Major Depressive Disorder With Psychosis-Like Symptoms Among Latinos

The DSM-IV warns clinicians not to overdiagnose psychotic disorders among members of some racial and ethnic minority groups and to consider the cultural manifestations of illness as alternative explanations for psychosis-like symptoms (1). Yet clinicians often face the dilemma of distinguishing what may constitute a legitimate psychotic symptom from what is possibly a culturally derived phenomenon.

The differential diagnosis of psychosis-like symptoms among Latinos is even more complex. Psychosis-like symptoms are common among Latinos, both in clinical and in epidemiological samples. In addition, because of the paucity of Latino psychiatrists in the United States, most clinicians do not share the cultural background of their Latino patients. Even for Latino psychiatrists, the diagnosis of psychosis-like symptoms among Latino patients remains a challenge, perhaps, in part, because of the wide range of cultural and ethnic characteristics of different Latino subpopulations.

Many clinicians argue that isolated psychosis-like symptoms in an otherwise well-adapted and functioning person do not necessarily demand treatment, especially if the symptoms are ego syntonic (2). An example might be hearing or seeing deceased relatives during bereavement. Clinicians, however, typically are not confronted with such a clear example, which might explain the high prevalence of psychosis-like symptoms among epidemiological samples of Latinos.

Determining whether a genuine psychotic component exists has important treatment implications, given that the presence of psychosis might indicate the use of antipsychotics. Full-blown psychotic syndromes can be clearly classified and demand treatment regardless of an individual's ethnic or racial background. The clinical dilemma occurs when an individual with an axis I mood or anxiety disorder presents with psychotic symptoms (3).

When treating Latinos, psychiatrists are often confronted by the dilemma of redefining the boundaries for the subtype of major depressive disorder with psychotic features. Perhaps understandably, the DSM-IV definition of psychotic symptoms reflects a general ambiguity: “The narrowest definition of psychotic is restricted to delusions or prominent hallucinations, with the hallucinations occurring in the absence of insight into their pathological nature. A slightly less restrictive definition would also include prominent hallucinations that the individual realizes are hallucinatory experiences. Broader still is a definition that also includes other positive symptoms of schizophrenia (i.e., disorganized speech, grossly disorganized or catatonic behavior)” (1).

The challenge of diagnosing psychosis-like symptoms among Latinos with axis I disorders will only become more acute. In the United States, depression is the third leading cause after cardiovascular diseases and lung and breast cancers of years lost because of disability or death (4). In one large study, Latinos were found to be significantly more likely than members of other ethnic or racial groups to be diagnosed as having major depressive disorder (5). Lifetime prevalence of major depressive disorder is 15% among Latinos as a group and 19% among Puerto Ricans and Cubans, 15% among Mexicans, and 14% among other Latino subgroups (6). In addition, the mean duration of a major depressive episode among Latinos as well as among African Americans is longer (7).

Further, Latinos are the nation's largest minority group and account for more than half of the U.S. population growth in the past decade. In 2010, the Latino population in the U.S. was estimated to be over 50.5 million, representing approximately 16% of the country's total population (8).

This article presents a selective review of the literature intended to help clinicians with the differential diagnosis of psychosis-like and genuinely psychotic symptoms among Latinos and to raise awareness of the actual significance of these atypical manifestations. We believe that the literature indicates that atypical psychotic symptoms among Latinos are in fact nonpsychotic manifestations and that their resemblance to psychotic phenomena is misleading. Nonetheless, such psychosis-like symptoms are pathological in nature and might help guiding treatment decisions for underlying mood and anxiety disorders.

Methods

This article is based on the literature published through February 2011 and derived from the National Library of Medicine (PubMed) by searching with the keywords “Latino” and “Hispanic” in combination with “depression and psychotic,” “idiom of distress,” and “psychotic.” Additional articles were found among the references cited by the publications of interest and, subsequently, by more targeted searches of PubMed. In our selection of the literature, we have emphasized large epidemiological studies but still included publications derived from small clinical samples because they can provide valuable and practical clinical insights. We use the term “psychosis-like” when quoting studies that questioned the actual psychotic nature of the symptoms. In addition, for simplicity, we omitted mention of ethnicity when referring to non-Latino whites and non-Latino blacks.

Results

The literature search found 37 articles that were reviewed for this article.

