What Clinical and Symptom Features and Comorbid Disorders Characterize Outpatients With Anxious Major Depressive Disorder: A Replication and Extension

Transcription

1 Original Research What Clinical and Symptom Features and Comorbid Disorders Characterize Outpatients With Anxious Major Depressive Disorder: A Replication and Extension Maurizio Fava, MD 1, A John Rush, MD 2, Jonathan E Alpert, MD, PhD 3, Cheryl N Carmin, PhD 4, GK Balasubramani, PhD 5, Stephen R Wisniewski, PhD 6, Madhukar H Trivedi, MD 7, Melanie M Biggs, PhD 8, Kathy Shores-Wilson, PhD 8 Objective: We previously found that 46% of the first 1450 outpatients with depression participating in the multicentre Sequenced Treatment Alternatives to Relieve Depression (STAR*D) project qualified for the designation of anxious depression. This study was designed to replicate and extend our initial findings in a subsequent, larger cohort of outpatient STAR*D participants with nonpsychotic major depressive disorder (MDD). Methods: Baseline clinical and sociodemographic data were collected on 2337 consecutive STAR*D participants. A baseline 17-item Hamilton Depression Rating Scale Anxiety-Somatization factor score of 7 or higher was designated as anxious depression. We identified concurrent Axis I disorders with the Psychiatric Diagnostic Screening Questionnaire (PDSQ), using a 90% specificity threshold. Depressive symptoms were assessed by clinical telephone interview with the 30-item Inventory of Depressive Symptomatology Clinician-Rated (IDS-C30). Results: The prevalence of anxious depression in this population was 45.1%. Patients with anxious MDD were significantly more likely to be in primary care settings and to be women, nonsingle, unemployed, Hispanic, less educated, and suffering from severe depression, both before and after adjustment for overall depression severity. Patients with anxious depression were significantly more likely to meet PDSQ thresholds for generalized anxiety disorder, panic disorder, obsessive compulsive disorder, posttraumatic stress disorder, agoraphobia, hypochondriasis, and somatoform disorder, both before and after adjusting for baseline depression severity. Individuals with anxious depression were also significantly less likely to endorse IDS-C30 items concerning atypical features and were significantly more likely to endorse items concerning melancholic-endogenous depression features, both before and after adjusting for baseline depression severity. Conclusions: This study clearly replicates our previous STAR*D findings and supports the notion that anxious depression may be a valid diagnostic subtype of MDD, with distinct psychiatric comorbidities and clinical and sociodemographic features. (Can J Psychiatry 2006;51: ) Information on funding and support and author affiliations appears at the end of the article. Clinical Implications Anxious depression, defined as depression with high levels of anxiety, is a relatively common subtype of MDD. The presence of high levels of anxiety in depression is associated with distinctive sociodemographic and clinical features. Compared with nonanxious depression, patients with anxious depression are more severely ill and have significantly higher rates of comorbid anxiety disorders. Limitations This was not a random sample of patients. The inclusion and exclusion criteria may have led to a selection bias. Many of the study assessments were based on patients self-reports and not on clinician-rated scales. Can J Psychiatry, Vol 51, No 13, November

2 The Canadian Journal of Psychiatry Original Research Key Words: anxiety, depression, anxious depression, subtype, features Anxiety, nervousness, and their somatic correlates are common symptoms among patients with MDD. For example, a study by Fawcett and Kravitz (1) found that patients with MDD (n = 200) reported high rates of moderateto-severe worrying (72%), psychic anxiety (62%), and somatic anxiety (42%). Although the DSM-IV classification does not specifically address the depressive subtype of anxious depression, there is a fair amount of evidence in the literature that this may indeed be a valid diagnostic subtype. From a nosological perspective, researchers have debated for decades whether anxious depression is best regarded as anxiety disorder comorbidity in the presence of MDD or as major depressive episodes with prominent anxiety. However, the more traditional definition of anxious depression as MDD with high levels of anxiety symptoms is often regarded as having greater ecological validity, and this condition has been shown to be associated with greater illness severity and functional impairment (2), greater chronicity (3), delayed response to treatment (4), and an increased risk of suicidality (5). Individuals with anxious depression (defined as experiencing MDD with high levels of anxiety symptoms) were also found to be less likely to respond to antidepressant treatment than those without anxious depression in some (6,7) but not all (5,8) short-term studies, regardless of the type of antidepressant used. In addition, no significant differences in Abbreviations used in this article CIRS HDRS17 IDS-C30 IPAQ MDD MDE OCD OR PDSQ PTSD QIDS-C16 QIDS-SR16 RC ROA SD STAR*D Cumulative Illness Rating Scale 17-item Hamilton Depression Rating Scale 30-item Inventory of Depressive Symptomatology Clinician-Rated Income and Public Assistance Questionnaire major depressive disorder major depressive episode obsessive compulsive disorder odds ratio Psychiatric Diagnostic Screening Questionnaire posttraumatic stress disorder 16-item Quick Inventory of Depressive Symptomatology Quick Inventory of Depressive Symptomatology Self-Report regional centre Research Outcomes Assessors