause of mental illness among the elderly found that Hispanics were less likely than Caucasians or African Americans to endorse biopsychosocial causes of depression, and this lower endorsement predicted lower use of mental health services (67). Poor health literacy also seems to be higher among Hispanics with low English proficiency (68).

Moreover, once Hispanic patients access services, poor English proficiency is a risk factor for lower quality of care, delayed services, poorer follow-up, longer hospital stay, higher resource utilization, lower adherence to psychotropic medications, and patient dissatisfaction with provider and treatment (70–80). The authors of a study in Connecticut concluded that culturally sensitive clinics are not enough to address the difficulties monolingual Hispanics have in managing medications, which may be due to the persistent effects of the language barrier and low education levels (81).

There is some debate on the frequency of the most common mental disorders, especially depression, in Hispanics (82), but the larger studies show lower caseness rates for Hispanics as a whole and for Mexican and Cuban Americans than for non-Latinos. A study using data from the National Health and Nutrition Examination Survey III (83) found that the prevalence of major depression was higher in Caucasians than African and Hispanic Americans; the opposite pattern was found for dysthymic disorder. The National Comorbidity Survey Replication (84) showed lower odds ratios for Hispanics compared with non-Hispanic whites for anxiety, mood, impulse control, and substance use disorders as well as for any disorder. Recently a large survey, the National Epidemiologic Survey on Alcoholism and Related Conditions (NESARC) (85), found 12-month prevalences of major depressive disorder of 4.27% for Hispanics, compared with 5.53% for Caucasians and 4.52% for African Americans. In the Hispanic Health and Nutrition Examination Survey 1982–1984 (14), however, Puerto Ricans had a significantly higher frequency of depression than Mexican and Cuban Americans. In a recent pooled analysis of the National Surveys on Drug Use and Health, 2001–2003 (86), similar findings were noted for Hispanics and Puerto Ricans. However, according to the Youth Risk Behavior Survey of the Centers for Disease Control and Prevention, in 1999, Latino youths were significantly more likely than young African and European Americans to have attempted suicide, made a suicide plan, or seriously considered suicide (87). Finally, some disadvantaged Hispanic subpopulations seem to be at higher risk of mental illness—for example, the elderly, illegal aliens, and segregated or impoverished sectors (88–91).

However, even if Hispanics do have a lower risk of mental disorders, they seem to have more persistent disorders (92, 93). This finding seems logical, given the less favorable environment for Hispanics, their lower rates of access to and utilization of mental health care, and the lower quality of the care they receive.

We have no reliable data on the frequency among Hispanics of mental disorders that are difficult to detect in the usual surveys, such as bipolar disorder and schizophrenia (94). For bipolar disorder, an analysis of the NESARC data found that the lifetime odds ratio of DSM-IV bipolar I disorder was 0.6 for Hispanics compared with Caucasians (95). Half a century ago, Malzberg (96, 97) found that the rate of first admissions to mental hospitals, especially for dementia praecox, for Puerto Ricans in New York State largely exceeded their proportion in the population. At the same time, Jaco (98), in Texas, found lower rates of psychiatric admissions for Mexican Americans than for white and black Americans. Similar data were reported in California (99–100). In Puerto Rico, the figures from the Epidemiologic Catchment Area (ECA) study are in line with national data, and data from the Los Angeles ECA study show no statistical differences between Mexican Americans and non-Hispanic whites (101–103). However, the methodology of this ECA study is highly suspect (104). The question of the prevalence of schizophrenia among Hispanic Americans is of further interest because studies in countries with national registries consistently reveal a higher incidence of schizophrenia in disadvantaged immigrant populations (105). A recent meta-analysis concluded that a history of migration is an important risk factor for schizophrenia, suggesting a role for psychosocial adversity in the etiology of the disease (106).

Studies of dementia suggest that its frequency is significantly higher among Hispanics than Caucasians and that Hispanics are at higher risk of developing both Alzheimer’s disease and vascular dementia (107–110).

Hispanics seem to be at higher risk of not having mental health problems detected in primary care settings (111). In a study with 496 Mexican Americans, primary care providers noted depression in 21% of the patients who were found to have depressive disorders in a systematic evaluation (112). In a pediatric clinic, only 24% of Hispanic mothers with depression were correctly identified as depressed, compared with 31% of black mothers and 38% of other mothers (113). A study of elderly Medicare beneficiaries found that those in the “Hispanic or other” ethnic groups had an odds ratio of 0.72 for diagnosed depression compared with Caucasians (114). Several studies have suggested that rates of recognition and treatment of attention deficit hyperactivity disorder are lower in Hispanic children (115–118). This underrecognition seems to be influenced by not only clinician factors but also patient and environmental factors. Hispanic adolescents are less likely than Caucasians to have their mental health problems reported by caregivers (119). A study in California found high levels of maternal depressive symptoms among low-income Hispanic women, but only half of these women identified themselves as needing help with depression (120).

