Data from Major Surveys

Most current data on the dually diagnosed population derive from two surveys conducted and analyzed over the past two decades—the Epidemiologic Catchment Area (ECA) Survey, the first wave of which was administered from 1980 to 1984,⁵ and the National Comorbidity Survey (NCS), administered between 1990 and 1992.⁶ Each explored the phenomenon of comorbidity among the general population. While the two surveys yielded somewhat different findings on the prevalence of specific mental disorders—which is not surprising since (among other differences) the ECA was focused on five geographic areas while the NCS had respondents in 34 states—they yielded remarkably consistent findings about comorbidity.

Each survey’s sample was limited to nonaged adults and those in their later teens. (The ECA’s population was limited to those 18 to 64, while the NCS survey was restricted to those 15 to 54.) The NCS was a household survey in the strictest sense of the word; only the ECA dealt with institutionalized and homeless individuals. Except as noted, however, the data in this issue brief are from the NCS, the more recent of the two surveys and the one based on a more representative national sample.

To begin with, the NCS found that psychiatric disorders of all kinds are more prevalent than previously believed; 48 percent of respondents reported a lifetime history of at least one disorder, and 29 percent indicated at least one disorder within the past 12 months. (A lifetime disorder is defined as one present at least once during the course of a person’s life.) While the presence of a disorder does not necessarily connote major disability or impairment of daily functioning, the NCS finding does indicate that a sizable portion of the population have experienced some level of mental illness.

According to the NCS, the most common disorders are major depression and alcohol dependence, reportedly experienced on a lifetime basis by 17.1 percent and 14.1 percent of the population, respectively. Next most common are social and simple phobias (13.3 percent and 11.3 percent). Collectively, substance abuse and anxiety disorders (including phobias) are somewhat more common than affective disorders (such as depression). Those reporting any lifetime substance abuse or dependence constituted 26.6 percent of the population, those with anxiety disorders 24.9 percent, and those with affective disorders 19.3 percent.

With respect to comorbidity—the co-occurrence of more than one disorder, not necessarily limited to mental illness and alcohol or drug disorders—56 percent of those with any disorders had two or more. Of these, 27 percent had only two disorders, while 29 percent had three or more. Total disorders were heavily concentrated—over half of all disorders found among those surveyed were concentrated in the group with three or more disorders, which constituted 14 percent of the total survey population. Extrapolated to the population as a whole, these findings mean that three million people with co-occurring disorders have at least three disorders, and one million have four or more.

Turning specifically to co-occurring mental illness and substance use disorders, 52 percent of those with lifetime alcohol abuse or dependence also had a lifetime mental disorder. (Thirty-six percent also had a lifetime nonalcohol drug use disorder.) Of those with a lifetime history of drug abuse or dependence, 59 percent had a lifetime mental disorder. (Seventy-one percent also had a lifetime alcohol use disorder.) Viewed from another angle, the NCS and other studies indicate that people with mental disorders are at least twice as likely to abuse alcohol and other drugs as people with no mental disorder. In other words, those with mental illness or substance use disorders are at high risk of co-occurring disorders that bring them into the dually diagnosed population.

Regier, who is a leader in the development and interpretation of the ECA, and his colleagues placed these relationships in a broader epidemiological context, which helps explain the interrelationships at work here:

Since in the general population mental disorders are more prevalent than alcohol disorders, which in turn are more prevalent than other drug disorders, there is a natural statistical tendency for the rates of co-occurring disorders to be higher in alcohol treatment than in mental health treatment settings, and highest in other drug abuse patient populations.⁷

The NCS indicates that comorbid psychiatric orders are both more severe and more chronic than purely psychiatric disorders. Nonetheless, only a minority of those with co-occurring disorders are in treatment. (Partly because they are so seriously ill, however, they are more likely to get treatment than those who experience only one disorder within a 12-month period.) Even among the group with three or more disorders, only half are receiving any kind of treatment. Furthermore, most of those in treatment are receiving only general medical attention, not specialized mental health or substance abuse treatment.

As noted above, the NCS also seems to indicate a significant sequencing in the comorbidity of mental illness and substance use disorders. For almost 90 percent of NCS respondents with such comorbidity, the mental disorder develops first. Furthermore, for the vast majority of dually diagnosed individuals, the mental illness sets in during their adolescent years. Typically, the subsequent addictive disorder develops within five to ten years. Median age of onset for the mental illness is 11, while the median age of subsequent onset for an addictive disorder is 21. This period represents the “window of opportunity” for preventive intervention noted above. Obviously, the size of the window varies with the individual.