Clinical features

When treating Latinos who have major depressive disorder, it is not uncommon to find out, almost incidentally, during a general review of psychopathology that the patient experiences psychosis-like symptoms (3,9). The most common experiences are auditory hallucinations, often described in the literature as “simple,” “vague” and “not elaborate,” accompanied by visual hallucinations (3,10,11) and illusions. Some individuals might hear knocking at the front door, doorbells ringing, footsteps, a voice mumbling, or the voice of a deceased relative calling their name (3,9,12,13), and others might see a celajes, or passing shadow, at the corner of their eye or feel a presence behind them (3,9,13). Paranoid ideation, such as a belief that one is being spied on or followed, is also common, but ideas of reference, such as feeling that one is being looked at by everybody in the street, are less common (10,11,14). Olfactory and tactile hallucinations have been reported as well (12,14). First-rank signs and symptoms are uncommon (14). A box on this page lists psychosis-like symptoms reported by Latinos.

The following clinical vignette exemplifies these clinical scenarios.

A 54-year-old married woman, from Colombia, who had lived in the United States for 21 years, presented with depression, pronounced anxiety, forgetfulness, frequent crying spells, and somatic features including palpitations, tension headaches, and tremors. She described fantasizing being dead (“I feel that I am dead and people are crying for me”) and reported hearing voices calling her (“I can distinguish the person—my brother, my sister, my spouse—but nobody is there”). At times she heard simple noises such as a phone ringing. She denied having visual disperceptions and paranoid ideation. Her history was noteworthy for a past abusive relationship with a man, which lasted for 17 years.

She was treated with sertraline up to 200 mg per day, zolpidem up to 10 mg per day, and risperidone up to 1.5 mg per day. Within four weeks, her depression responded, and her score on the Patient's Health Questionnaire-9 (PHQ-9) for depression dropped from 18 to 7. After 32 weeks of treatment, her depression had remitted (PHQ-9 score of 3). She continued experiencing the voices until week 44, but they stopped two weeks after risperidone was replaced with aripiprazole up to 5 mg per day. At week 68 of follow-up, aripiprazole was stopped and clonazepam .5 mg per day was started. At her last visit, at week 80, the patient continued to deny any perceptual disturbances and had a PHQ-9 score of 1.

This challenging clinical scenario is somewhat unusual for the population of Latino patients with psychosis-like symptoms. However, it contains several features often indicative of psychosis-like symptoms. The patient had anxious depression associated with forgetfulness—suggestive of a possible episode of dissociation—as well as numerous somatic symptoms and a history of trauma. The patient's beliefs did not reach delusional quality, and the disperceptions were simple in nature. In addition, it is not uncommon for these patients to believe that they hear the voice of a familiar person calling their name. The role of the antipsychotics in this case is unclear, given that the treatment benefits were maintained after switching to a benzodiazepine.

Epidemiology

In a nationally representative sample of Latinos living in the United States, the lifetime prevalence of one or more psychosis-like symptoms was found to be 9.5% (11). Yet only 7% of Latinos with lifetime endorsement of psychosis-like symptoms met DSM-IV criteria for a lifetime psychotic disorder or mood disorder with psychotic features. A majority of them had major depressive disorder with psychotic features (11). A total of 38% received a diagnosis of depressive disorder without psychotic features, 16% of substance use disorder, and 6% of anxiety disorder. Some individuals received more than one diagnosis, and 34% received no diagnosis at all (11).

In a university-affiliated internal medicine practice serving 1,005 predominantly low-income Caribbean immigrants, 22.6% of Latinos, who accounted for 73% (N=735) of the sample, exhibited one or more current psychosis-like symptoms (10). Psychosis-like symptoms were reported by 17% of blacks, who accounted for 22% (N=218) of the sample and by 13.5% of whites, who accounted for 5% (N=52). In the same study, 42.4% of patients with current psychosis-like symptoms had major depressive disorder versus 12.6% of patients without such symptoms.

In a large sample of 19,291 admissions to outpatient or inpatient psychiatric services, compared with whites, Latinos were disproportionately diagnosed with major depressive disorder (odds ratio [OR]=1.74) (5). Self-reported psychosis-like symptoms, but not psychotic disorders, were also more common among Latinos. In a psychiatric tertiary clinic in Boston, the percentage of Caribbean Latinos who reported having hallucinations was as high as 46% (3).

Among Vietnam veterans admitted to a posttraumatic stress disorder (PTSD) rehabilitation unit, psychotic symptoms were significantly more common among Latinos and blacks than among whites and were associated with major depressive disorder comorbidity (15). In the same group of Vietnam veterans with PTSD, 73% of the veterans who were Latino or black, versus 26% of the white veterans, experienced psychotic symptoms. Members of racial and ethnic minority groups were also more likely than whites to have major depressive disorder (87% versus 53%, respectively). Notably, the psychotic symptoms reported by the veterans were for the most part auditory (95%) and visual (86%) hallucinations. Delusions were present among 33% of the veterans, and thought disorder was less common (15). All the veterans with psychotic symptoms had hallucinations that reflected combat themes, such as hearing voices of dead comrades calling for help or seeing the faces of the dead.

