Изображения страниц
PDF
EPUB

CHAPTER

2

PSYCHIATRIC COMORBIDITY WITH ALCOHOL USE DISORDERS

Introduction

The term "comorbidity" refers to the pres

ence of two or more illnesses-medical

or psychiatric conditions, including alcohol and other drug use disorders—in the same person. Over the last decade, comorbidity involving psychiatric and alcohol use disorders has generated intense interest. This interest is fueled by the idea that understanding comorbidity will lead to the development of better etiology, prevention, and treatment models for alcohol use and psychiatric disorders, either as single entities or as component illnesses.

Feinstein (1970) was the first to elaborate on the idea that the presence of two or more disorders in the same person may create hybrid clinical and conceptual dilemmas. Using cancer as an index condition, he systematically outlined ways that other medical conditions can affect diagnostic findings, point of illness detection, illness course, treatment planning, treatment outcome, and mortality risk associated with the index diagnosis. Feinstein cautioned that appropriate treatment and study of an index condition are compromised when comorbid conditions are neglected.

This chapter provides an overview of the main issues and empirical data dealing with the comorbidity of alcohol and psychiatric disorders. Among the topics discussed are the prevalence of comorbidity, models that link alcohol disorders with other psychiatric disorders, implications of comorbidity for disorder course, and treatment issues.

The term "comorbidity" refers to the presence of two or more illnesses-medical or psychiatric conditions, including alcohol and other drug use disorders-in the same person.

Prevalence of Comorbidity

Interpretational Issues

Epidemiologic surveys assessing the presence of alcohol and other psychiatric disorders in circumscribed samples offer a means of estimating the extent of comorbidity in specific populations. Before considering these studies, however, it is useful to examine the methodological factors that can affect estimates, including the definition of significance, diagnostic methods, and sampling strategies.

Determining significance

In studies measuring the extent of comorbidity, "statistical significance" refers to the finding that the presence of one disorder increases the risk for a second disorder that goes beyond the risk for the latter when the former is absent. However, if risk for alcoholism is the same whether a psychiatric disorder such as depression is present or not, cases involving co-occurrence of these disorders would be attributed to the chance overlap of unrelated disorders.

The "clinical significance" of comorbidity refers to the extent to which the combined presence of comorbid conditions affects the clinical course or optimal treatment for either disorder. Even if a comorbid association occurs by chance, the association can be clinically significant if the co-occurring disorder alters the clinical course or optimal treatment of either condition. Clinical significance is an important consideration for more common psychiatric conditions, such as anxiety and affective disorders: Such conditions would be expected to co-occur with alcohol use disorders more often by chance alone than would less common psychiatric disorders, such as schizophrenia.

Diagnostic issues

Several factors that relate to assigning diagnoses can affect resulting comorbidity rates. Various studies have demonstrated that comorbidity rates differ with the use of different diagnostic methods clinical interview, research interview, chart review, and clinical consensus (Drake et al. 1990; Ford et al. 1989; Griffin et al. 1987; Hasin and Grant 1987). Although identifying one diagnostic method over others that produces a “correct" diagnosis is unjustifiable, it is generally agreed that structured diagnostic interviews produce the most reliable diagnostic results.

A potential source of error in estimating comorbidity rates is the overlap between diagnostic criteria for alcohol use and other psychiatric disorders. For example, there is overlap between several Diagnostic and Statistical Manual of Mental Disorders, Third Edition, Revised (DSM-III-R) diagnostic criteria for alcohol abuse and antisocial personality disorder (ASPD). This feature of the DSM-III-R diagnostic scheme can potentially lead to inflated comorbidity rates if various overlapping symptoms are counted toward a diagnosis for each comorbid disorder.

Findings that suggest that alcohol and other drug use disorders can cause symptoms that resemble psychiatric syndromes may be another potential source of inflated comorbidity rates. Some researchers (Blankfield 1986; Brown et al. 1991) emphasize that such symptoms are alcohol related and transient and should not constitute independent psychiatric syndromes or comorbid diagnoses. It is a diagnostic challenge to differentiate symptoms associated with transient alcohol effects from those associated with an independent psychiatric disturbance.

Assessments of the comorbidity rate may also be compromised by variations in the willingness

or ability of people to provide accurate information about psychiatric or alcohol-related symptoms (Griffin et al. 1987; Ridgely et al. 1990). Surveys that do not control for these methodological concerns run the risk of adding error to the resulting comorbidity rate estimates.