Comorbidity, Causality, and Relapse

The fact that co-occurring substance use disorders typically follow the onset of mental disorders should not be interpreted to connote a causal relationship. As Ronald Kessler, Harvard professor of health care policy and coordinator of the NCS, and his colleagues cautioned in a recent article: “It is important to recognize that neither temporal order nor prediction can be taken to imply causal priority.”⁸ Ultimately, only longitudinal surveys can confirm or disprove causality, and they have not yet been undertaken.

In the National Advisory Council report to SAMHSA, it is noted that at least four relationships are possible between co-occurring disorders:

• One may directly cause the other.
• One may indirectly lead to the other (for example, through self-medication).
• They may develop from different causes, but interact with each other.
• A common, independent factor (such as childhood emotional trauma) may cause both.

While more research is needed to determine the causal relationship between co-occurring mental and substance use disorders, there seems to be sufficiently strong evidence to indicate the importance of mental health intervention as a strategy to prevent subsequent onset of substance dependence or abuse and to reduce the likelihood that the underlying mental illness will be worsened by concurrent substance use disorders. The importance of preventive intervention seems especially clear for the adolescent population.

Research has shown that dually diagnosed people have a tendency to relapse. Drake and colleagues have pointed out: “Severely mentally ill patients followed in the community for one year had higher rates of rehospitalization if they were dually diagnosed. Substance abuse tends to persist among those with severe mental illness.” Similarly, the SAMHSA National Advisory Council report on comorbidity underscores the risk of relapse in this population:

The most common cause of psychiatric relapse today [in the dually diagnosed population] is the use of alcohol, marijuana, and cocaine. The most common cause of relapse to substance use/abuse today is untreated psychiatric disorder.

Any discussion of the seemingly recent comorbidity phenomenon would be incomplete without Kessler’s cautionary note that it may in part actually reflect the increasing sophistication of the diagnostic tools available to clinicians:

Although a number of recent studies…have been consistent with the NCS in showing that there is substantial lifetime and episode comorbidity among psychiatric disorders, controversy exists concerning the extent to which these results are, at least in part, artifacts of the recently developed diagnostic systems used to operationalize psychiatric disorders.… These systems… dramatically increased the number of diagnostic categories and reduced the number of exclusion criteria, resulting in the assignment of multiple diagnoses to many people who would previously have received only a single diagnosis.⁹

Dual Diagnoses, Homelessness, and Crime

Not surprisingly, having co-occurring mental and substance use disorders places an individual at greater jeopardy of a marginal social existence. Kessler has identified the following proximate risk factors associated with general comorbidity of any type included in the NCS:

Among the social role impairments significantly associated with comorbidity are marital separation and divorce, social isolation, and exposure to conflictual social relationships. Among the work role impairments significantly associated with comorbidity are low educational attainment, unemployment, and chronic financial difficulties.¹⁰

It is estimated that between 10 and 20 percent of the homeless population have co-occurring severe mental illness and substance use disorders. Typically, they have had a history of mental illness and/or substance abuse problems since they were children. Their very homelessness often reflects the failure of the treatment system to address their dual disorders. As Robert Drake and Kim Mueser of the New Hampshire-Dartmouth Psychiatric Research Center recently observed:

Patients with dual diagnoses of severe mental illness and AUD [alcohol use disorder] are particularly prone to unstable housing arrangements and homelessness.… One reason for this increased risk appears to be that dually diagnosed clients often are excluded from housing and treatment programs designated specifically for people with single disorders.… Compared with other homeless subgroups, those with co-occurring severe mental illness and AUD are more likely to experience harsh living conditions, such as living on the streets rather than in shelters; suffer from psychological distress and demoralization; grant sexual favors for food and money; be picked up by the police; become incarcerated; be isolated from their families; and be victimized.¹¹

Given the transience of dually diagnosed homeless people, their poor access to health care of any kind, and their inability to afford public transportation, experts in the field maintain that it is especially important for this population to be able to use mental health and substance abuse services which are at least provided in the same location—if not fully integrated.

On March 5, 1998, the front page of the New York Times featured an article by Fox Butterfield entitled, “Prisons Replace Hospitals for the Nation’s Mentally Ill.” The reporter described the Los Angeles County Jail as “by default… the nation’s largest mental institution… [with] 1,500 to 1,700 inmates who are severely mentally ill, most of them detained on minor charges, essentially for being public nuisances.”

