Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2023 May 12.
Published in final edited form as: J Affect Disord. 2011 May 23;133(3):467–476. doi: 10.1016/j.jad.2011.04.032

Randomized comparison of selective serotonin reuptake inhibitor (escitalopram) monotherapy and antidepressant combination pharmacotherapy for major depressive disorder with melancholic features: A CO-MED report

William V Bobo a,*, Helen Chen b, Madhukar H Trivedi c, Jonathan W Stewart d, Andrew A Nierenberg e, Maurizio Fava e, Benji T Kurian c, Diane Warden c, David W Morris c,f, James F Luther g, Mustafa M Husain c, Ian A Cook h, Ira M Lesser i, Susan G Kornstein j, Stephen R Wisniewski g, A John Rush k, Richard C Shelton a
PMCID: PMC10177662  NIHMSID: NIHMS295409  PMID: 21601287

Abstract

Background:

The clinical effects of antidepressant combinations vs. monotherapy as initial treatment for major depression with melancholic features (MDD-MF) are unknown.

Methods:

Outpatients with chronic or recurrent major depression (MDD) were randomized to initial treatment with escitalopram+placebo (the MONO condition), bupropion-sustained release+escitalopram, or venlafaxine-extended release+mirtazapine (the COMB conditions) in the Combining Medications to Enhance Depression Outcomes (CO-MED) trial. Secondary data analyses were conducted to compare demographic and clinical characteristics, and contrast clinical responses according to drug treatment, in patients with MDD-MF (n=124) and non-melancholic MDD (n=481).

Results:

While numerically lower, remission rates in MDD-MF did not differ significantly from those with non-melancholic MDD either at 12 (33.1% vs. 41.0%, aOR 1.16, p=0.58) or 28 (39.5% vs. 46.8%, aOR=1.02, p=0.93) weeks of treatment. Remission rates did not differ significantly between combination and monotherapy groups in either MDD-MF or non-melancholic MDD patients at either time point. Similar conclusions were reached for response rates, premature study discontinuation, and self-rated depression symptom severity.

Limitations:

This is a secondary analysis of data from the CO-MED trial, which was not designed to address differential treatment response in melancholic and non-melancholic MDD.

Conclusions:

We found no evidence of differential remission or response rates to antidepressant combination or monotherapy between melancholic/non-melancholic MDD patients, or according to antidepressant treatment group, after 12 and 28 weeks. Melancholic features may not be a valid predictor of more favorable response to antidepressant combination therapy as initial treatment.

Keywords: Escitalopram, Venlafaxine, Bupropion, Mirtazapine, Major depressive disorder, Melancholic features

1. Introduction

Melancholia is considered a severe subtype of major depressive disorder (MDD) (Rush and Weissenburger, 1994). While the prevalence of MDD with melancholic features (MDD-MF) is unknown in the general population (Kessing, 2007), melancholic features are present in 20–30% of outpatients and as many as 75% of inpatients with MDD (Hildebrandt et al., 2003; Khan et al., 2006; Zimmerman and Spitzer, 1989). Melancholic features in MDD are of significant nosologic interest—a considerable body of empirical evidence suggests that MDD-MF is a qualitatively distinct and valid depressive subtype in terms of biological functioning, treatment response, and suicidal risk (Leventhal and Rehm, 2005). Numerous studies have documented greater clinical responses to pharmacotherapy, lower placebo response rates, and poorer overall responses to psychosocial treatments in patients with melancholic- vs. non-melancholic MDD (Frank et al., 1992; Leventhal and Rehm, 2005; Parker et al., 1992; Prusoff et al., 1980; Thase and Friedman, 1999), although not all studies are in agreement. These findings raise the question of whether MDD-MF may also respond preferentially to pharmacotherapy with different antidepressants and antidepressant combinations which engage multiple mechanisms of action.

Studies conducted over the last three decades document superior response to tricyclic antidepressants (TCAs), particularly tertiary amine agents, and monoamine oxidase inhibitors (MAOIs) in patients with MDD-MF, compared to placebo-treated MDD-MF patients and to patients with non-melancholic MDD treated with these same agents (Angst and Stabl, 1992; Brown, 2007; Coryell and Turner, 1985; Peselow et al., 1992). TCAs and MAOIs are generally considered second- or third-line options due to their relatively adverse tolerability and safety profiles. There are considerably fewer studies of the selective serotonin reuptake inhibitors (SSRIs) and newer dual-mechanism antidepressants in MDD-MF, many of which are considered first-line pharmacotherapies for MDD broadly. Some controlled trial evidence supports the efficacy of SSRIs (Amsterdam, 1998), venlafaxine (Benkert et al., 1996; Gentil et al., 2000; Guelfi et al., 1995), duloxetine (Mallinckrodt et al., 2005), and mirtazapine (Guelfi et al., 2001) in the treatment of MDD-MF, with some studies documenting greater treatment responses with TCAs and venlafaxine than with SSRIs (Amsterdam, 1998; Anderson, 1998; Clerc et al., 1994; Sheehan et al., 2009). These results suggest that antidepressant medication monotherapy or combinations that interact with multiple neurotransmitter systems may be more effective for treating MDD-MF than monotherapy with single-mechanism agents.

In a recent secondary analysis of the Sequenced Treatment Alternatives to Relieve Depression (STAR*D) trial data, citalopram monotherapy resulted in significantly lower remission rates in patients with MDD-MF than those with non-melancholic MDD (McGrath et al., 2008). Remission rate differences were no longer significant, however, after controlling for study site, and differences in baseline demographic characteristics, clinical symptoms and suicide risk measures. Because MDD-MF has been linked with dysfunction in multiple neurotransmitter and neurobiological systems (Amsterdam, 1998; Leventhal and Rehm, 2005), combining antidepressants that interact with multiple CNS targets may result in superior clinical response for patients with MDD-MF than non-melancholic MDD. No prospective, randomized study has examined the clinical effects of antidepressant combinations as initial therapy for treating MDD-MF.

To further address this question, we conducted an analysis of data from the Combining Medications to Enhance Depression Outcomes (CO-MED) trial (Rush et al., 2011) to compare demographic and clinical characteristics, and contrast clinical responses to escitalopram monotherapy and two antidepressant combination pharmacotherapies, in patients with MDD-MF and non-melancholic MDD.

2. Method

2.1. Study overview

CO-MED was a randomized, single-blind, placebo-controlled trial designed to compare the effectiveness of escitalopram+placebo (ESC+PLC) vs. two antidepressant combinations as first-step MDD treatment during acute (12 weeks) and long-term (28 weeks) treatment. CO-MED study procedures and selection criteria have been previously published (Rush et al., 2011) and are available online at www.co-med.org, but will be reviewed briefly here. The main effectiveness results for all 665 randomized subjects have also been published (Rush et al., 2011). This report focuses on differences in clinical characteristics and outcomes between melancholic vs. non-melancholic CO-MED participants.

2.2. Patients

Outpatients (ages 18–75 years) meeting DSM-IV-TR criteria for non-psychotic MDD with a recurrent or chronic (current episode ≥2 years) illness course were enrolled between March 2008 and February 2009. MDD diagnoses were established by clinical interview and a DSM-IV-TR-based symptom checklist. Eligible subjects had index major depressive episodes of ≥2 month’s duration and a Hamilton Rating Scale for Depression (HSRD17) (Hamilton, 1960) score of ≥16 at study entry. Those with any psychotic illness or bipolar disorder, or who were in need of psychiatric hospitalization, were excluded.

