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

Randomized, Double-Blind, Placebo-Controlled Trial of Fluoxetine Treatment for Elderly Patients With Dysthymic Disorder

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

Objective: The authors compared the efficacy and side effects of fluoxetine and placebo in elderly outpatients with dysthymic disorder. Methods: Patients were randomly assigned to fluoxetine (20 mg-60 mg/day) or placebo for 12 weeks in a double-blind trial. Results: Of 90 randomized patients, 71 completed the trial. In the intent-to-treat sample, random regression analyses of the Hamilton Rating Scale for Depression (Ham-D; 24-item) and Cornell Dysthymia Rating Scale (CDRS) scores at each visit produced significant time treatment group interactions favoring the fluoxetine group. Analysis of percentage change in Ham-D scores yielded no effect for treatment group, but a similar analysis of percentage change in CDRS scores yielded a main effect for treatment group, favoring fluoxetine over placebo. In the intent-to-treat sample, response rates were 27.3% for fluoxetine and 19.6% for placebo. In the completer sample, response rates were 37.5% for fluoxetine and 23.1% for placebo. Conclusion: Fluoxetine had limited efficacy in elderly dysthymic patients. The clinical features of elderly dysthymic patients are typically distinct from those of dysthymic disorder in young adults, and the findings suggest that treatments effective for young adult dysthymic patients may not be as useful in elderly dysthymic patients. Further research is needed to identify efficacious treatments for elderly patients with dysthymic disorder, and investigative tools such as electronic/computerized brain scans and neuropsychological testing may help identify the factors that moderate antidepressant treatment response and resistance. (Am J Geriatr Psychiatry 2005; 13:59-68)

Dysthymic disorder, a chronic depressive illness of mild-to- moderate severity, affects approximately 2%-4% of the elderly population and is more prevalent in this age-group than major depressive disorder.1-4 In the first comprehensive, descriptive study of elderly outpatients with dysthymic disorder, we reported that most patients had late age at onset of dysthymia and limited comorbid psychiatric pathology.5,6 These features are similar to those reported in late-life major depression7,8 and differ markedly from young adults with dysthymic disorder, who typically have an early age at onset, with considerable comorbid psychiatric pathology.5,6

In young adults, placebo-controlled antidepressant treatment trials in dysthymic disorder have found efficacy for tricyclic antidepressants9,10 and selective serotonin-reuptake inhibitors (SSRIs).11 However, since young adults and elderly patients with dysthymic disorder differ in key clinical features, they may also differ in antidepressant treatment response. The only reported double-blind, placebo-controlled treatment study in elderly dysthymic patients was conducted in primary care, and it found a small but significant advantage for paroxetine over problem-solving therapy or placebo on the Hopkins Symptom Checklist, with similar response rates (40% to 51%) among the three conditions.12 Earlier, we reported an open, pilot trial of fluoxetine in elderly dysthymic patients, in which moderate efficacy was observed, with relatively few side effects.13 We now report on the efficacy and side effects of fluoxetine in a double-blind, placebo-controlled trial in elderly patients with dysthymic disorder.

Methods

Subjects

Patients at least 60 years old were recruited by physician or therapist referral, and by radio or newspaper advertisements (80% of patients) that offered free evaluation by experienced clinicians for participation in clinical trials in the Late-Life Depression Clinic of the New York State Psychiatric Institute. After a telephone screening to rule out major exclusionary factors for enrollment in the clinic, the initial evaluation was conducted by a psychiatrist, who used standardized forms to obtain a detailed medical (CIRS-G)14 and psychiatric history. Patients with a provisional clinical diagnosis of major depression or dysthymic disorder were interviewed by a research rater (social worker or nurse), who completed the SCID- P. On the basis of the psychiatrist's evaluation and the SCID-P interview, a DSM-IV diagnosis was made at a consensus staff conference.

