Tick marks show censored observations. NNT indicates number needed to treat; NNTB, number needed to treat to benefit; and NNTH, number needed to treat to harm.
eTable 1. Neuropsychological Functioning at Baseline for Groups Receiving Sertraline or Placebo
eTable 2. Rehabilitation Treatment
eTable 3. Rehabilitation Treatment Services Use
eTable 4. Medication Use During the Course of the Study
eTable 5. Number Needed to Treat for Study Time Points With In-Person Evaluations
Customize your JAMA Network experience by selecting one or more topics from the list below.
Jorge RE, Acion L, Burin DI, Robinson RG. Sertraline for Preventing Mood Disorders Following Traumatic Brain Injury: A Randomized Clinical Trial. JAMA Psychiatry. 2016;73(10):1041–1047. doi:10.1001/jamapsychiatry.2016.2189
Prevention is more effective than treatment to decrease the burden of significant medical conditions such as depressive disorders, a major cause of disability worldwide. Traumatic brain injury (TBI) is a candidate for selective strategies to prevent depression given the incidence, prevalence, and functional effect of depression that occurs after TBI.
To assess the efficacy of sertraline treatment in preventing depressive disorders following TBI.
Design, Setting, and Participants
A double-blind, placebo-controlled, parallel-group randomized clinical trial was conducted at a university hospital from July 3, 2008, to September 17, 2012, with 24 weeks of follow-up. A consecutive sample of 534 patients aged 18 to 85 years, hospitalized for mild, moderate, or severe TBI, was eligible for the study. Ninety-four patients consented to participate and were randomized (46 to placebo and 48 to sertraline), of whom 79 (84%) completed the study. Intention-to-treat data analysis was conducted from July 1, 2014, to December 31, 2015.
Placebo or sertraline, 100 mg/d, for 24 weeks or until development of a mood disorder.
Main Outcomes and Measures
Time to onset of depressive disorders, as defined by the DSM-IV, associated with TBI.
Of the 94 patients in the study (38 female and 56 male; 92 white), the number needed to treat to prevent depression after TBI at 24 weeks was 5.9 (95% CI, 3.1-71.1; χ2 = 4.6; P = .03) for sertraline treatment vs placebo. The influence of sertraline in the course of neuropsychological variables was not detected. The intervention was well tolerated, and adverse effects were mild in both the sertraline and placebo groups.
Conclusions and Relevance
Sertraline appears to be efficacious to prevent the onset of depressive disorders following TBI. Future studies should replicate these findings in a large sample of patients with TBI and depict their long-term physical, cognitive, behavioral, and functional outcomes.
clinicaltrials.gov Identifier: NCT00704379
The incidence of traumatic brain injury (TBI) in the United States is approximately 1.7 million cases per year.1 The prevalence of disability associated with TBI ranged from 2.5 million to 6.5 million people, and the total annual cost of TBI has been estimated to be $60 billion.2 These figures attest to the importance of TBI as a public health concern.
Quiz Ref IDDepressive disorders are a common event after TBI. The reported frequency of depression following TBI has varied from 6% to 77%.3 Such heterogeneity is owing to differences in recruitment settings, the methods used to assess depression, and the timing of these assessments with respect to the index TBI.4-16
Several prospective studies have observed patients during the first year following TBI. Two studies reported that 58 of 157 patients (36.9%) developed a depressive disorder during this period.14,17 A larger study of 559 patients with TBI estimated the 1-year frequency of major depression as 53.1%,18 while another study of 102 patients with TBI found that the frequency of major depression was 29.4%.19
The prevalence of depression continues to be high in more chronic stages of TBI, with only a few patients recovering from depression at 2 years of follow-up.20 Furthermore, the prevalence of major depression was estimated to be 26.7% in a group of 60 patients assessed 30 years after incident TBI.21 These findings suggest that depressive disorders are not only a frequent complication of TBI; they are likely to have a chronic and refractory course.
One study initially reported that depressive disorders increase disability and have a negative effect on reintegration into the community after TBI.22 This finding has been extensively replicated by different investigators in independent samples of patients with TBI.18,23-27 Overall, these studies emphasize the need to identify depression and treat it accordingly.
