• E-mail
• Print
• Comment
• Font Size
• Digg
• del.icio.us
• Discuss article

Impact of Symptoms of Generalized Anxiety Disorder on the Course of Late-Life Depression

Posted on: Thursday, 20 January 2005, 03:00 CST

Objective: The authors sought to examine the effect of symptoms of generalized anxiety disorder (GAD) on acute course of depression in a group of older depressed patients. Methods: Authors assessed GAD symptoms in a sample of 204 patients age 60 and older with unipolar major depression. Patients were treated naturalistically by study geriatric psychiatrists using a treatment guideline for geriatric depression and assessed with the Montgomery-Asberg Depression Rating Scale. They were dichotomized as meeting or not meeting GAD symptom criteria. The groups were analyzed by Cox proportional-hazards models. Results: There were 138 remitters and 66 nonremitters. After analysis controlled for benzodiazepine use, stressful life events, social support, and functional status, having GAD symptoms was associated with longer time-to-remission. Conclusions: Older depressed patients with GAD have a worse outcome than those without anxiety. Future studies will need to determine the appropriate role of benzodiazepines and other anxiolytics in the treatment of older depressed patients with symptoms of GAD. (Am J Geriatr Psychiatry 2005; 13:40-47)

Anxiety is frequently encountered in older depressed patients,1 with generalized anxiety disorder (GAD) being the most prevalent comorbid anxiety disorder.2 Several studies have shown that either psychological or somatic anxiety symptoms are common in elderly patients with major depression.3-5 The prevalence of symptoms that meet criteria for specific anxiety disorders in older depressed patients varies across studies, with some studies showing a low prevalence of diagnosable comorbid anxiety disorders,3,5 but one study showing as many as one-quarter of depressed elderly persons meeting current criteria for a comorbid anxiety disorder.5 An even higher prevalence of comorbid anxiety disorder (48%) had also been reported in a large epidemiological study.6 Higher depression ratings are associated with more severe ratings of GAD.7

Despite numerous studies documenting co-occurrence of depression and anxiety in elderly patients, data are scarce on the impact of comorbid anxiety on the course of late-life depression.8 Mulsant et al.4 found that, among 336 older depressed inpatieiits and outpatients, lifetime comorbid anxiety disorders did not affect the rate of response to depression, but they were associated with a higher use of benzodiazepines and a greater response time. Flint and Rifat9 reported that older anxious-depressed patients were less responsive to nortriptyline and more likely to discontinue treatment. Others have found that higher baseline anxiety symptoms predicted poorer response to nortriptyline and interpersonal therapy among older depressed patients.10 One study did report that baseline anxiety symptoms did not predict relapse; however, subjects experienced a shorter time-to-relapse if they had residual anxiety symptoms after remission of depression.11 More recently, Lenze et al.l2 found that presence of anxiety symptoms did not influence outcome. Thus, with one exception, results of studies over the past decade suggest that concurrent symptoms of anxiety may have prognostic importance in late-life depression.

In the present study, we sought to examine the role of anxiety on outcome of late-life depression. We hypothesized that individuals with late-life depression and comorbid anxiety would be less likely to experience remission of symptoms and require more time to achieve remission, compared with non-anxious depressed patients.

METHODS

Design and Sample

This study used a prospective cohort design. Subjects consisted of 204 depressed individuals enrolled in the NIMH-sponsored Mental Health Clinical Research Center (MHCRC) at Duke University Medical Center between December 1994 and September 2001. All subjects met DSM-IV criteria for major depression and were at least 60 years of age at baseline enrollment. Exclusion criteria for the MHCRC study include the following: another major psychiatric illness, including bipolar disorder, schizophrenia, schizoaffective disorder, dementia (see below), and alcohol or drug abuse or dependence; other clinically diagnosed primary neurological illness; and medications, medical illness, or physical disability that profoundly affect cognitive functioning and would interfere with ability to consent. Patients with psychotic depression or with comorbid anxiety disorders were not excluded. After giving a complete description of the study to the subjects, we obtained written informed consent.