standard deviation Sequenced Treatment Alternatives to Relieve Depression short-term (6- to 16-week) efficacy trials have typically been shown among antidepressants of the same (9) or different (5) class, except for a pooled analysis showing significantly higher rates of remission with venlafaxine, compared with fluoxetine (7). The association between anxious depression and poorer responses to antidepressant treatment may account for the results of a recent study showing that the concomitant use of anxiolytics hypnotics was a significant predictor of treatment resistance in older adults with depression (10). Finally, residual symptoms of anxiety have been reported to be associated with greater risk of relapse in patients with MDD (11,12). There is also clear evidence for a substantial overlap between anxious depression, defined in terms of degree of anxiety symptom severity, and anxiety disorder comorbidity in MDD; this overlap may be due to high levels of anxiety in MDD being a reflection of specific patterns of psychiatric comorbidity. In 2 separate studies among 197 and 255 (respectively) outpatients with MDD, the lifetime comorbidity for anxiety disorders was 41.6% (13) and 50.6% (14). In the latter study, the overall rate of concurrent comorbid anxiety disorders (excluding PTSD) was only slightly lower (44.7%) than the lifetime comorbidity. The most common concurrent comorbidities were social phobia (26.2%), simple phobia (14.9%), generalized anxiety disorder (10.2%), and panic disorder (8.2%) (14). Similarly, the Vantaa Depression Study (15) found that 57% of 269 patients with a new episode of MDD suffered from one or more concurrent anxiety disorders. We recently carried out a preliminary analysis of the first 1450 primary and specialty care patients with MDD participating in the STAR*D project to examine the clinical correlates and symptom features associated with anxious depression (16). The prevalence of anxious depression, defined as a baseline HDRS17 Anxiety-Somatization factor score of 7 or higher, was 46%. Patients with anxious MDD were significantly more likely to be older, unemployed, less educated; to have severe depression; and to have suicidal ideation both before and after adjustment for depression severity. Patients with anxious depression were significantly more likely to endorse symptoms related to various disorders, including OCD, panic disorder, PTSD, agoraphobia, hypochondriasis, somatoform disorder, and generalized anxiety disorder, both before and after adjustment for depression severity. Individuals with anxious depression were also significantly less likely to endorse IDS-C30 items concerning atypical features and were significantly more likely to endorse items concerning melancholic-endogenous depression features. This study was conducted to replicate and extend these findings in a subsequent, larger cohort of STAR*D participants. 824 Can J Psychiatry, Vol 51, No 13, November 2006

3 What Clinical and Symptom Features and Comorbid Disorders Characterize Outpatients With Anxious Major Depressive Disorder Methods Study Design STAR*D was designed to prospectively define which of several treatments are most effective for participants with nonpsychotic MDD with an unsatisfactory clinical outcome to an initial treatment and, if necessary, subsequent treatments (see references 17 and 18 for a detailed description of the STAR*D protocol). Study Organization The STAR*D infrastructure included the National Coordinating Center in Dallas, the Data Coordinating Center in Pittsburgh, and 14 RCs across the United States. Each RC oversaw the protocol implementation at 2 to 4 primary or specialty care sites in either the public or private sector. Nearly one-half of the sites (18 of 41) were primary care settings. Research outcome data were collected by telephone interviews with a small team of trained ROAs who were masked to treatment and by a telephone-based interactive voice response system. The ROAs received extensive training in the administration of the study s primary efficacy measures through live and videotaped interviews. Interrater reliability was periodically assessed throughout the study. Study Population Recruitment. STAR*D recruited 4041 participants into the study. This report presents data from the second 2541 consecutively recruited participants enrolled into STAR*D. The study protocol was reviewed and approved by the ethical review board at each site, in accordance with the principles of the Declaration of Helsinki. All risks, benefits, and adverse events associated with each treatment within the randomized treatments were explained to study participants, who were required to provide written informed consent prior to study participation. Advertising to recruit participants was not allowed, since it might have attracted a less representative spectrum of participants (19). Overall Inclusion and Exclusion Criteria. Every effort was made to recruit a broad spectrum of participants that represented all ethnic groups and both sexes. Men and women (aged 18 to 75 years) with nonpsychotic MDD and a baseline HDRS17 score of 14 or higher could enter the study, providing that the treating clinician had determined that outpatient treatment with an antidepressant was both safe and called for. Participants with schizophrenia, schizoaffective disorder, bipolar disorder, OCD, anorexia nervosa, or a primary diagnosis of bulimia nervosa were excluded. Participants with current substance abuse issues or dependence were eligible, as long as inpatient care for detoxification was not required clinically at study entry, although participation in a substance abuse program was encouraged by their clinician. Also excluded were