Prevention is more effective than treatment to decrease the burden of major medical conditions such as infectious illnesses, cardiovascular disease, and cancer. This is also true in the case of depression.28 Traumatic brain injury is a candidate for selective strategies to prevent depression given the incidence, prevalence, and functional effect of depression occurring after TBI.29
Our study examines the efficacy of pharmacologic treatment to prevent the onset of depression following TBI. We hypothesized that the time from baseline to onset of depressive disorders would be greater in a group of patients randomized to receive sertraline treatment vs a group of patients randomized to receive placebo. We also hypothesized that, when compared with patients receiving placebo, patients receiving sertraline would show better performance in a set of neuropsychological tests after 6 months of treatment.
Question Does prophylactic administration of sertraline prevent depressive disorders following traumatic brain injury?
Findings In this double-blind, placebo-controlled, parallel-group randomized clinical trial that included 94 adult patients, the hazards for developing depression for participants receiving placebo were about 4 times the hazards of participants receiving sertraline treatment.
Meaning Preventive treatment may have a decisive effect in reducing the burden of depressive disorders and improving the recovery and quality of life of patients with traumatic brain injury.
This double-blind, placebo-controlled, parallel-group randomized clinical trial was conducted at the Departments of Psychiatry and Neurosurgery of the University of Iowa. Participants were randomized to receive identical tablets of sertraline, 100 mg/d, or placebo once daily for 24 weeks. The study protocol is detailed in Supplement 1. The protocol was approved by the University of Iowa institutional review board. Written informed consent was obtained from every participant.
Participants were enrolled between July 3, 2008, and September 17, 2012. Patients between 18 and 85 years of age with a diagnosis of mild, moderate, or severe TBI were included in the study. They were required to have complete recovery of posttraumatic amnesia within 4 weeks of the traumatic episode. Patients with ongoing depression were excluded. To be eligible for the study, participants with a history of mood disorders should have been in full remission for at least 1 year after discontinuation of treatment.
Patients were randomized using a permuted blocks randomization scheme. Other exclusion criteria and further details on randomization can be found in the study protocol (Supplement 1).
Quiz Ref IDThe primary outcome of this study was time to onset of a depressive disorder. Diagnosis of a depressive disorder was made using the Mini-International Neuropsychiatric Interview30 and DSM-IV criteria. Participants were evaluated in person at the baseline evaluation as well as at the 2-, 4-, 8-, 12-, 16-, 20-, and 24-week follow-up visits. In addition, the Mini-International Neuropsychiatric Interview was administered via telephone on weeks 6, 10, 14, 18, and 22. If, during a telephone interview, the participant reported the occurrence of the anchor symptoms of depressive disorders in the previous 2 weeks, he or she was evaluated in person using the same structured interview. The presence of mood disorders was ascertained by experienced psychiatrists (R.E.J. and R.G.R.) who were unaware of treatment allocation. In the case of patients with a history of depression, the latter should have been in full remission for at least 1 year after discontinuation of antidepressant treatment. Demographics, injury characteristics, disability, alcohol use, and psychiatric and psychosocial assessments are detailed in the study protocol (Supplement 1).
Neuropsychological functioning was assessed with a comprehensive battery of tests tapping the following cognitive domains: focused and selective attention, working memory, episodic memory inhibition and executive control, and general IQ. A full description of the neuropsychological tests that form part of each domain is included in the study protocol (Supplement 1).
Sertraline and placebo were administered in a double-blind fashion via an equal number of identical tablets. Sertraline, 25 mg/d, was given for 5 days, 50 mg/d for the next 5 days, and 100 mg/d thereafter (Supplement 1).
Participants were asked about adverse events (AEs) at 2-week intervals (or sooner if an individual reported an AE) using a standardized checklist developed for this study. A data and safety monitoring board comprising investigators not involved in this trial assessed aggregate AE data associated with the 2 treatments. Sexual dysfunction was assessed using the Arizona Sexual Experience Scale (ASEX).31
Intention-to-treat data analysis was conducted from July 1, 2014, to December 31, 2015. A priori sample size calculations and analyses used to describe the sample are included in Supplement 1. To analyze the data on the time to onset of depressive disorders, we built a Cox proportional hazards regression model using time to onset of mood disorders as the dependent variable and treatment group as the independent variable. This analysis included all patients who completed the study as well as those who withdrew after randomization (ie, 94 patients). The latter were considered in the model as censored observations at the time they stopped participating in the study. The proportional hazards assumption was tested and met. The frequency of AEs happening at least once during the study was compared between the groups using the Fisher-Boschloo statistic for an exact unconditional test as recommended by Lydersen and colleagues.32 Adverse events found to be different between the groups (at α = .10) were further examined using logistic regression marginal models. Dependency between the AE measures was modeled using an autoregressive variance-covariance structure. The effect of treatment on the ASEX scores was modeled using a similar model with an identity link function.