Inclusion criteria were a DSM-IV diagnosis of dysthymic disorder, 24-item Hamilton Rating Scale for Depression (Ham-D) score ≥8 and ≤25, and Clinical Global Impression (CGI) severity score ≥3. Exclusion criteria were a diagnosis of major depression at evaluation or at any earlier period during the index episode, medical contraindication to fluoxetine, history of allergy to fluoxetine, and lack of response in the current episode to a minimum 6-week trial of any SSRI at adequate dosage (fluoxetine a40 mg/day, sertraline ≥150 mg/day, paroxetine ≥40 mg/day, or citalopram ≥40 mg/day). Other exclusion criteria were active suicidal ideation or plan, Mini-Mental State Exam score ≤23 out of 30, alcohol or substance abuse or dependence in the past 6 months, bipolar disorder, schizophrenia or other psychotic disorder, stroke (clinically manifest), dementia, and other major neurological disorders or insult. Physical examination; electrocardiogram; and blood work, including complete blood count, electrolytes, and liver, renal, and thyroid function tests were completed before study entry.

The protocol was approved by the Institutional Review Board of the New York State Psychiatric Institute. All patients provided written informed consent.

Fluoxetine/Placebo Trial

A minimum 1-week psychotropic medication, washout was required, with a washout of 3 weeks for patients receiving monoamine oxidase inhibitors (MAOIs) or fluoxetine. St. John's Wort and SAM-E were also subject to washout. Zolpidem (up to IO mg prn at bedtime) for insomnia and lorazepam (up to 2 mg/day) for anxiety were permitted during the trial. Psychotherapy of any type at a frequency greater than once/month was not permitted.

After a 1 week, single-blind placebo lead-in, patients who still met entry criteria were randomized to fluoxetine or placebo. The randomization was stratified by the presence or absence of a history of major depression in a previous episode. A randomized permuted block procedure with a block size of 6 within each stratum was used for treatment-group assignment.15 These assignments were generated from a closed list of random numbers by an investigator (HAS) not involved in conducting the trial, and transmitted to a research nurse not otherwise involved in the clinical trial. This nurse was responsible for allocation, preparation, and accounting of all medication capsules. Thus, treatment-group assignments were concealed from all patients, physicians, other care providers, and research raters.

Fluoxetine and placebo capsules were identical in appearance and taste, and administered as a morning dose. During the 12-week randomized phase, the dose regimen guideline was fluoxetine 10 mg/ day or placebo for the first week, followed by fluoxetine 20 mg/day or placebo (1 capsule) for the next 3 weeks, fluoxetine 40 mg/day or placebo (2 capsules) for the next 4 weeks (if responder criteria were not met), and fluoxetine 60 mg/day or placebo (3 capsules) for the final 4 weeks (if responder criteria were not met). Throughout the trial, the study psychiatrist had the option of deviating from this sequence, blindly raising or lowering dosage according to clinical response and side effects. Patients returned unused capsules at each visit, and pill counts were conducted to monitor adherence. Non-adherence was defined as two or more visits at which there was >25% excess in the number of capsules over the prescribed dosage level.

Patients were evaluated at 1-week intervals for the first 3 weeks and then at 2-week intervals. At each visit, the research rater, who remained blind to side effects, completed the 24-item Ham-D and the Cornell Dysthymia Rating Scale (CDRS),16 and the study psychiatrist completed the Clinical Global Impression scales (CGI-Severity and - Improvement). The CDRS was developed in young adult patients to provide a more comprehensive assessment of the symptoms of dysthymia than is obtained with traditional scales.16 Partly based on the Ham- D, the CDRS does not have items for retardation and agitation, but has items for poor concentration, indecisiveness, low productivity, social withdrawal, and related symptoms typical of dysthymia (0-4 range per CDRS item; total: 22 items). Strong interrater reliability has been demonstrated.16

The study psychiatrist (DPD or MSN) conducted clinical management based on standard therapeutic principles in antidepressant clinical trials,17 and completed side-effect ratings (Treatment-Emergent Symptom Scale, TESS, modified for commonly reported SSRI side effects by adding items for sexual dysfunction and restlessness/ jitteriness) at each visit. The patient completed the Beck Depression Inventory-II (BDI) at each visit. The self-report Quality of Life Enjoyment and Satisfaction Questionnaire (Q-LES-Q)18 was completed at the start and end of the 12-week trial.

