Approximately one-half of all patients diagnosed with major depressive disorder (MDD) have clinically meaningful levels of anxiety, which complicates clinical management and can affect treatment outcomes (Rao and Zisook, 2009; Wiethoff et al., 2010). Anxious depression is associated with more severe depressive symptoms, greater duration of current episode, suicidal ideation and history of suicide attempt, and more medical comorbidities (Rao and Zisook, 2009). Patients with both depressive and anxiety symptoms are less likely to achieve treatment response or remission (Fava et al., 2008; Kennedy, 2008), more likely to have delayed treatment response (Clayton et al., 1991), and are more likely to report adverse events during antidepressant treatment, including bursts of anxiety or agitation (Fava et al., 2006). Residual anxiety symptoms are also associated with an increased risk of MDD relapse (Ramana et al., 1995; Flint and Rifat, 1997).
Although anxious depression was not historically recognized as a separate diagnostic entity, the recently published Diagnostic and Statistical Manual of Mental Disorders, 5th ed. (DSM-V) includes ‘anxious distress’ as a category specifier for patients with depressive disorders (American Psychiatric Association, 2013). The symptoms listed for anxious distress include tenseness, restlessness, difficulty concentrating because of worry, fear that something awful may happen, and feelings of loss of control. Severity is based on the number of symptoms that patients exhibit.
Prior to the release of the DSM-V, clinical studies used several methods to identify patients with mixed depressive and anxiety symptoms (Ionescu et al., 2013). In the Sequenced Treatment Alternatives to Relieve Depression (STAR*D) study (Fava et al., 2008), anxious depression was defined as patients meeting the DSM-IV-TR (4th ed., text revision) criteria (American Psychiatric Association, 2000a) for MDD and having a score of 7 or greater on the Anxiety/Somatization subscale of the 17-item Hamilton Depression Rating Scale (HAMD17) (Hamilton, 1960). Slightly more than half of the STAR*D patients fulfilled these criteria (53.2%) and, when compared with nonanxious patients, those with anxious depression had poorer treatment outcomes across the first and second levels of therapy.
The availability of a pharmacologic treatment that effectively manages anxious depression would be an important development for the therapeutics of MDD. As many of the selective serotonin reuptake inhibitors (SSRIs) have established efficacy for anxiety disorders, they are commonly considered a first-line treatment for anxious depression (Nutt, 2005). For patients who do not respond to SSRIs, alternatives include other newer generation antidepressants, such as the serotonin and norepinephrine reuptake inhibitors, and adjunctive therapy with atypical antipsychotics and benzodiazepines. Adjunctive therapy with the alternate anxiolytic agent buspirone is also sometimes used to treat anxious depression that has not responded adequately to SSRIs (American Psychiatric Association, 2000b). From a mechanistic standpoint, buspirone is a strong 5-HT1A partial agonist that is considered to decrease serotonergic autoinhibition by desensitizing 5-HT1A autoreceptors (Albert and Francois, 2010). Buspirone also shows moderate affinity at dopamine D2 receptors and a short half-life (2–3 h for 10-40 mg/day), requiring twice-daily dosing (Bristol-Myers Squibb Company, 2010). The therapeutic potential of combining the actions of an SSRI and a 5-HT1A partial agonist is of clinical interest, but the value of such a strategy has not yet been shown conclusively. Several preliminary clinical studies found that adjunctive use of buspirone with an antidepressant agent may help improve symptoms in patients with inadequate response to antidepressant monotherapy (Nelson, 2000; Fleurence et al., 2009). However, the benefits of this adjunctive strategy in patients with anxious depression were not confirmed in the STAR*D study (Fava et al., 2008), although the results from this study must be interpreted with caution as inadequate response following buspirone augmentation may have been confounded by the study design or the specific pharmacokinetic or pharmacodynamic profile of buspirone.