The presence of melancholic features was determined by using the 30-item Inventory of Depressive Symptomatology-Clinician Rating (IDS-C30) (Rush et al., 1996) at baseline, using criteria identical to those of McGrath et al. (2008). To meet our criteria for melancholic features, the patient had to score 2 or 3 in the IDS-C30 mood reactivity or pleasure item and meet at least 3 of the following criteria: quality of mood, diurnal mood variation, psychomotor retardation, psychomotor agitation, appetite decrease or weight decrease, or self-outlook. The IDS-C30 scores psychomotor retardation or agitation based on patient report with minimal reliance on current latency or speed of speech.

2.3. Procedures

The study protocol was approved by the institutional review boards of the CO-MED National Coordinating Center (University of Texas Southwestern Medical Center, Dallas), the Data Coordinating Center (University of Pittsburgh), all participating clinical sites, and the Data Safety and Monitoring Board of the National Institutes of Health (Bethesda, MD). All subjects provided written informed consent prior to study enrollment.

Eligible subjects were randomized (1:1:1) to one of three allocation groups: a) ESC (10–20 mg/day)+PLC; b) bupropion-sustained release (BUP-SR, 150–400 mg/day) +ESC (10–20 mg/day); or c) venlafaxine-extended release (VLFX-XR, 37.5–300 mg/day)+mirtazapine (MRT, 15–45 mg/day). The first medication in each group was given on an open-label basis. Patients were blinded to the second medication. All medications were initiated at the low end of the specified dosage ranges. Doses were increased every 1–4 weeks according to a symptom measurement-based care protocol (Operations Manual available at www.co-med.org) to provide individualized but vigorous dosing (Trivedi et al., 2007). Dose adjustments were based on clinical response (as measured by the 16-item Quick Inventory of Depressive Symptomatology-Clinician-rated [QIDS-SR16]) (Rush et al., 2003) and tolerability (as measured by the Frequency, Intensity and Burden of Side Effects Rating) (Wisniewski et al., 2006) and the Systematic Assessment for Treatment Emergent Events—Systematic Inquiry (Levine and Schooler, 1986) at each visit. Patients were not allowed to receive non-protocol antidepressants, anxiolytics, sedative hypnotics, adjunctive antipsychotics, or depression-targeted, empirically-validated psychotherapies. However, other therapies were allowed, including supportive, couples/marital, and occupational therapies. Clinicians were alerted about non-protocol medications for which dose adjustments may be needed if coprescribed with bupropion-SR.

Prior to treatment allocation, trained clinical research coordinators collected clinical and demographic data, reviewed inclusion/exclusion criteria, and assessed current general medical illnesses using standard forms. These forms were also used to obtain additional psychiatric history, including past abuse or trauma and past psychiatric treatment. The self-report Psychiatric Diagnostic Screening Questionnaire was used to establish the presence of current Axis I disorders (Rush et al., 2005). Baseline comorbid psychiatric and general medical illnesses were additionally assessed using the Concise Associated Symptom Tracking scale (CAST) (Trivedi et al., 2010), and the Self-administered Comorbidity Questionnaire (SCQ) (Sangha et al., 2003).

Clinical outcomes were ascertained at baseline and at all post-randomization visits (weeks 1, 2, 4, 6, 8, 10, 12, 16, 20, 24, and 28). Depression severity measures included the QIDS-SR16 and IDS-C30. Measures of time-dependent psychiatric comorbidity included the Suicidality Rating Scale (SRS), Suicidality Rating Scale-Associated Symptoms (SRS-AS), and Altman Self-Rating Mania Scale (ASRM) (Altman et al., 1997). Quality of life and psychosocial functioning were assessed using the Quality of Life Inventory (QOLI) (Frisch et al., 2005), and the Work and Social Adjustment Scale (WSAS) (Mundt et al., 2002).

Clinical outcomes, medication adverse events and adherence were recorded by patients (self-report measures), study clinicians (clinician-rated instruments) or blinded clinical raters, using a Tablet PC data entry system.

2.4. End points

The primary study endpoint was symptom remission (at least two consecutive QIDS-SR16 ratings of <8, with at least one <6). Although the main study endpoint was remission at week 12, remission status was also assessed at week 28 as a secondary outcome. Those who left the study prior to week 12 were classified as remitters if their last two consecutive QIDS-SR16 scores satisfied these criteria. Individuals without two or more post-baseline measures were ineligible to be remitters.

Secondary outcomes included positive response (≥50% reduction in QIDS-SR16 score); changes in depressive psychopathology and anxiety symptoms (assessed using the IDS-C30), quality of life, adverse effect measures, and attrition, at 12 and 28 weeks.

2.5. Statistical procedures

Summary statistics used for these comparisons are presented as means and standard deviations for continuous measures, and as proportions for discrete variables. Clinical, demographic, and treatment outcome measures were compared according to diagnostic group (MDD-MF vs. non-melancholic MDD). Selected outcome measures were then compared by diagnostic and treatment group (ESC+PLC, BUP-SR+ESC, VLFX-XR+MRT). Chi-square tests were used to compare remission rates, the primary endpoint, across treatment groups. Fisher’s exact test was used when expected cell frequencies were <5. For binary outcomes including remission, bivariate logistic regression models were fit to estimate the effect of treatment on outcome. Multivariate models were used to adjust for Regional Center and baseline characteristics that were not balanced across treatment groups. Polytomous logistic regression was used for discrete outcomes with more than two levels. T-tests were used to compare continuous measures when data were normally distributed; otherwise, data were analyzed using the Kruskal–Wallis test. Linear regression models were used to compare means, adjusting for Regional Center and baseline characteristics not balanced across treatment groups. A general linear model with a negative binomial distribution and log link was estimated for outcomes with severely non-normal distributions (e.g., ICS-C30 anxiety subscale).

3. Results

3.1. Demographic and clinical characteristics

Of 665 participants enrolled in the main CO-MED study, 605 met inclusion criteria for this analysis, of which 124 (20.5%) were classified as MDD-MF. Baseline socio-demographic characteristics are presented in Table 1. The study sample was predominantly middle-aged, female, and Caucasian. Racial/ethnic minority groups were well-represented. There were no significant differences between patients with MDD-MF and non-melancholic MDD with regard to age, gender, race/ethnicity or level of education. Significantly fewer patients in the MDD-MF group were employed at study entry, compared to those with non-melancholic MDD.

Table 1.

Demographic measures by melancholic features.

Measure Melancholic features p
Yes (N = 124) No (N = 481)
Mean SD Mean SD
Age 41.1 12.2 43.1 13.3 0.1278
Education 13.5 3.2 13.8 3.0 0.3753
Monthly household income 2418 8476 2815 4483 0.0016
N % N %
Sex 0.0939
 Male 32 25.8 162 33.7
 Female 92 74.2 319 66.3
Race 0.5875
 White 82 67.2 314 67.8
 Black 35 28.7 120 25.9
 Other 5 4.1 29 6.3
Hispanic 21 16.9 73 15.2 0.6298
Employed 51 41.1 250 52.0 0.0313
Open in a new tab

Baseline clinical characteristics are presented in Table 2. Baseline depressive symptom severity was in the moderate-to-severe range for the entire sample. Significantly higher (worse) mean depressive symptom ratings and higher proportions of subjects with severe or very severe depressive symptoms were observed in the MDD-MF group, compared with non-melancholic depressives. Proportions of subjects with anxious depressive features and sleep disturbance were significantly higher in the MDD-MF group. A significantly greater proportion of MDD-MF subjects had histories of abuse and suicidal thoughts or behaviors.

Table 2.

Clinical measures by melancholic features.