participants with a well-documented history of nonresponse or clear intolerability, in the current major depressive episode, to one or more treatments required in the protocol, and delivered at an adequate dosage (17); those with severe, unstable concurrent psychiatric conditions likely to require hospitalization within 6 months from study entry (for example, participants with severe alcohol dependence who had a history of recent admissions aimed at detoxification); and those with certain concurrent psychiatric or medical conditions considered to be relative or absolute contraindications to the use of more than one treatment option within the protocol, which would nullify the possibility of randomization to any of the strategies or substrategies of STAR*D (17). Participants taking any concomitant nonpsychotropic medications (or anxiolytics and sedative hypnotics) could enter the study, as long as their clinician determined that antidepressant treatments in the protocol were appropriate and safe. Participants who were pregnant or trying to become pregnant, and those already receiving a targeted psychotherapy aimed at their depression, were also excluded. Research Outcome Assessments After participants had provided written informed consent, clinical and demographic information was collected, including prior course of depressive illness; current and past substance abuse; prior suicide attempts; family history of MDD, bipolar disorder, substance abuse, and suicide; current general medical illnesses; and prior history of treatment (both medication and psychotherapy) in the current major depressive episode. Participants completed a modified paper-andpencil version of the PDSQ (20) at the clinical site. The types and degrees of concurrent psychiatric symptoms and comorbidity were measured with the PDSQ by recording the number of items endorsed by study participants for each diagnostic category. We established rates of current psychiatric comorbidity with a threshold of 90% specificity for each diagnostic category (21). Clinical research coordinators at each site administered the HDRS17 (at baseline to establish inclusion exclusion criteria) and the 16-item QIDS-C16 and also reviewed inclusion exclusion criteria. The QIDS-C16 is a clinician-rated scale assessing the 9 diagnostic symptoms domains of MDD (22,23). Current general medical conditions were assessed with the 14-item CIRS (24,25), which was completed with the use of a manual (26) to guide scoring and to gauge the morbidity of general medical conditions relevant to different organ systems. ROAs completed a telephone interview within 72 hours of the baseline visit to administer the HDRS17 (27,28), the IDS-C30 (29,30), and the 5-item IPAQ. The IPAQ measures the participant s monthly income, (which in this study is Can J Psychiatry, Vol 51, No 13, November

4 The Canadian Journal of Psychiatry Original Research defined as the net monthly income after taxes and benefits are deducted) and the source of monthly income (for example, employment wages or public assistance). For the purpose of our analyses, the ROAs QIDS-C16 scores were extracted from the ROA IDS-C30 interviews. Statistical Analyses As in previous studies from our group (9) and the preliminary report from the STAR*D study (16), anxious depression was defined as MDD with high levels of anxiety symptoms (HDRS Anxiety-Somatization factor score 7). The Anxiety- Somatization factor, derived from a factor analysis of the HDRS17 scale conducted by Cleary and Guy (31), includes 6 items from the original 17-item version: Anxiety (psychic), Anxiety (somatic), Somatic Symptoms (gastrointestinal), Somatic Symptoms (general), Hypochondriasis, and Insight. The HDRS17 obtained at baseline by the ROAs was used to define anxious depression. We used descriptive statistics (for example, mean, median, and percentages) to describe the study population. Bivariate logistic regression models were used to assess the association between the independent variables of interest (for example, ethnicity, sex, and length of illness) and the presence of anxious depression. Logistic regression models were also used to assess the association between the independent variables of interest and the presence of anxious depression, independent of the effect of the depression severity (as measured by the ROA-administered HDRS17 score, not including the items used to identify anxious depression). Statistical significance was defined as a 2-tailed P value of less than As with our preliminary report (16), no adjustments to P values were made for multiple comparisons. Results Table 1 summarizes the clinical and sociodemographic characteristics of our sample, showing that a majority of the 2541 subjects were women (62.5%), that most (58.2%) were recruited in specialty care clinics, and that the severity of their depression was moderate to marked. Additionally, Table 1 provides the rates of those meeting criteria for comorbid psychiatric disorders (according to a 90% specificity criterion for each item within the PDSQ). The distribution of the Anxiety-Somatization factor scores of the HDRS17 scale appears to be normal, with a range of 0 to 15 and a median score of 6. Of the 2541 patients, 2337 had a baseline ROA assessment available with an adequate number of items of the HDRS17 scale and could therefore be classified as having anxious depression (n = 1053; 45.1%) or nonanxious depression (n = 1284; 54.9%). Tables 2 and 3 provide the results of the nonadjusted and adjusted (for baseline severity of depression, as measured by the HDRS17 score, excluding items used to identify anxious depression) associations among the presence or absence of anxious depression, sociodemographic and