To assess the statistical heterogeneity between studies, we used the I2 statistic, which provides an estimate of the percentage of variability due to heterogeneity rather than chance alone . The chi-squared test was used for significance testing. A visual inspection of the funnel plots was used to detect publication bias . The robustness of the conducted analyses was examined by conducting sensitivity analyses including only highquality studies. Furthermore, meta-regressions for potential moderators were planned . To our knowledge, this is the first meta-analysis of observational studies that focuses on the link between negative symptoms, cannabis and nicotine. Novel results are revealed based on a pool of more than ten thousand patients. Current non-abstinent cannabis users were not different from nonusers with regard to negative symptoms.Since nicotine was consumed by all participants in all cannabis studies, it is important to note that no association was found between nicotine use and negative symptoms. Thus, our findings support a specific effect of cannabis abstinence on negative symptoms. Regarding positive symptoms, we observed borderline significant results with a very small effect size, suggesting a possible association with current cannabis use; however, this effect became nonsignificant when including only high-quality studies. In addition, we found that nicotine users reported higher levels of positive symptoms than nonusers, a result that had a very small effect size and was essentially driven by the ‘nicotine as main drug of choice’ group. For cannabis, the overall results revealed the absence of a specific association between current cannabis use and the severity of negative symptoms. No difference was found between the subgroup using only cannabis and nicotine and the subgroup using cannabis as a main drug of choice. These findings are consistent with the meta-analysis by Large and colleagues that also reported no significant differences between current cannabis users and nonusers. The similarity of the results is of note because there were considerable differences in the methods and included studies. Overall, we included a larger set of studies, but we restricted the included groups to patients with schizophrenia and excluded patients with broad first-episode psychosis. An earlier meta-analysis by Potvin and colleagues also restricted the inclusion criteria to patients with schizophrenia and suggested that cannabis users would show less negative symptoms than nonusers. However, only three cannabis studies could be included in that meta-analysis, and in one study, subjects had to be abstinent for at least three weeks.
Overall, our study does not confirm Potvin and colleagues findings and shows that for patients with schizophrenia, there does not seem to be a significant difference in the severity of negative symptoms between current cannabis users and nonusers. A key finding of the present study is the observation that patients with schizophrenia and chronic cannabis use who have recently stopped using cannabis show less severe negative symptoms than patients with schizophrenia who do not use cannabis grow racks. To the best of our knowledge, this finding has not yet been reported in a meta-analysis of cross-sectional data. Some evidence comes from a meta-analysis of longitudinal studies that showed a small trend-level effect, suggesting that cannabis discontinuers show less severe negative symptoms than continuous users and nonusers . A recent poster by Ihler and colleagues reported that experiential negative symptoms related to amotivation improved after 12 months of follow-up in the group who discontinued cannabis compared to continued use . The findings for recent cannabis abstainers have to be considered with caution, because most studies had small sample size and the only large study reported a small effect size . However, the study was rated as having a high risk of bias and the low effect size in this study can be explained by less restrictive criteria for abstinence, in particular the lack of urine drug screening to verify abstinence. Importantly, all of the abovementioned results remained largely unchanged when restricting the analyses to high-quality studies. An important limitation of our results concerns the fact that it was not possible to conduct a meta-regression with the amount of cannabis used as an independent variable because only a few studies reported the grams of cannabis used per day, and only one study estimated the THC content. However, the dose dependence of cannabis effects remains of major importance and needs further evaluation with respect to negative symptoms .Our results suggest that patients with schizophrenia who use nicotine do not report different levels of negative symptoms than nonusers, with the effect size being close to zero. This finding is consistent with a recent meta-analysis by Huang and colleagues that also found no association between nicotine and negative symptoms . Here, we expand the findings by Huang et al. by using a larger sample of studies to show that this absence of an association occurs in studies including patients using nicotine only as well as in studies including patients using nicotine as a main drug of choice. The available data did not allow us to perform a meta-analysis of nicotine abstainers compared to nonsmokers. Boggs and colleagues report a small nonsignificant decrease in negative symptoms following one week of nicotine abstinence , indicating that further research is needed. Here, we report that patients with schizophrenia who current use cannabis do not differ from nonusers regarding the severity of negative symptoms. One interpretation of the absence of an association in the current users would be that cannabis simply does not exert any effects on negative symptoms. While this interpretation would not be in line with the hypothesis concerning the amotivational effects of cannabis use, it must be considered that the evidence for the negative effects of cannabis on motivation remains heterogeneous .
