Addiction has been around seemingly forever. However, how we have conceptualized it, has changed considerably over time.

The release of the fifth edition of the American Psychiatric Association’s Diagnostic and Statistical Manual of Mental Disorders (DSM-5) in May 2013 marks yet another shift in how clinicians diagnose substance use disorders (SUDs). This article will trace the evolution of the criteria sets in the DSM, starting with the first edition published way back in 1952.

DSM-I

DSM-I was a small handbook consisting of only 132 pages. SUDs were included under the section on personality disorders. Within this category, “alcoholism” and “drug addiction” were considered “sociopathic personality disturbances.”

DSM-I was very light on specifics (formal criteria sets didn’t appear until DSM-III). Alcoholism was simply defined as a “well-established addiction to alcohol without [a] recognizable underlying disorder” (Washington, DC: American Psychiatric Association 1952;39). The diagnosis was apparently self-evident or left to the discretion of the treating clinician. The manual further stated that “drug addiction is usually symptomatic of a personality disorder” but didn’t indicate how one would arrive at a diagnosis.

DSM-II

DSM-II was published in 1968 and was even shorter—only 119 pages. Addiction was still considered a personality disturbance but no longer fell under the umbrella of antisocial or “sociopathic” personality disorder.

The text states that alcoholism is present when, “alcohol intake is great enough to damage [a person’s] physical health, or their personal or social functioning, or when it has become a prerequisite to normal functioning” (Washington, DC: American Psychiatric Association, 1968;45). Within this category, three distinctions were made: episodic excessive drinking, habitual excessive drinking, and alcohol addiction. The first two diagnoses hinged on how frequently intoxication occurred, while the latter was suggested by withdrawal symptoms, daily drinking, or heavy drinking.

DSM-II also introduced a category for drug dependence. It noted that the “diagnosis requires evidence of habitual use or clear sense of need for the drug.” It also recognized various classes of substances associated with addiction: opioids, barbiturates, sedative-hypnotics, cocaine, cannabis, and hallucinogens. Tobacco and “ordinary caffeine-containing beverages” were excluded, as were prescription medications, so long as “intake is proportionate to the medical need.”

DSM-III

DSM-III, a watershed document, was published in 1980. The manual ballooned to 494 pages, focused on symptoms and behaviors (rather than putative causes of mental disorders), and introduced the now-familiar format involving criteria sets.

The manual made major changes when it came to addiction. For starters, the word “addiction” was edited out of the manual. Problematic substance use was removed from the realm of personality disorders and placed in a new category called “substance use disorders.” Within this category, two criteria sets were created—“substance abuse” and “substance dependence”—which were the progenitors for everything that followed in subsequent editions of the DSM.

Substance abuse was characterized by a pattern of pathological substance use and social or occupational impairment due to the substance (Washington, DC: American Psychiatric Association, 1980;163–167). Symptoms were required to be present for at least one month. Substance dependence was considered a more severe form of addiction and required “physiological dependence, evidenced by either tolerance or withdrawal.”

DSM-III also introduced various course specifiers—the extensions to a diagnosis that further clarify the course, severity, or special features of a disorder: continuous, episodic, and remission. Continuous was described as “more or less regular maladaptive use for over six months” whereas remission represented abstinence from the target substance.

DSM-III-R

DSM-III-R was published in 1987 and provided various updates. The category for addiction was retitled “psychoactive substance use disorders” and the criteria sets for abuse and dependence were further developed.

The core features of abuse were “continued use despite knowledge of having a persistent or recurrent social, occupational, psychological, or physical problem that is caused or exacerbated by the use of the psychoactive substance” and/or “recurrent use in situations in which use is physically hazardous” (Washington, DC: American Psychiatric Association, 1987;169).

The dependence criteria set was substantially revised to include non-physiologic features. These were the larger/longer, cut down/control, time, reduced activities, and use-despite-harm criteria that were propagated to subsequent editions. Tolerance and withdrawal were no longer required to make the diagnosis.

DSM-III-R also added some dimension to dependence with severity specifiers and further qualified remission as either partial or full. Finally, it recognized three new classes of addictive substances: amphetamines, phencyclidine, and tobacco.

DSM-IV

DSM-IV was published in 1994 and resulted in some minor housekeeping. Addiction was now described as “substance-related disorders.” The manual preserved the abuse and dependence criteria in DSM-III-R with minor tweaks and some rearrangements. For example, failure to fulfill role obligations due to substance use, which was previously a feature of dependence, crossed over to abuse. Substance-related legal problems was also added as a new abuse criterion (Washington, DC: American Psychiatric Association, 1994;182).

DSM-IV also added some granularity to the remission specifiers—early versus sustained—and added course specifiers for agonist therapy (eg, methadone) and for patients residing in controlled environments (eg, prison). For reasons that aren’t entirely clear, the severity specifiers from DSM-II-R were dropped.

DSM-IV-TR

DSM-IV-TR was published in 2000. The abuse and dependence criteria sets were unchanged compared to DSM-IV.

DSM-5

The much-anticipated DSM-5 was released in late May, 2013. Depending on your perspective, the changes were either relatively modest or something close to seismic. On face value, there were more incremental refinements: the addiction category was re-titled “substance-related and addictive disorders” and abuse and dependence were collapsed into a single criteria set (Arlington, VA: American Psychiatric Association, 2013;481–589). A deeper look, however, reveals a shift from an artificial categorical construct—either abuse or dependence—to a dimensional model involving illness severity.
The combined criteria set is called “substance use disorder,” where the patient’s drug of choice replaces “substance” when a diagnosis is rendered (eg, alcohol use disorder).

The abuse and dependence criteria from DSM-IV-TR were preserved with the exception of the substance-related legal problems criterion, which was eliminated. It was replaced with a criterion for “craving, or a strong desire or urge” to use a substance.


The threshold for diagnosis is that a patient must have at least two criteria during the same 12-month period. Severity, which is described as a “general estimate,” consists of mild (2–3 criteria), moderate (4–5 criteria) and severe (? 6 criteria).

Remission specifiers also have been simplified to early and sustained (the additional qualifiers, partial and full, which were confusing and clinically dubious, were eliminated). Early remission now represents the complete absence of symptoms (except craving) for at least three months; for sustained remission, no criteria (except craving) have been met for at least 12 months.

Implications of Changes

DSM-5 should simplify the diagnosis and longitudinal care of patients with SUDs. Change, however, is never easy, and some sticky issues can immediately be anticipated.

Probably the biggest unknown is how and when third party payers—commercial insurers and government agencies—will operationalize the changes. Funding algorithms have been predicated on “abuse” and “dependence” for decades and this isn’t going to change overnight. Until payers recalibrate, clinicians could reasonably ask why they should use the new DSM-5 nomenclature.

Even if clinicians jump on board right away, they will need to reverse engineer their work for coding and billing purposes. In the United States, for example, we are still using the ninth revision of the International Classification of Diseases (ICD-9). As there are no ICD-9 codes of “alcohol use disorder,” clinicians will need to convert back to abuse or dependence depending on severity (DSM-5 recommends coding “mild” as abuse and “moderate” and “severe” as dependence).

CATR’s Take: DSM-5 reminds us of the following adage: “Be not the first by whom the new is tried, nor yet the last to set the old aside.” Although the changes related to addiction appear clinically sound, readers could reasonably defer using the new criteria until third party payers give us further direction.