clinical variables, and rates of items endorsed by patients on the IDS-C30 (ROA). Anxious depression was significantly more common among women (47.7%), compared with men (40.7%); among Hispanics (57.3%), compared with non-hispanics (43%); among those in primary care (49.3%), compared with those in specialty care (42.2%); among those unemployed (54.8%), compared with employed (38.3%); among those with lesser schooling (mean number of years in school 12.7 compared with 14); and among married, divorced, or widowed individuals (47.3% to 49.3%), compared with those who never married (37.1%), both before and after adjusting for baseline depression severity. Patients were significantly more likely to have anxious depression, after adjusting for the baseline HDRS17 score, if they had more severe depression (as measured by the IDS-C30 [ROA] and QIDS-SR16) or experienced a greater duration of their current episode. In addition, MDD with anxious depression was associated with a greater likelihood of reporting suicidal ideation on the IDS-C30 (55% compared with 41.6%), but not on the HDRS17, and of having a personal history of attempted suicide (58.6% compared with 42.4%), both before and after adjusting for baseline depression severity. Finally, MDD patients with anxious depression were significantly more likely than those with nonanxious depression to report a greater degree of medical comorbidity, as evidenced by their CIRS scores both before and after adjusting for depression severity. Both before and after adjusting for depression severity, there was no significant difference between patients with anxious, and those with nonanxious, depression in age of onset, family history of depression or mood disorder, and number of episodes. There was a difference in age at the time of study entry between anxious and nonanxious depression. This difference was not significant before adjusting for depression severity, but it did become significant after adjusting. After adjusting for depression severity, there was no significant difference between anxious and nonanxious depression in ethnicity or length of illness, although before adjustments, this difference was statistically significant. Patients with anxious depression were significantly more likely than those without anxious depression, both before and after adjusting for baseline depression severity, to report low energy or fatigability (98% compared with 81.5%), leaden paralysis or lack of physical energy (58.8% compared with 32%), psychomotor slowing (71.9% compared with 53.6%), impaired concentration or decision making (95.7% compared with 85.9%), and negative view of self (85.1% compared with 75.8%). Compared with those without anxious depression, patients with anxious depression were also significantly more 826 Can J Psychiatry, Vol 51, No 13, November 2006

5 What Clinical and Symptom Features and Comorbid Disorders Characterize Outpatients With Anxious Major Depressive Disorder Table 1 Baseline characteristics of the STAR*D confirmatory analysis sample (n = 2541) Baseline Characteristics Mean (SD) Median (observed range) Age, years 40.5 (13.3) 40 (18 75) Education, years of schooling 13.3 (3.2) 13 (0 26) Income, in dollars 2403 (3311) 1500 ( ) CIRS Categories endorsed 2.9 (2.3) 2 (0 11) Total score 4.2 (3.8) 3 (0 30) Severity index 1.2 (0.7) 1.1 (0 4) Clinical Course Age at onset of first MDE 25.6 (14.6) 21 (2 74) Number of MDEs 6.0 (12.5) 3 (1 99) Length of current MDE, in months 24.0 (50.2) 8 (0 586) Length of illness, in years 14.9 (13.1) 11 (0.5 63) HDRS17 (ROA) 19.6 (6.5) 20 (1 38) IDS-C30 (ROA) 35.4 (11.4) 35 (3 70) QIDS-SR (4.4) 16 (2 27) Setting n % Primary care Specialty care Ethnicity White Black or African American Hispanic Other Sex-female Marital status Never married Married Divorced Widowed Employment status Employed Unemployed Table 1 continued Psychiatric comorbidities (90% specificity) n % Anxiety disorder OCD Absent Present Absent Present Panic disorder Absent Present Social phobia PTSD Absent Present Absent Present Agoraphobia Absent Present Alcohol abuse Absent Present Drug abuse Absent Present Somatoform disorder Absent Present Hypochondriasis Bulimia Absent Present Absent Present Retired Family history of depression Can J Psychiatry, Vol 51, No 13, November

6 The Canadian Journal of Psychiatry Original Research Table 2 Baseline characteristics associated with anxious depression (absence and [or] presence) Nonanxious depression n = 1284 (54.9%) Anxious depression n = 1053 (45.1%) Unadjusted OR P Adjusted OR a P a Characteristics n n % n % Setting Primary Specialty Ethnicity White Black or African American Hispanic < < No Yes Other Sex Male Female Marital status < Never married Married Divorced Widowed Employment status < < Employed Unemployed Retired Family history of depression No Yes Family history of mood disorder No Yes History of attempted suicide No < Yes Age at onset 18 years years Can J Psychiatry, Vol 51, No 13, November 2006