To be classified as a responder, patients had to have ≥50% decrease in Ham-D scores at final assessment relative to baseline (visit after placebo lead-in and before randomization), and a CGI impr\ovement score of 1 (Very Much Improved) or 2 (Much Improved).

Interrater reliability was assessed in 20 patients between the research raters for Ham-D and CDRS total scores, with intraclass correlation coefficients ranging from 0.90 to 0.98.

Blood levels for fluoxetine and nor-fluoxetine were drawn at Weeks 4, 8, and 12.

Statistical Analyses

The results are expressed as mean (standard deviation [SD]). All statistical tests were two-tailed, with a significance level of α = 0.05. For categorical and continuous measures, χ^sup 2^ and t-tests were used, respectively, to compare the fluoxetine and placebo groups on demographic and clinical features. All patients who participated in at least one evaluation after randomization and the start of treatment were included in the intent- to-treat sample (last observation carried forward). The completer sample comprised the patients who were evaluated after the full 12- week treatment period.

The primary analyses were conducted in the intentto-treat sample. Mixed-model, random regression was applied to the serial Ham-D (and CDRS) scores with treatment group and history of major depression (stratification variable) as the between-subject factors. ANOVAs were then conducted on percentage change in Ham-D (and CDRS) scores, with treatment group and history of major depression as the between- subject factors. Also, logistic-regression analyses were conducted on responder classification as the dependent variable, with treatment group and history of major depression as the independent variables.

Secondary efficacy analyses were conducted in the intent-to- treat sample. Mixed-model, random regression was applied to the serial BDI scores, with treatment group and history of major depression as the between-subject factors. Repeated-measures ANOVA was conducted on the four Q-LES-Q subscale scores, with treatment group and responder status as the between-subject factors. These were completer analyses because the Q-LES-Q was given only at baseline and end-trial.

Analyses of somatic side effects were conducted in the intent-to- treat sample. Mixed-model, random regression was applied to the serial total TESS scores, with treatment group and history of major depression as the between-subject factors. ANOVAs were conducted on percentage change in TESS scores, with treatment group and history of major depression as the between-subject factors.

RESULTS

Of the 91 patients with dysthymic disorder, one patient improved in the initial 1-week, single-blind, placebo phase and was not randomized. Of the remaining 90 patients, 44 were randomized to fluoxetine and 46 to placebo. The ethnic distribution in the fluoxetine group was 86.4% white, 2.3% African American, and 11.4% Hispanic; in the placebo group, it was 89.1% white, 6.5% African American, and 4.4% Hispanic. The majority of patients were male; cardiovascular disease was common, and more than four-fifths of the sample had an age at onset of dysthymia after 50 years (Table 1). The majority of patients had not received any treatment for depression in the index episode: 53.5% in the fluoxetine group and 55.6% in the placebo group. In the fluoxetine group, 16.3% had received antidepressant medication; 9.3% had received psychotherapy; and 20.9% had received both medication and psychotherapy. The comparable rates in the placebo group were 20% medication; 13.3% psychotherapy; and 11.1% both medication and psychotherapy. As Table 1 illustrates, the fluoxetine and placebo groups did not differ significantly in any baseline demographic or clinical feature.

TABLE 1. Demographic and Clinical Features of Dysthymic Patients Who Were Randomized

Blinding

At the last visit, the psychiatrist, research rater, and patient independently made their best guess as to whether fluoxetine or placebo was given. Best-guess accuracy rates were as follows: psychiatrist: fluoxetine 63.4%, placebo 37.1%; research rater: fluoxetine 65.9%, placebo 42.9%; patient: fluoxetine 63.4%, placebo 52.9%.