This is the latest evidence of a phenomenon described by Bert Pepper, M.D., and Hilary Ryglewicz in the following terms: “As psychiatric beds have decreased, prison beds have increased, and the problem of social control of mentally ill people has thereby been transferred to the criminal justice system by transinstitutionalization.”¹² The number of public psychiatric beds in the country (primarily in state and county mental hospitals) dropped from 559,000 in 1955 to 69,000 in 1995, largely as a result of deinstitutionalization. Over the past 25 years, the nation has greatly expanded its prison and jail capacity; the number of those incarcerated has jumped about tenfold—from 200,000 in 1972 to an estimated 2 million in 1995.

The population of the country has grown by 100 million in the past 40 years. Thus, the number of public psychiatric beds per 1,000 population has declined sharply, while the number of prison beds per 1,000 has soared. As one system has expanded, the other has shrunk. New York, a state with only 5,800 adult patients in its public mental hospitals, has an estimated 6,000 state prison inmates with serious mental disorders.

Butterfield’s article adds the following perspective:

With voters willing to spend freely to fight rising crime rates, states were building more jails and prisons. Jails became the only institutions left open to the mentally ill 24 hours a day.… In many states, so many public hospitals have closed, or the laws regulating admission to hospitals have been made so tight, that sometimes the only way to get care is to be arrested. Resources are especially scarce for juveniles.

Each of these developments has important implications for the dually diagnosed population. Any direct personal involvement with illicit drugs is by definition criminal activity. Further, by causing a decrease in social and functional status, co-occurring mental and substance use disorders may predispose some individuals to crime. It is estimated that more than one million prisoners in the United States—over half of the incarcerated population—have dual mental illness and substance use disorders.¹³ In light of this, effective diagnosis and treatment of dual disorders can be seen as a crime prevention as well as a public health strategy.

Providing effective treatment to prison and jail inmates with dual disorders would seem to make sense, if for no other reason than that most will eventually return to the community and the likelihood of their recidivism without treatment is high. Yet the lack of any behavioral health treatment for most of the nation’s incarcerated population—much less integrated treatment for dually diagnosed inmates—has been well documented. For example, the Center for Substance Abuse Treatment has reported: “Less than 10 percent of Federal inmates who are addicted have treatment available to them, despite the fact that research over the past decade confirms that intensive prison treatment programs can reduce recidivism by half after release.”¹⁴

Factors Contributing to Increased Comorbidity

Over the past several decades, a number of social and economic developments have combined to create an environment that seems to have increased people’s susceptibility to the co-occurrence of mental illness and substance abuse. These phenomena include the growth in single-parent families, the increase in poverty among children, the decline in the number of unskilled and semi-skilled jobs, and the greater availability and acceptability of illicit drugs such as marijuana and cocaine.

The growth of the “drug culture” has had a profound impact on the expansion of this population, especially in a context where people who might formerly have been institutionalized now remain in the community. Youth exposure to drugs has been on the rise for some time now. In 1962, only 2 percent of those over 12 years old had ever used an illicit drug. By two decades later, almost half the adult population had at least experimented with an illicit substance. The age of first drug use is steadily declining to the point where even elementary school children need to be educated about drugs.¹⁵

Further complicating the situation is the frequency with which dual disorders go undetected by general practitioners as well as mental health and substance abuse specialists. When the co-occurring disorders are not detected, clinicians tend to provide inadequate or inappropriate treatment, further exacerbating the patient’s condition. Drake and colleagues have described this tendency in mental health programs as follows:

Several factors account for the high rates of nondetection, including mental health clinicians’ inattention to AOD [alcohol and other drug] abuse; patients’ denial, minimization, or inability to perceive the relationships between AOD use and their medical and social problems; and the lack of reliable and valid detection methods for this population. Failure to detect AOD abuse in psychiatric settings can result in misdiagnosis; overtreatment of psychiatric syndromes with medications; neglect of appropriate interventions, such as detoxification, AOD education, and AOD abuse counseling; and inappropriate treatment planning.

