These studies also suggest that measures of exceptionally poor performance are perhaps better indicators of impairment than continuous performance scores or dichotomous classifications of samples based on arbitrary cutpoints. Despite evidence that patients with psychiatric comorbidity comprise the majority of patients in clinical settings and have greater neurocognitive deficits than patients with substance dependence only previous studies have not explored the impact of neurocognitive impairment on treatment outcomes for patients with comorbidity. The current study addresses this area of need by examining the effects of neurocognitive impairment on substance use outcomes in patients treated for substance dependence and MDD. Greater neurocognitive impairment at baseline predicted lower self-efficacy and lower 12-step affiliation, and our analyses confirmed that these proximal variables mediated the effects of impairment on future drinking and drug use. Previous studies found similar relationships between impairment and self-efficacy , and our study extends these findings to patients with comorbid MDD. Self-efficacy is typically defined as the confidence to abstain from substance use in high-risk situations, and patients with greater impairment apparently had lower confidence they could manage these situations without using substances. The specific mechanisms underlying these effects are unclear, but it is possible that impaired patients engaged less fully in TSF and ICBT or experienced less perceived benefit from therapy sessions. In one recent study poorer cognitive ability predicted lower acquisition of coping skills in CBT . Future studies might explore whether reduced coping skills or difficulty engaging in other aspects of group treatment explain lower self-efficacy for cognitively impaired patients.
Patients with greater impairment had lower levels of 12-step affiliation,mobile grow systems suggesting they had greater difficulty engaging in 12-step practices that common in both therapy conditions but were only specifically targeted in TSF. This may help explain why patients with poorer neurocognitive functioning had better long-term substance use in ICBT than in TSF . While it was originally hypothesized that patients with poorer neurocognitive functioning would fare worse in ICBT due to the cognitive demands of the treatment, the current study shows that impaired patients had difficulty frequently attending 12-step meetings or engaging in 12-step behaviors. The direction of this finding contradicts those of previous studies, which found that levels of AA affiliation were significantly greater for impaired patients . It is possible that for patients with substance dependence and MDD, neurocognitive impairment represents an additional and especially disabling risk factor leading to limited engagement in 12-step practices. Patients with greater neurocognitive impairment were also found to have greater depressive symptoms during the course of the study. Similar results were found in a sample of depressed, hazardous drinkers, as patients with better cognitive functioning had greater reductions in depressive symptoms during CBT . We hypothesized that neurocognitive effects on depressive symptoms would impact substance use outcomes, given that egative affect is a frequent precursor to relapse in patients with comorbid MDD , and is consistently tied to substance use over time , suggesting that greater depression severity within these patients is frequently a potential trigger for future substance use. This hypothesis was confirmed by mediation analyses demonstrating that greater depressive symptoms predicted greater future alcohol and drug use and mediated the effects of neurocognitive impairment on these outcomes. Our previous work tied reductions in depression during treatment to 12-step meeting attendance, and it is possible that reduced engagement in 12- step contributed to greater depression for cognitive impaired patients. Alternatively, these patients may have had difficulty with other elements of treatment that limited their improvement in depressive symptoms.
Investigation of neurocognitive impairment as a moderator revealed complex interactions with therapeutic process variables in the prediction of drinking and drug use.Greater 12-step affiliation predicted lower future drinking to a greater extent for impaired patients. Although the direction of this finding is contrary to effects observed in previous, similar studies , other research found the effects of baseline social support on future drinking were greater for impaired patients . Our sample differed from these studies in that all patients had comorbid MDD, and our results suggest at greater levels of depression severity, the benefits of engaging with external sources of support are especially useful. That is, 12-step affiliation was most predictive of future drinking for patients with severe impairment, especially when they were severely depressed. Although prior research has found relatively reduced benefits of AA practices for patients with MDD , our results suggest the opposite may be true for patients with neurocognitive impairment, as they may experience relatively greater benefits from 12- step affiliation. Neurocognitive impairment also moderated the effects of depression on future drug use, but in the opposite direction than expected: relations between depressive symptoms and future drug use were stronger for patients with less impairment. Although rates of drug use in the sample were low overall, patients with little or no impairment had a greater tendency to use drugs frequently following periods of more severe depressive symptoms. Because previous studies of neurocognitive effects on treatment have largely focused on alcohol users without MDD or depressive symptoms, the mechanisms underlying this unexpected finding are unclear. Others have speculated that cognitively impaired patients may be relatively less capable of the planning required to re-initiate substance use . Relations between depression and drinking were not moderated by impairment, suggesting this could be the case for drug use but not necessarily for alcohol. This interaction was also less robust, as it was not statistically significant when controlling for prior self-efficacy and 12-step affiliation. Future studies are needed to determine whether this finding is consistent in other samples of drug dependent patients, with or without comorbid MDD. Limitations of this study should be noted. The results may not be immediately generalizable to the broad population of patients with psychiatric comorbidity, as we only studied patients with MDD, and our sample was comprised of veterans who were predominantly male and Caucasian.