7 What Clinical and Symptom Features and Comorbid Disorders Characterize Outpatients With Anxious Major Depressive Disorder Table 2 continued Nonanxious depression n = 1284 (54.9%) Anxious depression n = 1053 (45.1%) Unadjusted OR P Adjusted OR a P a Units n n Mean SD n Mean SD Age (units = 5) Education (years of schooling) (units = 3) Clinical course CIRS Age at onset of 1st MDE (units = 5) Number of episodes (units = 5) Length of episodes (in months) (units = 5) Length of illness (in years) (units = 5) Categories endorsed < < < < Total Score < < Severity Index < < Suicide-HDRS-17 (ROA) < IDS-C30 (ROA) < < QIDS-SR16 (units = 5) < a Adjusted for the ROA-administered HDRS17 score, not including the items used to identify anxious depression likely, both before and after adjusting for baseline depression severity, to endorse the IDS-C30 items concerning irritable mood (89.5% compared with 75%), anxious mood (93.8% compared with 75.2%), panic or phobic symptoms (55.2% compared with 24.5%), sympathetic nervous system arousal (85.6% compared with 55.1%), sleep-onset insomnia (79.9% compared with 56.4%), somatic pain (89.3% compared with 67.4%), and gastrointestinal complaints (51.8% compared with 31%). Finally, compared with those without anxious depression, patients with anxious depression were significantly more likely, both before and after adjusting for baseline depression severity, to report a distinct quality of mood (78.2% compared with 72.5%), decreased appetite (63.2% compared with 27.6%), and weight loss (43% compared with 18.1%). Only the items concerning increased appetite (23.7% compared with 18%) and increased weight (25.4% compared with 20.9%) were significantly more common in nonanxious depression than in anxious depression, both before and after adjusting for depression severity. After adjusting for severity of depression, there was no significant difference between anxious and nonanxious depression in the items concerning mid-nocturnal and early morning insomnia, hypersomnia, mood reactivity and sad mood, outlook on the future, involvement, pleasure and enjoyment, sexual interest, psychomotor agitation, and interpersonal sensitivity. Both before and after adjusting for depression severity, there was no significant difference between anxious and nonanxious depression in the item concerning mood variation. As for concurrent psychiatric comorbidity, Table 4 shows that individuals with anxious MDD were more likely to meet criteria (according to a 90% specificity criterion for each item in the PDSQ) for generalized anxiety disorder, panic disorder, OCD, PTSD, agoraphobia, hypochondriasis, and somatoform disorder, both before and after adjusting for baseline depression severity. Both before and after adjusting for depression severity, there was no significant difference between anxious and nonanxious depression in rates of bulimia nervosa or drug and alcohol abuse. After adjusting for depression severity, Can J Psychiatry, Vol 51, No 13, November

8 The Canadian Journal of Psychiatry Original Research Table 3 Presenting symptoms in anxious and (or) nonanxious based on the IDS-C30 (ROA) Nonanxious depression n = 1284 (54.9%) Anxious depression n = 1053 (45.1%) Absent Present Absent Present IDS-C30 (ROA) items n % n % n % n % P P a Insomnia Sleep onset < Midnocturnal < Early morning < Hypersomnia < Mood Sad < Irritable < < Anxious < < Reactivity of mood < Mood variation Quality of mood Appetite Weight Decreased < < Increased < Decreased < < Increased Concentration and (or) decision making Outlook < Self < Future < Suicidal Ideation < < Involvement < Energy and (or) fatigability Pleasure and (or) enjoyment < < < Sexual interest < Psychomotor slowing < Psychomotor agitation < Somatic (pain) complaints < < Sympathetic arousal < < Panic and (or) phobic symptoms < < Gastrointestinal < < Interpersonal sensitivity < Leaden paralysis and (or) physical energy < < a Adjusted for the ROA-administered HDRS17 score, not including the items used to identify anxious depression 830 Can J Psychiatry, Vol 51, No 13, November 2006

9 What Clinical and Symptom Features and Comorbid Disorders Characterize Outpatients With Anxious Major Depressive Disorder Table 4 Association of anxious and (or) nonanxious with PDSQ psychiatric comorbidity (using 90% specificity) PDSQ Nonanxious depression n = 1284 (54.9%) Anxious depression n = 1053 (45.1%) Unadjusted OR P Adjusted OR a P a GAD 2.8 < < Absent Present OCD 2.5 < < Absent Present Panic disorder 4.9 < < Absent Present Social Phobia 1.7 < Absent Present PTSD 2.6 < Absent Present Agoraphobia 3.6 < < Absent Present Alcohol Abuse Absent Present Drug Abuse Absent Present Somatoform disorder 4.5 < Absent Present Hypochondriasis 3.4 < < Absent Present Bulimia Absent Present a Adjusted for the ROA-administered HDRS17 score, not including the items used to identify anxious depression GAD = generalized anxiety disorder Can J Psychiatry, Vol 51, No 13, November

10 The Canadian Journal of Psychiatry Original Research there was no significant difference between anxious and nonanxious depression in rates of comorbid social phobia. Discussion This is the largest sample used to examine clinical and demographic correlates and presenting symptom features of outpatients diagnosed with MDD. According to our a priori definition of anxious depression (9,16), nearly one-half of the sample (45.1%) qualified for this designation. This finding is consistent with the 46% rate observed in our preliminary report based on the first 1450 STAR*D subjects (16), as well as with the results of previous studies among outpatients with MDD, where the lifetime comorbidity for anxiety disorders was 41.6% (13) and 50.6% (14). Our study confirmed our preliminary observation that anxious depression had distinct clinical correlates before and after adjustment for depression severity. As in the preliminary report with the first 1450 STAR*D patients (16), major depressive episodes in those with high anxiety ratings on the HDRS Anxiety-Somatization factor were linked with a greater representation of more vulnerable and disadvantaged sociodemographic profiles (for example, those with unemployment and more limited education) and with older age, before and after controlling for depression severity at baseline. This may suggest that anxious depression impedes functioning at work and school, resulting in a more disadvantaged sociodemographic status, or that those with such a disadvantage are