David J. Hellerstein

Can people with nonsevere major depression benefit from antidepressant medication

BACKGROUND Several meta- or mega-analyses suggest antidepressant medications should be given only to severely depressed patients. In our experience, mild depression benefits from medication. We reanalyzed 1 clinics randomized placebo-controlled antidepressant studies, limiting analyses to patients with major depressive disorder (MDD) without severe illness, to determine whether nonsevere depression responds to antidepressant medication. DATA SOURCES Archives of the Depression Evaluation Service outpatient clinic of the New York State Psychiatric Institute were searched for randomized, placebo-controlled antidepressant studies that were conducted between 1977 and 2009 and included patients having MDD and pretreatment Hamilton Depression Rating Scale (HDRS) scores < 23. STUDY SELECTION Six placebo-controlled studies were found, including 8 active treatment arms and 1,440 patients. 825 patients were randomized and had MDD and an HDRS score < 23. DSM-III, DSM-III-R, or DSM-IV diagnostic criteria contemporary to each study were employed. DATA EXTRACTION Treatments were compared within study and via a patient-level meta-analysis using analysis of covariance (ANCOVA) of HDRS end point scores adjusted for pretreatment score. The number needed to treat (NNT) was calculated from remission rates (HDRS end point score ≤ 7), which were compared by χ². Effect sizes were calculated from change in HDRS scores. Secondary analyses investigated the effect of chronicity and atypical features on treatment response. DATA SYNTHESIS Three of 6 studies showed significant (P < .001) treatment effects by ANCOVA, and 4 of 6 studies showed significant (P < .04) differences in remission. The NNT ranged from 3 to 8. Effect sizes ranged from -0.04 to 0.8, with 4 of 8 greater than 0.4. The patient-level meta-analysis confirmed these results; neither chronicity nor atypical features significantly affected outcome. Secondary analyses utilizing global ratings and self-report mimicked the main findings. CONCLUSIONS Several studies demonstrated significant antidepressant efficacy for patients having nonsevere MDD. Efficacy was not trivial, as NNT ranged from 3 to 8, a range accepted by researchers as sufficiently robust to recommend treatment. These findings suggest mild-moderate MDD can benefit from antidepressants, contrary to findings by several other meta- or mega-analyses.

Predicting Falls Among Psychiatric Inpatients: A Case-Control Study at a State Psychiatric Facility

OBJECTIVE The purpose of the study was to add to the research on risk of falling in an understudied population of psychiatric inpatients in an acute setting. METHODS Five years of fall data in an inpatient psychiatric facility, where falls were frequent but benign, were examined for patient and treatment characteristics that might be associated with falls. This was a retrospective analysis, which matched 1:1 the medical records of fallers and nonfallers on primary psychiatric diagnoses. The total sample consisted of 148 patients. Statistical analysis was conducted on patient demographic characteristics, summed medical history items reported, summed physical complaints on the day of the fall, the number and types of medications taken within a 24-hour period of the fall, and the patients vital signs. Multivariate logistic regression was used to identify the most salient associations with faller status. RESULTS Univariate analyses revealed that fallers were prescribed significantly more medications and complained of more physical symptoms on the day of their fall. Fallers were more likely to have a current acute medical condition and to be currently prescribed clonazepam or antihypertensive medication. Multivariate logistic regression analysis revealed that current physical complaints and current clonazepam treatment had significant associations with faller status. CONCLUSIONS Risk factors identified in this study should be assessed in replication studies. Psychiatric clinicians can use such risk factors to create evidence-based fall prevention programs.

Does Negative Affectivity Predict Differential Response to an SSRI Versus a Non-SSRI Antidepressant?

OBJECTIVE This work tested the hypothesis that patients with high negative affectivity (NA) would have a better response to a serotonergic agent (escitalopram) than to one not thought to act directly on serotonin (bupropion). METHOD Data from a study conducted between August 2007 and July 2011 were reanalyzed retrospectively. Patients (N = 245) meeting criteria for major depressive disorder (MDD), diagnosed with DSM-IV-TR, were randomly assigned to double-blind treatment with bupropion extended-release, escitalopram, or the combination. Negative affectivity score was estimated using the guilt, hostility/irritability, and fear/anxiety items of the Hamilton Depression Rating Scale, the Montgomery-Asberg Depression Rating Scale, the Quick Inventory of Depressive Symptoms, and the Social Adjustment Scale. We felt that these items captured published descriptions of the NA construct. A Clinical Global Impressions-Severity of Illness (CGI-S) score ≤ 2 defined response. Because combined treatment addressed both serotonin and non-serotonin systems, patients treated with both medications did not test the hypothesis and so were excluded from the analyses. RESULTS Analysis of covariance with treatment as a grouping variable, NA as covariate, and CGI-S as dependent variable showed a significant 2-way interaction between treatment and NA (F₁,₁₅₆ = 4.82, P < .03). In the low-NA group, response rates were similar between treatments (escitalopram: 28/42 [67%]; bupropion: 23/40 [58%]; NS), while there was a significant advantage for escitalopram in patients with high NA (escitalopram: 24/40 [60%]; bupropion = 14/41 [34%]; P = .017). CONCLUSIONS These data suggest that patients with high negative affectivity respond preferentially to antidepressants that selectively enhance serotonin neurotransmission. Although patients with low NA appear to benefit from serotonin enhancement as well, they also improved with bupropion, an antidepressant not thought to directly affect serotonin neurotransmission. These findings come from retrospective analyses using unproven approximation of NA, so no clinical inferences should be made before independent replication utilizing accepted NA measurement. TRIAL REGISTRATION ClinicalTrials.gov identifier: NCT00519428.