Assessment Procedures

At baseline, subjects received standardized clinical assessments, including the Hamilton Rating Scale for Depression (Ham-D),13 the Montgomery-Asberg Depression Rating Scale (MADRS),14 the Clinical Global Impression scale (CGI), the Cumulative Illness Rating Scale (CIRS),15 a measure of medical illness burden in geriatric populations, and the Mini-Mental State Exam (MMSE).16 A trained interviewer administered the Duke Depression Evaluation Schedule (DDES),17 which assesses major depression and GAD by means of the NIMH Diagnostic Interview Schedule,18 as well as number of depression episodes, age at onset, cognitive status, physical health, social support, stressful life events, and activities of daily living (ADLs). The interrater reliability for most DIS items, including depression and anxiety measures, was 1.0, reflecting the extremely structured format of the DDES. Reliability was based on a blind reassessment of videotaped interviews.19 Social support measures included the Social Interaction Scale, the Social Network Scale, the Instrumental Social Support Scale, and the Subjective Social Support Scale from the Duke Social Support Index.17'20 The DDES includes 16 self-report items, assessing two domains of self- reported physical function. To determine the frequency and severity of stressful life events, we used a 19-item life-events checklist, previously used in the Epidemiologie Catchment Area surveys.20 Seven items, derived with some modification from Katz et al.,21 Branch et al.,22 and Nagi,23 address basic ADLs through assessment of physical self-maintenance skills: the ability to eat, dress, groom, ambulate, bathe, use the toilet, and bend down to pick up an object on the floor. Nine items, modified from Fillenbaum24 and Rosow and Breslau,25 assess performance of instrumental ADLs: getting around the neighborhood, shopping for basic necessities, preparing meals, cleaning house, doing yardwork or gardening, keeping track of money and bills, walking one-quarter mile, walking up and down one flight of stairs, and taking care of or watching children. Decreased social support, stressful life events, and functional impairment were examined because they have been associated with occurrence of depression in late life and may interact to affect outcome.26

At baseline clinical assessment, the geriatric psychiatrist confirmed depression diagnoses. Clinical assessments were repeated every 3 months and when contact was clinically indicated. For data analyses involving 3-month time intervals, the MADRS score closest to each 3-month time-point was used to determine MADRS score for that period. For the present study, all subjects had a baseline MADRS score above 15. The lead investigators trained all geriatric psychiatrists on use of the MADRS, and there were high interrater reliabilities (kappa: >0.9).

In the present study, we have labeled "symptoms of GAD," rather than "comorbid GAD" in recognition of the difficulty in clinically distinguishing anxiety from depression in elderly patients. We also did not determine whether the symptoms occurred outside of mood disorder symptoms, as required by DSM-IV criteria.

Baseline Cognitive Screening

Patients were excluded if they had dementia or suspected dementia at baseline. MHCRC geriatric psychiatrists clinically examined all subjects, reviewed medical records, and conferred with referring physicians for all patients. At baseline, although most subjects had MMSE scores above 24, some severely depressed patients had scores below 25 (range: 18-24). MHCRC protocol is to follow such patients through an acute (8-week) phase of treatment to determine whether cognition improves. Subjects whose MMSE scores remain below 25 are not followed longitudinally in the MHCRC. Thus, in the clinical judgment of study geriatric psychiatrists and by established MHCRC protocol, dementia was effectively excluded at or close to baseline in all elderly depressed MHCRC subjects.

Clinical Follow-Up With the Duke STAGED Approach

MHCRC clinicians followed patients in both inpatient and outpatient settings, using the somatic antidepressant treatment guidelines established in the Duke Mood Disorders Program.27 This STAGED approach takes into account past treatment and current severity. In general, never-treated patients are initially prescribed a selective serotonin re-uptake inhibitor (SSRI). If adequate doses of the SSRI do not bring about sufficient response after 8 to 12 weeks, the recommendation is either to augment with bupropion or switch to venlafaxine. Treatment options after inadequate response at this point include tricyclic antidepressants and lithium augmentation. Electroconvulsive therapy (ECT) remains a treatment option at each level of the algorithm, depending on symptom severity and number of failed \antidepressant trials. Patients were not routinely referred for psychotherapy, although some patients were already engaged in ongoing psychotherapy, although others were referred for individual and/or group psychotherapy, usually cognitive-behavioral psychotherapy. In the present study, all patients initially received somatic treatment, either antidepressant medications or ECT. MHCRC investigators monitored treatment to ensure that the clinical protocol was being followed. Anxiety was treated naturalistically, as a clinical decision between the patient and the study geriatric psychiatrist. Use of benzodiazepines was allowed. Patients were evaluated when clinically indicated and at least every 3 months while they were in the study. Each patient was thus ensured, to the best of our ability, the most appropriate care. To accommodate patients' schedules, we allowed a return appointment to be scheduled within 6 weeks of each 3-month follow-up period. All patients with between 3 months and 2 years of follow-up data were included in the present analyses.

