Unlike previous studies examining COMT genotyping
Unlike previous studies examining COMT genotyping in healthy controls seeking to explain impulsive behaviors, this study sought to examine COMT genotypes in a large sample of individuals with varying levels of gambling behavior to determine whether COMT Tenovin-6 was associated with differences in gambling phenomenology and cognitive functioning. We hypothesized that those Val/Val homozygote gamblers, compared to Met carriers, would exhibit greater problem gambling severity, relative impairment in spatial working memory, relative impairment in decision-making, and no differences with respect to response inhibition (since the latter appears to be largely under noradrenergic control) (Bari and Robbins, 2013).
Material and methods
Results The Val/Val group had the most subjects with gambling disorder (31.8%), a rate significantly different from the Val/Met (13.2%) group (p = 0.001).The Val/Val group was also associated with significantly more gambling disorder diagnostic criteria (reflected by the total number of SCI-GD criteria met) and greater frequency of gambling behavior compared to the met/met group (Table 1). In addition, comparison of the Val/Val group with the Val/Met group showed that they had significantly worse gambling symptoms (reflected by the subscales and total score of the PG-YBOCS) and met more SCI-GD criteria. We analyzed the frequency of current comorbid psychiatric disorders to see whether genotyping was associated with different distributions of these conditions (Table 2). The association of COMT with gambling disorder did not appear to have any significant relationship with other impulsive disorders based on the MIDI, bipolar disorder, or alcohol or substance use disorders. Table 3 shows the differences between COMT groups for the cognitive tasks. Although no differences were found on the Stop Signal Task (SST), the Val/Val group exhibited significantly worse cognitive performance on the Cambridge Gamble Task (CGT: risk adjustment and delay aversion) as well as on the Spatial Working Memory task (SWM: total errors).
Discussion This is the first study to explore clinical and cognitive associations between gambling behavior and COMT genotype in a large sample of young adults who gamble to varying degrees. Unlike many prior studies, we examined COMT genotypes in a representative non-treatment seeking community sample enriched for gambling behaviors. In this group of young adult gamblers, we found that a higher rate (31.8%) of the Val/Val group met criteria for gambling disorder compared to rates of 13.2% and 19.7% in the Val/Met and Met/Met groups, respectively. This finding was further reflected by a greater number of gambling disorder diagnostic criteria met by the Val/Val subjects as well as greater weekly frequency of gambling behavior. The finding of worse gambling symptomatology in the Val/Val subjects appears to be consistent with what one would expect in gamblers based on the mechanism of the COMT enzyme. Continued gambling behavior may be related to less efficient prefrontal neural signaling and problems with executive cognitive functioning (van Holst et al., 2010, Ledgerwood et al., 2012). Our findings, however, conflict with a recent study of healthy controls which found that the Met/Met homozygotes were more likely to be at-risk gamblers than Val/Val homozygotes (Guillot et al., 2014). The Guillot study, however, enrolled subjects based on alcohol use (and excluded heavier drinkers) and screened for gambling as a secondary aspect of the research. Our study specifically focused on gambling pathology and did not exclude subjects based on any issue other than ability to consent. The inconsistent findings, however, raise issues regarding how lower levels of gambling pathology are best measured and whether a range of impulsive behaviors (and which ones) may be associated with COMT genotype. The Val/Val group also exhibited greater problems in spatial working memory than the Met/Met group, in terms of the number of errors made. These data are consistent with those found in mice using a spontaneous alternation task as well as data from human studies which have shown better performance by Met carriers compared with Val carriers on certain tasks of working memory (Barnett et al., 2007, Diaz-Asper et al., 2008). Working memory influences self-regulation as those with low working memory capacity show more automatic behavior than individuals with high working memory capacity. Deficits in working memory may result in impaired self-regulation and by extension worse gambling behavior.