Citalopram in the treatment of dysthymic disorder

This study aimed to provide preliminary data on the tolerability and effectiveness of citalopram for patients with dysthymic disorder. Twenty-one adult subjects meeting DSM-IV criteria for dysthymic disorder were enrolled in this 12-week open-label study, of whom 15 had pure dysthymia (e.g. no major depression in the past 2 years). Citalopram was initiated at 20 mg/day, and increased to a maximum of 60 mg/day. Response was defined as 50% or greater drop in score on the Hamilton Depression Rating Scale (HDRS) and a Clinical Global Impressions-I score of 1 (‘very much improved’) or 2 (‘much improved’). Of these 15 pure dysthymic disorder subjects, all completed the trial, and 11 (73.3%) were treatment responders. All paired sample t-tests were highly significant, demonstrating significant average improvement on all measures of symptomatology and functioning. Scores on the 24-item HDRS decreased from 22.3±4.3 at baseline to 9.1±7.8 at week 12 [t(14)=6.1, P<0.001]. In addition, improvement was noted in self-reported measures of temperament and social functioning. The average final dose of citalopram was 39 mg/day. Side-effects were reported by nine of 15 subjects (60%), most frequently gastrointestinal symptoms (n=5), dry mouth (n=5) and sexual side-effects (n=3). These findings suggest the effectiveness and tolerability of citalopram in treating dysthymic disorder. Double-blind prospective studies are needed comparing citalopram both to placebo and to other medications, assessing both initial and sustained response to treatment.

Bupropion sustained-release for the treatment of dysthymic disorder : An open-label study

Many studies of antidepressants in the treatment of dysthymic disorder (DD) have been conducted, but none has included bupropion sustained-release (SR). The aim of this study was to provide preliminary data on the tolerability and effectiveness of bupropion SR for patients with DD. Twenty-one adult subjects meeting DSM-IV criteria for DD were enrolled in this 8-week open-label study. Bupropion SR was initiated at 150 mg/day and was increased to a maximum of 200 mg, twice daily. Response was defined as a 50% or greater decrease in score on the Hamilton Rating Scale for Depression (HAM-D). Of these 21 subjects, 15 (71.4%) responded to treatment. All paired sample t-tests were highly significant, demonstrating average improvement on all measures of symptomatology and functioning. Subject scores on the HAM-D decreased from 21.7 ± 5.6 at baseline to 5.9 ± 3.6 at week 8 (t [19] = 12.74, p < 0.001). The average final dosage was 364 mg/day. None of the subjects dropped out during the trial. Patients with a history of alcohol or chemical abuse were significantly less likely to respond to bupropion. Side effects were reported by eight subjects (38.1%), and the most frequently reported effects were headache, decreased appetite, insomnia, gastrointestinal problems, restlessness, and tremulousness. These findings suggest the effectiveness and high tolerability of bupropion SR for the treatment of DD. Double-blind prospective studies are needed for the comparison of bupropion SR to both placebo and other medications, assessing both initial and sustained responses to treatment.

Does personality disorder decrease the likelihood of remission in early-onset chronic depression?

BACKGROUND The impact of personality disorders (PD) on the course of depression has been gaining interest among clinical researchers over the past decade. Recent observational studies have found that PD was associated with impaired social functioning and reduced likelihood of depression recovery. Elevated rates of PD have been noted in early-onset and chronic forms subtypes of depression. However, scant data exist regarding the link between PD and outcome for this depression subtype. METHODS The National Epidemiological Survey on Alcohol and Related Conditions database was analyzed. This survey included 43 093 respondents, 18 years and older, conducted in 2001 through 2002. Logistic regression was used to identify demographic and clinical predictors of remission in early-onset chronic depression. RESULTS The absence of PD, having more years of education, and being married considerably improved the likelihood of remission. Paranoid personality disorder and obsessive-compulsive disorder were the only specific PD found to be associated with a reduced probability of remission. LIMITATIONS Depression remission status may have biased the recollection of PD symptoms. Borderline personality disorder, narcissistic personality disorder, and schizotypal personality disorder were not assessed. CONCLUSIONS This study suggests that PD are significant predictors of remission in early-onset chronic depression.