Comorbidity of Substance Use and Axis I Psychiatric Disorders

Kathleen T. Brady, MD, PhD


Medscape Psychiatry & Mental Health eJournal. 1998;3(4) 

In This Article

Anxiety Disorders

The NCS[45] found that approximately 36% of individuals with anxiety disorders also have a substance use disorder. Because anxiety disorders are such a heterogeneous group and the issues seem to differ substantially among disorders, each disorder will be discussed separately.

The active use of some substances (eg, marijuana, stimulants) is associated with anxiety symptoms, and withdrawal from other substances (eg, opiates, benzodiazepines) is marked by anxiety states. The necessary period of abstinence for diagnostic purposes is likely to vary by diagnosis and by the substance being used. For long half-life drugs (eg, some benzodiazepines, methadone), withdrawal symptoms may be quite protracted, and several weeks of abstinence may be essential for accurate diagnosis. For shorter-acting substances (eg, cocaine, short half-life benzodiazepines), both the acute intoxication and withdrawal duration are likely to be briefer, and it may be possible to make diagnoses with shorter periods of abstinence. In cases in which the diagnosis remains unclear, a family history of anxiety disorder, the onset of anxiety symptoms before the onset of substance abuse and dependence, and/or sustained symptoms of anxiety disorder during lengthy periods of abstinence in the past all weigh in favor of making a diagnosis of an anxiety disorder.

In the ECA survey, 36% of individuals with panic disorder had lifetime substance use disorder.[1] In the NCS, the odds ratio for co-occurring drug and alcohol dependence with panic disorder was 2.0.[45] The estimated prevalence of panic disorder and agoraphobia in treatment-seeking samples of alcoholics ranged from 5% to 42%.[46]

Although self-medication with alcohol and/or drugs to lower the anxiety associated with panic has been posited by some to explain the high comorbidity of panic and substance use disorders, some substances of abuse (eg, cocaine, marijuana, other stimulants) may actually induce panic attacks during periods of acute intoxication. Several reports have noted that both marijuana and cocaine can precipitate panic attacks in patients without previous panic disorder.[47,48] Cocaine, amphetamine, and phencyclidine act on the noradrenergic systems, which may explain their ability to induce symptoms of panic. Eliciting a good history of panic symptoms in the absence of substance abuse will help differentiate symptoms of panic that are substance induced.

The most widely used classes of pharmacotherapeutic agents for panic disorder are TCAs, SSRIs, monoamine oxidase inhibitors (MAOIs), and benzodiazepines (Table I).[49] Despite their effectiveness in the immediate relief of panic symptoms, benzodiazepines are generally contraindicated in substance-using populations because of their abuse potential. There is controversy concerning this, however, and some investigators report successful use of benzodiazepines in this patient population without evidence of abuse.[50] Benzodiazepines may be considered as adjuncts during the early treatment phase, when activation or latency of onset of the antidepressants are issues of concern. If one chooses to prescribe a benzodiazepine to a patient with comorbid substance use, symptoms of relapse should be closely monitored and the amounts of medication limited. As a rule, it seems best to avoid benzodiazepines in this population.

MAOIs are not recommended for use in substance-using populations. Dietary restrictions are necessary because of the interaction with tyramine in the diet, which may result in a hypertensive crisis. Moreover, MAOIs in combination with stimulant drugs may precipitate a hypertensive crisis.

Evidence from clinical trials has also demonstrated the efficacy of SSRIs in the treatment of panic disorder in non-substance-using patients.[49] As mentioned earlier, the SSRIs have been shown by some investigators to have modest effects in lowering alcohol consumption and thus might be a logical choice for the patient with comorbid panic disorder and alcoholism. Because of difficulty in using benzodiazepines and MAOIs in the substance-abusing population and the proven efficacy of both tricyclics and SSRIs in treating panic disorder, TCAs and SSRIs should be considered the treatment of choice in patients with comorbid panic and substance use disorders.

Panic disorder is quite responsive to nonpharmacologic treatment. Cognitive-behavioral (CB) techniques, such as exposure and systematic desensitization, have been shown to be particularly effective in the treatment of panic disorder.[51,52] Relaxation therapy and supportive therapy can also be helpful in some cases.[53] It is particularly important to maximize these nonpharmacologic treatments in patients with substance use disorders. The ability to self-regulate subjective states and the confidence that can result from successful mastery through CB therapy can be helpful to individuals in recovery. Many of the CB techniques used in anxiety disorders overlap with CB therapies known to be successful in the treatment of substance use disorders. Finally, by learning therapeutic anxiety-reducing techniques, patients may be able to break out of the mindset of using external agents to combat intolerable subjective states and acquire alternative coping strategies.

