Finding order in the DSM-IV nicotine dependence syndrome: a Rasch analysis.

Little is known about the relative severity or typical sequence of Diagnostic and Statistical Manual (DSM-IV) symptoms of nicotine dependence. Using data from the National Comorbidity Study (NCS), the current study used a Rasch logistic item response model to assess the unidimensionality of the construct of nicotine dependence, to identify which symptoms are associated with high levels of dependence, and to determine whether the dependence symptoms and typical symptom patterns are influenced by age, gender, race and income level. The nicotine dependence symptoms provided reasonable coverage of the dependence syndrome and can be used to scale individuals reliably. However, attention to potential demographic differences in responding to particular symptoms is needed. Furthermore, additional symptoms that describe behavioral changes that support continued smoking would be useful in helping to map the lower and higher ranges of dependence continuum. Evaluation of the typical pattern of responses to the nicotine dependence symptoms suggested that most smokers responded in a predictable manner. However, the tendency for younger smokers to provide more atypical responses than other smokers suggests a need to develop more specialized items for this population. More precise assessment of nicotine dependence would improve the predictive value of these measures and ultimately help inform nicotine dependence treatments.