Measure Melancholic features p
Yes (N = 124) No (N = 481)
Mean SD Mean SD
Age at first episode 22.6 14.4 24.6 14.2 0.1078
Years since first episode 18.4 12.0 18.5 13.9 0.5586
N prior episodes 12.1 26.5 7.6 16.0 0.5256
N suicide attempts 0.20 0.59 0.26 1.51 0.1676
Current episode duration (mos) 69.5 112 60.3 105 0.5408
HRSD17 28.2 4.5 22.6 4.3 <0.0001
IDS-C30 47.6 7.0 35.1 8.0 <0.0001
QIDS-C16 19.2 2.5 14.7 3.1 <0.0001
QIDS-SR16 18.8 3.4 14.4 4.1 <0.0001
ASRMS 1.4 2.2 1.5 2.2 0.3084
N % N %
Ever attempted suicide 15 12.6 39 8.4 0.1542
Lifetime severity of suicidality 0.0053
 None 20 16.8 163 35.0
 Thoughts of dying 36 30.3 123 26.4
 Suicidal thoughts 22 18.5 65 13.9
 Specific method 11 9.2 43 9.2
 Plan/gesture 12 10.1 22 4.7
 Preparation 3 2.5 11 2.4
 Attempt 15 12.6 39 8.4
Neglected before age 18 51 41.1 166 34.6 0.1756
Emotionally abused before age 18 59 47.6 176 36.7 0.0263
Physically abused before age 18 30 24.2 89 18.5 0.1584
Sexually abused before age 18 34 27.4 97 20.3 0.0845
Chronic/recurrent depression 0.6260
 Chronic only * 26 21.0 106 22.1
 Recurrent only 53 42.7 221 46.1
 Both 45 36.3 152 31.7
QIDS-SR16 <0.0001
 0–10 (none/mild) 2 1.7 79 16.9
 11–15 (moderate) 21 17.4 200 42.8
 16–20 (severe) 64 52.9 163 34.9
 21–27 (very severe) 34 28.1 25 5.4
IDS-C30 Lethargic depression 114 91.9 285 59.3 <0.0001
Anxious features 114 91.9 337 70.1 <0.0001
IDS-C30 Sleep disturbance 116 93.5 418 86.9 0.0403
CHRT-SR Suicidal thoughts/plans 32 25.8 64 13.3 0.0007
Clinical setting 0.8771
 Primary 64 51.6 252 52.4
 Specialty 60 48.4 229 47.6
Open in a new tab

Abbreviations: ASRMS Altman self-rated mania scale; CHRT-SR self-rated concise health risk tracking scale; HRSD Hamilton rating scale for depression; IDS-C clinician-rated inventory of depressive symptomatology; QIDS-C, -SR clinician-, self-rated Quick inventory of depressive symptomatology.

*

Chronic depression defined as current episode duration ≥24 months.

There were high rates of psychiatric and general medical comorbidity in the sample as a whole (Appendix 1). Significantly higher rates of comorbid anxiety and somatoform spectrum disorders were observed in patients with MDD-MF than non-melancholic MDD. Patients with MDD-MF had significantly more comorbid psychiatric disorders and heart disease. However, there were no significant between-group differences in the occurrence of substance use disorders or number of treated medical illnesses.

3.2. Treatment and follow-up

Medication treatment and follow-up data at weeks 12 and 28 are summarized in Table 3. Mean duration of post-baseline follow-up was 18.1 weeks in the MDD-MF group and 20.5 weeks in the non-melancholic MDD group (p=0.09). MDD-MF subjects had modestly but significantly fewer post-baseline visits than did non-melancholic MDD subjects. Mean and maximal antidepressant doses were in clinically acceptable ranges for all agents in both diagnostic groups. Higher mean and maximal VLFX-XR doses were observed at week 12 in the MDD-MF group. Strong statistical trends towards higher mean MRT dose at week 12, and higher maximal ESC (open-label) and VLFX-XR doses at week 28, were also observed in the MDD-MF group. There were no significant differences in duration of treatment between diagnostic groups for any drug/drug combination, at either week 12 or week 28.

Table 3.

Treatment measures by study phase and melancholic features.

Measure Acute phase Continuation phase
Melancholic features Melancholic features
Yes (N = 124) No (N = 481) p Yes (N = 124) No (N = 481) p
Mean SD Mean SD Mean SD Mean SD
Weeks in treatment 9.4 4.1 10.0 3.8 0.0575 18.1 10.9 20.5 10.3 0.0880
N postbaseline visits 5.0 2.2 5.4 2.2 0.1392 7.1 3.8 7.9 3.7 0.0398
Open med dose (mg)
 Maximum bupropion 333 81 320 82 0.2962 336 82 325 83 0.4276
 Last bupropion 300 114 282 124 0.4439 289 123 267 141 0.5101
 Weeks on last bupropion 5.3 4.3 5.2 4.2 0.9454 13.7 11.6 13.0 10.6 0.7916
 Maximum escitalopram 16.6 5.3 17.9 4.3 0.0914 16.8 5.2 18.2 4.2 0.0748
 Last escitalopram 16.0 6.0 17.1 5.1 0.1865 15.2 6.9 15.6 7.1 0.5441
 Weeks on last escitalopram 5.4 3.7 6.6 3.4 0.1099 11.3 10.5 15.5 10.5 0.0971
 Maximum venlafaxine 228 63 200 69 0.0103 233 66 211 74 0.0575
 Last venlafaxine 214 82 182 82 0.0116 192 94 171 93 0.1375
 Weeks on last venlafaxine 4.4 3.4 5.5 4.2 0.2287 12.4 10.7 13.9 10.3 0.4292
Blinded med dose (mg)
 Maximum escitalopram 13.3 8.3 14.0 7.1 0.8167 13.3 8.3 14.2 7.2 0.6777
 Last escitalopram 12.8 8.5 12.4 8.2 0.7355 11.9 8.9 11.4 8.6 0.6878
 Weeks on last escitalopram 5.5 3.4 5.7 3.9 0.8653 15.1 10.1 15.7 10.2 0.6611
 Maximum placebo (pills) 1.3 0.7 1.5 0.7 0.1313 1.3 0.8 1.5 0.7 0.1412
 Last placebo (pills) 1.1 0.8 1.4 0.7 0.0498 0.71 0.84 0.64 0.86 0.5737
 Weeks on last placebo 5.1 3.8 5.9 3.5 0.2524 13.6 11.8 19.8 10.6 0.0064
 Maximum mirtazapine 26.4 13.9 22.3 13.4 0.1025 27.4 14.2 23.9 14.2 0.1819
 Last mirtazapine 23.2 15.7 18.4 15.3 0.0710 20.9 16.6 16.7 16.2 0.1200
 Weeks on last mirtazapine 5.2 4.2 6.0 4.4 0.1852 14.6 10.6 15.9 10.4 0.4106
N % N % N % N %
Weeks in treatment
 < 4 20 16.3 65 13.5 0.4395 … … … … …
 < 8 33 26.8 97 20.2 0.1111 … … … … …
Open in a new tab

3.3. Clinical outcomes

Approximately one-third of the study sample achieved remission by week 12, with modestly higher rates at week 28. Slightly more than 50% of the sample had at least a positive treatment response at week 12, with modest further increases at 28 weeks. There were no significant differences between diagnostic groups with regard to remission or response rates at 12 or 28 weeks in the main analyses, or in sensitivity analyses adjusting the definition of remission to last QIDS-SR16 score <6 (Table 4).

Table 4.

Outcome measures by study phase and melancholic features.

