This patient presents with a “dual disorder,” most likely schizophrenia and co-occurring alcohol and cannabis use disorders. This is not an uncommon clinical scenario. The literature suggests that nearly 50% of patients with schizophrenia have a co-occurring substance use disorder, most frequently alcohol and/or cannabis (at a rate about three times as high as that of the general population) (1). Patients with dual diagnoses are highly prone to adverse outcomes in several domains: increased symptom severity; increased rates of hospitalization, infectious illnesses, violence, victimization, homelessness, and nonadherence to medication; and poor overall response to pharmacologic treatment (1). Co-occurring substance use disorders contribute substantially to the financial costs and emotional burdens of schizophrenia—for patients, their families, and the mental health system.

The co-occurrence of schizophrenia and substance use disorder is not seen only in patients with chronic illness. Rates of cannabis use disorder as high as 53% have been reported in some studies of patients with first-episode schizophrenic psychosis (2), and cannabis use has been associated with an earlier age at onset of schizophrenia (2), an elevated risk of developing psychosis (3), and a higher relapse rate after remission of acute psychotic symptoms in the first episode (4). Strategies to help patients decrease substance use both in their first episode and over the course of their illness are essential if we are to achieve optimal outcomes for patients with schizophrenia.

A number of theories have been advanced to explain the elevated prevalence of substance use disorder in people with schizophrenia (5, 6). First, some authors have proposed neural diathesis-stress models, which suggest that a neurobiologic vulnerability interacts with environmental stressors (such as substance use) in vulnerable individuals to precipitate the onset of schizophrenia or relapse of psychosis (7). Support for this model has been found in studies indicating that substance abuse is associated with an earlier age at onset of schizophrenia (2, 3), the recent finding that use of cannabis by adolescents who have the “high output” catechol O-methyltransferase polymorphism (Val/Val) is associated with a higher risk of developing psychosis in young adulthood (8), and evidence that patients with schizophrenia experience negative clinical effects, such as relapse, after using small quantities of substances of abuse (9). Patients with schizophrenia thus appear to have a heightened vulnerability to the effects of psychoactive substances, with relatively modest levels of alcohol and drug use leading to adverse consequences.

Second, the accumulative risk factor hypothesis (5) suggests that people with schizophrenia may have a greater risk of substance use disorder because of the cumulative effects of poor cognitive, social, educational, and vocational functioning as well as poverty, victimization, and exposure to deviant and/or substance-using familial and social environments, all of which are known risk factors for substance abuse. While this model has strong face validity, research has not yet tested whether the accumulation of these risk factors underlies the higher rates of substance abuse in patients with schizophrenia.

Third, the self-medication hypothesis (see reference 10, for example) proposes that patients with schizophrenia use substances in order to reduce symptoms of the disorder or to lessen the side effects of antipsychotic medications. While this model has some commonsense appeal, most studies have reported no relationship or only a limited relationship between substance use and either positive or negative symptoms or medication-induced extrapyramidal side effects (see reference 11, for example). Studies demonstrating elevated rates of substance use disorder among patients with first-episode psychosis before any exposure to antipsychotic medication further argue against this hypothesis (2).

Our group (6, 12) and Chambers and colleagues (13) have proposed an alternative model, that of reward circuitry dysfunction. Animal and human studies indicate that the brain areas thought to be dysfunctional in schizophrenia are part of the dopamine-mediated brain reward circuitry (14). In patients with schizophrenia, substance use may modulate this dysfunctional brain reward circuitry by producing an increase in the neuronally based signal detection of these dopamine-rich systems (15). Thus, the basis of the use of these substances may be related to the difficulty these patients have in experiencing “normal” levels of reward from the environment and to the ability of substances of abuse to ameliorate this reward circuitry deficit (6, 12).

Co-occurring substance use disorders are often underdetected and undertreated in mental health settings, where the traditional separation between mental health and substance abuse training programs and service delivery systems results in a lack of knowledge about co-occurring disorders and a seemingly inconsistent commitment to the treatment of the substance abuse component. Given the substantial lifetime vulnerability of patients with schizophrenia to substance use disorder, this situation is clearly problematic. Clinicians must keep possible substance use clearly in mind when evaluating and treating patients with schizophrenia. Screening and assessment can be aided by the use of standardized measures, especially instruments specifically developed for patients with severe mental illness, such as the Dartmouth Assessment of Lifestyle Instrument for Screening (16; http://dms.dartmouth.edu/prc/instruments/dali.pdf); the Alcohol Use Scale and the Drug Use Scale (17); and the Stage of Treatment Scale for ongoing assessment (18, 19). The key ingredient, however, is for clinicians to be alert to the potential for substance abuse when working with patients who are unstable. Taking note of behaviors consistent with substance abuse—for example, frequent missed appointments, poor medication adherence, and financial or legal problems—and obtaining collateral information from family members, case managers, and significant others can be of great help. Honest reporting of actual use is most likely to occur if the clinician establishes a nonjudgmental therapeutic alliance when assessing a patient who may have a co-occurring substance use disorder.