Efficacy

Seventy-one of the 90 randomized patients completed the trial. Of the 19 patients who exited early, 12 were in the fluoxetine group and 7 in the placebo group. In the fluoxetine group, reasons for early protocol exit were insufficient clinical response (N = 2), somatic side effects (N = 3), intervening medical illness (N = I), protocol violation (N = I), hypomania (N = I), moving residence (N = I), and other reasons (N = 3). In the placebo group, reasons for early protocol exit were insufficient clinical response (N = 2), somatic side effects (N = 1), protocol violation (N = 1), and other reasons (N = 3). Patients who exited early did not differ from study completers in age, sex, ethnicity, years of education, medical or cardiovascular comorbidity, or baseline Ham-D scores.

In mixed model, random-regression analyses (auto-regressive type) on serial Ham-D scores (intentto-treat sample, all time-points), there was a main effect of Time (F^sub [7, 531]^ = 3.28; p <0.002) and a Treatment Group 7times; Time interaction (F^sub [7,531]^ = 2.21; p <0.04). The same model applied to CDRS scores also produced a main effect for Time (F^sub [7, 531]^= 2.66; p <0.02) and a Treatment Group 7times; Time interaction (F^sub [7, 531]^ = 2.74; p <0.01). The same model applied to BDI scores also produced a main effect for Time (F^sub [7, 531]^= 2.49; p <0.02) but no Treatment Group Time interaction (F^sub [7, 531]^ = 0.95; p <0.47). For both the Ham-D and the CDRS, the significant interactions involving Treatment Group were due to a modestly greater improvement over time in the fluoxetine group as compared with the placebo group (Table 2). History of Major Depression was not significant in either analysis.

ANOVA on percentage change in Ham-D scores yielded no significant effect for Treatment Group (F^sub [3, 861]^ = 1.66; p <0.21), no effect for History of Major Depression (F^sub [3, 86]^ = 0.09; p <0.77), and no significant interaction between Treatment Group and History of Major Depression (F^sub [3 861]^= 0.36; p <0.56). ANOVA on percentage change in CDRS scores yielded a significant effect for Treatment Group (F^sub [3 86]^ = 4.57; p <0.04; fluoxetine: mean 26.2%, SD 8.3% versus placebo: mean 4.6, SD 8.1%), with no effect for History of Major Depression (F^sub [3, 86]^ = 0.66; p <0.43), and no interaction between Treatment Group and History of Major Depression (F^sub [3, 86]^ = 0.15; p <0.71). Similar analyses on BDI percent-change scores yielded no significant effect for Treatment Group, History of Major Depression, or their interaction in the intent-to-treat sample.

TABLE 2. Baseline (End 1-Week Placebo) and End-Trial Efficacy, Quality of Life, and Side-Effect Scores in the Placebo and Fluoxetine Groups

In logistic-regression analyses on Responder Status as the dependent variable, there were no significant effects for Treatment Group (WaId χ^sup 2^; n = 1.1; p <0.3) or History of Major Depression (χ^sup 2^^sub [1]^ (= 0.02; p <0.88), and no significant interaction between Treatment Group and History of Major Depression (χ^sub 2^^sub [1]^ = 0.02; p <0.89).

In the intent-to-treat sample, 12 of 44 patients on fluoxetine (27.3%) were responders, versus 9 of 46 (19.6%) on placebo (χ^sup 2^^sub [1]^ = 0.75; p <0.4). In the completer sample, 12 of 32 patients (37.5%) on fluoxetine were responders, versus 9 of 39 (23.1%) on placebo (χ^sup 2^ = 1.8;p <0.2).