Measure Acute phase Continuation phase
Melancholic features Analyses Melancholic features Analyses
Yes (N = 124) No (N = 481) Unadjusted Adjusteda Yes (N = 124) No (N = 481) Unadjusted Adjusteda
Mean SD Mean SD β p β p Mean SD Mean SD β p β p
Last QIDS-SR16 9.5 6.4 7.7 5.1 1.873 0.0007 −0.210 0.7348 9.1 6.5 7.1 5.3 1.237 0.0202 1.005 0.9609
Percent QIDS-SR16 change −47 34.6 −45 34.8 −2.431 0.4922 −2.631 0.5269 −49 36.2 −50 35.3 −0.349 0.9239 2.041 0.6364
IDS-C30 Anxiety subscale 3.4 2.4 2.4 2.0 0.363 <.0001 0.148 0.1319 3.3 2.5 2.3 2.0 0.391 <.0001 0.190 0.0805
Last QOLI −.59 2.58 0.38 2.24 −1.051 <.0001 −0.290 0.2971 −.20 2.66 0.64 2.29 −0.886 0.0004 −0.186 0.5167
N % N % OR p OR p N % N % OR p OR p
Early termination 42 33.9 122 25.4 1.558 0.0457 1.524 0.1376 58 46.8 163 33.9 1.696 0.0118 2.136 0.0046
Last FIBSER frequency 1.439 0.0645 1.097 0.7049 1.330 0.1532 1.126 0.6321
 No side effects 41 34.7 202 44.2 52 43.7 237 51.9
 10–25% of the time 52 44.1 172 37.6 47 39.5 152 33.3
 50–75% of the time 19 16.1 61 13.3 15 12.6 47 10.3
 90–100% of the time 6 5.1 22 4.8 5 4.2 21 4.6
Last FIBSER intensity 1.360 0.1183 1.073 0.7733 1.386 0.1002 1.204 0.4527
 No side effects 42 35.6 202 44.2 52 43.7 236 51.6
 Minimal/mild 53 44.9 167 36.5 43 36.1 148 32.4
 Moderate/marked 15 12.7 73 16.0 18 15.1 58 12.7
 Severe/intolerable 8 6.8 15 3.3 6 5.0 15 3.3
Last FIBSER burden 1.778 0.0046 1.251 0.3784 1.672 0.0125 1.217 0.4482
 No impairment 56 47.5 264 57.8 61 51.3 285 62.4
 Minimal/mild 41 34.7 152 33.3 41 34.5 126 27.6
 Moderate/marked 14 11.9 32 7.0 11 9.2 39 8.5
 Severe/intolerable 7 5.9 9 2.0 6 5.0 7 1.5
At least 1 SAE 3 2.4 17 3.5 0.733 0.6280 0.721 0.6698 13 10.5 26 5.4 1.944 0.0723 1.957 0.1784
Remissionb 41 33.1 197 41.0 0.671 0.0691 1.164 0.5783 49 39.5 225 46.8 0.742 0.1553 1.023 0.9318
Last QIDS-SR16 <6 42 33.9 185 38.8 0.766 0.2218 1.380 0.2435 48 39.3 223 46.8 0.744 0.1628 1.004 0.9886
Percent QIDS-SR16 63 52.1 240 51.7 1.022 0.9170 1.064 0.8102 65 54.6 275 59.4 0.854 0.4533 0.674 0.1437
reduction ≥ 50
Last WSASc 2.618 <.0001 1.323 0.2330 1.888 0.0008 0.978 0.9228
 0 16 13.8 79 17.3 22 18.8 104 22.8
 1–10 19 16.4 141 30.9 26 22.2 131 28.7
 11–20 23 19.8 106 23.2 19 16.2 103 22.5
 21–30 20 17.2 88 19.3 19 16.2 67 14.7
 31–40 38 32.8 43 9.4 31 26.5 52 11.4
Open in a new tab

Abbreviations: FIBSER frequency, intensity, and burden of side effect rating scale; IDS-C clinician-rated inventory of depressive symptomatology; QIDS-SR self-rated Quick inventory of depressive symptomatology; QOLI quality of life inventory; SAE serious adverse event; WSAS work and social adjustment scale.

a

Adjusted for treatment, race, substance abuse, mania, and WSAS.

b

Defined as at least one of the last two consecutive QIDS-SR16 scores <6 and the other <8.

c

Non-normal distribution required binning.

Mean QIDS-SR16 scores were modestly – but significantly – higher (worse) in the MDD-MF group at weeks 12 (9.5±6.6 vs. 7.7±5.1, p=0.0007) and 28 (9.1±6.5 vs. 7.1±5.3, p=0.02), relative to those of the non-melancholic MDD group, in the main analysis (Table 4). However, these between-group differences were not statistically significant in the adjusted analyses. There were no significant differences in mean change from baseline in QIDS-SR16 scores between diagnostic groups at either time point.

More MDD-MF patients than those with non-melancholic MDD exited the study prematurely by week 12 (33.9% vs. 25.4%) and week 28 (46.8% vs. 33.9%). In adjusted analyses, patients in the MDD-MF group were approximately 50% more likely to exit the study prematurely by week 12 (aOR=1.52, p=0.14), and more than twice as likely to exit prematurely by week 28 (aOR=2.14, p=0.005), compared with those in the non-melancholic group (Table 4). Patients with MDD-MF were more likely to report poorer psychosocial functioning (as assessed by the WSAS) than non-melancholic MDD in unadjusted analyses at weeks 12 (unadjusted OR=2.62, p<0.0001) and 28 (unadjusted OR=1.89, p=0.0008); however, relationships between diagnostic group and psychosocial functioning were far less robust and were no longer statistically significant in the adjusted analyses (Table 4).

Compared with non-melancholic patients, those with MDD-MF had significantly higher mean IDS-C30 anxiety subscale scores and lower quality of life index scores at 12 and 28 weeks in the main analyses (Table 4). In adjusted analyses, ICS-C30 anxiety subscale scores were nearly significantly higher in the MDD-MF group at week 28 (p=0.08), while the remaining between-group comparisons for anxiety symptom (week 12) and quality of life measures (weeks 12 and 28) were no longer statistically significant.

Selected outcome measures were assessed according to diagnostic and antidepressant treatment groups, as shown in Table 5. Remission rates, defined as last QIDS-SR16 score <6, were numerically similar across drug treatment groups in both the MDD-MF and non-melancholic MDD groups at weeks 12 and 28. Remission or positive treatment response rates, QIDS-SR16 scores, and early study attrition rates did not differ significantly between groups. Similarly, mean QIDS-SR16 scores and mean QIDS-SR16 change were numerically similar between treatment groups for melancholic and non-melancholic subjects, and did not differ significantly.

Table 5.

Selected outcome measures by melancholic features and study phase.