Integrated Treatment

More than 50 controlled studies have established the importance of integrating the treatment of patients with co-occurring disorders—in essence, overcoming the problems inherent in the separate systems of care for these individuals (20). While the philosophical approaches of substance abuse treatment programs and community mental health centers may differ, patients with co-occurring disorders need to be treated in a coherent manner, ensuring that they receive the same message from all clinicians about the importance of treatment for both their mental illness and their substance use. Programs that coordinate pharmacotherapy, psychosocial treatments, and substance abuse counseling into a single comprehensive package are most likely to have good treatment outcomes. Integrated treatment programs for dual-diagnosis patients should include staged interventions tailored to the patient’s motivation for change (e.g., the use of assertive outreach to engage patients in treatment and motivational interviewing techniques to develop motivation to address substance use); comprehensive services (e.g., medication management, rehabilitation, and social support interventions); and a long-term perspective, since relapse is a common occurrence. One such well-described treatment program, integrated dual disorder treatment (21), involves multidisciplinary teams to provide a truly integrated level of care. The teams often include clinical case managers, who may serve as individual, group, or family therapists, as well as psychiatrists and other medical staff. One of the keys to success in such treatment strategies lies in ensuring that patients have access to residential services and vocational supports, which the majority of these patients need.

Patients with co-occurring disorders benefit from standard psychosocial interventions, such as assertive community treatment and supported employment, just as other patients do, but these interventions have a minimal impact on their substance use disorder. The specific integrated interventions that consistently reduce substance abuse include group counseling with cognitive behavior and motivational components (22), contingency management (23), and, for patients who do not respond to less intensive interventions, long-term residential programs (24). With treatment, the majority of these patients will become abstinent over time, but incarceration, homelessness, infection with hepatitis C and HIV, and early mortality are clear risks for those who do not become abstinent.

Pharmacologic Management

Antipsychotic Medication

While some controversy exists on the optimal antipsychotic medication for patients with schizophrenia, there is a general consensus that the first-generation (conventional) antipsychotics are not particularly helpful in the treatment of patients with schizophrenia and substance use disorder. Studies over many years have documented a high rate of substance use disorder in schizophrenia patients who have been treated with these agents and have shown that patients with dual diagnoses experience a more difficult course of illness when treated with these medications compared with schizophrenia patients who do not have a co-occurring substance use disorder. Several investigators have suggested that conventional antipsychotics may actually precipitate or worsen the abuse of substances in patients with schizophrenia (see reference 25, for example). The data suggesting that haloperidol may increase cigarette smoking in patients with schizophrenia are often cited to highlight the problems associated with the use of conventional antipsychotics in these patients (26).

Injectable Antipsychotics

Given the poor treatment adherence that is characteristic of patients with schizophrenia and co-occurring substance use disorder, it is not surprising that long-acting injectable conventional antipsychotics have often been used in these patients to enhance control over psychosis. Data on the efficacy of these medications are sparse, however. One unblinded study (27) suggests that the long-acting injectable second-generation antipsychotic risperidone may be of greater value than long-acting injectable conventional antipsychotics in these patients.

Atypical Antipsychotics

In contrast to concerns that conventional antipsychotics may worsen substance abuse, a number of (still preliminary) studies suggest that some of the second-generation (atypical) agents may be helpful for these patients (see reference 28 for detailed review). For example, there have been reports that for patients treated with clozapine and olanzapine, overall outcomes during treatment are as good among those who have a co-occurring substance use disorder as those who do not. Some researchers have suggested that the lower incidence of neurologic side effects produced by the atypical antipsychotics, along with the possibility that these agents may be more likely to decrease negative symptoms, make them a logical choice for patients with co-occurring substance use disorder (even though parameters of the metabolic syndrome must be monitored while using these agents).

Regarding nicotine, several controlled studies (reference 29, for example) have indicated that smoking cessation interventions may be more effective in patients with schizophrenia who are treated with atypical rather than conventional antipsychotics. McEvoy and colleagues prospectively studied patients switching from conventional antipsychotics to clozapine and observed spontaneous reductions in cigarette smoking, in both a small (N=12) and a larger (N=55) group of patients (review in reference 28).