Data Analysis

Summary statistics were derived for demographic, social, and clinical variables, with means and standard deviations (SD) reported for continuous variables and percentages for dichotomous variables. We used a MADRS cutoff of <7 to define remission, as has been reported previously.28'29 Benzodiazepine use was defined as present when patients were treated with a benzodiazepine at any time during the study, with the exception of patients treated with a benzodiazepine only after they achieved remission. For continuous variables, when variances between groups were unequal, we calculated the f statistic by use of a Satterthwaite approximation. We used Kaplan-Meier estimates to determine the percentage of patients achieving remission as listed in the text. Cox proportional-hazard models were used to estimate the effect baseline anxiety has on remission and time-to-remission, with co varia tes significant at p <0.15 included in the model. A final reverse stepwise proportional- hazard regression model was constructed to identify significant variables.

RESULTS

The sample consisted of 204 older depressed individuals who were followed between 3 months and 2 years (see Table 1). Subjects were about 70 years old, on average, and were mostly female and Caucasian. Among 23 individuals of non-white race, 20 were African American, 2 were Asian, and 1 was Native American. The sample's mean baseline MADRS score was 27.5, an average for moderate-to-severe depression. Mean CIRS score was 3.92, indicating moderate illness burden. Information about reported functioning, social support, and Stressors is also listed. Subjects reported an average of 0.62 impairments on basic ADLs and 4.02 impairments on instrumental ADLs. They reported a mean of 2.54 Stressors and 1.49 negative Stressors in the year before study entry. Their mean scores for each of the four scales of the DSSI were as follows: 5.77 on the Social Interaction Scale; 2.03 on the Social Network Scale; 9.39 on the Instrumental Social Support Scale (of total possible 13); and 23.14 on the Subjective Social Support Scale (of total possible 30). On all DSSI scales, higher scores indicate greater social support.

TABLE 1. Characteristics of the Entire Sample and Among Remitted and Non-Remitted Groups

Regarding treatment, all patients except one were on antidepressant medication during the study. Of the 203 patients on antidepressant medication, 33.5% were on monotherapy with the same medication during follow-up, and 66.5% were given trials of at least two medications. Also, 43.1% of individuals were prescribed a benzodiazepine sometime during their treatment in the study. Among patients who reported baseline symptoms of generalized anxiety, 58.7% were prescribed benzodiazepines, whereas 36.2% of patients not reporting generalized anxiety symptoms were prescribed benzodiazepines (χ^sup 2^^sub [1]^ = 9.0349; p <0.0026). In terms of remission status, as shown in Table 1, more non-remitters than remitters were prescribed benzodiazepines.

Also listed in Table 1 are comparisons between remitters (N = 138) and non-remitters (N = 66), on the basis of MADRS scores up to 2 years. Compared with non-remitters, the remitted group had participated for a shorter time in the study, had fewer IADL impairments, greater subjective social support, greater instrumental social support, more social interaction, fewer total Stressors, and more negative Stressors, were less likely to report symptoms of GAD, and had lower scores on the CIRS.

Kaplan-Meier curves comparing remission rates for depressed patients with and without generalized anxiety are shown in Figure I. A corresponding logrank test demonstrated that the two curves were significantly different (χ^sup 2^^sub [1]^ = 11.8322; p <0.0006).

In bivariate proportional-hazards models (see Table 2), shorter time-to-remission was associated with lower CIRS score, fewer IADL impairments, fewer negative Stressors, higher subjective social support, higher instrumental social support, fewer generalized anxiety symptoms, and non-use of benzodiazepines.