Clinical trials of treatment for alcohol and drug dependence frequently suffer these demographic limitations, and replication of these findings in a wider range of patients is needed before broader generalizations can be made. Although we examined complex mediated and moderated pathways to substance use, there are untested relationships among the variables in this study that merit further exploration. Greater self-efficacy has been associated with greater 12-step affiliation, which could be one factor explaining lower self-efficacy for patients with greater neurocognitive impairment. Because the effects of mediating variables were examined in separate HLMs, we did not test whether each variable relates independently to future substance use, which can be examined in “multiple mediator” models that help demonstrate which mediating processes may be most crucial for limiting future substance use. Although this study met several conceptual criteria for examining mechanisms of change , further criteria are required to conclude with greater confidence that these process variables are mechanisms of change or casual factors for limiting long-term substance use. Despite decades of clinical research developing and testing behavioral interventions for substance dependence, evidence-based interventions are only moderately effective ,cannabis grow supplies with a substantial number of patients returning to substance use following treatment . Furthermore, many of the largest randomized trials have found equivalent outcomes across distinct treatment conditions , and theoretically distinct treatments often fail to produce differential change on hypothesized mechanisms of action , suggesting we may know little about the underlying mechanisms of change within addiction treatment. Consequently, researchers have increasingly turned towards process-focused investigations aimed at identifying how treatments work, for whom they are effective, and which modifiable factors are most integral for maintaining positive change , with hopes of improving the overall likelihood that substance dependence treatment will result in long-term change. The past decade has seen vast growth in process-oriented studies in the general addictions treatment field . However, very little of this work has involved samples of patients with other psychiatric disorders, despite evidence that these patients comprise the majority in many treatment settings . Among individuals with substance dependence, major depression is the most common co-occurring Axis I disorder in the general population and in treatment settings . Patients with co occurring MDD typically cost more to treat and have poorer treatment outcomes than patients without this MDD , suggesting it is especially important to identify core therapeutic processes within these patients.
However, because little process-oriented research has involved patients with substance dependence and MDD, it is not known whether previous findings generalize to this highly prevalent, costly, and disrupted population. Studies of mediators of treatment outcome are essential to the general area of “treatment process” research. Within treatment studies, investigations of mediators typically examine an intermediate factor that may explain the effect of treatment “dose” on a clinical outcome . These studies have the potential to inform the development of more efficient and portable interventions by identifying the skills or behaviors targeted by interventions that are most responsible for producing positive change. In previous studies examining mediation, lower marijuana use in CBT + contingency management compared to contingency management alone was mediated by enhanced self-efficacy , and better abstinence rates in TSF were explained by greater commitment to abstinence . Importantly, statistical mediation is only one of several conditions to be met before declaring a variable is a mechanism of change, as temporal precedence, specificity, and experimental manipulation of the variable must also be demonstrated . In practice these criteria are rarely.Study 4 has been submitted for publication in Journal of Consulting and Clinical Psychology, as following: Worley, M.J., Tate, S.R., Tapert, S.F., Granholm, E.G., & Brown, S.A. Neurocognitive impairment interacts with 12-step affiliation and depression to predict future drinking in depressed, substance-dependent veterans. The dissertation author was the primary author of this manuscript under review. Despite decades of clinical research developing and testing behavioral interventions for substance dependence, evidence-based interventions are only moderately effective ,with a substantial number of patients returning to substance use following treatment . Furthermore, many of the largest randomized trials have found equivalent outcomes across distinct treatment conditions , and theoretically distinct treatments often fail to produce differential change on hypothesized mechanisms of action , suggesting we may know little about the underlying mechanisms of change within addiction treatment. Consequently, researchers have increasingly turned towards process-focused investigations aimed at identifying how treatments work, for whom they are effective, and which modifiable factors are most integral for maintaining positive change , with hopes of improving the overall likelihood that substance dependence treatment will result in long-term change. The past decade has seen vast growth in process-oriented studies in the general addictions treatment field . However, very little of this work has involved samples of patients with other psychiatric disorders, despite evidence that these patients comprise the majority in many treatment settings . Among individuals with substance dependence, major depression is the most common co-occurring Axis I disorder in the general population and in treatment settings . Patients with co occurring MDD typically cost more to treat and have poorer treatment outcomes than patients without this MDD , suggesting it is especially important to identify core therapeutic processes within these patients. However, because little process-oriented research has involved patients with substance dependence and MDD, it is not known whether previous findings generalize to this highly prevalent, costly, and disrupted population. Studies of mediators of treatment outcome are essential to the general area of “treatment process” research. Mediators are third variables that statistically explain an association between a predictor and outcome . Within treatment studies, investigations of mediators typically examine an intermediate factor that may explain the effect of treatment “dose” on a clinical outcome . These studies have the potential to inform the development of more efficient and portable interventions by identifying the skills or behaviors targeted by interventions that are most responsible for producing positive change. In previous studies examining mediation, lower marijuana use in CBT + contingency management compared to contingency management alone was mediated by enhanced self-efficacy , and better abstinence rates in TSF were explained by greater commitment to abstinence . Importantly, statistical mediation is only one of several conditions to be met before declaring a variable is a mechanism of change, as temporal precedence, specificity, and experimental manipulation of the variable must also be demonstrated .