In this study, we explored the association of COMTval/met (rs4680) with suicidal behavior. First a case-control study was conducted. Additionally, we performed a meta-analysis to assess the evidence of association between COMTval/met and suicidal behavior.
We could not find any association between COMTmet or COMTval allele and suicidal behavioral in a Mexican population. To our knowledge, this is the first study addressing the genetic association between COMTval/met alleles and suicidal behavior in a Mexican population. Our results are in agreement with recent reports in the literature stating the no association of COMTval/met and suicidal behavior [26
]. This result is not surprising considering that complex behaviors, such as suicidal behavior, are the result of a moderate number of genes that individually have small to modest effects on disease liability [26
Available evidence suggests that the effect of this polymorphism on suicidal behavior may be related to the lethality of suicide attempts rather than to the risk for attempting suicide per se [11
]. In our study we only performed an association with attempted suicide, because we wanted to establish which COMT polymorphisms were associated with suicide attempts; this could be considered as a limitation of our study. However, other study analyzing genotype differences with respect to lethality of suicide attempts or violent attempt methods reported results similar to ours [35
It is worth mentioning the evidence provided by other studies reporting a positive association of the COMTval allele when compared to the control group [5
]. But these results are controversial, since such association was observed in presence of the COMTmet allele in patients presenting alcohol dependency or other diseases [33
]. These differences among studies might be explained by the different diagnostic entities used. Some studies evaluated patients with alcohol dependence, while others included patients with schizophrenia, or schizoaffective or mood disorders [33
]. Other limitation could be that these association studies were conducted in various populations and different criteria were used to define the phenotypes. Other relevant factor is the difference in the size of the samples. We observed that almost all studies consisted of small samples (n < 200) and some even made subdivisions within samples (gender or affective status, for example). Hence, n was very small and had a low power of association. When we detected this limitation we decided to analyze the evidence in a meta-analysis.
We tested the probability of the association of COMTmet with suicidal behavior. However, we could not find any association between these two factors. A previous meta-analysis reported a Met association with suicidal behavior [11
]; however, this association is not strong because the results of this meta-analysis were highly dependent upon the inclusion of all the studies. When five of the six studies involved were individually removed from the analysis, the relationship between COMT and suicidal behavior was no longer significant [11
]. Our results are in accordance with other recently published meta-analysis in which no association was found between Met or Val allele and suicidal behavior [26
]. This evidence supports a lack of direct modulation of COMT on suicidal behavior.
Similarly to the results presented in a previous study, the sample sizes of the studies included in the present meta-analysis are in the low range compared to genetics studies for other diseases. Therefore, future studies comprising larger samples of completed suicide are important to determine this association. We also consider that given the small number of studies available the association is not observable.
In a first approach we observed heterogeneity; however this variation was due to four specific studies. We carried out a second analysis, in which the studies that gave rise to heterogeneity were discarded. However, no association between COMTmet and suicidal behavior was encountered. Finally, with the aim of establishing whether this association depended on completed suicide or suicide attempt, we performed a last analysis which only included suicide attempters. Once again, no association was confirmed.
Our study presents some limitations. Our case-control study lacks of a specific scale investigating suicide attempt. With regard to the meta-analysis, publication bias has to be considered, since negative studies are less likely to get published. Also, an overrepresentation of the results showing an association between the polymorphism and the investigated disorder is also possible [44
]. Although the contribution covering from genetic factors to personality traits may differ between male and female subjects, we did not analyze for gender. Other limitations are inherent in many meta-analysis of association (including this one) such as their retrospective nature and the inclusion of study-level data.