Sampling strategy

Beyond issues related directly to diagnosing comorbid disorders, population samples also affect comorbidity rates. Studies relevant to comorbidity can be divided into those conducted using patient samples and those conducted using community (general population) samples. The effect of sample population on comorbidity rates is considered in Berkson's (1949) fallacy: Because people with two or more disorders are more likely to enter treatment than those with one disorder, comorbidity rates tend to be higher in samples of patients in treatment settings. Moreover, different patient populations do not necessarily have equivalent comorbidity rates.

Comorbidity Prevalence From the Epidemiologic Catchment Area Survey

The Epidemiologic Catchment Area (ECA) survey, which surveyed 20,000 individuals in community and institutional settings at five sites across the country, is an important source of data on the prevalence and incidence of comorbidity in the general U.S. population (Helzer and Pryzbeck 1988; Regier et al. 1990). Although the ECA survey is larger than previous communitybased surveys, it is not free from methodological difficulties. For example, the derived comorbidity rates may be inflated by such methodologies as attributing the same symptoms to both comorbid conditions and defining transient symptoms (depression, anxiety) produced by alcohol withdrawal as independent psychiatric syndromes. In addition, some of the data should be interpreted with caution because estimates of comorbid conditions were based in part on combined data from the five sites-a procedure that is statistically problematic.

Sequencing criteria in the ECA study consisted of age at which the first symptom of alcohol and other drug use disorders appeared rather than age at onset of the syndrome. The ECA defined alcohol use disorders as the occurrence of enough symptoms to meet the associated diagnostic criteria over the life course. The sporadic occurrence of isolated symptoms, per

haps years apart, provides an insufficient basis for testing competing hypotheses related to comorbidity. Accordingly, this methodology should be considered when interpreting conclusions about the causal relationship between comorbid psychiatric disorders and alcohol and other drug use disorders that are based on ECA data.

General comorbidity rate

ECA survey findings suggest that the rate of comorbid psychiatric-alcohol disorders significantly exceeds rates that would be expected by chance associations alone (Helzer and Pryzbeck 1988; Regier et al. 1990). As shown in figure 1, 19.9 percent of the survey sample had at least one psychiatric disorder. Among those with alcohol disorders, the rate of psychiatric disorders increased to 36.6 percent; 13.5 percent of the sample received alcohol-related diagnoses (figure 2). This number increased to 22.3 percent among those with a psychiatric diagnosis. These findings indicate that the presence of either an alcohol or psychiatric disorder nearly doubles the risk for a comorbid diagnosis.

Comorbidity rates for specific psychiatric disorders

The ECA also provides information about specific types of psychiatric disorders with which alcohol can be comorbid. As shown in table 1, antisocial personality disorder (ASPD) has the strongest association with alcohol disorders (21fold increase in risk), followed by schizophrenia (3.3-fold increase in risk), affective disorders (1.9fold increase in risk), and anxiety disorders (1.5fold increase in risk). When expressed as an odds ratio,1 anxiety disorders show the smallest association with alcohol disorders (1.5). However, anxiety disorders show the greatest association with alcohol use disorders when expressed as a proportion of respondents with alcohol abuse and dependence diagnoses who have comorbid anxiety diagnoses (19.4 percent). This apparent contradiction in the data is explained by the fact that, compared with other assessed disorders, anxiety disorders are common whether or not an alcohol disorder is present. Table 1 shows that odds ratios remain the same whether they describe the change in risk for an alcohol

[blocks in formation]

1 An odds ratio represents the relative risk of a comorbid disorder, given that an alcohol use disorder is present, compared with the risk of the comorbid disorder occurring in the population as a whole. A proportion of comorbidity indicates the number of cases of a comorbidity disorder per 100 cases of individuals with an index diagnosis.

Table 1. Lifetime comorbidity of alcoholism and psychiatric disorders per 100 respondents in a community sample.*

[blocks in formation]

* Table data based on the combined five-site Epidemiologic Catchment Area (ECA) data weighted to the U.S. population (Helzer and Pryzbeck 1988; Regier et al. 1990 from JAMA 264(19):2511-2518).

+ Ratio of the odds of having a psychiatric disorder (rows) for those with an alcohol disorder (second column) to the odds for those in the total community sample (first column).

+ 13.5 percent represents the proportion of the total ECA community sample with a lifetime alcohol abuse or dependence diagnosis.

§ Ratio of the odds of having an alcohol disorder for those with a row psychiatric diagnosis (second column) to the odds for those in the total community sample (first column).

versus 5 percent); drug use and dependence; panic disorder (7 percent versus 2 percent); and mania (4 percent versus 1 percent). However, referring to current (versus lifetime) psychiatric diagnoses, Robins et al. (1989) reported that psychiatric disorders were slightly more prevalent among women with alcohol disorders than among men with alcohol disorders.