Liberalizing the response criteria to include all patients with a 50% decrease in Ham-D scores only, or to include "partial response" by using a criterion of ≥25% reduction in Ham-D scores,19 did not alter the findings. Rates of response also did not differ as a function of sex, total CIRS-G scores (cumulative medical burden), history of cardiovascular disease, age, age at onset of dysthymia, baseline Ham-D scores, benzodiazepine use (11.4% in the fluoxetine group and 6.5% in the placebo group), or other hypnotic use (20.5% in the fluoxetine group and 10.9% in the placebo group) during the trial.

Quality of Life

In repeated-measures ANOVAs on the four Q-LES-Q subscale scores, there was no significant effect of Treatment Group (F^sub [3, 67]^ = 0.52; p <0.48), a significant effect for response (F^sub [3, 67]^ = 4.37; p <0.05), and no Treatment Group by Response interaction (F^sub [3, 67]^ = 0.58; p <0.45). Compared with Non-Responders, patients who responded to either fluoxetine or placebo reported significant improvement in physical health (t^sub [69]^ = 2.7; p <0.01), subjective feelings (t^sub [69]^ = 2.9; p <0.01), and general activities (f^sub [69]^ = 4.7; p <0.001), but not social relationships (t^sub [69]^ = 1.1; p <0.28).

Somatic Side Effects

In mixed-model, random-regression analysis on serial total TESS scores (intent-to-treat sample), there was a main effect of Treatment Group (F^sub [1, 87]^ =5.14; p <0.03) and a main effect of time (F^sub [7, 531]^ = 3.30; p <0.002) but no Treatment Group X Time interaction (F ^sub [7,531]^= 0.40; p <0.1). ANOVA on percentage change in TESS scores yielded no effect for Treatment Group (F^sub [3, 85]^ = 1.41; p <0.24), no effect for History of Major Depression (F^sub [3, 85]^ = 0.07; p <0.80), and no significant interaction between Treatment Group and History of Major Depression (F^sub [3, 85]^ = 0.03; p <0.86). The only symptom that differed significantly between the treatment groups was yawning, which was more common in patients on fluoxetine (baseline: 2.5%, end- trial: 20%) than in patients on placebo (baseline: 6.3%, end-trial: 7.5%; percentage-change scores: t^sub [87]^=2.2; p <0.03).

Oral Dose and Blood Levels

The mean end-point fluoxetine dose was 45.5 mg/ day (SD 16.9). Final fluoxetine oral dose did not correlate with percent change in Ham-D or CDRS scores and was unrelated to Responder status. Fluoxetine and nor-fluoxetine plasma levels were detectable in all patients in the fluoxetine group. These levels (Week 12:fluoxetine mean: 143.5 ng/ml, SD: 233.7, and nor-fluoxetine mean: 106.5 ng/ml, SD: 156.8) did not correlate significantly with percent change in Ham-D (r = 0.1 with fluoxetine and r = 0.2 with norfluoxetine levels at Week 12) or CDRS scores (r = 0.1 with fluoxetine and r = 0.2 with nor-fluoxetine levels at Week 12). Plasma levels of fluoxetine and nor-fluoxetine were not associated with Responder status.

DISCUSSION

The magnitude of the efficacy differences between fluoxetine and placebo was modest, and the proportion of responders was low in both the intent-to-treat (fluoxetine: 27.3% versus placebo: 19.6%) and completer (fluoxetine: 37.5% versus placebo: 23.1%) samples. The only other published placebo-controlled clinical trial in elderly dysthymic patients was conducted in primary care. Williams et al.12 reported response rates (defined as final Ham-D <7) of 45.6% for paroxetine, 50.8% for problem-solving therapy, and 40.3% for placebo, with paroxetine showing a small but significant advantage over placebo in change scores on the Hopkins Symptom Checklist. Although the response rates in that trial were higher than in this study, the medication-placebo differences were small in both studies. In this study, the best guess ratings by the psychiatrist, research rater, and patient suggested that the blind was effectively maintained.