FIGURE 1. Effect of Baseline GAD on Time-to-Remission of Depression

We then constructed three proportional-hazards models examining variables associated with time-to-remission. An initial model (not shown) contained variables significant at p <0.15 at the bivariate level from Table 2. For this model, race, presence of baseline anxiety, and Subjective Social Support Scale (SSSS) score were significantly associated with time-to-remission, whereas use of benzodiazepines and total number of negative Stressors had p values less than 0.1. In the two remaining proportional-hazard analyses, we used backward stepwise procedures to construct models containing significant variables, again examining those variables in Table 2 significant at p <0.15. In the first model (not shown), we removed the benzodiazepine-use variable, given its potential clinical confounding with generalized anxiety. Here, the following variables were significantly associated with shorter time-to-remission: no GAD symptoms, non-white race, fewer IADL impairments, higher SSSS score, and fewer reported negative Stressors in the previous year. In the final model, four variables remained significantly associated with time-to-remission: GAD symptoms, benzodiazepine use, SSSS score, and race (see Table 3).

DISCUSSION

The major finding of this study was that older depressed patients with baseline symptoms of generalized anxiety disorder were less likely to remit than non-anxious depressed patients. Anxiety had a negative impact on outcome even after controlling for other factors that affect outcome, such as subjective social support, instrumental functional impairment, negative Stressors, and severity of comorbid medical conditions. Anxiety also affected outcome after controlling for use of benzodiazepines, which itself proved to be an independent predictor of poor outcome.

We found a frequency of symptoms of generalized anxiety disorder of 30.9% in our sample. This relatively high frequency is consistent with two recent studies of anxiety in depression.5'6 Other studies have reported a smaller prevalence of diagnosable comorbid anxiety symptoms.3'4 One issue related to differences in reported frequencies of anxiety and anxiety disorders is embedded in the DSM- IV criteria for GAD, which require a determination of whether anxiety symptoms occur outside of mood disorder symptoms. We note that the validity of this exclusion approach has been questioned by a recent study.30

Our finding that comorbid anxiety adversely affects depression treatment outcome in elderly patients is consistent with most previous studies.9'10 Older depressed patients with anxiety have been shown to take longer to respond to antidepressant treatment than non-anxious depressed patients.4 Anxiety has also been associated with a chronic course of depression in older depressed individuals.31

On the other hand, the results of our study stand in contrast to a recent report by Lenze et al.,12 in which the authors present findings from two controlled studies of late-life depression, one being a trial comparing nortriptyline and paroxetine (N = 116), and the other, an open-treatment trial of paroxetine and interpersonal therapy (N = 125). There were no differences between anxious and nonanxious groups in response rates or time-to-response. These studies and ours had similar sample sizes. One difference is that, in the Lenze study,12 patients with severe anxiety or insomnia were specifically allowed to receive lorazepam (0.5 mg-2.0 mg/day), and 64% of anxious subjects were treated with lorazepam. In our study, 43% of subjects were treated with a benzodiazepine, and this difference may have contributed to the divergent results of the two studies. The findings may also differ between the studies because of how anxiety was defined. The Lenze et al. study12 included several comorbid anxiety disorders, whereas ours focused on presence of GAD symptoms, as assessed by a standardized diagnostic interview.

TABLE 2. Cox Bivariate Proportional-Hazards Models of Shorter Time-to-Remission

TABLE 3. Proportional-Hazards Model of Time-to-Remlssion

Our finding that benzodiazepines were independently associated with time-to-remission may be an indication of the complexity of patients prescribed these medications. It has been suggested that co- administration of benzodiazepines may play an important role in treatment of anxious elderly patients with depression.32 However, investigators have also expressed concern about the relatively high use of benzodiazepines among older depression patients, even in those receiving care in mental health settings.33

We also found that non-white race was associated with faster remission than white race. In our sample, the non-white group almost entirely comprised African Amer\icans. There is scant literature on racial differences in outcomes of late-life depression, in part because of many barriers to participation of African Americans.34 A previous longitudinal study noted that African Americans, compared with whites, were more likely to be members of the group that remitted from depression despite continued functional impairment and were less likely to be in the group that experienced continued depression despite improved disability status.35 Clearly, more research is needed to better understand racial differences in occurrence and outcome of late-life depression.