In the literature on the diagnosis of mental disorders in Hispanic Americans, we have mainly retrospective or small studies scattered over several decades, with conflicting results. The way Hispanics with mental illness are perceived seems to differ from the way Caucasians are perceived and seems to fluctuate over time. For example, from the 1960s to the early 1980s, studies consistently reported a tendency for higher disorganization and impairment among Hispanic than among nonminority patients with schizophrenia (121–125). However, studies from the 1980s to date report either no difference or a tendency for the nonminority group to be more symptomatic (126–134).

Some studies suggest that Hispanics are more likely than Caucasians to receive a diagnosis of a psychotic disorder and less likely to receive one of a mood disorder. In the 1980s, a review of the records of 76 patients with bipolar disorder showed that Hispanic (Puerto Rican) and black patients were more likely than whites to be misdiagnosed as having schizophrenia, particularly if they were young and had auditory hallucinations (135). A 7-year longitudinal study (136) of 936 inpatients in Texas who had at least four hospitalizations found that 44% of Hispanics who were initially diagnosed as having schizophrenia received a change in diagnosis during the study period, double the rate for Caucasians and African Americans. A study of a national sample of elderly inpatient veterans (137) found that African Americans and Hispanics were more likely than Caucasians to have a diagnosis of a psychotic disorder. However, a retrospective study in New Jersey examining new admissions in a large behavioral health system found that Hispanics, in contrast to African Americans, were not more likely than Caucasians to receive a diagnosis of schizophrenia, although they were disproportionately diagnosed as having major depression, despite higher levels of psychotic symptoms by self-report (138). A study using the Structured Clinical Interview for DSM-IV found that half of Hispanic adolescents diagnosed as having major depressive disorder met the criteria for bipolar disorder (139). A high frequency of psychotic symptoms in the absence of formal psychosis and associated especially with depression has been reported in Hispanic Americans. At a general medicine practice in New York City, psychotic symptoms were more commonly observed in depressed Hispanics than in depressed non-Hispanics (140). A study in Rhode Island found that Hispanics with major depression were more likely to report psychotic symptoms (141). A study in Boston found that 46% of outpatient Caribbean Latinos reported having hallucinations, but only 9% of them were diagnosed as having a thought disorder (142). A retrospective study of applicants for state benefits (143) found that Hispanics were more likely to endorse hallucinations but not to receive a diagnosis of psychosis.

A tendency for Hispanics to somatize distress has been repeatedly reported in Latin America, although the methodology of the studies has been criticized (144). A U.S. study using a personality inventory in claims for workers’ compensation (145) found that Hispanics were more likely than Caucasians to somatize. In a study in California (146), depressed Latinas and African American women scored significantly higher than Caucasians on somatization. A recent study of anxiety and fears (147) found that parents of Mexican children in Mexico and Hispanic American children in the United States reported more worry and physiological symptoms for their children than Caucasian parents.

A recent publication reviewed various psychosocial interventions for ethnic minorities (184). Here we focus on those with the most evidence of clinical efficacy.

Our purpose in this review was to synthesize a broad bibliography and to gauge the status of the current knowledge base on Latino mental health research. Hopefully, the review has made apparent the areas of glaring weakness in the research “menu.” We have not attempted a deep analysis, nor an exacting interpretation of this literature beyond primary findings, and inevitably many articles have been excluded in our effort to focus on the most relevant work for our purposes. It is our earnest hope that the rapidly accumulating knowledge about Latino mental health will be examined by researchers and practitioners and that gaps in the knowledge will be addressed through high-quality research followed by rapid dissemination of findings to practitioners. Outstanding problems such as the low enrollment rates in intervention trials require urgent attention from federal research and regulatory agencies. As this review demonstrates, disparities in mental heath care for Latinos remain numerous and are reinforced by inadequate information, which itself could serve to impede progress.

Here are some of the gaps in the research that in our opinion urgently need to be addressed:

In closing, we offer providers who treat Hispanic patients some practical lessons drawn from our review:

CME Financial Disclosure The authors are affiliated with Department of Psychiatry, University of Medicine and Dentistry of New Jersey, Robert Wood Johnson Medical School, Piscataway, New Jersey.

Humberto Marin, M.D., and William A. Vega, Ph.D.: No disclosure of financial interests or affiliations to report.

Javier I. Escobar, M.D.: Occasional Lectures/Consultation: Eli Lilly, Pfizer, Bristol Myers, Otsuka. Stock ownership: Bristol Myers, Pfizer, J&J.