It is important to note that gender base-rate differences for specific psychiatric disorders in the ECA (Eaton et al. 1989) affect comorbidity findings. Alcohol disorders are more common in males than in females (about five to one in the ECA survey). It may be expected that female alcoholics would be more extreme than their male counterparts on various severity measures, including comorbidity. For example, although more male alcoholics than female alcoholics were diagnosed with ASPD, a female's risk for ASPD associated with an alcohol diagnosis is three times greater than for males. This effect also must be considered regarding conditions. such as depression and phobic disorders, which are more common in females in the general population.

Drugs and comorbidity rate

Helzer and Pryzbeck (1988) reported that 18 percent of the ECA sample diagnosed with an alcohol use disorder were also diagnosed with another drug use disorder. In contrast, respondents who did not meet diagnostic criteria for an alcohol disorder had lower rates of other drug

use disorders-3.5 percent, or a sevenfold difference in risk. Drug abuse appears to affect one's risk for psychiatric comorbidity; more than 50 percent of ECA respondents diagnosed with a nonalcohol drug disorder were also diagnosed with a non-substance abuse mental disorder. This finding contrasts with findings on respondents not meeting diagnostic criteria for a drug disorder; these individuals had lower rates of mental disorder (20 percent, an odds ratio of 4.5) (Regier et al. 1990).

These findings suggest that, in the community, alcohol disorders are complicated by nonalcohol drug abuse in about one-fifth of cases. Psychiatric comorbidity is likely to be inflated in this subgroup because alcohol and other drug disorders seem to increase the risk for a psychiatric diagnosis. It is not known to what extent risk factors for psychopathology associated with alcohol and other drug disorders overlap and, if independent, whether they operate additively or synergistically in those with multiple substance abuse diagnoses.

Comorbidity in Patient Populations

Comorbidity in community versus
institutional samples

As predicted by Berkson's fallacy, risk for comorbidity is generally greater in patient and institutional samples than in community samples.

Regier et al. (1990) reported that alcoholdisordered ECA respondents sampled from mental hospitals, nursing homes, and prisons had 3.8 times the risk for a current (i.e., 6-month) comorbid psychiatric diagnosis as did alcohol-disordered respondents sampled from a community setting (55 percent versus 24.4 percent) (figure 3). A further increase in current comorbidity was found in those sampled from an addiction treatment setting (65 percent) (Ross, Glaser, and Germanson 1988). Based on comparisons in figure 3, the risk for having a current comorbid psychiatric diagnosis in the addiction treatment sample is 5.8 times greater than that noted for alcohol-disordered ECA community respondents.

Comorbidity in specific patient samples

Over the past 10 years, several studies have assessed the comorbidity rate in patient samples. Most of these studies have focused on the rate of psychiatric disorders among substance abusers (Mirin et al. 1988) or the rate of substance abuse among general psychiatric patients (Galanter et al. 1988). Studies in these populations can be further subdivided. For example, Galanter et al. (1988) reviewed studies that assess comorbidity in emergency room settings, psychiatric acute inpatient services, veterans' hospitals, and public and private psychiatric hospitals. The consistent finding in each study is that comorbidity rates exceed base-rate (chance) expectations.

In addition to evaluating patient comorbidity based on place of presentation, other researchers have emphasized a specific category of psychiatric disorder in their reviews of the comorbidity literature. Kushner et al. (1990) reviewed comorbidity literature involving anxiety disorders; Mueser et al. (in press) and Schneier and Siris (1987) reviewed comorbidity studies involving schizophrenia; Goodwin and Jamison (1990) reviewed studies involving affective disorders; O'Farrell (1990) reviewed studies relevant to sexual functioning of male alcoholics; and Keane et al. (1988) reviewed studies relevant to alcoholism among those with posttraumatic stress disorder. Each psychiatric condition considered shows a significant statistical association with alcohol disorders.

Comorbidity in Addiction Treatment Patients

Because of the broad clinical implications, comorbidity issues germane to patients receiving treatment for alcohol-related disorders are of spe

cial interest. As discussed above, such factors as population sampled, diagnostic approach, specific psychiatric disorder, gender, and extent of drug abuse seem to affect comorbidity rates. Ross, Glaser, and Germanson (1988) and Ross, Glaser, and Stiasny (1988) published a study that goes beyond earlier studies in adequately accounting for these factors in the assessment of comorbidity in an addiction treatment sample. However, because Ross, Glaser, and Germanson (1988) assessed comorbidity in recently detoxified alcoholics, the possibility exists that withdrawal and other alcohol-related symptoms were confused with bona fide psychiatric syndromes.

[merged small][merged small][merged small][merged small][merged small][merged small][ocr errors][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small]
« ПредыдущаяПродолжить »