O'Brien and colleagues in the DSM-V workgroup have proposed draft diagnostic criteria for DSM-V Substance use disorders (SUDs) [1]. These draft criteria eliminate the diagnosis of ‘substance abuse’ and define a single SUD for various substance classes—such as ‘alcohol use disorder’ (AUD)—using a set of 11 symptoms. This commentary reviews the draft criteria in light of their application to adolescents.

Adolescence is a developmental period characterized by relatively high rates of substance use, as well as appreciably high rates of substance use disorders. Based on the 2008 NSDUH data set, 7.6% of 12–17-year-olds met criteria for at least one DSM-IV SUD in the past year, and an additional 17% of teenagers, while not meeting an abuse diagnosis, still reported one or two substance-dependence criteria [2]. As with adults, a SUD diagnosis is often needed for youth to be eligible to receive treatment for a substance use problem within the US health-care system.

The proposed DSM-V criteria include some positive changes for defining SUDs among adolescents. The use of a combined criterion set to diagnose a single SUD makes sense for adolescents. Specifically, the criterion sets for DSM-IV abuse and dependence have overlapping conceptual content [3], and do not differ systematically in prevalence, sensitivity, specificity, severity or age of onset [4,5]. Factor and latent class analyses indicate a single dimension of substance problems [4,6,7]. We also support the decision to eliminate the ‘legal problems’ symptom which, in adolescents, was related to coexisting conduct disorder, and tended to be less relevant for female and younger teenagers [4]. These changes have the potential to strengthen the validity of the SUD diagnosis for youth.

Potential limitations with proposed DSM-V SUD

There are, nonetheless, a number of limitations of the proposed changes. Specifically, some of the draft criteria have questionable validity when applied to adolescent substance users.


Tolerance to substances, particularly to alcohol, is a relatively mild symptom and can present without significant harm or distress. Indeed, it may be normative in adolescent and young adult drinkers [8], when individuals typically escalate their substance use from experimentation to more regular use. Tolerance implicitly references an undefined temporal comparison which makes it a developmentally sensitive criterion, such that it is much easier to meet this criterion earlier in the career of the substance user [9]. Also, neurodevelopmental changes during adolescence [10] may contribute to variability of developmental sensitivity to substances among teenagers who are only a few years apart in age [11].


Withdrawal symptoms are a fairly rare phenomenon in adolescents because they generally only emerge after years of heavy drug use [12], and withdrawal reflects severe substance problems and may have prognostic significance in the few adolescents who report it so its inclusion in DSM-V is useful. However, alcohol withdrawal has shown relatively high rates of endorsement and only a moderate association with levels of problem severity [6]. Also, the DSM-IV definition of alcohol withdrawal requires only two of eight subcriteria, some of which may be confused with symptoms of hangover, producing false positive symptoms.

Hazardous use

There is debate as to whether hazardous use reflects a compulsive pattern of use that is relevant across all substances, such as tobacco [13]. In addition, the hazardous use criterion is somewhat developmentally bound. Hazardous use is much more common in adults than among adolescents [2,4], probably because adolescents have less access to automobiles than adults (by far the most common way for adolescents to meet this criterion).


The value of including the craving criteria when diagnosing adolescents is uncertain. There are clinical observations that many youth who have escalated to regular use report strong cravings [2,14], but the way in which this symptom is defined and operationalized may have a large effect on its observed prevalence (e.g. sign of continued appetitive urges or behaviors/withdrawal-related phenomenon?). More research is needed to examine the validity of craving as a criterion of SUDs among adolescents.

The two-symptom threshold for SUD

The impact of the two of 11 threshold for SUD in youth deserves further research. Because some of the proposed DSM-V symptoms are mild, developmentally normative for teenagers and/or easily misunderstood and over-endorsed [15], the proposed algorithm may include many mild cases, which do not reflect the classic definition of a compulsive pattern of substance use. This may unnecessarily apply a stigmatized and loaded label to youth whose problem severity may be mild and whose substance use pattern may be more intermittent than regular and more likely to remit. A related problem is that the proposed threshold for SUD may include identify a very heterogeneous group of adolescents.


Despite some favorable changes, the proposed DSM-V SUD criteria do not go far enough in improving SUD diagnosis for youth. Some targeted developmental adjustments, including clearer operational symptom definitions [16], are needed to ensure that DSM-V validly identifies SUD across developmental periods. We recommend excluding the ‘hazardous use’ criterion, clarifying the operational definitions of ‘tolerance’, ‘withdrawal’ and ‘craving’, and critically evaluating the diagnostic threshold for SUDs in adolescents.

Declaration of interest