Anxiety is commonly linked to alcohol and drug withdrawal. However, there are few studies that examine the prevalence of comorbid substance use disorders and generalized anxiety disorder (GAD). In the NCS, the odds ratio for the co-occurrence of GAD with alcohol dependence was 3.7.[45] One complicating factor is that the Diagnostic and Statistical Manual of Mental Disorders (DSM-IV) criteria for GAD require that symptoms occur for at least 6 months without being directly related to the physiological effects of a substance or a general medical condition. Symptoms of GAD have substantial overlap with acute intoxication with stimulants and withdrawal from alcohol, sedative/hypnotics, and opiates, thus the 6-month period of abstinence often may be difficult to ascertain. While many substance-using individuals report anxiety with symptoms consistent with GAD, they may not meet diagnostic criteria for GAD because of difficulty in distinguishing symptoms of anxiety from substance-related symptoms.

The treatment of GAD complicated by a substance use disorder is challenging (Table I). Benzodiazepines are effective in the treatment of GAD; however, as previously discussed, their abuse potential limits their utility in the substance-abusing population. Buspirone is a non-benzodiazepine anxiolytic with no abuse potential that has been studied for the pharmacologic treatment of GAD with comorbid alcohol dependence. In 2 double-blind placebo-controlled trials, alcoholics with anxiety disorder had decreased alcohol consumption and improved symptoms of anxiety during treatment with buspirone.[54] In a third study, however, buspirone was found to have no effect on alcohol consumption or anxiety symptoms.[55] Although the data remain somewhat contradictory because of the low abuse potential and reports of success in well-controlled studies, buspirone remains a good choice in individuals with comorbid GAD and substance use disorders. There are no systematic trials of TCAs or SSRIs in the treatment of GAD in individuals with substance use disorders; however, these agents have been useful in non-substance abusing populations.[49]

As mentioned in the discussion of panic disorder, nonpharmacologic treatments for anxiety disorder can be very useful. GAD can be effectively managed using relaxation, coping skills, and CB therapy techniques.[51,52,53] Pharmacotherapy and psychotherapy are likely to complement one another in maximizing patient outcomes. Nonpharmacologic treatment strategies in conjunction with judicious pharmacotherapeutic management should be encouraged.

Social phobia is defined by the DSM-IV as a marked and persistent fear of situations in which an individual is exposed to unfamiliar people or to the scrutiny of others. The studies examining the interface of alcohol abuse and dependence with social phobia have found rates of comorbidity ranging from 8% to 56%.[46] Consistent with the self-medication hypothesis, patients with social phobia report the use of alcohol to reduce social anxiety in most studies, and the onset of social phobia occurs prior to the onset of alcohol abuse and/or dependence.[56] Social phobia in drug dependence is not well studied; but in one study, Myrick and Brady[57] found a lifetime prevalence of social phobia in a cocaine-dependent population to be 13.9%. In nearly all cases, the social phobia preceded the cocaine dependence.

Because social phobia may interfere with an individual's ability to engage effectively in treatment, early recognition is essential to ensure an improved chance of recovery. Frequently, the diagnosis is missed unless specific symptomatology is thoroughly assessed. A lengthy period of abstinence may not be needed as the fear of interaction in social situations is not a specific feature of substance use or withdrawal. However, the social fears that occur only during periods of intoxication with marijuana or stimulants should not be considered sufficient to meet diagnostic criteria for social phobia.

Although there are no studies examining the psychopharmacologic treatment of individuals with comorbid social phobia and substance use, many agents have been studied for the treatment of uncomplicated social phobia.[49] Of these agents, MAOIs, reversible inhibitors of monoamine oxidase (RIMs), SSRIs, and benzodiazepines have documented efficacy (Table I). Several other agents such as bupropion, ondansetron, buspirone, and venlafaxine may also have efficacy, but have not been well studied. In choosing a medication for the treatment of comorbid social phobia and substance abuse, SSRIs would probably be first choice because they are effective, require no dietary restrictions, and are without the potential toxic interactions of the MAOIs. As previously mentioned, SSRIs may have the additional benefit of producing modest decreases in alcohol consumption.