Vilazodone is a serotonin reuptake inhibitor and 5-HT1A receptor partial agonist that was approved by the US Food and Drug Administration (FDA) in January 2011 for the treatment of MDD in adults. Vilazodone potently and selectively inhibits reuptake of serotonin and binds selectively with high affinity to 5-HT1A receptors (Forest Laboratories Inc., 2013). With a terminal half-life of ∼25 h, vilazodone is administered once daily with food. It has been proposed that because vilazodone combines serotonin reuptake inhibition and a buspirone-like anxiolytic mechanism, it may be a useful treatment option for patients with MDD and symptoms of anxiety. The efficacy and safety of vilazodone have been established by the results of two phase III, 8-week, randomized, double-blind, placebo-controlled trials: NCT00285376 (RCT-1) (Rickels et al., 2009) and NCT00683592 (RCT-2) (Khan et al., 2011). In both trials, change from baseline to week 8 was statistically significant in favor of vilazodone versus placebo on the primary efficacy measure, the Montgomery–Asberg Depression Rating Scale (MADRS) (Montgomery and Asberg, 1979). Statistically significant improvements were also observed on secondary endpoints including the HAMD17 and Hamilton Anxiety Rating Scale (HAMA) (Hamilton, 1959). Further support for the safety and tolerability of vilazodone in patients with MDD was demonstrated in a 1-year open-label study (NCT00644358) (Robinson et al., 2011). In all three studies, diarrhea and nausea were the most common treatment-emergent adverse events associated with vilazodone therapy.
To further investigate the clinical benefits of vilazodone in patients with MDD and prominent anxiety symptoms, data from the two phase III vilazodone studies (RCT-1 and RCT-2) were pooled. Post-hoc analyses using the HAMD17 and MADRS scales were conducted to evaluate changes in depressive symptoms in anxious and nonanxious patients as well as evaluate changes in anxiety symptoms in the anxious subgroup. Several different anxiety-related measures were used to assess both the somatic and the psychic symptoms of anxiety found in patients with MDD.
RCT-1 and RCT-2 were carried out at multiple US study centers between 2006 and 2009 in full compliance with FDA guidelines for good clinical practice and the ethical principles of the Declaration of Helsinki. Final study protocols were approved by the appropriate institutional review board for each study site and all patients provided written informed consent before the initiation of any study-specific procedures.
Detailed study design, inclusion and exclusion criteria, and statistical methods from the two pivotal phase III, randomized, double-blind, placebo-controlled, multicenter studies of vilazodone studies have been published (Rickels et al., 2009; Khan et al., 2011). Adult patients with MDD were randomized (1 : 1) to receive vilazodone 40 mg/day or placebo for 8 weeks of double-blind treatment. Patients randomized to active treatment received once-daily vilazodone 10 mg for 7 days, 20 mg for 7 days, and 40 mg for the rest of the study.
Principal inclusion and exclusion criteria
Adult patients (aged 18–70 years) with MDD (single or recurrent episode) as defined by the DSM-IV-TR were included. Patients were required to have a current major depressive episode of 4 or more weeks’ and less than 2 years’ duration, a HAMD17 total score of at least 22, and HAMD17 item 1 (Depressed Mood) score at least 2. Psychiatric exclusions included an Axis I disorder other than MDD within 6 months of screening (generalized anxiety disorder, social phobia, or simple phobia were allowed), a history of psychotic or bipolar disorders, substance abuse (in the past 3 months) or dependence (in the past 6 months), and serious suicide or homicide risk. Patients with clinically significant medical conditions that might interfere with study participation were excluded at the discretion of the investigator; patients taking mediations with serotonergic mechanisms of action or who had not responded to two or more previous antidepressants from different classes were also excluded.
Patient subgroups and efficacy assessments
The intent-to-treat population, defined as all randomized patients who received at least one dose of study drug and had at least one postbaseline MADRS total score assessment, was divided into subgroups based on baseline HAMD17 Anxiety/Somatization subscale scores. The HAMD17 Anxiety/Somatization subscale comprises six individual items: item 10, Psychic Anxiety; item 11, Somatic Anxiety; item 12, Somatic Symptoms (gastrointestinal); item 13, General Somatic Symptoms; item 15, Hypochondriasis; and item 17, Insight. Patients with a baseline HAMD17 Anxiety/Somatization subscale score 7 or more were classified as having anxious depression; those with a score less than 7 were classified as having nonanxious depression.
