Diagnostic criteria for alcohol abuse and dependence have evolved over time. As new data become available, researchers revise the criteria to improve their reliability, validity, and precision (4,5). This Alcohol Alert traces the evolution of diagnostic criteria for alcohol abuse and dependence through the current standards of the American Psychiatric Association's Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition (DSM-IV) (6). For comparison, the criteria found in the World Health Organization's International Classification of Diseases, Tenth Revision (ICD-10) also are reviewed briefly, although these are not often used in the United States (7).

Evolution of Diagnostic Criteria

Early Criteria

At least 39 diagnostic systems had been identified before 1940 (2). In 1941 Jlinek first published what is considered a groundbreaking theory of subtypes of what was, until 1980, termed alcoholism (2,8). Jellinek associated these subtypes with different degrees of physical, psychological, social, and occupational impairment (2,9).

Formulations of diagnostic criteria continued with the American Psychiatric Association's publication of the Diagnostic and Statistical Manual of Mental Disorders, First Edition (DSM-I), and Second Edition (DSM-II) (10,11). Alcoholism was categorized in both editions as a subset of personality disorders, homosexuality, and neuroses (2,12).

In response to perceived deficiencies in DSM-I and DSM-II, the Feighner criteria were developed in the 1970's to establish a research base for the diagnostic criteria of alcoholism (5,13). These criteria were the first to be based on research rather than on subjective judgment and clinical experience alone (5). Though designed for use in clinical practice, they were primarily developed to stimulate continued research for the development of even more useful diagnostic criteria (5). Several years later, Edwards and Gross focused solely on alcohol dependence (8). They considered essential elements of dependence to be a narrowing of the drinking repertoire, drink-seeking behavior, tolerance, withdrawal, drinking to relieve or avoid withdrawal symptoms, subjective awareness of the compulsion to drink, and a return to drinking after a period of abstinence (8)

The DSM Criteria

Researchers and clinicians in the United States usually rely on the DSM diagnostic criteria. The evolution of diagnostic criteria for behavioral disorders involving alcohol reached a turning point in 1980 with the publication of the Diagnostic and Statistical Manual of Mental Disorders, Third Edition (14). In DSM-III, for the first time, the term "alcoholism" was dropped in favor of two distinct categories labeled "alcohol abuse" and "alcohol dependence" (1,2,12,15). In a further break from the past, DSM-III included alcohol abuse and dependence in the category "substance use disorders" rather than as subsets of personality disorders (1,2,12).

The DSM was revised again in 1987 (DSM-III-R) (16). In DSM-III-R, the category of dependence was expanded to include some criteria that in DSM-III were considered symptoms of abuse. For example, the DSM-III-R described dependence as including both physiological symptoms, such as tolerance and withdrawal, and behavioral symptoms, such as impaired control over drinking (17). In DSM-III-R, abuse became a residual category for diagnosing those who never met the criteria for dependence, but who drank despite alcohol-related physical, social, psychological, or occupational problems, or who drank in dangerous situations, such as in conjunction with driving (17). According to Babor, this conceptualization allowed the clinician to classify meaningful aspects of a patient's behavior even when that behavior was not clearly associated with dependence (18).

The DSM was revised again in 1994 and was published as the Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition (DSM-IV) (6). The section on substance-related disorders was revised in a coordinated effort involving a working group of researchers and clinicians as well as a multitude of advisers representing the fields of psychiatry, psychology, and the addictions (2). The latest edition of the DSM represents the culmination of their years of reviewing the literature; analyzing data sets, such as those collected during the Epidemiologic Catchment Area Study; conducting field trials of two potential versions of DSM-IV; communicating the results of these processes; and reaching consensus on the criteria to be included in the new edition (2,19).

DSM-IV, like its predecessors, includes nonoverlapping criteria for dependence and abuse. However, in a departure from earlier editions, DSM-IV provides for the subtyping of dependence based on the presence or absence of tolerance and withdrawal (6). The criteria for abuse in DSM-IV were expanded to include drinking despite recurrent social, interpersonal, and legal problems as a result of alcohol use (2,4). In addition, DSM-IV highlights the fact that symptoms of certain disorders, such as anxiety or depression, may be related to an individual's use of alcohol or other drugs (2).