Among patients with a primary diagnosis of major depressive disorder treated at an outpatient psychiatric clinic, Latinos were more likely than whites to experience psychosis-like symptoms, a difference that was still significant after adjustment for demographic and clinical variables, including depression severity and PTSD comorbidity (27.3% versus 5.1%, respectively, before adjustment, and 27.3% versus 6.8%, respectively, after adjustment) (12). Comparison of psychosis-like symptoms among Latinos and members of other racial and ethnic minority groups with major depressive disorder have been mixed. A smaller proportion of blacks (16%) with major depressive disorder than Latinos with the same disorder had psychosis-like symptoms, but the difference was not significant. However, rates of psychosis-like symptoms among Portuguese Americans (4.3%) with major depressive disorder were significantly lower than among Latinos with the same disorder (12).

The acculturation factor

The National Latino and Asian American Study (NLAAS) found that 19% of Latinos were Puerto Rican, 23% were Cuban, 34% were Mexican, and 24% belonged to other groups. Presence of lifetime psychosis-like symptoms among Latinos was associated with greater proportion of lifespan lived in the United States and higher English proficiency (11). Alegráa and colleagues (6) confirmed the immigration paradox in a larger sample that included both the NLAAS and the National Comorbidity Survey Replication—U.S.-born Latinos were at a significantly higher risk of having a major depressive episode and of having any depressive disorder compared with immigrant Latinos (19% versus 13% and 20% versus 15%, respectively).

Similarly, the Mexican-American Prevalence and Services Survey (MAPSS) found that lifetime prevalence of psychosis-like symptoms was 27% among U.S.-born Mexican Americans, 12% among immigrants, and 18% overall (14). The same study also showed that U.S.-born Mexican Americans had higher lifetime rates of a major depressive episode (14.8%) compared with Mexican immigrants (5.2%). The rate of major depressive episode among immigrants was similar to that among residents of Mexico City (3.2%) but was lower than that among immigrants who had lived in the United States. for at least 13 years (7.9%) (16). Acculturation has been postulated as the most likely determinant for this variance in rates of major depressive episode (16). These findings suggest that a stronger tie to the mother culture may have a protective effect against psychopathology.

Clinical significance

Although the nature of psychosis-like symptoms presented by Latinos is not clear, the actual clinical significance is documented. Specifically, Olfson and colleagues (10) have shown that at an internal medicine practice with predominantly Latino patients, the patients with psychosis-like symptoms were 7.5 times more likely than patients without psychosis-like symptoms to have panic disorder, 6.6 times more likely to have generalized anxiety disorder, 4.9 times more likely to have major depressive disorder, and three times more likely to have an alcohol use disorder. Similar associations between the lifetime presence of psychosis-like symptoms and of anxiety, mood, and substance use disorders were confirmed in an epidemiological sample of Mexican Americans (14).

Even after adjustment for psychiatric comorbidity and demographic variables, psychosis-like symptoms also predicted significantly greater self-reported impairment in work, family, and social functioning as well as greater suicidal ideation (OR=3.5). Similarly, psychosis-like symptoms were associated with greater utilization of outpatient and inpatient mental health services (OR=2.0) (10). Lewis-Fernandez and colleagues (11) have confirmed the results in the NLAAS study of U.S. Latinos. They found that lifetime history of psychosis-like symptoms not only increased the likelihood of having a diagnosis of depressive, anxiety, or substance use disorder (current and lifetime) but also of having chronic medical conditions and of rating one's health as fair or poor. Lifetime psychosis-like symptoms were also associated with greater utilization of outpatient mental health services and with mental disability and suicidal ideation. Endorsement of psychosis-like symptoms during the past 12 months was significantly associated with current suicidal ideation (OR=2.0) (12). Following a different approach, Schaffer and others (17) studied a diverse sample of patients with major depressive disorder with psychotic features and found that Latino background predicted higher suicidality.

Lifetime psychosis-like symptoms among Latinos have also been associated with self-ratings of history of trauma and ataque de nervios, a culturally specific syndrome that closely resembles a panic attack. (11). More recently, in a smaller cohort, they were also associated with dissociative phenomena (13). These data have strengthened the hypothesis that the altered perceptions described by Latinos could be better classified as trauma-related phenomena than as psychotic symptoms or mere cultural manifestations of distress (18). Yet it is possible that some Latinos with major depressive disorder experience subsyndromal psychotic symptoms that are part of a continuum that includes full-fledged psychotic syndromes (19,20). There is no consensus on the phenomenological differentiation of these two clinical scenarios.