MADRS and HAMD17 total scores were used to assess the effects of vilazodone on depressive symptoms in both the anxious and the nonanxious subgroups. Since the smaller nonanxious subgroup (n=155) did not have prominent symptoms of anxiety, analyses of anxiety-related measures were only carried out in the anxious subgroup (n=708). In addition to the HAMA total score, two HAMA-derived subscales were evaluated: the Psychic Anxiety subscale (item 1, Anxious Mood; item 2, Tension; item 3, Fears; item 4, Insomnia; item 5, Intellectual; item 6, Depressed Mood; and item 14, Behavior at Interview) and the Somatic Anxiety subscale [item 7, Somatic (muscular); item 8, Somatic (sensory); item 9, Cardiovascular Symptoms; item 10, Respiratory Symptoms; item 11, Gastrointestinal Symptoms; item 12, Genitourinary Symptoms; and item 13, Autonomic Symptoms]. Changes in anxiety symptoms were also measured using the HAMD17 Anxiety/Somatization subscale, Psychic Anxiety item (item 10), and Somatic Anxiety item (item 11), as well as the MADRS Inner Tension item (item 3).
Least squares mean (LSM) changes from baseline to week 8 for all post-hoc efficacy parameters were analyzed using a mixed-effects model for repeated measures with treatment group, study, study center, visit, and treatment group by visit interaction as fixed effects and baseline value and baseline value by visit as covariates. An unstructured covariance matrix was used to model the covariance of within-patient scores. The Kenward–Roger approximation was used to estimate denominator degrees of freedom (Kenward and Roger, 1997). Only observed cases of postbaseline scores were used without imputation of missing values; P values for the statistical tests in the post-hoc analyses are nominal and not adjusted for multiple comparisons. Treatment effect sizes for vilazodone versus placebo in anxiety-related measures were estimated using Cohen’s d formula.
Of the 863 patients included in the intent-to-treat population, 708 (82.0%) patients were classified as having anxious depression (HAMD17 Anxiety/Somatization subscale score≥7); 155 (18.0%) were classified as nonanxious. For both the anxious and the nonanxious depression subgroups, the demographic characteristics and history of MDD were generally similar between the vilazodone and the placebo treatment groups (Table 1). In the anxious depression subgroup, the mean duration of the current episode was 1 or less year for the majority of patients (>75%); approximately two-thirds of patients reported had a previous depressive episode. The mean baseline HAMD17 and MADRS total scores met or exceeded the threshold scores commonly used to indicate severe depressive symptomatology (Nemeroff, 2007). The mean baseline HAMA total scores were consistent with those observed in studies of generalized anxiety disorder (Matza et al., 2010).
Efficacy in depression measures
Mean improvements with vilazodone were found for MADRS and HAMD17 total scores, with statistically significant differences from placebo in the anxious depression subgroup. LSM differences between vilazodone and placebo [with 95% confidence interval (95% CI)] for changes in the MADRS total score were as follows: anxious, −3.6 (−5.2 to −2.0; P<0.001); nonanxious, −0.9 (−4.6 to 2.8; P>0.05). LSM differences for changes in the HAMD17 total score were as follows: anxious, −2.3 (−3.5 to −1.1; P<0.001); nonanxious, −0.5 (−3.2 to 2.2; P>0.05).
Efficacy in anxiety-related measures in patients with anxious depression
In the anxious depression subgroup, patients treated with vilazodone showed significantly greater improvements in anxiety-related outcomes than patients receiving placebo. After 8 weeks of treatment, the LSM difference between vilazodone and placebo in the HAMA total score was −1.82 (95% CI −2.81 to −0.83; P<0.001). Significant differences between treatment groups were observed by week 6 (Fig. 1a). On the HAMD17 Anxiety/Somatization subscale, significant mean improvements with vilazodone versus placebo were detected by week 2 (Fig. 1b). At the end of the double-blind treatment, the LSM difference between vilazodone and placebo in the HAMD17 Anxiety/Somatization subscale score was −0.75 (95% CI −1.17 to −0.32; P<0.001). As estimated by Cohen’s d, the treatment effect size was 0.25 for both the HAMA total and the HAMD17 Anxiety/Somatization subscale scores.
Statistically significant improvements with vilazodone were also detected on all other anxiety-related measures, except the HAMA Somatic Anxiety subscale (Table 2). The largest treatment effects were observed on the HAMA Psychic Anxiety subscale (0.31) and the MADRS Inner Tension item (0.27).
Identifying of effective therapeutic strategies for anxious depression is an important clinical and societal concern; anxiety symptoms in patients with MDD have been associated with increased disease severity, functional disability, treatment resistance, and increased healthcare costs (Rao and Zisook, 2009). We therefore examined the effects of vilazodone in patients with anxious MDD by pooling the samples of two phase III studies. In earlier papers describing the results of the individual studies, vilazodone therapy was reported to have significant effects on the HAMA total score (Rickels et al., 2009; Khan et al., 2011). In the pooled analyses, we could better characterize the types of anxiety symptoms and further evaluate the impact of vilazodone treatment in patients with anxious depression.