This study has both the limitations and advantages of a naturalistic follow-up design. There would have been obvious advantages to having all patients treated with the same medication. We did not control for the varying effects that antidepressant medications might have on anxiety level, nor did we control for concomitant anxiolytic use. On the other hand, this study provides information on a heterogeneous population of over 200 patients treated according to established treatment guidelines in a tertiary- care setting. In future studies that incorporate individual medication trajectories, we will be able to examine other important outcomes, such as time-to-therapeutic dose, time-to-switching of medication, and the percentage of responders and attrition at each point in the algorithm. With our sample being largely female, relatively young, Caucasian, and physically able to access outpatient psychiatric services, our findings may not generalize to more diverse, physically frailer, or cognitively impaired older populations.

Further studies are needed to improve treatment for co-existing depression and anxiety in elderly patients. Such treatment may involve optimizing single-medication or psychotherapy treatment, combining medications, or offering combined medication and psychotherapy. Research is also needed to improve our diagnostic understanding of anxiety and depression, particularly in elderly patients, where both types of symptoms frequently co-occur. Perhaps the term comorbid is an antiquated notion when applied to older anxious and depressed individuals. Studies aimed at clarifying our diagnostic nomenclature in this area are clearly warranted. Finally, the effect of anxiety symptoms in older depressed patients on a variety of outcomes such as functional status, medical illness, and healthcare utilization, are understudied and merit further attention.

This study was supported by NIMH grants P50 MH60451, ROl MH54846, and K24 MH70027.

References

1. Devanand OP: Comorbid psychiatric disorders in late-life depression. Biol Psychiatry 2002; 92:236-242

2. Mancla M, Katona C, Livingston G: How common are the anxiety disorders in old age? IntJ Geriatr Psychiatry 1996; 11:65-70

3. Parmelee PA, Katz IR, Lawton MP: Anxiety and its association with depression among institutionalized elderly. AmJ Geriatr Psychiatry 1993; 1:46-58

4. Mulsant BH, Reynolds CF Π?, Shear MK, et al: Comorbid anxiety disorders in late life depression. Anxiety 1996; 2:242-247

5. Lenze EJ, Mulsant BH, Shear MK, et al: Comorbid anxiety disorders in depressed elderly patients. Am J Psychiatry 2000; 157:722-728

6. lieekman AT, de Beurs E, van Balkom AJ, et al: Anxiety and depression in later life: co-occurrence and communality of risk factors. AmJ Psychiatry 2000; 157:89-95

7. Hopko DR, Bourland SL, Stanley MA, et al: Generalized anxiety disorder In older adults: examining the relation between clinician severity ratings and patient self-report measures. Depress Anxiety 200; 12:217-225

8. Lenze FJ, Mulsant BH, Shear MK, et al: Comorbidlty of depression and anxiety disorders in later life. Depress Anxiety 2001 ; 14:S6-93

9. Flint AJ, Rifat SL: Anxious depression in elderly patients: response to antidepressant treatment. Am J Geriatr Psychiatry 1997; 5:107-115

10. Dew MA, Reynolds CF III, Houck PR, et al: Temporal profiles of the course of depression during treatment: predictors of pathways toward recovery in the elderly. Arch (icn Psychiatry 1997; 54:1016- 1024

11. Flint AJ, Rifat SL: Two-year outcome of elderly patients with anxious depression. Psychiatry Res 1997; 66:23-31

12. Lenze FJ, Mulsant BH, Dew MA, et al: Good treatment outcomes in late-life depression with comorbid anxiety. J Affect Disord 2003; 77:247-254

13. Hamilton M: A rating scale for depression. J Neurol Neurosurg Psychiatry I960; 23:55-61

14. Montgomery SA, Asberg M: A new depression scale designed to be sensitive to change. BrJ Psychiatry 1979; 134:382-389

15. Miller MD, Paradis CF, Houck PR, et al: Rating chronic medical illness burden in geropsychiatric practice and research: application of the Cumulative Illness Rating Scale. Psychiatry Res 1992; 41:237-248