We screened 1053 patients aged 18 to 85 years with a diagnosis of closed TBI; 534 (50.7%) were found to be eligible. Reasons for ineligibility included the following: 37 patients had TBI severity that precluded evaluation or died soon after initial screening, 2 had evidence of penetrating head injury, 23 had evidence of spinal cord injury, 51 had a severe complicating medical illness, 14 had a neurodegenerative disorder, 18 had cerebrovascular illness or dementia, 145 had a diagnosis of alcohol or drug dependence, 37 had a concurrent mood or anxiety disorder, 20 had a concurrent psychotic disorder, 87 were already receiving antidepressants, 41 had received antidepressants sometime during the 6 months before the injury, 5 failed a previous trial with sertraline, 3 had AEs that caused discontinuation of previous sertraline treatment, 7 had nontraumatic subdural hemorrhage, 2 had a positive urine pregnancy test, 1 planned to become pregnant during the study, 7 had legal issues precluding study evaluation, 7 did not have TBI, 4 were not able to complete the screening evaluation, 3 were moving out of Iowa and follow-up visits would not have been possible, and 5 did not have a reason recorded. Of those eligible, 440 (82.4%) were not interested in participating in the study (66 patients were unwilling to commit to the time required for participation in the study, 20 lived too far for study visits, 52 had their families decline participation, 120 were unwilling to add other medication to the ones they were already taking, 30 decided to refuse placebo and to take antidepressants to improve recovery from TBI, 18 were going to a facility that does not accept external medication or the use of research drugs, 39 thought depression was not going to occur, 77 were out of reach, 2 gave legal reasons, 7 would not participate in research, 2 had jobs that do not allow them to take antidepressants, 3 did not want to be reminded of the accident owing to grief, 3 had too many health issues, and 1 patient did not have a reason recorded). The remaining 94 patients (17.6%) were randomized to receive either sertraline treatment or placebo (Figure 1). Most participants underwent their baseline evaluation between 3 and 4 weeks after TBI.
Demographic variables are summarized in Table 1. The sertraline and placebo groups were comparable with regard to variables such as age, sex, educational level, marital status, and occupational level.
Quiz Ref IDCharacteristics associated with injuries that correspond with the sertraline and placebo groups are summarized in Table 2. The mechanisms of injury that occurred most often in our patients were falls (45 [48%]) and motor vehicle crashes (36 [38%]).
Severity of TBI was determined using Glasgow Coma Scale scores and the findings on their admission computed tomographic scans. Seventy-four patients (79%) had Glasgow Coma Scale scores between 13 and 15. Of these 74 patients, 58 had intracranial abnormalities on the results of computed tomographic scans and were classified as having moderate TBI. Sixty-nine patients (73%) had moderate TBI, 16 (17%) had mild TBI, and 9 (10%) had severe TBI.
Baseline scores on psychiatric scales are depicted in Table 2. A history of mood disorders was present among 6 patients (12%) receiving sertraline treatment and in 4 patients (9%) receiving placebo. All of these patients were in full symptomatic remission for many years, and there was no history of recurrent mood episodes. Premorbid intellectual achievement, global measures of cognition, and scores in the 5 cognitive domains were comparable between the groups (eTable 1 in Supplement 2).
Descriptive statistics of rehabilitation interventions and other medications received during the study are included in eTables 2 through 4 in Supplement 2. The sertraline and placebo groups were similar regarding the frequency of these interventions. Of the 94 participants, 32 (34%) received some form of rehabilitation treatment, mostly physical therapy. None of the patients received cognitive rehabilitation therapy.
