In agreement with our first hypothesis, we observed a significant correlation of MCMI scores and impaired neuropsychological test performance in a group of subjects with a history of opiate-dependence treated with methadone. Contrary to our second hypothesis, the regression models indicate that personality pathology explains a greater amount of the variance in neuropsychological test performance than do any of the variables of drug use history. Lastly, the relationship of personality pathology and neuropsychological performance is significant for the abstinent former opiate addicts, but not for the methadone-maintained group. We suggest that these results are best explained by postulating a multi-factorial origin to the neuropsychological deficits associated with opiate dependence, as opposed to thinking that drug use alone causes such deficits. To our knowledge, this report is the first to identify a link between personality pathology and the measurement of neuropsychological impairment in opiate addiction.
Our findings are consistent with previous studies that demonstrated significantly increased MCMI scores in opiate-dependent subjects compared to healthy controls [21
]. Additionally, our results are consistent with those studies of non-addicted subjects that showed significantly worse neuropsychological test performance in individuals with personality disorders [16
]. We know of no previous study that has attempted to measure the correlation of MCMI scores and neuropsychological test performance. In every instance where we measured this association, we found a negative correlation of MCMI scores and neuropsychological test performance, indicating that greater character pathology is associated with impairment in neuropsychological function. This relationship was present when examining all subjects together, PA and MMT subjects combined, and amongst the PA subjects only, but did not achieve statistical significance when examining the MM and non-dependent comparison groups separately.
The absence of significant findings in MM subjects is unexpected and contrary to our third hypothesis. One possible explanation for the group differences in statistical significance in the correlation of MCMI scores and neuropsychological test performance is differences between groups in MCMI scores and/or neuropsychological test performance. We have previously reported comparable degrees of personality pathology and cognitive impairment in both MM and PA subjects, with some minor variations – PA subjects tend to have more Cluster A pathology and worse visuospatial recall [8
]. It is possible that we underestimated the importance of these minor variations, and that in fact, as Gruber et al. conclude [43
], methadone treatment does selectively improve verbal and visuospatial encoding and recall in opiate dependent subjects. Such an effect could obscure the relationship between cognition and personality measures, particularly in the MM group. Similarily, Calsyn [21
] reported decreases in total MCMI scores in opiate addicts after 18 months of methadone maintenance treatment. Both an improvement in neuropsychological test performance and/or a reduction in MCMI scores may result in a decrease in the correlation coefficient relating personality pathology to neuropsychological performance. The presence of methadone, possibly through the action of mediating factors, may confer a protective effect that is removed in subjects choosing to end methadone treatment. This might explain the difference in the correlation coefficient between the MM and PA subjects reported here.
We have used linear regression to examine the functional relationships of drug use history and character pathology on neuropsychological performance. In every regression model tested, the composite MCMI score accounted for a greater portion of the variance in neuropsychological performance than did any of the variables of drug use history. As a measure of the relative importance of predictor variables, the regression models suggest that drug use history has less impact on neuropsychological test performance than does personality pathology. In fact, none of the drug use history variables were any better than random chance at accounting for variance in neuropsychological scores.
Because the neuropsychological variables assess different cognitive processes, it is potentially important to distinguish which, if any, of them is most responsible for the observed relationship between neuropsychological function and personality. Interestingly, for the PA subjects only, strong negative correlations of the WAIS-R Vocabulary subscore were demonstrated with each of the MCMI Clusters scores, indicating that greater personality pathology of any sort is associated with worsened verbal function. The WAIS-R Vocabulary subscore is a test of verbal function [30
], and previous research has reported decreased WAIS-R Vocabulary scores in abstinent opiate addicts compared to healthy control subjects [3
]. We have shown that Stroop CW scores correlated negatively with both MCMI Cluster B and Cluster C scores in our PA subjects. This finding suggests that PA subjects who score higher in the dramatic, erratic, anxious, and fearful personality domains may have greater difficulty sustaining attention and inhibiting a habitual response. Using an emotional Stroop task to measure attentional bias, Marisssen et al [44
] reported that attentional bias in abstinent opiate addicts predicted relapse to addictive drug use. Low scores on the BVRT subscore were correlated with high MCMI Cluster C scores in our PA subjects. To date, there have been no other studies of the BVRT in abstinent formerly opiate dependent subjects. In sum, these results suggest that specific neuropsychological deficits may be preferentially associated with subtypes of personality disorders. More research is needed to fully understand these correlations.
Personality disorders are thought to be enduring and stable over time [41
]. Some authors have suggested that personality disorders predate substance use in opiate-addicted patients [23
]. The data presented here points to a strong association of neuropsychological performance and a measure of personality pathology. Because of the strong association of personality pathology and neuropsychological performance, it is possible that neuropsychological deficits present in opiate dependent subjects may also be enduring and stable over time. Indeed, our previous study [8
] showed no improvement in neuropsychological performance between subjects receiving methadone and those who completed a course of methadone treatment and were subsequently opiate abstinent. Similar results have been reported by Davis [46
] who reported no significant difference between former opiate abusers and healthy controls on a test of verbal fluency, and Verdejo [7
] who found no difference in the test performance of MM and PA subjects using a modified Controlled Oral Word Association test, the Stroop Interference test, and the Wisconsin Card Sorting test. These studies indicate that neuropsychological deficits do not necessarily resolve when drug abuse ceases.
Cognitive deficits associated with opiate dependence have traditionally been explained as arising due to drug use itself. However, based on the results reported here, we suggest that it may be more helpful to posit a multi-factorial origin to opiate-associated cognitive impairment. A variety of factors may result in neuropsychological deficits; including opiate abuse, additional concomitant illicit substance abuse, adulterants added to illicit substances, overdoses and head trauma, lifestyle and nutritional factors, infections, educational and occupational attainment, genetic and familial factors, and early-childhood experiences. It is also possible that the neuropsychological performance deficits identified in opiate-dependent subjects, like the enduring and stable personality disorders often associated with opiate dependence, may be a manifestation of a pre-existing condition that persists through drug use, abuse, and addiction treatment.
Strengths and limitations
Insofar as possible, we have attempted to limit confounding influences on our results. We have recruited the subjects with a history of opiate-dependence from a residential treatment setting, thereby providing for a standardization of nutrition and activities, along with frequent testing for illicit substance use. We have included only MM subjects who have experienced a significant period of methadone treatment, and PA subjects with a prolonged period of drug abstinence. Nonetheless, our results must be interpreted in light of the study limitations. Firstly, this cross-sectional study can only identify associations, and does not support causal interpretations. Because entrance into abstinence treatment was entirely voluntary, it is possible that the observed differences between MM and PA groups reflect a selection bias rather than an actual effect of methadone. We have not controlled for additional past illicit drug use in our sample, and this may impact our outcome measures in unexpected ways. We have used a simple battery of cognitive tests, which limits the scope of our findings. Additionally, unmeasured factors, such as substance-related lifestyle and familial variables, may have influenced our results.
Methadone has been used for years as a substitution treatment for opiate addiction [47
]. Despite the amply documented success of methadone maintenance programs [48
], there is significant public and social resistance to treating addiction with long-term prescription opiates. For their part, the patients receiving methadone also complain about the treatment and the way it is delivered. Some opponents of methadone maintenance therapy point to studies reporting opiate-induced cognitive deficits as an argument against MM treatment. In response to these concerns, Methadone-to-Abstinence treatment has received FDA approval and is considered a viable alternative to continued methadone maintenance [56
]. Our study has shown no correlation between neuropsychological performance and either opiate use or methadone treatment. However, we have demonstrated a significant correlation between personality pathology and worsened cognition. Because personality is understood to be stable over time [23
], we suggest cognitive deficits observed in methadone maintained patients may also be stable over time. Furthermore, there has been no published study that has established a causative link of methadone to neuronal pathology or cognitive deficits. Thus, citing opiate-induced cognitive deficits as a reason for cessation of methadone treatment makes for poor policy. There may well be good reasons to discontinue MMT in select patients. However, in the main, MMT has been shown to be effective and cost-efficient means to reduce the problems of addiction, and for the majority of patients the advantages far out-weigh purported and poorly documented disadvantages.
Current drug addiction treatment focuses on abstinence and/or harm reduction. The underlying assumption is that any existing neuropsychological and/or cognitive deficits will resolve with cessation of opiate use. If, as we have argued, neurocognitive difficulties do not correlate with opiate use, addiction treatment alone will not ameliorate impairments. Most studies show worsened treatment outcomes in addicts suffering from neurocognitive impairments [2
]. Therefore, cognitive rehabilitation should be an important adjunct to drug addiction treatment. Patients entering drug addiction treatment programs should be screened for the presence of cognitive deficits. Approaches to restoring cognitive abilities and functioning can include training in goal setting and planning, developing problem-solving skills, sustaining attention, and inhibiting habitual responses. Patients will benefit most from drug addiction treatment when they can attend and receive new information and incorporate it into behavioral response [60