A substantial portion of the available data concerning the effects of atypical antipsychotics on substance use in patients with schizophrenia is from clozapine studies. Initially, case reports and series noted reduced use (with frequent abstinence) and reduced craving for substances while taking clozapine. Two retrospective surveys of approximately 35 patients each (one by our group) reported significant reductions in alcohol and other substance abuse. These findings were supported by data from a prospective trial indicating that 11 of 16 patients (70%) with co-occurring substance use disorder and schizophrenia reduced or stopped using substances during a 12-week prospective trial of clozapine (review in reference 28).

Our group also reported other data supporting this unusual effect of clozapine. In a naturalistic study of 151 patients with schizophrenia and substance use disorder who were followed longitudinally for more than 3 years (30), we found that the 36 patients treated with clozapine were more than twice as likely to attain full remission of alcohol abuse than those treated with conventional antipsychotics (79% versus 34%). In a 10-year follow-up report on the same patient sample (31), we noted that in patients who attained remission, use of clozapine was associated with a markedly reduced relapse rate over the subsequent year (8% of the clozapine group versus 40% of the conventional antipsychotic group). We have proposed that the effect of clozapine on substance use in these patients is related to clozapine’s broad-spectrum pharmacologic effects (i.e., its weak antagonism at the dopamine D₂ receptor and its potent blockade of the noradrenergic α ₂ receptor, coupled with its ability to release norepinephrine in the brain), which result in an amelioration of the brain reward circuit deficiency in these patients (6, 12). However, despite the strong suggestions these studies offer on clozapine’s potential, none were prospective randomized controlled trials, and thus the evidence about clozapine’s value for these patients remains preliminary.

A number of other, more recently developed atypical antipsychotics have also been assessed in patients with co-occurring substance use disorder, but there is even less information about them than about clozapine. In a randomized, prospective study comparing risperidone and haloperidol, risperidone treatment decreased cue-craving and substance abuse relapse in patients with schizophrenia and co-occurring cocaine dependence (32). However, our group, in a retrospective study of 41 patients with schizophrenia and co-occurring alcohol or cannabis use disorder (33), noted that abstinence rates were considerably higher among patients treated with clozapine (54%) than among those treated with risperidone (12.5%). Moreover, a large (N=249) retrospective chart review of Department of Veterans Affairs (VA) patients (34) found that after confounding factors were controlled for, there were no differences in improvement on Addiction Severity Index scores between patients treated with atypical antipsychotics (mostly risperidone and olanzapine) and those treated with conventional antipsychotics.

Although case reports and one open-label trial (review in reference 28) have suggested that olanzapine may be associated with a reduction in substance use, results of randomized controlled trials have been mixed. Smelson and colleagues (35) reported reduced cue-craving and reduced cocaine-positive urine screens with olanzapine compared with haloperidol, but Sayers and colleagues (36) reported negative findings for olanzapine compared with haloperidol. Our group reported that although the rate of substance use appeared to decrease in patients treated with olanzapine, the same decreased rate was observed in patients treated with conventional antipsychotics (see reference 28). Finally, as noted above, in the large VA chart review, no differences were observed in improvement in substance abuse measures between those taking olanzapine or risperidone and those taking conventional antipsychotics (34).

In two open trials, treatment with quetiapine was associated with reductions in one of several measures of substance abuse in 24 patients with schizophrenia spectrum disorders and with decreased stimulant craving but not decreased stimulant use in 24 patients with co-occurring disorders whose medications were switched from conventional antipsychotics (see reference 28). In two small pilot studies, aripiprazole has been reported to decrease craving for and use of alcohol and cocaine (see reference 28). To our knowledge, there are no reports on the use of ziprasidone in this population.

Clearly, further studies of the effects of the various atypical antipsychotics will be needed before clear recommendations can be made on optimal pharmacologic management for patients with schizophrenia and co-occurring substance use disorder.

Adjunctive Agents

A number of medications have been used adjunctively with antipsychotics in patients with schizophrenia and co-occurring substance use disorder. For example, several recent studies have suggested that bupropion may be helpful for smoking cessation in patients with schizophrenia (reference 37, for example). The tricyclic antidepressants desipramine and imipramine may be helpful in reducing cocaine use in patients with a co-occurring cocaine use disorder (reference 38, for example). Although a few studies have suggested that valproic acid may reduce alcohol use in patients with bipolar disorder, this agent has not been studied in patients with schizophrenia and co-occurring alcohol use disorder.

Adjunctive Alcohol Use Disorder Treatments

Three medications have been approved in the United States for the treatment of alcohol use disorders—disulfiram, naltrexone, and acamprosate. While disulfiram has been available for decades as a treatment for alcoholism and has been used safely in patients with schizophrenia and alcohol use disorders, its use in this patient group has been quite limited. This may be due to concerns about its potential ability to increase psychosis and about its potential liver toxicity. However, a recent retrospective study of disulfiram in 33 patients with alcohol use disorder and severe mental illness found that 64% developed a sustained remission from their alcohol use disorder without evidence of exacerbation of psychosis (39). A recent report (40) on the use of disulfiram in a large group of patients (N=250) with a variety of mental illnesses and co-occurring alcohol dependence (66 of whom had a diagnosis of a psychotic spectrum disorder) demonstrated that disulfiram was well tolerated and more effective than placebo in reducing alcohol use in this population, without worsening psychosis or causing liver toxicity.