slower to pursue treatment (that is, illness severity must be higher for them to seek help). While our preliminary study had shown a slightly nonsignificantly higher rate of anxious depression in Hispanics than in non-hispanics (48% compared with 46%), the difference became statistically significant in our replication study sample (57.3% compared with 43%). Finally, our preliminary study had shown a significantly higher rate of anxious depression in primary care patients, compared with specialty care patients (49% compared with 44%), and in our replication study sample, this difference became even larger and of greater statistical significance (49.3% compared with 42.2%). This finding may indicate that primary care is the point of entry into psychiatric treatment for more disadvantaged populations with depression and may account for the fact that patients with anxious depression report a significantly greater degree of medical comorbidity than do those with nonanxious depression. As in our preliminary report (16), anxious depression was also associated with a significantly higher frequency of items concerning endogenous-melancholic features, such as distinct quality of mood, psychomotor retardation, decreased appetite and (or) weigh loss, as well as with a significantly lower frequency of items concerning atypical features, such as increased appetite and weight increase. This suggests that anxious depression may share some features with endogenous, melancholic subtypes of depression and is consistent with the view that anxious-retarded depression may represent a form of melancholia (32). Patients with anxious depression were also more likely to report greater illness severity, as measured by the IDS-C30 (ROA) and QIDS-SR16. This is consistent with the findings of our preliminary report (16) and of those from previous studies, which have shown that MDD with high levels of anxiety is associated with greater illness severity and functional impairment (2). The increased illness severity may be a function of either the greater psychiatric comorbidity (16), or of the more disadvantaged sociodemographic profile, or a combination of both. Our preliminary study had shown only a nonsignificantly higher rate of anxious depression in women, compared with men (48% and 43%, respectively) (16); however, in our replication study sample, the difference between sexes became statistically significant (47.7% compared with 40.7%). This is certainly consistent with the findings of a previous multicentre study (9,33), which had shown a higher rate for women of anxious depression (65%) than of nonanxious depression (56%). Our study did observe a significantly longer duration of the current episode among patients with anxious depression, and this is consistent with previous reports of an association between anxious depression and chronicity of depressive illness (3). Further, the significant association between anxious depression and the presence of both current suicidal ideation on the IDS-C30 (although not on the HDRS17) and a history of attempted suicide is consistent with the findings of our preliminary STAR*D report (16) and with those of a previous clinical trial linking anxious depression to an increased risk of suicidality (5). As in our preliminary STAR*D study (16), those subjects with anxious depression were also significantly more likely to meet criteria for, or endorse items related to, anxiety disorders on the PDSQ (generalized anxiety disorder, OCD, panic disorder, agoraphobia, and PTSD). Further, these patients were also more likely to report anxiety-related symptoms on the IDS-C30, such as anxious mood, panic and (or) phobic symptoms, sympathetic arousal, gastrointestinal symptoms, and somatic pain. These findings suggest that there is a substantial degree of overlap between definitions of anxious depression, specifically, between the dimensional (according to the degree of anxiety symptom severity) and the syndromal (according to the degree of anxiety disorder comorbidity), as these relations remained significant even after adjusting for depression severity. These findings are consistent with those of Malhi and others (34), who had observed that state anxiety 832 Can J Psychiatry, Vol 51, No 13, November 2006

11 What Clinical and Symptom Features and Comorbid Disorders Characterize Outpatients With Anxious Major Depressive Disorder was strongly linked with lifetime anxiety disorder prevalence in depression. The findings from both this and our preliminary STAR*D studies have several implications for the recognition and diagnosis of anxious depression in primary and specialty care populations. The presence of certain sociodemographic characteristics should alert us to the possibility that anxiety may be a prominent feature of MDD. In particular, even greater attention should be paid to the assessment and monitoring of suicidal ideation in such individuals. In addition, as recommended by Robins and Guze (35), a psychiatric diagnostic entity may be considered valid if it can be shown to have differentiating features, evidence of familiality, specific treatment responsivity, and a unique course. In this study, anxious depression appears to be associated with a characteristic clinical profile that is independent of the depression severity. The study also suggests a unique course of illness, as evidenced by the significantly longer duration of the current episode. In terms of familiality, we did not obtain any family history of anxious depression, so we do not know whether this subtype of depression runs in families. Since this report focuses on a cross-sectional assessment of patients with anxious depression, we cannot specifically address the issue