Quiz Ref IDThe number of patients needed to treat to prevent development of depression after TBI at 24 weeks was 5.9 (95% CI, 3.1-71.1; likelihood ratio test χ2 = 4.6; P = .03) (Figure 2) for sertraline vs placebo (eTable 5 in Supplement 2). All incident cases of depression had features of major depression. Although symptoms of anxiety were relatively frequent among patients with depression, there were no incident cases of anxiety disorders as defined by DSM-IV criteria. Suicide risk was assessed using the suicidality module included in the Mini-International Neuropsychiatric Interview. Suicidal ideation was observed in only 1 patient who developed major depressive disorder and was allocated to the placebo group. Thus, an effect of sertraline treatment on suicidal behavior was not detectable in this trial.
Given that selective serotonin reuptake inhibitors (SSRIs) can enhance neuroplasticity in multiple areas of the brain that participate in cognitive processing, we also investigated whether, among patients with TBI who did not develop a mood disorder during the 6-month follow-up period, those who were treated with sertraline had improved memory and executive function compared with those who received placebo. Thus, we examined whether sertraline had an effect on cognitive enhancement that was independent from its effect on mood.
Sixty-seven patients did not develop a mood disorder during the follow-up period and had neuropsychological evaluations at baseline and 24 weeks. The difference between the 24-week score and the baseline score was compared between treatments using the Wilcoxon rank sum test. Differences between the treatment group for the domains of attention, working memory, episodic memory, and executive control and inhibition were not detected. In addition, no influence of sertraline treatment on speed of information processing was detected. Similar results were obtained when we independently examined young and older (ie, ≥55 years) patients and when we examined the group of patients whose baseline scores on the Mini-Mental State Examination33 were more than 1 SD of the age- and educational level–adjusted normative mean.
Adverse events reported at least once during the study have been summarized in Table 3. Almost all patients experienced mild or moderate AEs. One patient in each treatment group experienced a serious AE. Both participants had a hip fracture requiring hospitalization. However, these AEs were adequately resolved and judged not to be associated with the study medication. After controlling for the effect of time, the odds of having dry mouth (odds ratio, 7.2; 90% CI, 1.9-27.6; P = .01) and diarrhea (odds ratio, 2.3; 90% CI, 1.0-5.5; P = .10) were higher for participants receiving sertraline treatment than for those receiving placebo. Overall, sexual AEs were mild and did not have a significant effect in the participants’ quality of life.
Quiz Ref IDOur findings suggest that sertraline given at a low dosage early after TBI is an efficacious strategy to prevent depression after TBI. The number of patients needed to treat with sertraline to prevent depression after TBI was approximately 6 when compared with placebo. Furthermore, the prophylactic effect was not restricted to the few participants with a remote history of depressive disorders. We did not observe a cognitive enhancement effect of SSRIs in this group of patients with TBI. The intervention was well tolerated, and AEs were mild in both the sertraline and placebo groups. However, the frequency of dry mouth and diarrhea was slightly higher among those receiving sertraline treatment.
The limitations of our study include the relatively small sample size, the fact that participants were recruited at a single emergency trauma center in Iowa City, the predominance of mild and moderate TBI in the study group, and the scarce representation of racial and ethnic minorities among the enrolled patients. In addition, follow-up was limited to 6 months after incident TBI. Although we are aware that longer trials might be associated with smaller effect sizes,34 previous studies on the incidence of depressive disorders during the first year following TBI indicate that approximately 80% of depressive episodes have their onset during the first 6 months.14,17 Furthermore, the first 6 months is the period when we observe most of the recovery from brain injury and when depression has a significant deleterious effect on functional outcome. Taken together, the aforementioned factors limit the generalization of these results.
Our results are consistent with previous reports on the efficacy of antidepressants to prevent depression after stroke.35 The fact that small doses of sertraline are efficacious to prevent depression after TBI stands in sharp contrast to the lack of efficacy of antidepressants to treat depression in the chronic stage of TBI. Ashman et al36 did not detect an effect of sertraline when compared with placebo to treat depression after TBI in a group of 41 participants enrolled in the trial at least 6 months after TBI. A meta-analysis of randomized clinical trials examining the efficacy of antidepressants to treat depression after TBI reported an overall effect size of Hedges g = 0.46 (95% CI, −0.44 to 1.36; P < .001), favoring groups receiving placebo in favor of those receiving active treatment.37 Overall, these results suggest that sertraline is efficacious to prevent the onset of depressive disorders when administered early after TBI but that it has a questionable antidepressant action when depressive disorders are already present and follow a chronic course. In addition, it seems likely that the smaller doses of sertraline used in preventive treatment are associated with fewer AEs than are the higher doses usually prescribed to treat depression.
