Page 78 - PhD GT

opiate using behaviour is narrow to begin with and possibly not likely to show much variation across different levels of severity. On the other hand, it may be that the question as posed was unable to tap this element of the syndrome. This question will be addressed in the present study with the differently worded question designed to measure this element in the Leeds Dependence Questionnaire.
Of particular interest in the debate on the nature of dependence is the fact that, in analysis of the internal structure of the questionnaire, two almost unrelated factors emerged, one accounted for 21% of the variance and the other for 17% while the next factor accounted for only 4%.
“All the items relating to withdrawal symptoms, both physical and affective, load on one factor; all the cognitive-behavioural items relating to withdrawal relief drug taking load on the other.” (Phillips et al. 1987 p. 693)
The authors offer three possible explanations: close examination of the relationship between the average scores on each of the three withdrawal symptoms and withdrawal relief sections revealed that though physical and affective withdrawal items correlate highly (as one would expect), a weaker relationship existed between physical withdrawal and withdrawal relief, and no relationship at all between affective withdrawal and withdrawal relief. Consistently high scoring on two of the withdrawal relief items may have accounted for the lack of a correlation as those withdrawal relief items showing more variability in the responses had a slightly higher correlation with physical withdrawal. Other possible explanations explored were to do with cultural differences and different interpretations of items in different cultural contexts. What about the possibility that withdrawal symptoms are withdrawal symptoms and cognitive-behavioural items are dependence?
A further validation study was conducted with another U.S. sample of 126 opiate addicts (Sutherland et al. 1988) and results found in the first validation study, conducted with a similar US sample, were, by and large, replicated. The relationship found between the overall SODQ score and the overall OSDQ (the five subjective items scale) score was found to be weaker, leading the authors on this occasion to wonder whether there was a separate dimension of withdrawal phenomena that were closely related to the behaviours triggered by them but might not be so closely related to cognitions and behaviours described as altered behavioural and altered subjective states in the syndrome formulation (Edwards et al. 1977 pp. 12-13) and later distinguished more explicitly (Edwards et al. 1982). Additionally the authors pointed out that, since British addicts had
66