The final sample was derived by using multiple inclusion and exclusion criteria to ensure higher levels of within-sample homogeneity. Data from male and female adult participants  were considered in the analyses if they consumed cannabis and showed at least one sign of harmful cannabis use in the past 12 months and had valid responses on all questionnaires which were included in the analyses. The six-item Cannabis Abuse Screening Test  was used to assess the presence of signs of harmful cannabis use . Only those individuals were included in the final sample who had at least 1 point on the total scale score of the CAST based on its dichotomized items . Consequently, the final sample size was 750. Table 1 presents the main sample characteristics. High proportions of the sample were males, had secondary educational attainment, worked in full-time, and did not study currently. The Hungarian translation of some of the applied questionnaires  was created in the present study in a step-by-step procedure based on the guidelines of Beaton et al.:  forward translation to Hungarian language, indoor cannabis grow system back translation of the Hungarian version to English,  comparing the original and the back translated English versions and revising the Hungarian version based on the mismatches between the two English versions,  pre-testing the clarity of the instructions and items and asking for written feedback,  finalizing the Hungarian version.

Bivariate correlations between the variables are presented in Table 2. Table 3 and Fig. 2 show the results of the mediation model. Female gender was significantly associated with elevated rates of brooding and reflection, and with lower levels of harmful cannabis use. Anxious-depressive symptoms had significant positive predictive effects on NU, reflection, brooding in addition to the significant and positive direct effect of anxious-depressive symptoms on harmful cannabis use. Anxious-depressive symptoms were also significantly and negatively linked to CRSE and CPBS. A significant and negative relationship was presented between brooding and cannabis use frequency. Reflection had positive predictive effects on CPBS and CRSE. NU was significantly and negatively related to CRSE and cannabis use frequency. Higher levels of CRSE and CPBS were significantly associated with less frequent and less harmful cannabis use. Total, direct and indirect effects are summarized in Table 4. In the case of cannabis use frequency, indirect effects were not estimated as the total effect of anxious-depressive symptoms on cannabis use frequency was non-significant.

Overall, five significant indirect effects were identified on harmful cannabis use. First, anxious-depressive symptoms had a negative effect on CPBS which in turn was negatively linked to harmful cannabis use. Second, higher levels of anxious-depressive symptoms were associated with lower levels of CRSE which subsequently had a negative effect on harmful cannabis use. Third,cannabis grow set up a double-mediation pathway indicated that anxious-depressive symptoms had a positive effect on reflection which in turn was positively associated with CPBS which subsequently contributed to lower levels of harmful cannabis use. Fourth, in a double-mediation pathway, higher levels of anxious-depressive symptoms were associated with increased rates of reflection which subsequently had a positive effect on CRSE which in turn was negatively associated with harmful cannabis use. Fifth, in a double-mediation pathway anxious-depressive symptoms were positively associated with NU which had a negative link with CRSE and higher levels of CRSE were associated with lower rates of harmful cannabis use. The present study aimed to examine the mediating functions of rumination, negative urgency , and constructs of cannabis use regulation  on the associations of anxious-depressive symptoms with cannabis use frequency and harmful cannabis use.

To the Authors’ best knowledge, this is the first time that a comprehensive mediation model was tested which simultaneously covers anxious-depressive symptoms, constructs of ER and cannabis use regulation and outcomes of cannabis use. It is important to highlight that the present sample was consisted of cannabis users showing signs of harmful cannabis use. Moreover, due to the recruitment procedure of the study, it might be possible that the self-selection of those individuals was presented who show high interest and motivation in cannabis- and illicit drug use. These characteristics can indicate that regular and problematic cannabis users might be over represented in the sample which might influenced the levels of other constructs as well. For example, it might be possible that the associations with CPBS and CRSE would have been different by having more recreational users in the sample with less frequent and less harmful consumption patterns .