Measure Melancholic features No melancholic features p *
BUP-SR + ESCIT (N = 36) ESCIT + PBO (N = 42) VEN-XR + MRT (N = 46) BUP-SR + ESCIT (N = 163) ESCIT + PBO (N = 164) VEN-XR + MRT (N = 154)
Acute phase Mean SD Mean SD Mean SD Mean SD Mean SD Mean SD
Last QIDS-SR16 10.0 6.3 7.7 5.0 9.6 6.7 7.3 4.6 9.0 6.4 8.1 5.6 0.4605
% QIDS-SR16 reduction −45 34.0 −43 35.6 −46 36.0 −48 33.0 −51 34.2 −45 35.8 0.6908
Acute phase N % N % N % N % N % N %
Early termination 14 38.9 49 29.9 16 38.1 34 20.9 12 26.1 39 25.3 0.3160
Last QIDS-SR16 <6 11 30.6 64 39.5 13 31.0 64 39.3 18 39.1 57 37.5 0.5922
Percent QIDS-SR16 reduction ≥ 50 17 48.6 83 51.9 23 54.8 80 51.0 23 52.3 77 52.4 0.8548
Last FIBSER Burden 0.8059
 No impairment 18 52.9 90 58.1 16 41.0 92 58.6 22 48.9 82 56.6
 Minimal/mild 9 26.5 54 34.8 17 43.6 52 33.1 15 33.3 46 31.7
 Moderate/marked 4 11.8 7 4.5 5 12.8 10 6.4 5 11.1 15 10.3
 Severe/intolerable 3 8.8 4 2.6 1 2.6 3 1.9 3 6.7 2 1.4
Continuation phase Mean SD Mean SD Mean SD Mean SD Mean SD Mean SD
Last QIDS-SR16 8.7 6.8 7.0 5.1 9.8 6.6 6.5 4.9 8.7 6.3 7.8 5.8 0.3868
% QIDS-SR16 reduction −51 40.0 −48 36.6 −45 35.6 −54 32.6 −53 33.8 −48 36.5 0.2312
Continuation phase N % N % N % N % N % N %
Early termination 15 41.7 61 37.2 23 54.8 48 29.4 20 43.5 54 35.1 0.1852
Last QIDS-SR16 <6 17 47.2 75 46.6 13 31.7 83 50.9 18 40.0 65 42.8 0.2328
Percent QIDS-SR16 reduction ≥ 50 20 57.1 91 57.2 20 48.8 97 61.8 25 58.1 87 59.2 0.5119
Last FIBSER Burden 0.5534
 No impairment 20 58.8 96 61.5 19 47.5 106 67.9 22 48.9 83 57.2
 Minimal/mild 9 26.5 47 30.1 17 42.5 38 24.4 15 33.3 41 28.3
 Moderate/marked 3 8.8 9 5.8 3 7.5 9 5.8 5 11.1 21 14.5
 Severe/intolerable 2 5.9 4 2.6 1 2.5 3 1.9 3 6.7
Open in a new tab

Abbreviations: FIBSER Frequency, intensity, and burden of side effect rating scale; QIDS-SR self-rated Quick inventory of depressive symptomatology.

*

Probability value associated with the treatment by melancholic features interaction term.

3.4. Safety and tolerability

The majority of patients (~80–90%) in both diagnostic groups experienced no, minimal, or mild adverse effect burden or intensity as assessed by the FIBSER, indicating that antidepressant treatment was well-tolerated in melancholic and non-melancholic patients (Table 4). There were no significant differences between diagnostic groups with regard to adverse effect intensity. In the main analysis, significantly greater burden from adverse effects was reported in the MDD-MF group at weeks 12 (unadjusted p=0.005) and 28 (unadjusted p=0.01). A higher proportion of patients with MDD-MF had moderate/marked (week 12, 11.9% vs. 7.0%; week 28, 9.2% vs. 8.5%) or severe/intolerable impairment (week 12, 5.9% vs. 2.0% vs. week 28, 5.0% vs. 3.3%) from adverse effects, compared with non-melancholic patients. In the adjusted analyses, the relationship between diagnostic group and adverse effect burden became smaller and was no longer statistically significant.

4. Discussion

In this study, approximately one-third of community-treated adults with recurrent or chronic, non-psychotic MDD-MF and non-melancholic MDD achieved remission. Remission rates were numerically lower in MDD-MF patients than those with non-melancholic MDD. However, remission rates did not differ significantly between melancholic/non-melancholic diagnosis groups, nor did they differ significantly according to antidepressant treatment group among those with MDD-MF or non-melancholic MDD, after 12 or 28 weeks of follow-up. Similar conclusions were reached for rates of positive treatment response, rates of premature study discontinuation, and self-rated depression symptom severity.

To date, several antidepressants have been studied for the treatment of MDD-MF. Comparative efficacy trials of TCAs and SSRIs in patients with MDD-MF have yielded mixed results, with early studies showing superiority of TCAs and later studies showing non-differential efficacy (Amsterdam, 1998; Anderson, 1998; Perry, 1996). A modest efficacy advantage of venlafaxine over fluoxetine monotherapy for remission induction was documented in one small study of melancholic depressed patients (Tzanakaki et al., 2000); however, these findings were not replicated in a larger randomized study (Sheehan et al., 2009). These studies, however, compared single agent therapies and did not focus on antidepressant combinations or augmentation strategies.

Use of antidepressant combination therapy as an initial strategy is motivated by disappointingly low remission rates with antidepressant monotherapy in clinical trials (Khan et al., 2005; Rush et al., 2006b). There is some evidence, mostly uncontrolled, suggesting that, in MDD samples with multiple depressive sub-types, the use of SSRI+bupropion can increase rates of clinical response and remission in patients who have not responded optimally to SSRI alone (Leuchter et al., 2008; Zisook et al., 2006). A recently published 6-week randomized study compared the effectiveness of fluoxetine monotherapy vs. MRT in combination with fluoxetine, VLFX or BUP as initial MDD treatment (Blier et al., 2010). Greater improvement in depressive symptoms and remission rates was observed with the three combination therapy groups compared with fluoxetine monotherapy. Interestingly, most of the patients in this study had melancholic features (Rush, 2010), raising the possibility that MDD-MF may be particularly responsive to antidepressant combination therapy relative to SSRI monotherapy. However, this study was not designed to address antidepressant effect modification by MDD subtype.

In our study, no advantage of antidepressant combination therapy over SSRI monotherapy for inducing remission or clinical response was found in either the MDD-MF or non-melancholic MDD groups, at least for the combinations tested. Our remission rates are consistent with those reported among STAR*D participants with MDD-MF and non-melancholic MDD (McGrath et al., 2008). In that data analysis, no differences in remission or response rates between melancholic and non-melancholic groups were observed after adjustment for selected clinical characteristics that differed between groups at baseline; however, the MDD-MF group had a significantly lower remission rate in an unadjusted analyses based on achieving a HSRD17 score of ≤7. Our definition of remission was more conservative, requiring at least two consecutive QIDS-SR16 ratings of <8, with at least one <6. We based our remission definition on clinical ratings at two consecutive study visits in hopes of limiting false-positives due to transient mood elevation. Additionally, the HSRD17 does not cover diurnal variation, and incompletely addresses anhedonia and psychomotor agitation, all of which are key factors in melancholia. While the QIDS-SR16 also does not measure all melancholic features, it is more relevant to clinical practice than the IDS-C (which does assess all DSM-IV melancholic symptoms), since the QIDS-SR16 is a self-rated instrument and takes only 5 to 7 min to complete.

Our sample was representative of those seeking care in community practice. However, our ability to identify differential treatment effects based on melancholic/non-melancholic features may have been limited by higher prevalence of concurrent general medical illness, psychiatric comorbidity (particularly anxiety and somatoform spectrum disorders), anxiety symptoms, and emotional abuse in the MDD-MF group, similar to findings from previous studies (McGrath et al., 2008; Rush et al., 2008). These factors have all been associated with poorer clinical response in MDD patients (Trivedi et al., 2006; Walker et al., 2000), and may have reduced the likelihood of response or remission in the MDD-MF group compared with the non-melancholic MDD group in our study.

Despite the aforementioned differences in methodology, we were able to confirm several findings from the STAR*D analysis (McGrath et al., 2008). In both studies, MDD-MF patients had higher depression severity scores and greater adverse effect burden than did the non-melancholic MDD patients in spite of similar study drug doses and attrition rates. Higher mean antidepressant doses in the MDD-MF group for some agents, VLFX-XR in particular, may have partially accounted for adverse effect differences between melancholic and non-melancholic groups. In STAR*D, we found little evidence that demographic and treatment features could be helpful for differentiating melancholic and non-melancholic MDD. Our results extend the STAR*D findings, however, by addressing the comparative effects of antidepressant combinations vs. SSRI monotherapy in melancholic and non-melancholic depressives.

