Side-effects of SSRIs disrupt multimodal treatment for pediatric OCD in a randomized-controlled trial

Highlights

• We use advanced statistical modeling to study Activation Syndrome.

• Activation Syndrome disrupts multimodal treatment for Obsessive-Compulsive Disorder.

• Irritability, akathisia, and disinhibition are most associated with worse outcome.

• Weekly changes in irritability can immediately disrupt treatment.

Abstract

Objective

Activation Syndrome (AS) is a side-effect of antidepressants consisting of irritability, mania, self-harm, akathisia, and disinhibition. The current study was conducted to analyze how AS may hinder treatment outcome for multimodal treatment for children and adolescents with Obsessive-Compulsive Disorder.

Methods

Fifty-six children or adolescents were recruited at two treatment sites in a double-blind randomized-controlled trial where participants received Cognitive-Behavioral Therapy and were randomized to slow titration of sertraline, regular titration of sertraline or placebo.

Results

Using a recently developed measure of AS, results suggested that higher average levels of irritability, akathisia, and disinhibition significantly interfered with treatment response and explained 18% of the variance in obsessive-compulsive symptoms during treatment. Interestingly, only session-to-session increases in irritability resulted in a session-to-session increase in obsessive-compulsive symptoms. The observed results were unchanged with the addition of SSRI dosage as a covariate.

Conclusions

Results provide empirical support for the proposed hypothesis that AS may hinder multimodal treatment outcome for pediatric OCD. These findings suggest that dosage changes due to AS do not explain why those with higher AS had worse multimodal outcome. Other possible mechanisms explaining this observed disruption are proposed, including how AS may interfere with Cognitive-Behavioral Therapy.

Introduction

In 2004, the United States Food and Drug Administration (FDA) issued a “Black Box Warning” on antidepressants such as selective serotonin reuptake inhibitors (SSRIs), cautioning that these medications were associated with increased suicidality in children and adolescents (FDA, 2004). According to this FDA report, suicidal thoughts and behaviors were twice as high in children taking antidepressants (4%) compared to placebo (2%) in a pooled analysis of clinical trials. Research spurred by the warning has identified several behavioral side-effects of SSRIs beyond suicidality, including anxiety, agitation, panic attacks, insomnia, irritability, aggressiveness, impulsivity, psychomotor restlessness, hypomania, and mania (Murphy et al., 2008, Sinclair et al., 2009).

A variety of terms refer to these symptoms, including jitteriness/anxiety syndrome (Sinclair et al., 2009) or activation syndrome (AS; Murphy et al., 2008), the latter being the title used in this manuscript. Activation Syndrome has been estimated to occur in 4–65% of children taking SSRIs (Sinclair et al., 2009). The variability of this estimate is largely due to an absence of any comprehensive assessment measure of AS. Recently, the Treatment-Emergent Activation and Suicidality Assessment Profile (TEASAP) was developed to remedy this problem, providing a psychometrically validated method to comprehensively assess AS in children and adolescents receiving SSRI treatment (Bussing et al., 2013). The measure identifies five clusters of SSRI-induced side effects, including irritability, akathisia, disinhibition, mania, and self-harm. Utilizing this new measure, clinicians are now better positioned to track AS during SSRI administration and may be more effective at modulating dosage if AS emerges.

Indeed, adverse side-effects from antidepressants have been repeatedly linked to treatment discontinuation (e.g. Goldstein and Goodnick, 1998, Safer and Zito, 2006, Bloch et al., 2010, Schirman et al., 2010). While SSRIs are both safer and more tolerable than older medications like tricylclic antidepressants, meta-analytic findings suggest that up to 12–26% of patients taking SSRIs may drop out due to side-effects (Anderson, 2000, Usala et al., 2008). In addition to being more likely to drop-out, some older research suggests that antidepressant side-effects may hinder pharmacological treatment outcome (e.g. Gorman et al., 1987, Pohl et al., 1988), though these studies were limited by small sample size (Gorman et al., 1987), investigated currently outdated antidepressants (Pohl et al., 1988), and lacked a psychometrically validated measure of these side effects. To the best of the authors' knowledge, no study to date has further investigated how side effects may interfere with pharmacological or multimodal treatment outcome beyond these two older studies. These types of analyses are likely underreported because of methodological complications. Without advanced statistical modeling, it could be difficult to have enough power to analyze the impact of side effects that are waxing and waning throughout treatment. As stated above, a comprehensive measure of AS has only recently been developed as a tool towards this endeavor.

Multimodal treatment, or combined pharmacology-psychotherapy, has been identified as an effective and efficacious treatment for a variety of internalizing disorders (Davidson, 2010, Walkup et al., 2008), including Obsessive-Compulsive Disorder (OCD; Geller and March, 2012, Koran et al., 2007). Considering that Cognitive-Behavioral Therapy with Exposure and Response Prevention (CBT-ERP; Franklin et al., 2015) and SSRIs (Sánchez-meca et al., 2014) each have moderate to large treatment effect sizes for pediatric OCD, it is logical that a multimodal approach would be the first-line treatment regimen for this population (see Jordan et al., 2012 for a review). Combined CBT-ERP with SSRI is currently recommended by numerous organizations, such as the Academy of Child and Adolescent Psychiatry (Geller & March, 2012). Due to its frequency of use and the high doses of SSRIs required to treat pediatric OCD (Bloch et al., 2010), pediatric OCD would be a logical population to investigate how AS may impact multimodal treatment outcome.

This study examined the extent to which symptoms of AS hinder outcome during multimodal treatment of pediatric OCD. It was hypothesized that higher average AS symptoms would be associated with less symptom reduction during treatment (i.e., moderate treatment response). Additionally, session-to-session increases in AS symptom severity was expected to predict session-to-session increases in obsessive-compulsive symptom severity. If expected results are observed, increased monitoring and faster dose adjustment when AS is noted could enhance multimodal treatment outcomes.