Applicability of DSM Criteria to Nicotine Dependence

The current DSM (i.e. DSM-IV) criteria for substance dependence arose from observations and theory regarding alcohol and opioid dependence [1]. Nicotine dependence differs from alcohol and opioid dependencies in several respects [2,3]. Nicotine use does not cause behavioral intoxication and, thus, the major harm from nicotine dependence is from physical, not behavioral or social, problems. As a result, many of the proposed generic DSM-V criteria (http://www.dsm5.org) are rarely endorsed by regular tobacco users (e.g. use causing failure to fulfill role obligations, use in physically hazardous situations and use despite social problems) [2]. In addition, the DSM-IV criteria possess less validity in predicting key consequences of nicotine dependence (e.g. relapse) than do nicotine-specific measures such as time-to-first-cigarette after awakening and cigarettes smoked/day (cigs/day) [4,5]. These simpler measures have good predictive validity; i.e. they predict ability to quit and response to treatment [4] and, surprisingly, have very little overlap with the DSM criteria [4,6]. Consequently, DSM criteria are very rarely used in nicotine dependence research; e.g. the US Public Health and Human Services (USPHS) tobacco treatment guidelines do not even mention the DSM [7]. One possibility is to abandon the use of generic criteria and substitute nicotine-specific criteria [2]. However, the number of tests comparing the DSM and non-DSM criteria has been limited [4], and therefore such a dramatic change is probably premature at this point. The proposed DSM-V includes changes that should increase the validity of the generic criteria when applied to nicotine dependence; i.e. the inclusion of craving (a robust predictor of ability to stop smoking) [8] and the ratings of a severity dimension, recognizing that dependence can occur on a continuum (http://www.dsm5.org) [3]. We suggest additional changes: Include measures of time-to-first-cigarette and cigs/day [5] in DSM field trials to allow comparisons with DSM criteria and consider parallel measures of urgency and heaviness of use for other drug dependencies; Include craving as a nicotine withdrawal symptom and include it in other drug withdrawal syndrome descriptions because withdrawal-associated craving is highly predictive of clinical outcomes [8]; Change the name from ‘nicotine use disorder’ to ‘tobacco use disorder’, because non-tobacco nicotine dependence is very rare [9] and the current terminology perpetuates misperceptions of smokers and regulators that tobacco products and nicotine medications are equally addictive [10]. The current DSM criteria for substance dependence were based not on existing data sets but on a ‘provisional description’ that was ‘a concept as stimulus to enquiry’[1,11]. No substantially new items have been added in the last 30 years (e.g. compare DSM-III-R and proposed DSM-V). We believe it is unlikely that the DSM found the optimal criteria in 1980; thus, we suggest research on the inclusion of completely new items. Several multi-item tobacco dependence scales have been developed and undergone extensive testing in the last decade [4] that include several items that appear valid but are not included in DSM-V. In addition, novel items such as awakening at night to smoke have been tested and appear to be valid [12]. One reason that the acceptance of DSM assessment of tobacco dependence has languished is that the existing criteria have rarely been tested against explicit, clinically meaningful outcomes [4]. Therefore, we recommend that research validates DSM and alternative criteria against benchmarks that are clear and clinically relevant; e.g. success at a quit attempt or response to dependence-based treatment. If future comparisons continue to show that these alternative scales/items outperform the DSM criteria, we suggest that the better-validated items be adopted, independent of whether they were included in the original conceptualization for DSM and independent of whether they apply to other substances. We believe such a program could produce the most empirically based, accurate and useful criteria to determine nicotine dependence. John Hughes and Saul Shiffman have received grants and consulting fees from many non-profit and for-profit organizations that develop or promote smoking cessation products or services. Tim Baker has served as an investigator on research sponsored by various pharmaceutical companies. Naomi Breslau has no interests to declare. Lirio Covey has received a grant from a for-profit organization that develops and promotes smoking cessation products. None.