Patients who seek treatment for anxiety disorders often have problems with alcohol or drug abuse and are otherwise at an increased risk for developing such problems. Therefore, it benefits mental health practitioners to be aware of the following:
• The specific prevalence of substance use disorders among those with anxiety disorders
• How to reliably identify comorbid drug and alcohol use disorders in patients with anxiety disorders
• The signs of elevated risk of substance use problems developing in patients with anxiety disorders
• How to respond clinically when comorbid drug or alcohol problems exist or when there is a high risk of such problems to develop
In this article, we attempt to leverage state-of-the-art research findings to provide empirically informed perspectives and practices related to these issues. Toward this end, we draw heavily on our own research (funded by the National Institute on Alcohol Abuse and Alcoholism [NIAAA] and National Institute on Drug Abuse) and our clinical experience, as well as on the National Epidemiologic Survey on Alcohol and Related Conditions (NESARC).1
Risk of comorbidity
Substance use disorders (SUDs) occur significantly more often in patients with anxiety disorders than in the general population. Table 1 summarizes the magnitude of these associations for current (ie, past 12 months) diagnoses expressed as odds ratios. The odds of alcohol dependence being diagnosed are 2 to 3 times greater among patients with an anxiety disorder; these correlations are even greater for drug dependence. (Substance abuse—unlike dependence—does not appear to be strongly correlated with anxiety disorders.)
Odds ratios are base-rate neutral by design (ie, they are insensitive to the raw number of individuals classified). Therefore, it is important to know the actual percentage of people with an anxiety disorder who also have an SUD. The Figure shows the 12-month prevalence rates of alcohol and drug dependence in patients with anxiety disorders and in persons in the general population. Table 1 shows greater correlations between anxiety disorders and drug dependence; the Figure shows greater absolute numbers for alcohol use disorder among patients with an anxiety disorder. This reflects the greater overall rate of alcohol-related disorders and suggests that clinicians are 2 to 3 times more likely to see alcohol dependence than drug dependence in their patients with anxiety disorders.
Both Table 1 and the Figure refer to current diagnoses because we considered this time frame to be most relevant to the clinical focus of the article. However, the prevalence of lifetime dependence is obviously higher than that of current diagnoses. In addition, these data do not include patients whose use of drugs or alcohol, while problematic, does not (yet) rise to the level of a DSM diagnosis. By some estimates, this rate could be 3 times greater than that of a diagnosable disorder.2
In addition to current problems associated with drugs and alcohol use among a subgroup of patients with anxiety disorders, it is important to know that there is an elevated risk of future SUDs in all patients with anxiety disorders. Using the student dataset collected by Kushner and colleagues,3 we found that the odds ratios for developing alcohol dependence within 4 to 7 years were 3 to 5 times higher in college freshman with an anxiety disorder than in students who did not have an anxiety disorder. Similarly, Christie and colleagues4 found that individuals between the ages of 18 and 30 who had an anxiety disorder experienced a 2- to 3-fold increased risk of later development of drug dependence.
Some patients with anxiety disorders are at greater risk for SUDs. For example, we found that the 20% of patients with anxiety disorders in the NESARC sample who endorsed using alcohol to self-medicate anxiety symptoms had a much greater risk than others with anxiety disorders of developing a new SUD within the next 3 to 4 years.5 We also found that the number of lifetime internalizing disorders (eg, common anxiety disorders, mood disorders) is a more powerful predictor of SUD risk than is the type of anxiety disorder.6 Finally, the risk of future SUDs among persons with anxiety disorders should also be expected to covary with all of the usual signs of risk, including a positive family history, binge drinking (ie, 5 or more drinks during a single episode) in the past year, a younger age, and being male (Table 2).
All patients with anxiety disorders should be screened for drug- and alcohol-related problems at the initial assessment.7 Because of their prospective risk for an SUD, patients with anxiety disorders seen on an ongoing basis should also have follow-up screening on a routine basis. The most widely used screening tools for alcohol problems include the 4-item CAGE questionnaire (Cut down, Annoyed, Guilty, Eye-opener) and the 10-item AUDIT (Alcohol Use Disorders Identification Test).8,9 A truncated 3-item version of the AUDIT is also recommended by the NIAAA.10 The items from these screening tools are shown in Table 3.
The World Health Organization and the NIAAA have developed the ASSIST (Alcohol, Smoking and Substance Involvement Screening Test) to quantify drug-related risk and behavior of patients (Table 4).11 The patient is first asked about lifetime and previous 3-month substance use by class (eg, marijuana, pain medication, prescription stimulants, cocaine). For each class of drug that is endorsed, additional ASSIST questions are asked.
Causes for comorbidity
Drinking to cope. The view that individuals with anxiety disorders are at risk for SUDs because of escalating substance use aimed at “self-medication” is ubiquitous relative to its limited empirical support. Although it is clear that alcohol’s acute effects can include attenuation of stress and anxiety responding, there is little evidence that this effect alone accounts for the progression to alcohol and drug dependence among those with anxiety disorders.12,13
We found that only about 1 of 5 patients with an anxiety disorder reports ever having used alcohol explicitly for anxiety relief.5 Indeed, patients with anxiety disorders who self-medicate were at increased risk for SUD; however, patients who did not self-medicate were still at greater risk for SUD than those with no anxiety disorder. The amount of alcohol consumed accounted for only about half of the increased risk of alcohol dependence for patients with anxiety disorders who self-medicate. For patients with anxiety disorders who do not self-medicate, the amount of alcohol use did not account for any of their increased alcohol use disorder risk. Calling into question the centrality of self-medication in maintaining the anxiety disorder–SUD association, Thomas and colleagues14 presented data that show that successfully treating anxiety disorders, even among a subgroup who report drinking to self-medicate, does not eliminate (or even significantly reduce) pathological drinking.
Substance-induced anxiety disorders. DSM-IV establishes an anxiety disorder as independent of a co-occurring SUD only if it began before the onset of any SUD and persisted during periods of abstinence from drug and alcohol use that lasted at least 4 weeks. The second of these criteria indicates that substance-induced anxiety disorders remit in the absence of ongoing drug or alcohol use and, therefore, do not require specific treatment.15
Grant and colleagues1 used DSM-IV criteria to evaluate the extent of independent versus dependent anxiety disorders and mood disorders in individuals with SUDs from the NESARC dataset. They reported that “only a few individuals were classified as having only substance induced-disorders.”1(p807) Furthermore, because anxiety disorders begin after SUDs for up to half of those with both disorders, it is the second criterion (the resolution of anxiety disorder symptoms with abstinence) that was extremely rare in the NESARC sample.16 On the basis of these findings, Grant and colleagues1 concluded that clinicians should rarely withhold treatment for anxiety disorders simply because an individual has a comorbid SUD, even if the SUD began before the anxiety disorder.
However, the conclusion that anxiety disorders are not likely to resolve when SUDs are treated should not be taken to mean that pathological use might not contribute to the maintenance of anxiety disorder symptoms. In fact, there is evidence to suggest that pathological substance use does exacerbate anxiety symptoms, and it renders anxiety disorders more persistent and less responsive to treatment.17,18 Thus, as with the self-medication view, the idea of substance-induced anxiety disorders as an explanation for the anxiety disorder–SUD correlation is only partially supported; ie, while SUDs worsen anxiety disorders and anxiety disorder outcomes, simply eliminating problematic drug and alcohol use does not routinely resolve established anxiety disorders.
Dependence susceptibility and the “telescoping” of SUD risk. Telescoping refers to an accelerated transition from a nonpathological to a pathological state.19 Basic neuroscience has long linked disruptions in neurosystems, such as the hypothalamic-pituitary-adrenal axis and the basolateral amygdala, to the development of both anxiety disorders and physical substance dependence.20-22 On the basis of this etiological overlap, we hypothesized that individuals with anxiety disorders should be especially susceptible to developing physical dependence on alcohol, manifested in a telescoping of the time from early drinking landmarks (eg, first regular drinking) to alcohol dependence landmarks (eg, first withdrawal symptoms). This telescoping effect has been found in alcohol-dependent patients who have an anxiety disorder compared with those who do not have an anxiety disorder.23 Furthermore, the telescoping effect was observed whether the anxiety disorder preceded or followed the SUD.
Treatment implications and recommendations
The data reviewed paint a picture of anxiety disorder symptoms and substance use locked into a vicious circle of positive feedback. Anxiety symptoms can lead to more use via attempts at self-medication while heavy and long-term use can exacerbate anxiety symptoms. Moreover, treating either disorder alone is unlikely to resolve the untreated problem and runs a high risk of treatment nonresponse and relapse. Therefore, both problems must be treated simultaneously to maximize the chance of good clinical outcomes.
Unfortunately, as Watkins and colleagues24 have noted in a comprehensive review of the relevant literature, little empirical evidence exists to guide treatment in persons with comorbid anxiety and SUDs. In the absence of randomized controlled trials, expert opinion and consensus panels (ie, American Psychiatric Association [APA] guidelines25 and Substance Abuse and Mental Health Services Administration’s Treatment Improvement Protocol Series26) provide rational and experience-based treatment guidelines.
Low and moderate risk of SUDs. Each of the screening instruments comes with its own algorithms describing scores that correspond to low, medium, and high levels of drug/alcohol use and SUD risk. For patients at low risk for SUDs, we recommend that the clinician provide education about the association between anxiety disorders and SUDs, ie, patients with an anxiety disorder are at higher risk for SUDs and can develop SUDs with less overall consumption compared with patients who do not have an anxiety disorder. For patients at medium risk, ie, patients who self-medicate or score in the elevated but not high-risk range of the screening instruments, we recommend that the treating clinician provide a brief inter-vention that involves advice and assistance focused on the substance use (Table 5). If individuals fail to modify their use after the brief intervention, we suggest their risk assessment be elevated to high.