16. Folstein MF, Folstein SE, McHugh PR: Mini-Mental State, a practical method for grading the cognitive state of patients for the clinician. J Psychiatr Res 1975; 12:189-198

17. Landerman R, George LK, Campbell RT, et al: Alternative models of the stress-buffering hypothesis. Am J Community Psychol 1989; 17:626-642

18. Robins N, Heizer JE, Croughan J, et al: National Institute of Mental Health Diagnostic Interview Schedule. Arch Gen Psychiatry 1981; 38:381-389

19. Krishnan KRR, Hays JC, HlazerDG: MRl-deuncd vascular depression. Am J Psychiatry 1997; 154:497-501

20. George LK, Blazer DG, Hughes DC, et al: Social support and the outcome of major depression. BrJ Psychiatry 1989; 154:478-485

21. Katz S, Down TD, Cash HR, et al: Progress in the development of an index of ADL. Gerontologist 1970; 10:20-30

22. Branch LG, Katz S, Kniepmann K, et al: A prospective study of functional status among community elders. Am J Public Health 1984; 74:266-268

23. Nagi SZ: An epidemiology of disability among adults in the United States. Milbank Memorial Fund Q Health Soc 1976; 54:439-468

24. Pillcnbaum GG: Screening the elderly: a brief Instrumental Activities of Daily Living measure. J Am Geriatr Soc 1985; 33:698- 706

25. Rosow I, Breslau N: A Guttman Health Scale for the aged. J Gerontol 1966; 21:556-559

26. Hays JC, Steffens DC, Flint EP, et al: Does soeial support buffer functional decline in elderly patients with unipolar depression? AmJ Psychiatry 2001; 158:1850-1855

27. Steffens DC, McQuoicl DR, Krishnan KRR: The Duke Somatic Treatment Algorithm for Geriatric Depression (STACiED) approach. Psychopharmacol Hull 2002; 36:58-68

28. Snaith RP, Harrop FM, Newby DA, et al: Grade scores of the Montgomery-Asberg Depression and the Clinical Anxiety Scales. BrJ Psychiatry 1986; 148:599-601

29. Steffens DC, McQuoid DR, Krishnan KR: Partial response as a predictor of outcome in geriatric depression. Am J Geriatr Psychiatry 2003; 11:340-348

30. Zimmerman M, Chelminski I: Generalized anxiety disorder in patients with major depression: is DSM-IVs hierarchy correct? Am J Psychiatry 2003; 160:504-512

31. Schoevers RA, Beekman AT, Deeg DJ, et al: The natural history of late-life depression: results from the Amsterdam Study of the Elderly (AMSTEL). J Affect Disord 2003; 76:5-14

32. Lenze EJ, Mulsant BH, Shear MK, et al: Anxiety symptoms in elderly patients with depression: what is the best approach to treatment? Drugs Aging 2002; 19:753-760

33. Valenstein M, Taylor KK, Austin K, et al: Benzodiazepine use among depressed patients treated in mental health settings. Am J Psychiatry 2004; 161:654-661

34. Steffens DC, Artigues DL, Ornstein KA, et al: A review of racial differences in geriatric depression: implications for care and clinical research. J Natl Med Assoc 1997; 89:731-736

35. Koenig HG, George LK: Depression and physical disability outcomes in depressed medically ill hospitalized older adults. Am J Geriatr Psychiatry 1998; 6:230-247

David C. Steffens, M.D., M.H.S.

Douglas R. McQuoid, B. S.

Received December 11, 2003; revised March 15, April 16, 2004; accepted May 18, 2004. From the Department of Psychiatry and Behavioral Sciences, Duke University Medical Center, Durham, NC. Send correspondence and reprint requests to David C. Steffens, M.D., M.H.S., Associate Professor of Psychiatry and Medicine, Head, Division of Geriatric Psychiatry, Duke University Medical Center, Box 3903, Durham, NC 27710. e-mail: steff001@mc.duke.edu

2005 American Association for Geriatric Psychiatry

Copyright American Psychiatric Press, Inc. Jan 2005

Source: American Journal of Geriatric Psychiatry, The

More News in this Category