**6. Conclusions**

As far as substance used is concerned, our findings are consistent with the literature and with clinical observations, especially regarding the increased use of cannabinoids, cocaine, and "other" drugs on the one hand and the decreased consumption of heroin on the other.

In both 2-year periods, female gender and being ≥61 years old appear to be associated with a decreased risk of SMI-SUD comorbidity. Both results are consistent with the existing litera‐

In the 2-year period 2003–2004, having a university degree was associated with a decreased risk of comorbid SMI-SUD, whereas having a diagnosis of personality disorder was associat‐ ed with an increased risk of comorbid SMI-SUD, but 10 years later we found educational level and diagnosis having no impact on comorbidity. As already described above, this may suggest possible changes in the pattern of SUD as far as problematic family issues are concerned; notwithstanding the fact that, in 2013–2014, educational level and diagnosis no longer represented risk factors, it would have been interesting to assess whether individuals with different cultural levels as assessed by schooling of with different diagnosis share the same pathways towards SUD and similar choices regarding type of substance and use. As far as educational level is concerned, in a recent study, we found that, although having a universi‐ ty degree was associated with a decreased risk of DD for males, it was associated with an increased risk of DD in females [65]. We hypothesized a different pattern of social function‐ ing and performance in male and female SMI-SUD patients [44] and that males and females may access substances via different pathways and choose different types of substances as

As far as diagnosis is concerned, the same study mentioned above, which assessed the period 2003–2012, found affective and "other" disorders associated with an increased risk of comorbid SUD, compared to personality disorders, which according to Baigent [94] would be more likely than Axis I disorders to be associated with chronic SUD. On the contrary, reports from the literature show mixed results about this issue, and recent studies suggest that the frequency of comorbid SUD is similar in schizophrenic psychoses and in personality disor‐ ders [37] and that primary mood and/or anxiety disorders are at high risk for comorbid SUD

Last, in the 2-year period 2013–2014, we found that living in therapeutic rehabilitation centers was associated with a decreased risk (compared to living alone) of comorbidity with SUD. This result is encouraging and may support the effectiveness of such therapeutic settings in

Some limitations should be underscored. The retrospective design and data gathering through clinical charts entail some limitations. Some information could not be retrieved, for example,

Surprisingly, we found polyabuse to be relatively stable even after 10 years.

182 Recent Advances in Drug Addiction Research and Clinical Applications

ture and could be expected according to clinical experience [66,85–87,91].

well [99–102], with a variable impact on their lives.

protecting patients from exposition and/or relapse into SUD.

**4.3. Multivariate analysis**

as well [96,97].

**5. Limitations**

This study adds to the scant literature about this issue in our country, and the large sample size is a strength of this research. Both SMI and SUD are predictors of underachievement and failure in educational and occupational settings, difficulty facing family responsibilities, violent and abusing behaviors, poverty, legal problems, and scarce compliance to treatment [103]. Acute settings may be particularly appropriate for the development of targeted interventions [104], and the treatment of patients with comorbid psychiatric disorders and SUD should begin early during hospitalization [105].

Changes in the pathways leading to drug abuse and in the patterns of addiction should not be overlooked.