Our study had several limitations in addition to those already discussed. First, this was a planned post-hoc analysis of data from a clinical trial not designed to address our study question. CO-MED randomization was not stratified by melancholic status. Second, we identified MDD-MF patients using a symptom checklist and the IDS-C30. Melancholic features were therefore not ascertained via a structured interview, which may have resulted in some misclassification. Third, even though the QIDS-SR16 and IDS-C show high concordance (Rush et al., 2006a), the QIDS-SR16 does not completely address all DSM-IV melancholic features. Thus, our ability to measure improvement in melancholic symptoms may have been limited to some degree by our a priori definitions of outcome. Finally, although aggregate mean values suggested generally adequate antidepressant dosing, a substantial number of individual patients may not have received doses sufficient to elicit full benefit, particularly for combination therapy. Mean VLFX-XR doses were particularly modest, and may not have been at the level where putative multiple neurotransmitter systems are affected (Harvey et al., 2000).

In conclusion, we found no evidence of differential remission or response rates between melancholic/non-melancholic MDD patients after 12 and 28 weeks. Our results further indicate that antidepressant combination therapy with BUP-SR+ESC or VLFX-XR+MRT does not result in higher remission rates or differential improvement in other clinical response indicators relative to SSRI monotherapy (ESC+PLC) in patients with melancholic or non-melancholic MDD. Melancholic features may not be a valid predictor of more favorable response to antidepressant combination therapy as initial treatment.

Acknowledgments

The authors would like to thank the clinical staff at each clinical site for their assistance with this project; all of the study participants who contributed to this project; and Eric Nestler, MD, PhD, Lou and Ellen McGinley Distinguished Professor and Past Chairman, Department of Psychiatry, University of Texas Southwestern Medical Center as well as Carol A. Tamminga, MD, Communities Foundation of Texas, Inc. Chair in Brain Science, and Interim Chair, Department of Psychiatry, University of Texas Southwestern Medical Center for their administrative support. We appreciate the support of Forest Pharmaceuticals, Inc., GlaxoSmithKline, Organon, Inc., and Wyeth Pharmaceuticals in providing medications at no cost for the CO-MED trial.

Role of funding source

This project was funded by the National Institute of Mental Health under Contract N01MH90003 to UT Southwestern Medical Center at Dallas (PIs A.J. Rush, M.H. Trivedi). The content of this publication does not necessarily reflect the views or policies of the Department of Health and Human Services, nor does mention of trade names, commercial products, or organizations imply endorsement by the U. S. Government. The NIMH had no role in the drafting or review of the manuscript, nor in the collection or analysis of the data.

Conflict of interest

William V. Bobo, MD, MPH, has received research/grant support from Cephalon, Inc., the National Alliance for Research on Schizophrenia and Depression (NARSAD), and the National Institute of Mental Health (NIMH); and has served in the past on speaker bureaus for Janssen Pharmaceutica and Pfizer, Inc.

Helen Chen, MD, has received research support from Lundbeck, and speaker fees from GlaxoSmithKline.

A. John Rush, MD, has received consultant fees from Advanced Neuromodulation Systems, AstraZeneca, Best Practice Project Management, Bristol-Myers Squibb/Otsuka, Cyberonics, Forest Pharmaceuticals, Gerson Lehrman Group, GlaxoSmithKline, Jazz Pharmaceuticals, Magellan Health Services, Merck & Company, Neuronetics, Novartis Pharmaceuticals, Ono Pharmaceuticals, Organon, Otsuka Pharmaceuticals, Pamlab, Pfizer, Transcept Pharmaceuticals, Urban Institute and Wyeth Ayerst; speaking fees from Cyberonics Inc., Forest Laboratories, GlaxoSmithKline and Otsuka; royalties from Guilford Publications, and Healthcare Technology Systems; and research support from National Institutes of Mental Health and the Stanley Medical Research Institute. He has owned stock in Pfizer.

Benji T. Kurian, MD, has received research support from Targacept, Inc., and Pfizer, Inc.

Diane Warden, PhD, MBA, has previously owned stock in Pfizer, Inc., and Bristol-Myers Squibb.

David W. Morris, PhD, has no disclosures.

Ian A. Cook, MD, has served as an advisor and consultant for Ascend Media, Bristol-Myers Squibb, Cyb eronics, Janssen Pharmaceutica, Neuro-Sigma, and the U.S. Departments of Defense and of Justice. He has served on the speaker bureaus for Bristol-Myers Squibb, Neuronetics, and Wyeth/Pfizer. He has received research support from Aspect Medical Systems/Covidien, Cyberonics, Eli Lilly and Company, Neuronetics, Novartis, Pfizer, Sepracor, and the National Institutes of Health. His patents on biomedical devices are assigned to the Regents of the University of California.

Richard C. Shelton, MD, has served as a consultant for Eli Lilly and Company, Evotec AG, Forest Pharmaceuticals, Gideon Richter PLC, Janssen Pharmaceutica, Medronic, Inc., Otsuka Pharmaceuticals, Pamlab, Inc., Pfizer, Inc., Repligen, Inc., and Sierra Neuropharmaceuticals. He has received research support from Eli Lilly and Company, Forest Pharmaceuticals, Janssen Pharmaceutica, Novartis Pharmaceuticals, Otsuka Pharmaceuticals, Pamlab, Pfizer, Inc., Repligen, Inc., and St. Jude Medical.

Ira M. Lesser, MD, has received research support from the National Institute of Mental Health and Aspect Medical System/Covidian.

Susan G. Kornstein, MD, has served as a consultant for Bristol-Myers Squibb, Forest Laboratories, Eli Lilly and Company, Pfizer, Inc., Rexahn, Dey Pharma, PGxhealth, Takeda, and Wyeth. She has received research support from the National Institute of Mental Health, Boehringer-Ingelheim, Bristol-Myers Squibb, Eli Lilly and Company, Forest Laboratories, Wyeth, Novartis, Pfizer, Inc., Otsuka, and Takeda. She has received book royalties from Guildord Press.

Appendix 1. Comorbidity measures by melancholic/non-melancholic features

Measure Melancholic features p
Yes (N = 124) No (N = 481)
Mean SD Mean SD
No. treated health problems 3.3 3.1 3.4 3.7 0.68
N % N %
Agoraphobia 22 17.1 43 8.9 0.005
Alcohol abuse 10 8.1 51 10.6 0.40
Bulimia 12 9.7 59 12.3 0.42
Drug abuse 9 7.3 25 5.2 0.37
Generalized anxiety 38 30.6 81 16.8 0.0006
Hypochondriasis 11 8.9 15 3.1 0.005
Obsessive-compulsive 21 16.9 51 10.6 0.052
Panic 17 13.7 43 8.9 0.11
Post-traumatic stress 21 16.9 48 10.0 0.03
Social phobia 46 37.1 115 23.9 0.003
Somatoform 8 6.5 11 2.3 0.04
Substance abuse, other 14 11.3 66 13.8 0.47
No. comorbid psychiatric disorders 0.002
 0 41 33.1 227 47.3
 1 29 23.4 118 24.6
 2 19 15.3 63 13.1
 3 12 9.7 35 7.3
 ≥ 4 23 18.5 37 7.7
No. treated comorbid general 0.33
medical conditions
 0 70 56.5 233 48.6
 1 22 17.7 119 24.8
 2 16 12.9 68 14.2
 ≥ 3 16 12.9 59 12.3
Specific medical conditions
 Chronic back pain 22 17.7 78 16.3 0.69
 Diabetes mellitus 13 10.5 55 11.4 0.77
 Heart disease 1 0.8 36 7.5 0.006
 Menopause 19 22.1 110 35.5 0.02
Open in a new tab