A recent meta-analysis of randomized clinical trials of psychological interventions to prevent depression included 32 randomized clinical trials with 6214 participants.34 The analysis of the 32 trials estimated that the number needed to treat was approximately 20. In our study, the number needed to treat decreased from 14 to 5.9 as follow-up time increased. This finding compares favorably with the previous research.
We can only speculate about the biological substrate of sertraline’s preventive effect. It is increasingly recognized that enhanced neuroplasticity in mood regulation areas, such as the hippocampus and prefrontal cortex, is an important part of the antidepressant effect of SSRIs. The early stages of TBI are a significantly stressful period in which the individual must cope with the physical and psychological challenges associated with brain injury. It is plausible that sertraline can modify the effect of stress on vulnerable mood regulation circuits, modeling them in a way that promotes more adaptive responses. Furthermore, SSRIs may influence the balance between excitatory and inhibitory neurotransmission and normalize glucose metabolism in mood-regulating networks, an effect that might be associated in part with stimulation of 5HT-2B receptors.38
Findings from animal models support the hypothesis that antidepressants produce structural and functional changes in limbic networks mediating emotional processing.39 Furthermore, in a murine model of TBI, mice receiving sertraline since the seventh day after TBI and for 3 weeks had significantly fewer depressive behaviors and increased social interaction than did mice receiving placebo.40 These findings are consistent with what we observed clinically in our group of patients.
In contrast to the effects observed among patients with stroke treated with escitalopram,41 we did not observe a cognitive enhancement effect of sertraline. This finding might be associated with multiple factors, including differences in the pathophysiologic conditions of stroke and TBI, the type of SSRI prescribed, or the duration of therapy (escitalopram was given for 12 months to patients with stroke). The more parsimonious hypothesis, however, argues for the presence of a ceiling effect, as most of our patients with TBI had neuropsychological performance within normal limits.
This is the first trial, to our knowledge, indicating that antidepressants are efficacious to prevent the onset of depression following TBI. Given the prevalence and functional effect of depression among patients with TBI, these findings have profound therapeutic implications. However, although our findings are novel and provocative, recommending a change in the guidelines to treat patients with TBI requires replication of these findings in multicenter studies. In addition, it would be important to study whether combining antidepressants with behavioral interventions, such as psychotherapy or cognitive rehabilitation protocols, will optimize long-term functional outcomes.
Corresponding Author: Ricardo E. Jorge, MD, Menninger Department of Psychiatry and Behavioral Sciences, Baylor College of Medicine, One Baylor Plaza, Houston, TX 77030 (firstname.lastname@example.org).
Accepted for Publication: July 25, 2016.
Published Online: September 14, 2016. doi:10.1001/jamapsychiatry.2016.2189
Author Contributions: Dr Jorge had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.
Study concept and design: Jorge, Acion, Robinson.
Acquisition, analysis, or interpretation of data: Jorge, Acion, Burin.
Drafting of the manuscript: Jorge, Acion, Burin.
Critical revision of the manuscript for important intellectual content: Jorge, Robinson.
Statistical analysis: Jorge, Acion.
Obtaining funding: Jorge.
Administrative, technical, or material support: Jorge, Robinson.
Study supervision: Jorge, Robinson.
Conflict of Interest Disclosures: Dr Jorge reported receiving lecture honoraria from Xiang-Jansen Pharmaceuticals. Dr Robinson reported receiving compensation for participation in an advisory committee meeting sponsored by Avanir Pharmaceuticals, lecture honorarium from Xiang-Janssen Pharmaceuticals, research funding from the Senator Financial Group, and royalties from Cambridge University Press and serving as a consultant to Otsuka Pharmaceuticals. No other disclosures were reported.
Funding/Support: This study was funded by grant R01 NS055827 from the National Institutes of Health.
Role of the Funder/Sponsor: The funding source had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication.
Create a personal account or sign in to: