Stephen Ross, MD Associate Professor of Psychiatry and Child & Adolescent Psychiatry at New York University School of Medicine. Director of Addiction Psychiatry, NYU Tisch Hospital.
Dr. Ross has disclosed that he has no relevant financial or other interests in any commercial companies pertaining to this educational activity.
CATR: Why do you think clinicians should pay attention to psychiatric disorders co-occurring with addiction? Dr. Ross: The reason is that they’re very common. For example, about 80% of people with schizophrenia have nicotine use disorder, and 50% have a non-nicotine substance use disorder (SUD) (Miller SC, Fiellin DA, Rosenthal RN, Saitz R, eds. ASAM Principles of Addiction Medicine. 6th ed. Philadelphia, PA: Wolters Kluwer, 2019:1401–1417). And just about any psychiatric disorder has a higher rate of addiction comorbidity than in the general population. The rates are even higher in treatment settings—more than 50% of people in addiction treatment settings will have a co-occurring psychiatric problem, and more than 50% in psychiatric settings will typically have a co-occurring SUD (https://tinyurl.com/y3fly9u2). The treatment outcome is going to be poor if you’re not addressing the addiction part, which is often missed if clinicians don’t look for it. It’s important to recognize this multiplicity of diagnoses so you can form a treatment plan that addresses both psychiatric and addiction issues using all available tools.
CATR: Any thoughts on why comorbidity is so common? Dr. Ross: There are different plausible explanations. One is the self-medication hypothesis. Drugs, if used acutely, can be very potent psychopharmacologic tools. They make people feel good, and they are used in an attempt to treat an underlying symptom. However, they have enormous side effects, including addiction. You can also argue that people with certain psychiatric disorders may have dopaminergic deficiencies in the parts of the brain involved in the reward system. There are also certain substances that cause long-term psychiatric issues. So, if people drink alcohol very heavily over decades, they can develop enough damage to cause a chronic cognitive or psychotic disorder, even if they stop drinking.
CATR: Does the self-medication hypothesis mean that by treating the psychiatric problem early, we may prevent addiction from developing? Dr. Ross: The self-medication hypothesis is not perfect, but there is something to it. Let’s say you take ADHD—40% to 50% of people with ADHD will develop a drug problem. Even more will develop tobacco addiction, and there are good data that, if you treat somebody with ADHD with a stimulant, it either decreases the patient’s longitudinal risk of developing an SUD or does not increase the risk (Miller SC, Fiellin DA, Rosenthal RN, Saitz R, eds. ASAM Principles of Addiction Medicine. 6th ed. Philadelphia, PA: Wolters Kluwer, 2019:1418–1435). Another example is co-occurring major depressive disorder (MDD) and alcohol use disorder (AUD). The data show that treatment with antidepressants may not independently get the patient to drink less, but if the antidepressant works for the depression, then secondarily the drinking gets better. But none of this means that by just getting at the psychiatric problem, everything will flow downhill and the addiction will go away. You would hope there would be some benefit, but you also independently must treat the addiction. You would never treat co-occurring disorders with just a psychopharm agent for the psychiatric problem and hope the whole thing gets better.
CATR: Any advice on diagnosing comorbid psychiatric disorders? And specifically, how do you differentiate between primary and substance-induced disorders? Dr. Ross: The main thing is you want to take a very careful longitudinal history, and you want to establish temporally what came first. If you get a clear history of patients who had depression in their 20s and addiction in their 30s, and they were depressed for many years without ever being addicted, you can rule in that they have an independent depressive spectrum disorder. But, it’s often more complicated, and you may get patients who say, “I’ve been depressed and drinking forever.” In that situation you would want to help them get abstinent and observe them for a period. Family history can also help; if there’s strong family history for MDD, that points to a primary disorder. Another thing is symptom severity—if somebody meets 9 out of 9 criteria of major depression and gets suicidal, or gets psychotic depression, the severity there speaks to the fact that the patient probably has underlying MDD independently. So, careful history, family history, severity of illness, and following somebody longitudinally can all help sort out whether this is a co-occurring or a substance-induced illness.
CATR: Are there general medication strategies you’d recommend for treating co-occurring disorders? Dr. Ross: You want to match the medications to the patient’s needs to maximize adherence. This is true for all patients, but it may be even more pressing in people with both addiction and psychiatric issues. So, if somebody has psychosis and tends to not take medications, you’d want to give the patient a long-acting injectable formulation if you could. You can also think of an anti-addictive agent with an extended duration of action, such as extended-release naltrexone (Vivitrol) or the new formulation of subcutaneous buprenorphine. Another strategy is to directly observe taking the medication. Antabuse is an amazing medicine, but patients often stop taking it. So, if you build in that somebody will take it in the office, 3 times a week for example, you can enhance adherence. And the other thing is to treat early and treat aggressively.
CATR: Do you have advice on choice of meds for specific psychiatric disorders co-occurring with addiction? Dr. Ross: Overall, you want something that can affect symptom domains in both disorders. For instance, if somebody has MDD and tobacco addiction, you’d think of bupropion, which can treat both. If you have bipolar disorder in adults, lithium may not be the best choice in patients that have co-occurring addiction. This may be because people with addiction are more likely to have mixed and rapid cycling features, and they seem to respond better to valproic acid (Cipriani A et al, Cochrane Database Syst Rev 2013;10:CD003196). If somebody has bipolar disorder and co-occurring substance use, I would think of valproic acid over lithium as a first-line treatment.
CATR: Are there data on treating co-occurring schizophrenia? Dr. Ross: If you have schizophrenia that co-occurs with an SUD, you would want to consider starting with an atypical over a typical antipsychotic. This is because although the data are somewhat mixed, there is evidence that the atypicals are superior to typical antipsychotics in patients with co-occurring schizophrenia and an SUD, and there are data to suggest that typical antipsychotics are either not associated with decreased substance use or may actually make it worse (Miller SC, Fiellin DA, Rosenthal RN, Saitz R, eds. ASAM Principles of Addiction Medicine. 6th ed. Philadelphia, PA: Wolters Kluwer, 2019:1401–1417). In contrast, there are data for several atypical antipsychotics (ie, clozapine, risperidone, olanzapine, and aripiprazole) being associated with both a reduction in psychotic symptoms and substance use in this dually diagnosed cohort. Of the atypicals, the best data are for clozapine. Of course, you may not start with clozapine because of the side effects. But overall, you’d want to go with atypicals before you would use a typical.
CATR: What about comorbid anxiety? Should we avoid benzos? Dr. Ross: Yes. If you have an anxiety disorder and a co-occurring substance problem, you do not want to use benzos as first-line agents. You would want to use an antidepressant like an SSRI or an SNRI and try a course of cognitive behavioral therapy (CBT). You’d also want to try psychotherapy, especially for PTSD, because none of the meds work particularly well for PTSD.
CATR: Are there situations when you would add a benzo? Dr. Ross: You’d want to get the patient abstinent before considering a benzo. But let’s say you’ve tried SSRIs and CBT, and the patient is still highly symptomatic and is abstinent. In this situation, you could potentially treat the patient with clonazepam, which is a benzo with relatively low addictive liability, and conduct close monitoring. There are no hard and fast rules. Some people say you should never give an addictive substance to somebody with addiction, and that’s just not true because there are certain situations like the one I mentioned.
CATR: There’s also the issue of whether to prescribe stimulants for ADHD co-occurring with addiction. What are your thoughts on that? Dr. Ross: The first step would be to get the patient abstinent. You might want to start with something like atomoxetine or clonidine. But if that doesn’t work and the patient is abstinent, then you could, with careful monitoring and a contract, prescribe the stimulant. You’d also give the patient an extended-release formulation, to lower the risk of misuse.
CATR: What about patients who continue to use cocaine, for example, and yet do have ADHD? Would you then intervene with a stimulant, or would you prescribe it only if the patient is sober? Dr. Ross: I would make it contingent on the patient being sober, because if someone has ADHD and has, let’s say, an alcohol and a cocaine problem, this patient is at high risk of having a seizure. And if you’re adding another stimulant on top of that, you could do a lot of harm. I would say to the patient, “I will not give it to you until we have a stable recovery plan.” If the patient is in stable recovery and still highly symptomatic after everything you’ve tried for the ADHD, at that point I would consider a stimulant. You’d also want to provide close monitoring: one week at a time, regular urine drug screens, and a treatment contract.
CATR: Going back to the antidepressants, there are some data on the effect of SSRIs on drinking based on age of onset of AUD. Should we take that into consideration when choosing an antidepressant? Dr. Ross: Yes, there is an old literature that people with early-onset alcoholism have more family history of addiction and tend to have antisocial traits, and that SSRIs can worsen their drinking. Not all the studies show that, but there are enough to suggest something might be going on there. And there’s also the literature on antisocial personality and serotonergic dysfunction, so we have a plausible biological explanation for how SSRIs may be further disrupting serotonergic pathways that are making the antisocial issues and the drinking worse. So, in somebody with early-onset alcoholism, I may not use an SSRI as a first-line medication. I would try something with a different mechanism of action, like mirtazapine.
CATR: We’ve mostly spoken about meds so far. Any advice on choosing the right psychotherapy for co-occurring addiction and psychiatric disorders? Dr. Ross: Similar to choosing the right psychopharm agent, ideally you want to pick something that targets symptoms of both. For example, for patients with borderline personality disorder and co-occurring addiction, there are data from controlled trials that dialectical behavior therapy can decrease symptoms of both. This approach has not always panned out, though—there was dual focus schema therapy that Sam Ball developed at Yale, with the idea of treating the personality problem and the addiction, but it did not end up showing effectiveness for both. But even if the data aren’t too robust for these combined psychotherapies, you want to think of a psychotherapy strategy that can address both diagnoses. So, if someone has depression and addiction, you can use CBT principles to target both problems. And you want to make sure the patient is in the appropriate level of care and is getting all psychopharm and psychosocial treatments in the appropriate dose over a long-enough period. Addiction is a chronic illness, and you need a long-term treatment plan.
CATR: Yes, we often hear about integrating treatment. What do you think about doing it sequentially instead, by first treating one disorder then the other one? Dr. Ross: It can be hard to access, but the literature’s very clear that integrated treatment outperforms sequential or parallel treatment (https://store.samhsa.gov/system/files/theevidence-itc.pdf). In the integrated model, you don’t treat the addiction first and then hope to treat the psychiatric disorder later, or treat the psychiatric disorder first and then the addiction—you treat both in the same setting aggressively early on. The data show that patients do better with integrated care models than with treatment delivered sequentially or in parallel. Parallel means going to one place to get addiction treatment and then to a different place to get psychiatric treatment. It’s better if everything is delivered in the same setting at the same time.
CATR: Any additional advice for clinicians treating co-occurring disorders? Dr. Ross: Clinicians should remember that patients with co-occurring illnesses are very treatable—the trick is getting the right components in an integrated setting. Patients labeled as “treatment refractory” usually have not gotten anywhere close to the correct diagnosis or integrated care. And when they do get that correct care, they tend to do well. We have so many effective tools, and it’s important that we integrate them in the right ways.