of specific treatment responsivity, although there is some evidence from a large pooled analysis that serotonin norepinephrine reuptake inhibitors such as venlafaxine may be superior to selective serotonin reuptake inhibitors in anxious depression (36). Therefore, although further data are needed to provide information on the familiality of this condition and on its treatment responsivity, the findings of previous studies, as well as our own, are certainly consistent with the view that anxious depression is a valid diagnostic subtype of MDD. There are several limitations to this study that should be considered in interpreting the results and recommendations. Our definition of anxious depression is based on the severity of anxiety symptoms, as measured by the HDRS Anxiety-Somatization factor. Although the HDRS does include anxiety items and its Anxiety-Somatization factor has been used in several previous studies, only a limited number of anxiety symptoms are captured by the HDRS, and therefore, the possibility of a significant risk for misclassification cannot be ruled out. Conversely, the significant relation between our study s definition of anxious depression and the degree of anxiety disorder comorbidity seems to suggest that such a risk may be relatively low. In addition, this study did not use a clinician-administered structured diagnostic interview, which might have provided a more accurate diagnostic picture. Since only outpatients with MDD were enrolled into the study, it is possible that the clinical correlates and symptom patterns associated with anxious depression may be different for inpatients treated for depression. Similarly, factors related to treatment seeking may have affected these results. Since our STAR*D subjects were drawn from patients seen in primary and specialty care settings, our results can therefore be taken as a reasonable and generalizable estimate of both the prevalence of this depressive subtype in clinical samples and its associated characteristics, but they do not necessarily represent the underlying epidemiology of this condition. Despite the fact that the presence of primary OCD and bulimia nervosa was one of the exclusion criteria, a substantial proportion of our patients (14.2% and 12.3%, respectively) met PDSQ thresholds for OCD and bulimia nervosa, suggesting that these 2 comorbid psychiatric disorders were often secondary to MDD. Other limitations include the fact that this was not a random sample of patients and that many of the study assessments, including the PDSQ assessment of psychiatric comorbidity, were based on patients self-reports. Finally, participants were enrolled from several sites in a nonrandom manner. In summary, this study clearly replicates our previous STAR*D findings and supports the association in MDD between high levels of anxiety (anxious depression) and distinct clinical and sociodemographic features. Funding and Support This project was funded with federal funds from the National Institute of Mental Health, National Institutes of Health, under Contract N01MH 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 US Government. We also thank all the STAR*D investigators for all their help in making this large and complex multicentre study possible, and for generating the data for this report. References 1. Fawcett J, Kravitz HM. Anxiety syndromes and their relationship to depressive illness. J Clin Psychiatry 1983;44: Joffe RT, Bagby RM, Levitt A. Anxious and nonanxious depression. Am J Psychiatry 1993;150: VanValkenburg C, Akiskal HS, Puzantian V, Rosenthal T. Anxious depressions. Clinical, family history, and naturalistic outcome comparisons with panic and major depressive disorders. J Affect Disorders 1984;6: 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: Tollefson GD, Holman SL, Sayler ME, Potvin JH. Fluoxetine, placebo, and tricyclic antidepressants in major depression with and without anxious features. J Clin Psychiatry 1994;55: Flint AJ, Rifat SL. Anxious depression in elderly patients. Response to antidepressant treatment. Am J Geriat Psychiat 1997;5: Davidson JR, Meoni P, Haudiquet V, Cantillon M, Hackett D. Achieving remission with venlafaxine and fluoxetine in major depression: its relationship to anxiety symptoms. Depress Anxiety 2002;16: Russell JM, Koran LM, Rush AJ, Hirschfeld RM, Harrison W, Friedman ES, and others. Effect of concurrent anxiety on response to sertraline and imipramine in patients with chronic depression. Depress Anxiety 2001;13: Fava M, Rosenbaum JF, Hoog SL, Tepner RG, Kopp JB, Nilsson ME. Fluoxetine versus sertraline and paroxetine in major depression: tolerability and efficacy in anxious depression. J Affect Disorders 2000;59: Can J Psychiatry, Vol 51, No 13, November

12 The Canadian Journal of Psychiatry Original Research 10. Bosworth HB, Hays JC, George LK, Steffens, DC. Psychosocial and clinical predictors of unipolar depression outcome in older adults. Int J Geriatr Psych 2002;17: 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: Flint AJ, Rifat SL. Two-year outcome of elderly patients with anxious depression. Psychiat Res 1997;66: Sanderson, WC, Beck AT, Beck J. Syndrome comorbidity in patients with major depression or dysthymia: prevalence and temporal relationships. Am J Psychiatry 1990;147: Fava M, Rankin MA, Wright EC, Alpert JE, Nierenberg AA, Pava J, and others. Anxiety disorders in major depression. Compr Psychiatry 2000;41: Melartin TK, Rytsala HJ, Leskela US, Lestela-Mielonen PS, Sokero TP, Isometsa ET. Current comorbidity of psychiatric disorders among DSM-IV major depressive disorder patients in psychiatric care. J Clin Psychiatry 2002;63: Fava M, Alpert JE, Carmin CN, Wisniewski SR, Trivedi MH, Biggs