We believe that our pooled analysis has yielded several noteworthy findings. Of great relevance to the aims of the current analysis, 82% of the pooled study group met the criteria for anxious depression compared with only about 50% of STAR*D participants (Fava et al., 2004, 2006, 2008). Although the reasons for this variance cannot be ascertained definitively, it is possible that site investigators in the vilazodone studies preferentially enrolled patients with prominent anxiety. Alternatively, differences in the entry criteria could also be an important factor. Specifically, the studies for this pooled analysis required a much higher HAMD17 total score for entry (≥22) than STAR*D (≥14). As moderately high correlations between anxiety and depressive symptoms are found in MDD (Fava et al., 2004), it is plausible that the greater severity of patients at the time of entry into the vilazodone studies accounted for the high prevalence of patients with anxious symptoms.
Only 18% of the pooled vilazodone study population had nonanxious depression at baseline. Despite this relatively small number of patients, we analyzed changes in the overall depression measures (i.e. MADRS and HAMD17 total scores) in this subgroup as well as the anxious subgroup to investigate potential differences in treatment effects. The results suggested larger mean improvements with vilazodone versus placebo in the anxious subgroup relative to the nonanxious subgroup, which was somewhat unexpected as similar analyses with other antidepressants generally failed to find a difference or found lower efficacy among patients with anxious depression (Fava et al., 2008; Nelson, 2010). Although caution is advised when comparing the anxious and nonanxious groups as the sample size of nonanxious patients was relatively small, resulting in larger deviations (i.e. larger CIs), these results suggest that vilazodone may have additional benefits in treating depression with prominent anxiety symptoms. Given its activity as a 5-HT1A receptor partial agonist, vilazodone may have had some preferential effects on anxiety symptoms, which are prevalent symptoms associated with MDD and assessed by both the MADRS and HAMD17. Given the statistical limitations of the present analysis, further research may be warranted to evaluate these issues more thoroughly.
At baseline, the mean scores for the HAMA subscales and HAMD17 items suggest that psychic symptoms (e.g. irritability, tension, worrying) were more pronounced in this patient population than somatic symptoms (e.g. headaches, palpitations, gastrointestinal disturbances, hyperventilation) (Table 2). These results are consistent with other studies that have reported higher baseline scores for psychic anxiety than for somatic anxiety in patients with MDD (Russell et al., 2007; Nelson et al., 2010).
After 8 weeks of treatment, statistically significant improvements in favor of vilazodone versus placebo were observed in the anxious depression subgroup for HAMA total and HAMA Psychic Anxiety subscale scores (P<0.001 for each measure). Although numerical improvements with vilazodone were also observed on the HAMA Somatic Anxiety subscale, the difference from placebo did not reach statistical significance. This may have been partly because of baseline Somatic Anxiety subscale scores being markedly lower than Psychic Anxiety subscale scores (∼6 and 13, respectively), which might have limited the potential to observe statistical improvements for somatic anxiety. It is also possible that as the HAMA was developed primarily as a rating scale for anxiety disorders, it may have relatively low sensitivity to detect somatic anxiety symptoms associated with depression. As the HAMD17 was designed to evaluate symptoms in patients with a primary diagnosis of depression, the HAMD17 Anxiety/Somatization subscale and the Somatic Anxiety item may be more appropriate than the HAMA for evaluating of somatic anxiety in patients with MDD. On both of these anxiety measures, as well as the HAMD17 Psychic Anxiety subscale and the MADRS Inner Tension item, mean improvements were significantly greater with vilazodone than placebo in patients with anxious depression.
Although treatment effect sizes for anxiety-related measures were generally modest in this analysis (d<0.4), the significant improvements found with vilazodone warrant some consideration. It is not possible to directly compare the vilazodone studies with the STAR*D study because of considerable differences in patient selection, treatment, and study design. However, the STAR*D study did show that in patients who did not achieve remission with the SSRI citalopram, the addition of buspirone was a relatively ineffective strategy for patients with anxious depression. Given that patients in the vilazodone studies were immediately randomized to treatment, rather than provided an adjunctive treatment after failing initial antidepressant therapy as in the STAR*D study, the favorable results found with vilazodone on various anxiety measures suggest that the effective management of anxious depression may require early detection of anxiety symptoms and prompt administration of a medication with potential anxiolytic mechanisms.