References

  1. Altman EG, Hedeker D, Peterson JL, Davis JM, 1997. The Altman self-rating mania scale. Biol. Psychiatry 42 (10), 948–955. [DOI] [PubMed] [Google Scholar]
  2. Amsterdam JD, 1998. Selective serotonin reuptake inhibitor efficacy in severe and melancholic depression. J Psychopharmacol 12 (3 Suppl B), S99–111. [DOI] [PubMed] [Google Scholar]
  3. Anderson IM, 1998. SSRIS versus tricyclic antidepressants in depressed inpatients: a meta-analysis of efficacy and tolerability. Depress. Anxiety 7 (Suppl 1), 11–17. [PubMed] [Google Scholar]
  4. Angst J, Stabl M, 1992. Efficacy of moclobemide in different patient groups: a meta-analysis of studies. Psychopharmacology (Berl) 106 Suppl, S109–S113. [DOI] [PubMed] [Google Scholar]
  5. Benkert O, Grunder G, Wetzel H, Hackett D, 1996. A randomized, double-blind comparison of a rapidly escalating dose of venlafaxine and imipramine in inpatients with major depression and melancholia. J. Psychiatr. Res 30 (6), 441–451. [DOI] [PubMed] [Google Scholar]
  6. Blier P, Ward HE, Tremblay P, Laberge L, Hebert C, Bergeron R, 2010. Combination of antidepressant medications from treatment initiation for major depressive disorder: a double-blind randomized study. Am. J. Psychiatry 167 (3), 281–288. [DOI] [PubMed] [Google Scholar]
  7. Brown WA, 2007. Treatment response in melancholia. Acta Psychiatr. Scand. Suppl (433), 125–129. [DOI] [PubMed] [Google Scholar]
  8. Clerc GE, Ruimy P, Verdeau-Palles J, 1994. A double-blind comparison of venlafaxine and fluoxetine in patients hospitalized for major depression and melancholia. The venlafaxine French inpatient study group. Int. Clin. Psychopharmacol 9 (3), 139–143. [DOI] [PubMed] [Google Scholar]
  9. Coryell W, Turner R, 1985. Outcome with desipramine therapy in subtypes of nonpsychotic major depression. J. Affect. Disord 9 (2), 149–154. [DOI] [PubMed] [Google Scholar]
  10. Frank E, Kupfer DJ, Hamer T, Grochocinski VJ, McEachran AB, 1992. Maintenance treatment and psychobiologic correlates of endogenous subtypes. J. Affect. Disord 25 (3), 181–189. [DOI] [PubMed] [Google Scholar]
  11. Frisch MB, Clark MP, Rouse SV, Rudd MD, Paweleck JK, Greenstone A, Kopplin DA, 2005. Predictive and treatment validity of life satisfaction and the quality of life inventory. Assessment 12, 66–78. [DOI] [PubMed] [Google Scholar]
  12. Gentil V, Kerr-Correa F, Moreno R, D’Arrigo BE, De Campos JA, Juruena MF, Lafer B, Moreno DH, De Cassia Rodrigues RL, Tiosso A, Benedictis E, 2000. Double-blind comparison of venlafaxine and amitriptyline in outpatients with major depression with or without melancholia. J Psychopharmacol 14 (1), 61–66. [DOI] [PubMed] [Google Scholar]
  13. Guelfi JD, Ansseau M, Timmerman L, Korsgaard S, 2001. Mirtazapine versus venlafaxine in hospitalized severely depressed patients with melancholic features. J. Clin. Psychopharmacol 21 (4), 425–431. [DOI] [PubMed] [Google Scholar]
  14. Guelfi JD, White C, Hackett D, Guichoux JY, Magni G, 1995. Effectiveness of venlafaxine in patients hospitalized for major depression and melancholia. J. Clin. Psychiatry 56 (10), 450–458. [PubMed] [Google Scholar]
  15. Hamilton M, 1960. A rating scale for depression. J. Neurol. Neurosurg. Psychiatry 23, 56–62. [DOI] [PMC free article] [PubMed] [Google Scholar]
  16. Harvey AT, Rudolph RL, Preskorn SH, 2000. Evidence of the dual mechanisms of action of venlafaxine. Arch. Gen. Psychiatry 57 (5), 503–509. [DOI] [PubMed] [Google Scholar]
  17. Hildebrandt MG, Stage KB, Kragh-Soerensen P, 2003. Gender differences in severity, symptomatology and distribution of melancholia in major depression. Psychopathology 36 (4), 204–212. [DOI] [PubMed] [Google Scholar]
  18. Kessing LV, 2007. Epidemiology of subtypes of depression. Acta Psychiatr. Scand. Suppl (433), 85–89. [DOI] [PubMed] [Google Scholar]
  19. Khan A, Kolts RL, Rapaport MH, Krishnan KR, Brodhead AE, Browns WA, 2005. Magnitude of placebo response and drug-placebo differences across psychiatric disorders. Psychol. Med 35 (5), 743–749. [DOI] [PubMed] [Google Scholar]
  20. Khan AY, Carrithers J, Preskorn SH, Lear R, Wisniewski SR, John RA, Stegman D, Kelley C, Kreiner K, Nierenberg AA, Fava M, 2006. Clinical and demographic factors associated with DSM-IV melancholic depression. Ann. Clin. Psychiatry 18 (2), 91–98. [DOI] [PubMed] [Google Scholar]
  21. Leuchter AF, Lesser IM, Trivedi MH, Rush AJ, Morris DW, Warden D, Fava M, Wisniewski SR, Luther JF, Perales M, Gaynes BN, Stewart JW, 2008. An open pilot study of the combination of escitalopram and bupropion-SR for outpatients with major depressive disorder. J Psychiatr. Pract 14 (5), 271–280. [DOI] [PMC free article] [PubMed] [Google Scholar]
  22. Leventhal AM, Rehm LP, 2005. The empirical status of melancholia: implications for psychology. Clin. Psychol. Rev 25 (1), 25–44. [DOI] [PubMed] [Google Scholar]
  23. Levine J, Schooler NR, 1986. SAFTEE: a technique for the systematic assessment of side effects in clinical trials. Psychopharmacol. Bull 22 (2), 343–381. [PubMed] [Google Scholar]
  24. Mallinckrodt CH, Watkin JG, Liu C, Wohlreich MM, Raskin J, 2005. Duloxetine in the treatment of Major Depressive Disorder: a comparison of efficacy in patients with and without melancholic features. BMC Psychiatry 5, 1. [DOI] [PMC free article] [PubMed] [Google Scholar]
  25. McGrath PJ, Khan AY, Trivedi MH, Stewart JW, Morris DW, Wisniewski SR, Miyahara S, Nierenberg AA, Fava M, Rush JA, 2008. Response to a selective serotonin reuptake inhibitor (citalopram) in major depressive disorder with melancholic features: a STAR*D report. J. Clin. Psychiatry 69 (12), 1847–1855. [DOI] [PubMed] [Google Scholar]
  26. Mundt JC, Marks IM, Shear MK, Greist JH, 2002. The work and social adjustment scale: a simple measure of impairment in functioning. Br. J. Psychiatry 180, 461–464. [DOI] [PubMed] [Google Scholar]
  27. Parker G, Hadzi-Pavlovic D, Brodaty H, Boyce P, Mitchell P, Wilhelm K, Hickie I, 1992. Predicting the course of melancholic and nonmelancholic depression. A naturalistic comparison study. J. Nerv. Ment. Dis 180 (11), 693–702. [DOI] [PubMed] [Google Scholar]