It may also be difficult to detect the effects of cannabis consumption on negative symptoms because patients with schizophrenia will often suffer from a combination of primary and secondary negative symptoms. Thus, the potential amotivational effects of cannabis as a secondary symptom may account for only a part of the overall negative symptomatology shown by the individual patient. Alternatively, it is conceivable that in chronic cannabis users with schizophrenia, the stimulating and blocking effects of cannabis on the reward system offset each other and result in the absence of an effect on negative symptoms . However, these considerations cannot account for the second important finding reported here, i.e., that recent cannabis abstainers show less severe negative symptoms than nonusers. This finding is consistent with the reduced susceptibility of developing negative symptoms among patients with schizophrenia who use cannabis grow system. Previous reports have suggested that cannabis-using patients with schizophrenia have better cognitive functioning than nonusers, particularly with respect to premorbid cognitive functioning . Furthermore, a recent study conducted by Mallet and colleagues has found that patients with heavy cannabis use before the onset of psychosis showed significantly less neurological soft signs, less negative symptoms and better cognitive functioning in different domains than their non-heavy user counterparts . It has therefore been hypothesized that cannabis-using patients constitute a subgroup that has lower biological vulnerability, which also results in a reduced susceptibility to developing negative symptoms. Another explanation suggests that cannabis-using patients might more easily access cannabis due to better premorbid social functioning, which could also be related to a reduced susceptibility to developing negative symptoms . Thus, cannabis-using patients with schizophrenia have less severe negative symptoms when they abstain from the drug. When using the drug, this difference may be obscured by the amotivational effects of cannabis . Importantly, the abstinence duration of at least three weeks required in the abstinence studies seems to be sufficient to alleviate the negative effects of cannabis on motivation. This timeframe is consistent with positron emission tomography studies in healthy cannabis users that show reduced dopamine release in the associative striatum in current users . Interestingly, in earlier studies with the same minimum duration of abstinence as in our abstainer studies, dopamine release in the striatum was not different from nonusing controls . Therefore, the time period of abstinence required in our recent abstainer group is consistent the normalization of dopamine release in the striatum.
We found a borderline significant association of current cannabis use with positive symptoms without significant subgroup differences between cannabis and nicotine groups and cannabis as the main drug of choice groups. However, the effect size was very small and might be of questionable clinical significance. These results differ to some extent from those reported by Large and colleagues, who found a medium effect size for the association of current cannabis use with positive symptoms . Several differences between the two meta-analyses must be noted. First, we included a larger number of studies than Large, but we did not include studies with patients experiencing broad first-episode psychosis, which may lead to stronger effects of cannabis on positive symptoms. Second, in the first-episode studies included in the Large meta-analysis, not all patients were receiving antipsychotics. The fact that we were focusing on stabilized populations treated with antipsychotic medication might have led to a weaker association of cannabis with positive symptoms. Third, our analysis allowed us to differentiate the groups ‘cannabis and nicotine’ from cannabis as the main drug of choice, but this differentiation did not have an impact on effect size. Furthermore, we were able to specifically address the recent cannabis abstainer group. In contrast to the findings for negative symptoms, this group did not significantly differ from the nonuser group. Overall, our results show only a very limited cross-sectional association between cannabis use and positive symptoms. Our results show that overall nicotine use is associated with more severe positive symptoms, although the effect size was very small. In the subgroup analysis, the effect was significant only for the ‘nicotine as a main drug of choice’ group. Importantly, this subgroup effect remained significant when considering only high-quality studies, while the overall effect became nonsignificant. Huang and colleagues reported a somewhat larger effect size for the association of nicotine use with positive symptoms . Our meta-analysis included a larger number of important studies and allowed a clear distinction of ‘nicotine only’ and ‘nicotine as a main drug of choice’ subgroups. Our subgroup analysis suggests that the observed association with positive symptoms might be more strongly related to the concomitant use of other drugs along with nicotine. However, we cannot exclude the possibility that nicotine use alone could increase the severity of positive symptoms to some extent. Our results only suggest a very limited cross-sectional association of continued cannabis use with positive symptoms. The acute psychosisinducing effects of THC have been well documented and seem to at least be partially related to increased dopamine release in the striatum . However, the long-term effects seem to depend on a large number of parameters, including the duration and intensity of the exposition as well as the proportion of THC and cannabidiol .
Our data suggest that the effects of ongoing cannabis consumption are to some extent offset by ongoing antipsychotic drug treatment. Although our data provide little evidence for a specific association of nicotine use with positive symptoms, it has to be noted that there is some evidence that psychotic-like experience have been associated with the smokers’ status in the general population after adjustment for confounding factors . Moreover nicotine has been suggested to increase positive symptoms via increased dopamine release and the increased metabolism of antipsychotic drugs . There was a small but highly significant association in the nicotine as the main drug of choice group. A mechanistic interpretation of the finding is difficult because substance use in this subgroup was very heterogeneous across the different studies. The main limitation of our meta-analysis is the nature of the included studies, which employed heterogeneous methods. It should be noted that the number of patients included was much higher in the nicotine groups than for the cannabis groups. Therefore, future studies with large sample size could change the results for cannabis users and recent abstainers. Nevertheless, this is the largest meta-analysis on the topic so far, and we were able to conduct sensitivity analyses including only high-quality studies that confirmed the main findings. Several limitations concern the case and control definition, such as differences in population, inclusion criteria and methodology across the included studies.