The current analyses were limited by the retrospective nature of the evaluations, which were not corrected for multiple comparisons. In addition, although the HAMD17 definition of anxious depression has been used in other studies, subgrouping patients using another criterion might have resulted in different outcomes for vilazodone versus placebo. Moreover, neither study was designed to specifically recruit or compare patients with anxious versus nonanxious depression, and the observed differences in treatment effects might be related to other differences between these groups in baseline characteristics. Finally, the patients included in the vilazodone studies are not representative of all patients with MDD and the analyses presented in this report may not be generalizable to a wider population.
In these post-hoc analyses of pooled data from two pivotal, phase III clinical trials, efficacy across broad measures of anxiety and in a subgroup of patients with anxious depression suggests that vilazodone is an effective treatment option for the large proportion of patients with MDD who have prominent anxiety symptoms. The promising results from these analyses indicate the need for further research evaluating various strategies for the management of patients with anxious depression.
Writing assistance and editorial support was provided by Mildred Bahn at Prescott Medical Communications Group (Chicago, Illinois, USA), with support from Forest Research Institute Inc.
Conflicts of interest
The clinical studies analyzed in this report were supported by Forest Laboratories Inc. M. Thase has served as an advisor/consultant for Alkermes, AstraZeneca, Bristol-Myers Squibb Company, Eli Lilly & Co., Dey Pharma, L.P., Forest Laboratories Inc., Gerson Lehman Group, GlaxoSmithKline, Guidepoint Global, H. Lundbeck A/S, MedAvante Inc., Merck and Co. Inc. (formerly Schering Plough and Organon), Neuronetics Inc., Novartis, Otsuka, Ortho-McNeil Pharmaceuticals (Johnson & Johnson), Pamlab, L.L.C., Pfizer (formerly Wyeth Ayerst Pharmaceuticals), PGx Health Inc., Shire US Inc., Supernus Pharmaceuticals, Takeda, and Transcept Pharmaceuticals. He has received grant support from Agency for Healthcare Research and Quality, Eli Lilly and Company, Forest Pharmaceuticals Inc., GlaxoSmithKline, National Institute of Mental Health, Otsuka Pharmaceuticals, and Sepracor Inc. He has served on the Speakers Bureau for AstraZeneca, Bristol-Myers Squibb Company, Dey Pharmaceutical, Eli Lilly & Co., Merck and Co. Inc., and Pfizer (formerly Wyeth Ayerst Pharmaceuticals). D. Chen and J. Edwards are full-time employees of Forest Research Institute, a wholly owned subsidiary of Forest Laboratories Inc. A. Ruth is a full-time employee of Prescott Medical Communications Group, a contractor for Forest Research Institute Inc.
Albert PR, Francois BL. Modifying 5-HT1A receptor gene expression as a new target for antidepressant therapy. Front Neurosci 2010; 4:35.
American Psychiatric Association. Diagnostic and statistical manual of mental disorders: DSM-IV-TR 2000a.Washington, DC: American Psychiatric Association.
American Psychiatric Association. Practice guideline for the treatment of patients with major depressive disorder 2000b.Washington, DC: American Psychiatric Association.
American Psychiatric Association. Diagnostic and statistical manual of mental disorders: DSM-5 2013.Washington, DC: American Psychiatric Association.
Bristol-Myers Squibb Company. BuSpar (buspirone HCl). Prescribing information 2010.Princeton, NJ: Bristol-Myers Squibb Company.
Clayton PJ, Grove WM, Coryell W, Keller M, Hirschfeld R, Fawcett J. Follow-up and family study of anxious depression
. Am J Psychiatry 1991; 148:1512–1517.
Fava M, Alpert JE, Carmin CN, Wisniewski SR, Trivedi MH, Biggs MM, et al.. Clinical correlates and symptom patterns of anxious depression
among patients with major depressive disorder in STAR*D. Psychol Med 2004; 34:1299–1308.
Fava M, Rush AJ, Alpert JE, Carmin CN, Balasubramani GK, Wisniewski SR, et al.. What clinical and symptom features and comorbid disorders characterize outpatients with anxious major depressive disorder: a replication and extension. Can J Psychiatry 2006; 51:823–835.