  28. Perry PJ, 1996. Pharmacotherapy for major depression with melancholic features: relative efficacy of tricyclic versus selective serotonin reuptake inhibitor antidepressants. J. Affect. Disord 39 (1), 1–6. [DOI] [PubMed] [Google Scholar]
  29. Peselow ED, Sanfilipo MP, Difiglia C, Fieve RR, 1992. Melancholic/endogenous depression and response to somatic treatment and placebo. Am. J. Psychiatry 149 (10), 1324–1334. [DOI] [PubMed] [Google Scholar]
  30. Prusoff BA, Weissman MM, Klerman GL, Rounsaville BJ, 1980. Research diagnostic criteria subtypes of depression. Their role as predictors of differential response to psychotherapy and drug treatment. Arch. Gen. Psychiatry 37 (7), 796–801. [DOI] [PubMed] [Google Scholar]
  31. Rush AJ, 2010. Combining antidepressant medications: a good idea? Am. J. Psychiatry 167 (3), 241–243. [DOI] [PubMed] [Google Scholar]
  32. Rush AJ, Carmody TJ, Ibrahim HM, Trivedi MH, Biggs MM, Shores-Wilson K, Crismon ML, Toprac MG, Kashner TM, 2006a. Comparison of self-report and clinician ratings on two inventories of depressive symptomatology. Psychiatr. Serv 57 (6), 829–837. [DOI] [PubMed] [Google Scholar]
  33. Rush AJ, Gullion CM, Basco MR, Jarrett RB, Trivedi MH, 1996. The Inventory of Depressive Symptomatology (IDS): psychometric properties. Psychol. Med 26 (3), 477–486. [DOI] [PubMed] [Google Scholar]
  34. Rush AJ, Trivedi MH, Ibrahim HM, Carmody TJ, Arnow B, Klein DN, Markowitz JC, Ninan PT, Kornstein S, Manber R, Thase ME, Kocsis JH, Keller MB, 2003. The 16-Item Quick Inventory of Depressive Symptomatology (QIDS), clinician rating (QIDS-C), and self-report (QIDS-SR): a psychometric evaluation in patients with chronic major depression. Biol. Psychiatry 54 (5), 573–583. [DOI] [PubMed] [Google Scholar]
  35. Rush AJ, Trivedi MH, Stewart JW, Nierenberg AA, Fava M, Kurian BT, Warden D, Morris DW, Luther JF, Husain MM, Cook IA, Shelton RC, Lesser IM, Kornstein SG, Wisniewski SR, 2011. Combining Medications to Enhance Depression Outcomes (CO-MED): acute and long-term outcomes: a single-blind randomized study. Am. J. Psychiatry 168, 689–701 Ref Type: Generic. [DOI] [PubMed] [Google Scholar]
  36. Rush AJ, Trivedi MH, Wisniewski SR, Nierenberg AA, Stewart JW, Warden D, Niederehe G, Thase ME, Lavori PW, Lebowitz BD, McGrath PJ, Rosenbaum JF, Sackeim HA, Kupfer DJ, Luther J, Fava M, 2006b. Acute and longer-term outcomes in depressed outpatients requiring one or several treatment steps: a STAR*D report. Am. J. Psychiatry 163 (11), 1905–1917. [DOI] [PubMed] [Google Scholar]
  37. Rush AJ, Weissenburger JE, 1994. Melancholic symptom features and DSM-IV. Am. J. Psychiatry 151 (4), 489–498. [DOI] [PubMed] [Google Scholar]
  38. Rush AJ, Wisniewski SR, Warden D, Luther JF, Davis LL, Fava M, Nierenberg AA, Trivedi MH, 2008. Selecting among second-step antidepressant medication monotherapies: predictive value of clinical, demographic, or first-step treatment features. Arch. Gen. Psychiatry 65 (8), 870–880. [DOI] [PubMed] [Google Scholar]
  39. Rush AJ, Zimmerman M, Wisniewski SR, Fava M, Hollon SD, Warden D, Biggs MM, Shores-Wilson K, Shelton RC, Luther JF, Thomas B, Trivedi MH, 2005. Comorbid psychiatric disorders in depressed outpatients: demographic and clinical features. J. Affect. Disord 87 (1), 43–55. [DOI] [PubMed] [Google Scholar]
  40. Sangha O, Stucki G, Liang MH, Fossel AH, Katz JN, 2003. The Self-Administered Comorbidity Questionnaire: a new method to assess comorbidity for clinical and health services research. Arthritis Rheum 49 (2), 156–163. [DOI] [PubMed] [Google Scholar]
  41. Sheehan DV, Nemeroff CB, Thase ME, Entsuah R, 2009. Placebo-controlled inpatient comparison of venlafaxine and fluoxetine for the treatment of major depression with melancholic features. Int. Clin. Psychopharmacol 24 (2), 61–86. [DOI] [PubMed] [Google Scholar]
  42. Thase ME, Friedman ES, 1999. Is psychotherapy an effective treatment for melancholia and other severe depressive states? J. Affect. Disord 54 (1–2), 1–19. [DOI] [PubMed] [Google Scholar]
  43. Trivedi MH, Morris DW, Wisniewski SR, Fava M, Gollan J, Nierenberg AA, Gaynes BN, Husain MM, Luther JF, Zisook S, and Rush AJ Concise Associated Symptoms Tracking (CAST) Scale and Concise Health Risk Tracking (CHRT) Scale: brief self and clinician ratings of suicidality and associated symptoms. Poster presented at the 50th National Clinical Drug Evaluation Unity (NCDEU) Annual Meeting, Boca Raton, FL June 14–17, 2010. 2010. Ref Type: Generic [Google Scholar]
  44. Trivedi MH, Rush AJ, Gaynes BN, Stewart JW, Wisniewski SR, Warden D, Ritz L, Luther JF, Stegman D, veaugh-Geiss J, Howland R, 2007. Maximizing the adequacy of medication treatment in controlled trials and clinical practice: STAR(*)D measurement-based care. Neuropsycho-pharmacology 32 (12), 2479–2489. [DOI] [PubMed] [Google Scholar]
  45. Trivedi MH, Rush AJ, Wisniewski SR, Nierenberg AA, Warden D, Ritz L, Norquist G, Howland RH, Lebowitz B, McGrath PJ, Shores-Wilson K, Biggs MM, Balasubramani GK, Fava M, 2006. Evaluation of outcomes with citalopram for depression using measurement-based care in STAR*D: implications for clinical practice. Am. J. Psychiatry 163 (1), 28–40. [DOI] [PubMed] [Google Scholar]
  46. Tzanakaki M, Guazzelli M, Nimatoudis I, Zissis NP, Smeraldi E, Rizzo F, 2000. Increased remission rates with venlafaxine compared with fluoxetine in hospitalized patients with major depression and melancholia. Int. Clin. Psychopharmacol 15 (1), 29–34. [DOI] [PubMed] [Google Scholar]
  47. Walker EA, Katon WJ, Russo J, Von KM, Lin E, Simon G, Bush T, Ludman E, Unutzer J, 2000. Predictors of outcome in a primary care depression trial. J. Gen. Intern. Med 15 (12), 859–867. [DOI] [PMC free article] [PubMed] [Google Scholar]
  48. Wisniewski SR, Rush AJ, Balasubramani GK, Trivedi MH, Nierenberg AA, 2006. Self-rated global measure of the frequency, intensity, and burden of side effects. J. Psychiatr. Pract 12 (2), 71–79. [DOI] [PubMed] [Google Scholar]
  49. Zimmerman M, Spitzer RL, 1989. Melancholia: from DSM-III to DSM-III-R. Am. J. Psychiatry 146 (1), 20–28. [DOI] [PubMed] [Google Scholar]
  50. Zisook S, Rush AJ, Haight BR, Clines DC, Rockett CB, 2006. Use of bupropion in combination with serotonin reuptake inhibitors. Biol. Psychiatry 59 (3), 203–210. [DOI] [PubMed] [Google Scholar]

RESOURCES