One study gives particularly striking evidence of how insidiously factors conspire to inhibit the diagnosis of co-occurring disorders. Fifty-six consecutive people admitted to a psychiatric hospital were tracked, of whom 62 percent had positive urine tests for at least one abused drug. Admitting physicians failed to detect drug use in 66 percent of those with positive urinalyses—frequently because they credited patient denials of drug involvement. Of the 26 patients with positive test results who denied their recent drug use, only one received a positive assessment for drugs by admitting physicians.¹⁶

Reformers have called for more aggressive diagnostic techniques to detect co-occurring disorders. In particular, they have recommended that clinicians routinely consider the possibility that patients presenting with either mental illness or substance abuse might have a co-occurring disorder.

Speakers

Bert Pepper, M.D., executive director of the Information Exchange, Inc., in New City, New York, will begin the meeting with an overview of co-occurring mental illness and substance abuse disorders. He will also present the findings included in the SAMHSA National Advisory Council report, which he wrote. Dr. Pepper has been director of the Consultation Service of the American Psychiatric Association since 1984 and has engaged in the private practice of psychiatry since 1962. From 1975 to 1988, he served as director of the Rockland County (New York) Mental Health Center and commissioner of community services. Prior to that, he served for three years as Maryland state commissioner of mental hygiene and for four years as associate commissioner of the New York State Department of Mental Hygiene. Dr. Pepper received his medical education at the New York University School of Medicine, served his internship at the U.S. Public Health Service Hospital in Staten Island, and completed his psychiatric residency at New York State Psychiatric Institute and Rockland State Hospital, where he also served as senior psychiatrist.

Darren Skinner, M.S.W., C.S.W., C.A.S.A.C., social worker and chemical dependency counselor at Staten Island University Hospital, will follow and tell about his personal experience working with and living among people who are dually diagnosed. He will also describe the unique program of Harbor House, which provides integrated services to dually diagnosed homeless people. Mr. Skinner has over eight years of experience in the field of mental health and substance abuse treatment and serves as an independent training consultant with the Information Exchange, Inc. An adjunct professor at Audrey Cohen College in New York City, he received the 1995 Certified Addiction Counselor of the Year Award for the New York City region.

Next, Arthur Cox, D.S.W., L.C.S.W., program director of the Florida Center for Addictions and Dual Disorders, will describe his 50-bed residential treatment program for people with co-occurring mental illness and substance abuse disorders. He will focus on what it means to provide integrated treatment as well as the cost-effectiveness of such treatment. Dr. Cox is also CEO of the Mid-Florida Center for Mental Health and Substance Abuse Services, Inc., a consulting organization specializing in dual disorders. Previously, he served as superintendent of Florida’s first and only JCAHO-accredited mental health institution, the Florida Addictions Treatment Center. He has also served as interim superintendent of a mental health center in Boston and as dean of two schools of social work. Dr. Cox received his D.S.W. from Columbia University.

Stephen Mayberg, Ph.D., director of the California Department of Mental Health, will follow with a description of the challenges he faces as director of the largest state mental health agency in the nation in serving dually diagnosed patients and the steps California has taken to provide innovative services in this area. He will also tell about his recent experience negotiating with SAMHSA to secure approval to fund integrated services to dually diagnosed patients from the separate federal mental health and substance abuse block grants. Appointed to his current position in February 1993, Dr. Mayberg has responsibility for a budget of almost $2 billion as well as nearly 7,500 employees. Previously, he served as director of the Yolo County Mental Health program. Throughout his public service career, he has continued to provide clinical services. Dr. Mayberg received his doctorate in clinical psychology from the University of Minnesota, completed his internship at the University of California, Davis, and has worked in the California mental health system since that time.

Kevin Hennessy, Ph.D., health policy analyst in the Office of the Assistant Secretary for Planning and Evaluation (ASPE), DHHS, will conclude with a discussion of the public policy issues that arise from serving people with dual diagnoses as well as an analysis of the role of SAMHSA in serving this population. His work with ASPE focuses on issues of mental health, disability policy, and research. He also serves as policy liaison between the Office of the Secretary and SAMHSA. Prior to joining DHHS, he served as director of psychological services at a private psychiatric hospital in New Orleans. A licensed psychologist in Maryland, he maintains a small outpatient psychology practice. He received both a doctorate in clinical psychology and a master’s degree in public policy from the University of Rochester.

Special thanks for helpful comments on this issue brief to Kevin Hennessy, Ph.D.; Mary Knipmeyer, Ph.D.; Bert Pepper, M.D.; Clark Ross, D.P.A.; and Max Schneier, J.D. Wherever possible, the brief has accommodated their input. Nonetheless, none of them is responsible for the final content.