Discussion

A conceptual framework

On the basis of the findings reported in this review and of our clinical experience, we propose a set of criteria, summarized in a box on this page, to characterize psychosis-like manifestations among Latinos. We believe that these symptoms are distinct from the DSM-IV psychotic spectrum because they have a unique phenomenological presentation; they have been associated with trauma and dissociative experiences; their response to antipsychotics is uneven; and they are highly prevalent among Latinos. These symptoms are much more prevalent among Latinos than would be expected given the rates of psychotic disorders and disorders with psychotic features among Latinos in general.

We expect that these criteria may help to more easily differentiate psychosis-like symptoms from genuine psychotic symptoms among Latinos. Clinicians should be aware that in clinical scenarios an overlap of both types of manifestations is still possible. At times, the patient and the family will rationalize the perceptual experiences as somewhat expected in specific circumstances, such as hearing the voice of a recently deceased family member. In fact, such an experience may be suggestive of a culturally appropriate reaction to a life event and, although similar in nature to psychosis-like symptoms, may have a different clinical relevance.

Proposed treatment strategies

There is no consensus about treatment strategies for major depressive disorder with psychosis-like symptoms among Latinos. Guidelines for treatment of major depressive disorder apply to all racial and ethnic groups. Yet it is unclear how to treat psychosis-like symptoms frequently associated with major depressive disorder among Latinos. Nevertheless, even when these phenomena are accepted and seem mostly ego syntonic, it is not uncommon for patients to feel a certain sense of relief when the symptoms disappear. Patients do not report a positive attitude toward the abnormal perceptions (3).

When treating patients with depression and psychosis-like symptoms, clinicians are typically divided between those who favor early use of antipsychotic augmentation and those who use this strategy only if antidepressant monotherapy fails. In a sample of internal medicine patients with psychosis-like symptoms from New York, 42.4% had major depressive disorder but only 7.6% had been prescribed an antipsychotic in the past month (10). In a small community sample of Latino patients with mood and anxiety disorders, mostly major depressive disorder with psychosis-like features, 45% of the patients received an antipsychotic (9).

The limited data available about treatment outcome do not suggest a best treatment option (10). We recommend following the treatment guidelines for major depressive disorder and delaying the use of an antipsychotic whenever the psychosis-like symptoms meet the criteria listed in the box on page 485. However, the use of antidepressant monotherapy for major depressive disorder has been questioned because it yields limited remission rates (21).

As mentioned, psychosis-like symptoms are an indicator of severity and might justify adopting a more aggressive initial approach to treatment of major depressive disorder among Latinos. We favor the use of antidepressant augmentations or combinations with medications other than antipsychotics (21,22) because antipsychotics have a worse side-effect profile (23). When prescribing antidepressants to their Latino patients, clinicians should carefully consider also prescribing adjunctive hypnotics and anxiolytics, as long as concerns about their safety have been addressed. These drugs are prescribed less often among Latinos than among other racial and ethnic groups (24). Psychotherapy and antidepressants could also be a valuable combination. Electroconvulsive therapy is reserved for the most refractory cases of major depressive disorder, and it appears to be underutilized by Latinos.

Research directions

Future studies should seek to validate the criteria we propose for psychosis-like symptoms among different subgroups of Latinos, such as those from Mexico, the Caribbean, and Central America. The clinical meaningfulness of the psychosis-like features, defined by our criteria, should be tested as well. Although response to treatment is never a diagnostic criterion in the DSM-IV, response to antipsychotics could be used by researchers as an external validator for establishing clinically meaningful parameters. Comparing different treatment approaches would also be very informative.

Conclusions

When treating Latinos with major depressive disorder, clinicians should always investigate the presence of psychosis-like symptoms, given that they are common and useful prognostic factors. Ignoring psychosis-like symptoms, just because they are not mentioned specifically in treatment guidelines, could lead to an underestimation of the severity of illness. Also, a poor differential diagnosis might lead to under- or overdiagnosis of major depressive disorder with psychotic features. For example, the presence of psychosis-like symptoms may obscure the fact that a patient is also experiencing more typical psychotic symptoms. On the other hand, clinicians who fail to distinguish psychosis-like symptoms may be at risk of unnecessarily treating depressed Latinos with antipsychotics and exposing them to potentially harmful short- and long-term side effects. It is important, therefore, to carefully reassess the actual benefit of antipsychotics whenever they are prescribed as an augmentation to antidepressants and to consider the option of discontinuing them if they are not effective.