Fava M, Rush AJ, Alpert JE, Balasubramani GK, Wisniewski SR, Carmin CN, et al.. Difference in treatment outcome in outpatients with anxious versus nonanxious depression
: a STAR*D report. Am J Psychiatry 2008; 165:342–351.
Fleurence R, Williamson R, Jing Y, Kim E, Tran QV, Pikalov AS, Thase ME. A systematic review of augmentation strategies for patients with major depressive disorder. Psychopharmacol Bull 2009; 42:57–90.
Flint AJ, Rifat SL. Two-year outcome of elderly patients with anxious depression
. Psychiatry Res 1997; 66:23–31.
Forest Laboratories Inc. Vibryd (vilazodone HCl). Prescribing information 2013.New York, NY: Forest Laboratories Inc.
Hamilton M. The assessment of anxiety states by rating. Br J Med Psychol 1959; 32:50–55.
Hamilton M. A rating scale for depression
. J Neurol Neurosurg Psychiatry 1960; 23:56–62.
Ionescu DF, Niciu MJ, Henter ID, Zarate CA. Defining anxious depression
: a review of the literature. CNS Spectr 2013; 18:1–9.
Kennedy SH. Treating each and every depressed patient. J Psychopharmacol 2008; 22:19–23.
Kenward MG, Roger JH. Small sample inference for fixed effects from restricted maximum likelihood. Biometrics 1997; 53:983–997.
Khan A, Cutler AJ, Kajdasz DK, Gallipoli S, Athanasiou M, Robinson DS, et al.. A randomized, double-blind, placebo-controlled, 8-week study of vilazodone, a serotonergic agent for the treatment of major depressive disorder. J Clin Psychiatry 2011; 72:441–447.
Matza LS, Morlock R, Sexton C, Malley K, Feltner D. Identifying HAM-A cutoffs for mild, moderate, and severe generalized anxiety disorder. Int J Methods Psychiatr Res 2010; 19:223–232.
Montgomery SA, Asberg M. A new depression
scale designed to be sensitive to change. Br J Psychiatry 1979; 134:382–389.
Nelson JC. Augmentation strategies in depression
2000. J Clin Psychiatry 2000; 61Suppl 213–19.
Nelson JC. Anxiety does not predict response to duloxetine in major depression
: results of a pooled analysis of individual patient data from 11 placebo-controlled trials. Depress Anxiety 2010; 27:12–18.
Nelson JC, Mankoski R, Baker RA, Carlson BX, Eudicone JM, Pikalov A, et al.. Effects of aripiprazole adjunctive to standard antidepressant treatment on the core symptoms of depression
: a post-hoc, pooled analysis of two large, placebo-controlled studies. J Affect Disord 2010; 120:133–140.
Nemeroff CB. The burden of severe depression
: a review of diagnostic challenges and treatment alternatives. J Psychiatr Res 2007; 41:189–206.
Nutt DJ. Overview of diagnosis and drug treatments of anxiety disorders. CNS Spectr 2005; 10:49–56.
Ramana R, Paykel ES, Cooper Z, Hayhurst H, Saxty M, Surtees PG. Remission and relapse in major depression
: a two-year prospective follow-up study. Psychol Med 1995; 25:1161–1170.
Rao S, Zisook S. Anxious depression
: clinical features and treatment. Curr Psychiatry Rep 2009; 11:429–436.
Rickels K, Athanasiou M, Robinson DS, Gibertini M, Whalen H, Reed CR. Evidence for efficacy and tolerability of vilazodone in the treatment of major depressive disorder: a randomized, double-blind, placebo-controlled trial. J Clin Psychiatry 2009; 70:326–333.
Robinson DS, Kajdasz DK, Gallipoli S, Whalen H, Wamil A, Reed CR. A 1-year, open-label study assessing the safety and tolerability of vilazodone in patients with major depressive disorder. J Clin Psychopharmacol 2011; 31:643–646.
Russell J, Raskin J, Wiltse C, Walker D, Brawman-Mintzer O. Efficacy and tolerability of duloxetine treatment in elderly patients with major depressive disorder and concurrent anxiety symptoms. Psychiatry (Edgmont) 2007; 4:33–45.
Wiethoff K, Bauer M, Baghai TC, Moller HJ, Fisher R, Hollinde D, et al.. Prevalence and treatment outcome in anxious versus nonanxious depression
: results from the German Algorithm Project. J Clin Psychiatry 2010; 71:1047–1054.