Page 180 - PhD GT

(Spearman’s rho correlation coefficient .21, 2-tailed significance p<.01). However, when both sex and drug group were entered into a linear regression analysis, sex was found significantly to predict number of days per week using whereas drug group was not. This relationship was further examined in those with high dependence in the whole t1 sample (LDQ score equal to or greater than 24); sex and drug group were again entered in a regression analysis and for this sample, drug group was a significant predictor of number of days per week using while sex was not. This lends support to the suggestion that heroin users with high dependence may be more likely to use heroin on seven days of the week than are alcohol users with high dependence.
8.4.1 Changes in substance use and dependence in Sample 6c
The sample was subdivided by main substance in order to examine substance specific changes and a number of significant differences were detected over the follow-up period: for the heroin group, significant reductions in number of days per week using, in mean amount used on the heaviest day, in mean dependence score and in mean general health score were detected between t1 and t2. An additional reduction in the mean number of days per week using in the previous week between three and twelve month follow-up was found to be significant and there was a significant increase in the proportion of the sample who were abstinent at three months. There was a significant decrease in the proportion of the sample who were using on 7 days of the week and a further significant decrease in this proportion between three months and twelve months. There was an increase in the proportions using between one and three days of the week between t1 and t2 and between t2 and t3 and the proportion using between four and six days of the week was identical at the three data collection points. Reductions in quantity of heroin used on the heaviest day, in dependence, psychological ill-health and social dissatisfaction which were found between three months and twelve months did not reach significance; significant differences between t1 and t3 scores for these variables were found (see Table 8.12).
For the alcohol group, significant reductions in the mean number of days per week drinking occurred between the first data collection point (t1) and the three month follow-up (t2) and between the first data collection point and the twelve month follow-up (t3). However no significant changes occurred between the three and the twelve months follow-up and this was the case for mean dependence score and for psychological health score. Social satisfaction score, however showed no significant differences between t1 and three months, nor between three months and twelve months; however there was a significant reduction in social dissatisfaction over the whole follow-up period, that is between t1 and the twelve month follow-up. There was a reduction in amount drunk on the
175