The ICD Criteria

While the American psychiatric community was formulating its editions of diagnostic criteria for mental disorders, the World Health Organization was developing diagnostic criteria for the purpose of compiling statistics on all causes of death and illness, including those related to alcohol abuse or dependence, worldwide (1,4,20). These criteria are published as the International Classification of Diseases (ICD). The first ICD classification of substance-related problems, published in 1967 in ICD-8 (21), classified what was then called alcoholism with personality disorders and neuroses, as had DSM-I and DSM-II. In ICD-8, alcoholism was a separate category that included episodic excessive drinking, habitual excessive drinking, and alcohol addiction that was characterized by the compulsion to drink and by withdrawal symptoms when drinking was stopped (1).

Although ICD-9 (22,23) included separate criteria for alcohol abuse and dependence, this revision defined them similarly in terms of signs and symptoms (1). According to Babor, an important assumption in ICD-9 was that alcohol use in the absence of dependence "merits a separate category by virtue of its detrimental effects on health" (1, p. 87).

The category of alcohol dependence was central to the current revision, ICD-10 (1,2,7). Alcohol dependence is defined in this classification in a way that is similar to the DSM. The diagnosis focuses on an interrelated cluster of psychological symptoms, such as craving; physiological signs, such as tolerance and withdrawal; and behavioral indicators, such as the use of alcohol to relieve withdrawal discomfort (1). However, in a departure from the DSM, rather than include the category "alcohol abuse," ICD-10 includes the concept of "harmful use." This category was created so that health problems related to alcohol and other drug use would not be underreported (1). Harmful use implies alcohol use that causes either physical or mental damage in the absence of dependence (1).

Moving Toward Agreement Between Diagnostic Criteria

The DSM diagnostic criteria for psychiatric disorders are the criteria primarily used in the United States. The ICD is an international diagnostic and classification system for all causes of death and disability, including psychiatric disorders (4). Earlier editions of these two major diagnostic criteria dealing with alcohol abuse and dependence were criticized for being too dissimilar (2). Therefore, the DSM-IV and the ICD-10 were revised in a coordinated effort among researchers worldwide to develop criteria that were as consistent with one another as possible (1,2).

Although some differences between the two major diagnostic criteria still exist, they have been revised by consensus as to how alcohol abuse and dependence are best characterized for clinical purposes (18). Clinicians, international health agencies, and researchers are now better able to categorize people with alcohol dependence, abuse, and harmful use to plan treatment, collect statistical data, and communicate research results (18).

Diagnostic Criteria--A Commentary by
NIAAA Director Enoch Gordis, M.D.

The research community has long found standardized diagnostic criteria useful. Such criteria provide agreement as to the constellation of symptoms that indicate the alcohol dependence syndrome and allow researchers all over the world to communicate clearly as to what kinds of disorders are being studied.

Standardized diagnostic criteria are equally important and useful to clinicians. In the alcohol field, there have been many different ways by which clinical staff might arrive at a diagnosis--sometimes differing among staff within the same program. Although the use of standard diagnostic criteria may seem somewhat burdensome, it provides many benefits: more efficient assessment and placement, more consistency in diagnoses between and within programs, enhanced ability to measure the effectiveness of a program, and provision of services to people who most need them. As we move more and more into a managed health care arena, third-party payors are requiring more standardized reporting of illnesses; they want to know what conditions they are paying for and that these conditions are the same from program to program. The standardized diagnostic criteria presented in this Alert are based on the newest research, have been developed based on field trials and extensive reviews of the literature, and are continually revised to reflect new findings. Although clinical judgment will always play a role in diagnosing any illness, alcohol treatment programs that use standardized diagnostic criteria will be in the best position to select appropriate treatment and to justify their selection to third-party payors.