Naltrexone, which has been shown to reduce drinking in patients with primary alcohol use disorders, has shown promise in patients with co-occurring schizophrenia or other psychotic disorders and alcohol use disorder. In two randomized controlled trials, one with 31 patients with schizophrenia, the other with 66 patients with a psychotic spectrum disorder, alcohol use decreased more in patients treated with adjunctive naltrexone than in those treated with adjunctive placebo (40, 41). Continued studies will be needed to fully delineate the extent of this effect and to ascertain whether naltrexone’s potential liver toxicity will be problematic for routine use in this population. The availability of long-acting injectable naltrexone will provide another potential therapeutic option for these patients.

Two other medications may be of interest in the future for this population. Acamprosate, an agent with glutamatergic effects, was recently approved for use in the treatment of alcoholism. However, we are not aware of any studies of this medication in patients with schizophrenia. Finally, the anticonvulsant topiramate has shown promising results in patients with alcoholism. At the time of this writing, case reports suggest that topiramate may also have a beneficial effect on alcohol use in patients with schizophrenia (review in reference 28).

As we noted earlier, nearly 50% of patients with schizophrenia develop a substance use disorder in their lifetime, and this co-occurring disorder substantially worsens the course of schizophrenia by destabilizing the illness, impeding treatment adherence, and adding the problems of psychosocial instability, legal entanglements, and medical illnesses to the challenge of managing psychotic symptoms. While the basis of the high rate of substance abuse in this patient group is uncertain, it appears that these patients are highly vulnerable to the effects of alcohol and other drugs and that they suffer from an accumulation of known risk factors for the development of substance abuse disorder. We and others have hypothesized that a brain reward circuit dysfunction may underpin substance use in these patients. The emerging data suggesting that patients with schizophrenia may have a biologic basis for their substance abuse may be helpful therapeutically. The notion that patients have a desire to use substances in order to achieve some feelings of normalcy can help clinicians suspend judgment about substance use and adopt a pragmatic attitude that they will need if they are to work effectively with these patients and apply the techniques recommended by clinical experts in behavioral substance abuse counseling (21). In addition, careful prescription of medications to treat the psychotic illness and the substance use disorder is likely to enhance the treatment of patients with co-occurring disorders.

Although the patient described in the vignette at the beginning of this article was homeless, psychotic, and intoxicated, clinicians should be optimistic about the potential for effective treatment of his co-occurring conditions. While he is in the hospital, clinicians should conduct an assessment (in a nonjudgmental fashion), including contacting the outpatient treatment team and family or support persons for information. Psychiatric stabilization with medication should be initiated immediately, with the patient being offered those antipsychotic medications that are least likely to exacerbate substance abuse and most likely to enhance adherence, such as one of the atypical agents (recognizing that no published data are as yet available on ziprasidone in this population). Assessment for common comorbid medical conditions, such as infectious diseases and diabetes, and comorbid psychiatric conditions, such as posttraumatic stress disorder, should also be conducted and any comorbid conditions treated.

Since this patient is likely to have a very brief hospital stay, a community-based, stagewise clinical approach to treatment would be optimal. Outpatient clinicians treating a patient who has recently suffered a significant relapse, as this patient has, should reach out to him while he is in the hospital and immediately after discharge to offer practical assistance, such as help with regaining housing, reaccessing entitlements, and other concerns the patient may have. Once the patient becomes reengaged in treatment, clinicians can provide motivational counseling to enhance his motivation to participate in treatment and to consider the impact of alcohol use on his ability to attain his own personal goals. Efforts to stabilize the patient’s psychiatric symptoms should continue. Ensuring that he continues to take his antipsychotic medication and engaging him in rehabilitation, such as supported employment services and social skills training, are often addressed first. As the patient becomes motivated to address his substance use, he should be encouraged to participate in a dual-disorder treatment group or a contingency management program, since these have the strongest evidence of efficacy. The treatment team should encourage the patient to develop friendships with sober peers and to become involved in activities with substantial reinforcement value that are alternatives to substance use, such as work, exercise, or a hobby. Use of an adjunctive medication aimed at decreasing substance use, such as naltrexone, should also be considered. Once remission is attained, clinicians should focus on maintaining remission over the long term while they address other key issues, such as overall physical health, employment, and ongoing symptom management.