Psychotherapeutic treatment is important. It may be difficult for such patients to participate in group therapy or 12-step programs. A treatment plan that includes cognitive and behavioral therapies may prove to be more effective. Although there are no studies examining these treatments in patients with comorbid social phobia and substance use, several types of nonpharmacologic treatments such as systematic desensitization, imaginal flooding, graduated exposure, social skills training, and cognitive approaches have proven effective.[58]

The prevalence of comorbid posttraumatic stress disorder (PTSD) and substance use disorders is high. The NCS study indicates that approximately 30% to 50% of men and 25% to 30% of women with lifetime PTSD have a co-occurring substance use disorder.[59] Individuals with PTSD had a 2.5 odds ratio for a co-occurring lifetime substance use disorder, the highest odds ratio of any of the anxiety disorders for co-existence with a substance use disorder.[45]

In samples of treatment-seeking substance abusers, the lifetime prevalence of PTSD is between 36% and 50%, and the current prevalence of PTSD is between 25% and 42%.[60,61,62] In 1 study, cocaine-dependent individuals were significantly more likely to experience a traumatic event and to suffer from PTSD than were alcohol-dependent individuals.[62]

It is likely that drug use (in particular, cocaine use) and repeated withdrawal (in particular, alcohol, sedative-hypnotic, and opiate withdrawal) exacerbate symptoms of PTSD. Cocaine use is associated with paranoia, hypervigilance, sleep disturbance, and autonomic arousal, all of which are features of PTSD. Sedative-hypnotic and opiate withdrawal are marked by feelings of anxiety and autonomic nervous system hyperactivity, which are believed to have as their origins excessive firing of neurons in the locus ceruleus.[63] It is possible that common pathophysiologic mechanisms are responsible for the symptom overlap and exacerbation in individuals with comorbid PTSD and substance dependence.

Little is known about the effective treatment of patients with comorbid PTSD and substance abuse/dependence (Table I). One approach often suggested in the literature is to treat the substance use disorder first and to defer treatment of trauma-related issues. The rationale for this approach includes the premise that unchecked substance abuse will impede therapeutic efforts directed at other problems. Although it is the standard of care in the many substance abuse treatment programs, the approach of deferring treatment of trauma-related issues in substance-abusing individuals has never been systematically compared with a treatment approach that targets both disorders simultaneously.

In this light, it is noteworthy that a number of case studies in the literature indicate that successful treatment of symptoms of PTSD can lead to reductions in alcohol and drug abuse.[64,65] For some patients relief of PTSD symptoms may improve the chances of recovery from substance abuse. It is important to note that CB therapy has demonstrated efficacy in the treatment of PTSD.[66]

Pharmacotherapy is playing an increasingly important role in the treatment of PTSD. The range of PTSD symptoms is broad and can include symptoms of autonomic arousal, depression, anhedonia, impulse dyscontrol, and dissociative and psychotic features. The most rational pharmacotherapeutic regimen depends on the symptom constellation exhibited by the individual.

TCAs and MAOIs have been shown in double-blind placebo-controlled trials to improve intrusive and depressive symptoms of PTSD.[67] There are also uncontrolled studies reporting positive effects of carbamazepine, beta blockers, clonidine, benzodiazepines, and lithium. A recently published placebo-controlled study found fluoxetine (an SSRI) to be an effective treatment for PTSD.[68] This is of particular interest because SSRIs may also have a modest effect in reducing alcohol consumption, are well tolerated, and are relatively nontoxic if combined with alcohol. There is one report of sertraline treatment in a small group of individuals with comorbid PTSD and alcohol dependence that resulted in decreased alcohol consumption and symptoms of PTSD.[69] This promising preliminary finding warrants further investigation.

The coexistence of obsessive-compulsive disorder (OCD) and substance use disorders is an understudied area. Rasmussen and Tsuang[70] reported that 12% of OCD probands in a clinical population had a lifetime history of alcoholism. Eisen and Rasmussen[71] screened 50 patients with a diagnosis of alcohol dependence or abuse and found that 6% met DSM-III-R criteria for OCD -- 3 times the lifetime prevalence of OCD in the general population.

There are no controlled trials or case reports of the treatment of comorbid OCD and substance use (Table I). Clomipramine and SSRIs are both efficacious in the treatment of OCD.[49] However, toxic interactions with alcohol, stimulants, and CNS depressants are also more likely to occur with clomipramine. Consequently, SSRIs are likely to be the first line of treatment in individuals with OCD and a substance use disorder because there are fewer side effects or risks for toxic interactions.

The use of psychotherapeutic techniques in combination with pharmacotherapy is particularly important in the treatment of OCD.[72] CB therapies, including thought-stopping, exposure, and response prevention, have convincingly and reliably been shown to be extremely effective in the treatment of OCD.[73,74] Again, a synergistic effect of pharmacotherapy and psychotherapy might be expected.


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.
Post as: