2010 in Review

The study by Helen M. Pettinati and colleagues (1) is a straightforward approach to the comorbidity of alcoholism and depression. Psychiatrists have always recognized the high likelihood of co-occurrence of these two conditions and often assumed that they were both manifestations of the same illness. Surprisingly, the recognition of comorbidity rarely translated into a well-articulated joint treatment. Instead, advocates for alcoholism treatment seemed to believe that vigorous treatment of alcoholism would result in resolution of a secondary depression, and vice versa, advocates for mood disorder treatment seemed to believe that vigorous treatment of the mood disorder would resolve the alcoholism. Pettinati et al. clearly document that neither assumption is true, and in the process they validate a new treatment standard, the simultaneous prescription of naltrexone for alcoholism and the selective serotonin reuptake inhibitor sertraline for depression. Patients receiving the combination had double the likelihood of abstinence and double the duration of abstinence with fewer adverse side effects compared to patients who received only one treatment. In a period when we have been disappointed by the effectiveness of treatments for mood disorders, it is heartening and instructive to consider the value of considering comorbidity as a new therapeutic target, rather than a nuisance.

Both the role of cannabis as a medicinal agent and the risks associated with illicit cannabis use remain topics of debate in our society. During 2009, 7.3% of 12–17-year-olds reported using marijuana, an increase of nearly 10% from the year before; at the same time, the proportion of this age group who thought smoking marijuana carried a great risk of harm declined. The article by Daniel J. Foti and colleagues (2) brings an important perspective to these issues by highlighting the frequency and impact of cannabis use in a subset of the population. Specifically, these investigators found that two-thirds of individuals with schizophrenia had a lifetime history of cannabis use prior to first hospitalization and that cannabis use was associated with both an earlier onset of psychotic symptoms and more severe symptoms. In turn, individuals with more severe psychotic symptoms were more likely to use cannabis in the future. These findings strengthen the suggestion that interventions to reduce cannabis use during adolescence in vulnerable individuals could be disease modifying. In addition, given the substantial recent advances in our understanding of cannabinoid signaling in the brain, they may inform the development of novel therapeutic approaches for the treatment of psychosis.

Mental health care in the United States has changed over the last decade. Mark Olfson and Steven C. Marcus (3) analyzed data between 1998 and 2007 to describe those changes, with a special emphasis on psychotherapy. Several things have not changed— the overall percentage of persons using psychotherapy, who is most likely to receive psychotherapy (female, white, single, well educated, and unemployed), and the fraction of psychotherapy provided by psychologists or social workers. Several other things have changed, some dramatically: 1) the number of individuals receiving any mental health services (a 44% increase), 2) the percentage receiving psychotherapy alone without medication (a 34% decrease), 3) the percentage receiving medication alone (a 30% increase), 4) the mean number of visits per psychotherapy patient (an 18% decrease), and 5) expenditure per psychotherapy visit (a 23% decrease).

The economic data are striking when viewed on a national basis. During this period, expenditures for outpatient medical care, adjusted for inflation, increased 64%. Expenditures for outpatient mental health care increased 4%. However, this almost constant figure is composed of two radically disparate components: psychotherapy decreased 35% while nonpsychotherapy mental health services increased 98%. This means that expenditures for nonpsychotherapy outpatient mental health care increased at a greater rate than did costs for general medical care, while psychotherapy expenses decreased significantly.

In summary, individuals are receiving psychotherapy from the same provider mix with the same socioeconomic biases as in the past, but many more individuals receive nonpsychotherapy mental health services. The number of visits per psychotherapy patient, the cost per visit, and the percentage of mental health expenditures devoted to psychotherapy have all decreased significantly.

Olfson and Marcus have mapped dramatic changes in outpatient mental health care. Their study does not extend to causes, but they note a more than tripling of the number of individuals covered by managed care during this period. The most critical question remains unanswered: Has the cost-benefit ratio of mental health expenditures increased or decreased?

I find myself “rooting” for randomized, controlled trials. Not only are they rare, important examples of clinically relevant “experiments,” but they hold hope for improving how we treat our patients. Sadly, few chances arise to celebrate randomized, controlled trials. Relatively few are submitted to the Journal. Moreover, perhaps because they are so expensive and burdensome to complete, those we do receive often appear to “play it safe”: some contrast novel treatments against minimally effective control conditions; others select novel treatments that resemble known, only moderately effective therapies. Because Marylene Cloitre and colleagues went beyond these hurdles, I am thrilled to select their article as my 2010 “favorite” (4).

In the study, 104 patients with posttraumatic stress disorder (PTSD) were randomly assigned to one of three treatments. Compared to two control treatments, the study showed superiority for a third one that combined the well-worn features of exposure therapy with novel techniques that teach patients emotion-regulatory and interpersonal skills. While almost any properly performed randomized, controlled trial is noteworthy, two features of this one set it apart. First, the study showed that a novel treatment was superior both to a minimally effective control condition and to a standard treatment. Second, the novel treatment emerged from observational research on core deficits in PTSD, thereby providing a guide for developing other novel treatments. As research in genetics and neuroscience generates increasingly deep insights on pathophysiology for a range of mental illnesses, this “translational” strategy adopted by Cloitre and colleagues can be applied to focus on an expanding wealth of treatment targets.

Terry E. Goldberg and colleagues (5) call for a reconceptualization of the relationship between cognition and function in mild cognitive impairment. Mild cognitive impairment has typically been considered a condition free of functional decline, but the authors have now demonstrated that they can detect early functional deficits in these patients by using the University of California, San Diego (UCSD) Performance-Based Skills Assessment (UPSA). In a sample of 55–85-year-olds, Goldberg et al. detected a large effect size in the difference between those with mild cognitive impairment and a cognitively healthy group in scores on the UPSA. These subtle functional changes may provide benchmarks for early detection of illness. These early-appearing deficits may also represent a new target for treatment efforts whose benefits can prolong the quality of life for older adults. This concern will reach greater public health significance in the next two decades, when the percentage of people over age 65 in the United States will exceed 20%. The continuum of normal aging to mild cognitive impairment is thus relevant to an already large population that will grow dramatically.

It is always refreshing to see a clinical trial designed and executed by academic psychiatrists to answer real-world treatment questions. The article by Pierre Blier and colleagues (6) is one of those studies, asking which antidepressant strategy is best for serious depression. Blier et al. carried out a complex but clear study that answers a simple question about the efficacy of antidepressant monotherapy versus combined therapy in depression. The main study was a 6-week trial of monotherapy (fluoxetine/placebo) versus three combination therapies (fluoxetine/mirtazapine versus venlafaxine/mirtazapine versus bupropion/mirtazapine) at usual doses. The 105 depressed volunteers were randomly assigned to the four groups, the study medications were blinded, and classical outcome measures were used. It may have been important that the majority of patients were “melancholic.” The answer was perfect and perfectly simple. Although depressive symptoms were reduced with the monotherapy, all three combination treatments performed better and were equally successful. Different analytic outcome analyses converged on the same conclusion. The article was not burdened with complex mechanistic explanations, only a simple but sensible conclusion (given our state of knowledge) that “the greater the number of neuronal elements recruited to enhance 5-HT and norepinephrine transmission, the greater the potential therapeutic benefit” (p. 286). These results reminded me of the early combined-therapy AIDS studies, which converted AIDS treatment from mono-therapy to combination treatment, and the real jump in good outcomes that followed.

Editorial accepted for publication September 2010

Disclosures of The American Journal of Psychiatry editors are published in each January issue. Dr. Pine is serving in a personal capacity; the views expressed are his own and do not necessarily represent the views of NIH or the U.S. government.