High risk of SUDs. We recommend referral to an SUD specialist for patients at high risk for an SUD based on screening and those at medium risk who failed to modify their substance use following the brief intervention. There are 3 treatment models for comorbid SUD and anxiety disorder: sequential, parallel, and integrated.
The sequential model is no longer the standard for treating comorbid disorders; simultaneous treatment is now recommended.24,27-29 In the parallel model, both the anxiety disorder and SUD are treated simultaneously, usually by 2 clinicians or even at different facilities. A further evolution of the parallel model is the integrated model, with 1 program for treating and monitoring both disorders simultaneously. Integrated treatment is currently considered to be the optimal option; however, such programs are, unfortunately, not available in all areas.
Jake is a 32-year-old who presents for help with his intense anxiety when giving presentations at work. He reports fear of embarrassment in many social situations. The patient is asked about his alcohol use and the pattern of drinking associated with his anxiety symptoms. He replies that he only drinks on the weekends, 6 beers on Fridays and Saturdays when he is faced with socializing with his coworkers. The patient does not report any problems related to his drinking.
This vignette is an example of excessive drinking to self-medicate anxiety. The physician recommends an SSRI for the patient’s social anxiety but also provides feedback about his drinking. This includes telling the patient about the NIAAA drinking guidelines for men being no more than 5 drinks in a single occasion and no more than 14 drinks in a week. The physician would follow this information with a brief intervention such as that shown in Table 5. If the patient is unable to cut back or stop drinking, further discussion and a potential referral are warranted.
Mary is a 42-year-old with a long history of panic attacks for which she has received various medications and psychotherapies. She is taking 4 mg of alprazolam daily but still reports intense ill-defined anxiety as well as panic attacks weekly. She admits to sometimes taking up to 6 mg of alprazolam daily and being preoccupied with her medication and obtaining it. She then runs out early and becomes shaky, sweaty, and more anxious.
This vignette points out some of the complexities in the assessment of persons who present to the psychiatrist’s office. The diagnosis of panic disorder seems clear, but the nature of the ill-defined anxiety needs to be elucidated. Is Mary’s anxiety a consequence of withdrawal from alprazolam or of another comorbid disorder, such as generalized anxiety disorder? Or is it a symptom and not a disorder? It appears that she meets criteria for sedative dependence on the basis of tolerance, withdrawal, loss of control, preoccupation.
Treatment should be directed not only at the panic attacks but also at her dependence on benzodiazepines. Given the moderate-high dose of alprazolam, it is likely that she will need inpatient stabilization for withdrawal management and alternative medication as the beginning of a substance use treatment. An integrated program that has the capacity to simultaneously manage the anxiety as well as the substance dependence is important.
Whether treatments are integrated or parallel, there are some unique concerns when treating anxiety disorders in patients with an SUD or who are in recovery from an SUD. First, because most empirically supported treatments for anxiety disorders were studied in samples that excluded SUDs, it is important to establish that these treatments are also effective in patients with an SUD. On the basis of their meta-analysis of 15 relevant randomized controlled trials, Hobbs and colleagues30 concluded that both cognitive-behavioral therapy and pharmacotherapy for anxiety and depression had significant effects in patients with concurrent SUDs. This suggests that standard anxiety treatments are effective in patients with comorbid SUDs. Table 6 presents approaches based on APA guidelines concerning pharmacotherapies for patients with anxiety disorders who have a current or past comorbid SUD.31
1. Grant BF, Moore TC, Kaplan K. Source and Accuracy Statement: Wave 1 National Epidemiologic Survey on Alcohol and Related Conditions (NESARC). Bethesda, MD: National Institute on Alcohol Abuse and Alcoholism; 2003.
2. Dawson DA, Grant BF, Li TK. Quantifying the risks associated with exceeding recommended drinking limits. Alcohol Clin Exp Res. 2005;29:902-908.
3. Kushner MG, Sher KJ, Erickson DJ. Prospective analysis of the relation between DSM-III anxiety disorders and alcohol use disorders. Am J Psychiatry. 1999;156:723-732.
4. Christie KA, Burke JD Jr, Regier DA, et al. Epidemiologic evidence for early onset of mental disorders and higher risk of drug abuse in young adults. Am J Psychiatry. 1988;145:971-975.
5. Menary KR, Kushner MG, Maurer E, Thuras P. The prevalence and clinical implications of self-medication among individuals with anxiety disorders. J Anxiety Disord. 2011;25:335-339.
Click here for a full list of references.