MM, and others. Clinical correlates and symptom patterns of anxious depression among patients with major depressive disorder in STAR*D. Psychol Med 2004;34: Fava M, Rush AJ, Trivedi MH, Nierenberg AA, Thase ME, Sackeim HA, and others. Background and rationale for the sequenced treatment alternatives to relieve depression (STAR*D) study. Psychiatr Clin North Am 2003;26: Rush AJ, Fava M, Wisniewski SR, Lavori PW, Trivedi MH, Sackeim HA, and others. Sequenced Treatment Alternatives to Relieve Depression (STAR*D): Rationale and Design. Control Clin Trials 2004;25: Bielski RJ, Lydiard RB. Therapeutic trial participants: Where do we find them and what does it cost? Psychopharmacol Bull 1997;33: Zimmerman M, Mattia JI. The reliability and validity of a screening questionnaire for 13 DSM-IV Axis I disorders (the Psychiatric Diagnostic Screening Questionnaire) in psychiatric outpatients. J Clin Psychiatry 1999;60: Rush AJ, Zimmerman M, Wisniewski SR, Fava M, Hollon SD, Warden D, and others. Comorbid psychiatric disorders in depressed outpatients: a preliminary report. J Affect Disorders 2005;87(1): Rush AJ, Trivedi MH, Ibrahim HM, Carmody TJ, Arnow B, Klein DN, and others. 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 2003;54: Trivedi MH, Rush AJ, Ibrahim HM, Carmody TJ, Biggs MM, Suppes T, and others. The Inventory of Depressive Symptomatology, Clinician Rating (IDS-C) and Self-Report (IDS-SR), and the Quick Inventory of Depressive Symptomatology, Clinician Rating (QIDS-C) and Self-Report (QIDS-SR) in public sector patients with mood disorders: a psychometric evaluation. Psychol Med 2004;34(1): Linn, BS, Linn MW, Gurel L. Cumulative Illness Rating Scale. J Am Geriatr Soc 1968;16: Miller MD, Paradis CF, Houck PR, Mazumdar S, Stack JA, Rifai AH, and others. Rating chronic medical illness burden in geropsychiatric practice and research: application of the Cumulative Illness Rating Scale. Psychiatry Res 1992;41: Miller MD, Towers A. A manual of guidelines for scoring the cumulative rating scale for geriatrics (CIRS-G). Pittsburgh (PA): Western Psychiatric Institute and Clinic. University of Pittsburgh School of Medicine; Hamilton, M. A rating scale for depression. J Neurol Neurosur Psychiatry 1960;23: Hamilton, M. Development of a rating scale for primary depressive illness. The British Journal of Social and Clinical Psychology 1967;6: Rush AJ, Giles DE, Schlesser MA, Fulton CL, Weissenburger JE, Burns CT. The Inventory for Depressive Symptomatology (IDS): Preliminary findings. Psychiatry Res 1986;18: Rush AJ, Gullion CM, Basco MR, Jarrett RB, Trivedi MH. The Inventory of Depressive Symptomatology (IDS): Psychometric properties. Psychol Med 1996;26: Cleary P, Guy W. Factor analysis of the Hamilton Depression Scale. Drug Exp Clin Res 1977;1: de Winter RF, Zwinderman KH, Goekoop JG. Anxious-retarded depression: relation to family history of depression. Psychiatry Res 2004;127(1 2): Fava M, Hoog SL, Judge RA, Kopp JB, Nilsson ME, Gonzales JS. Acute efficacy of fluoxetine versus sertraline and paroxetine in major depressive disorder including effects of baseline insomnia. J Clinical Psychopharm 2002;22: Malhi GS, Parker GB, Gladstone G, Wilhelm K, Mitchell PB. Recognizing the anxious face of depression. J Nerv Ment Dis 2002;190: Robins E, Guze SB. Establishment of diagnostic validity in psychiatric illness: its application to schizophrenia. Am J Psychiatry 1970;126: Fava M, Entsuah R, Tummala R. Venlafaxine Versus SSRIs and Placebo in the Treatment of Anxious Depression. Paper presented at the APA Annual Meeting; 2005; Atlanta (GA). Manuscript received January 2006, revised, and accepted July Associate Chief of Psychiatry for Clinical Research, Massachusetts General Hospital, Boston, Massachusetts; Director, Depression Clinical and Research Program, Massachusetts General Hospital, Boston, Massachusetts; Professor of Psychiatry, Harvard Medical School, Boston, Massachussets. 2 Vice Chair, Department of Clinical Sciences, The University of Texas Southwestern Medical Center, Dallas, Texas; Professor, Department of Psychiatry, The University of Texas Southwestern Medical Center, Dallas, Texas. 3 Associate Director, Depression Clinical and Research Program, Massachusetts General Hospital, Boston, Massachusetts. 4 Associate Professor, Stress & Anxiety Disorders Clinic,University of Illinois at Chicago, Chicago, Illinois; Director, Stress & Anxiety Disorders Clinic,University of Illinois at Chicago, Chicago, Illinois; Director, Cognitive Behavior Therapy Program, University of Illinois at Chicago, Chicago, Illinois. 5 Research Associate, Epidemiology Data Center, Graduate School of Public Health, University of Pittsburgh, Pittsburgh, Pennsylvania. 6 Associate Professor, Epidemiology Data Center, University of Pittsburgh, Pittsburgh, Pennsylvania; Deputy Director, Epidemiology Data Center, University of Pittsburgh, Pittsburgh, Pennsylvania; Associate Dean for Research, Graduate School of Public Health, University of Pittsburgh, Pittsburgh, Pennsylvania. 7 Professor of Psychiatry, Mood Disorders Research Program and Clinic, The University of Texas Southwestern Medical Center, Department of Psychiatry, Dallas, Texas; Director, Mood Disorders Research Program and Clinic, The University of Texas Southwestern Medical Center, Department of Psychiatry, Dallas, Texas. 8 Associate Professor, The University of Texas Southwestern Medical Center, Department of Psychiatry and Department of Family and Community Medicine, Dallas, Texas. Address for correspondence: Dr Maurizio Fava, Depression Clinical and Research Program, Massachusetts General Hospital, 55 Fruit Street Bulfinch 351, Boston MA 02114; mfava@partners.org 834 Can J Psychiatry, Vol 51, No 13, November 2006