SSRIs in Late-Life Depression

The efficacy findings raise the question of whether selective serotonin-reuptake inhibitors (SSRIs) like fluoxetine are efficacious in geriatric dysthymia to a clinically meaningful extent. Of note, the same question is not fully resolved for geriatric major depression. In a 6-week, double-blind, randomized trial comparing fluoxetine and placebo in 671 elderly patients with geriatric major depression, the intent-totreat response rates were 21% for fluoxetine and 13% for placebo, and completer response rates were 27% for fluoxetine and 16% for placebo,20 similar to the rates obtained here in elderly dysthymic patients. However, the short 6- week trial duration in the major depression study may have been a limitation.20

T here are few other published placebo-controlled studies of SSRIs in geriatric major depression. In a double-blind, randomized study in geriatric major depression, there was minimal advantage for sertraline over placebo.21 In another trial, neither venlafaxine nor fluoxetine showed robust superiority over placebo.22 Earlier, Scandinavian studies comparing citalopram and placebo in elderly patients with major depression showed a small but significant advantage for the SSRI in a few scale items, but response rates were not reported.23,24 Overall, the extant placebocontrolled studies of SSRIs in geriatric major depression and dysthymia consistently show low-to-moderate response rates with small or no advantages for the SSRI over placebo, with generally tolerable side effects. The relatively low response rates and the marginal advantages for the SSRIs over placebo raise the clinical question of whether other classes of medications should be considered as the first-line in the treatment of geriatric dysthymic disorder or major depression, particularly since tricyclic antidepressants have been shown to be superior to placebo in acute treatment25 and relapse prevention26 trials in geriatric major depression. However, tolerability and side effects in the context of comorbid medical illness need to be considered in making this decision.27'28 Of note, in geriatric major depression, response rates in studies not including a placebo and focusing on drug-drug comparisons of SSRIs27'29 have been higher than those reported in placebo-controlled trials.

The random regression analyses showed similar results for Ham-D and CDRS scores, but ANOVA on percentage-change scores showed a significant advantage for fluoxetine over placebo on the CDRS but not on the Ham-D. These results indirectly suggest that the Ham-D, with its heavy weighting for neurovegetative symptoms, may not be the ideal scale to use in elderly dysthymic patients, who often have symptoms such as social withdrawal, indecisiveness, poor concentration, and low productivity, which are items included in the CDRS but not in the Ham-D. In adults with dysthymic disorder, the CDRS has been shown to have superior content validity, compared with the Ham-D in the DSM-IV Mood Disorders Field Trial30 and in antidepressant treatment studies.31 Fluoxetine did not show any advantage over placebo on the BDI, which is consistent with the literature that shows less treatment-responsiveness on self-report than on objective measures.32

Fluoxetine Dosage and Blood Levels

The fluoxetine oral doses used were moderately high, and it is unlikely that the low response rates were caused by an inadequate dosing regimen. The results support the previously established finding that elderly depressed patients, on average, can tolerate a fluoxetine dose that is comparable to those used in young adults.20 The observed lack of association between fluoxetine blood levels and treatment response is consistent with previous studies of fluoxetine in major depression.20'33

Quality of Life

Several quality-of-life measures in these elderly dysthymic patients improved concomitantly with treatment response to either fluoxetine or placebo, with no difference in these measures between fluoxetine- and placebo-treated patients. Improvement in quality of life is an important indicator of treatment response, and these results indirectly help validate the responder criteria used in this study. Quality of life and social adjustment measures have been shown to improve with clinical response in treatment trials of young adult dysthymic patients12'34'35 and elderly patients with major depression.36

Young Versus Elderly Dysthymic Patients

In young adults with dysthymic disorder, several studies suggest moderate-to-strong efficacy for a variety of antidepressants.11'37"43 A metaanalysis of placebo-controlled antidepressant trials in adults with dysthymic disorder revealed a mean response rate of 55% on antidepressant versus 30% on placebo, giving a difference in response rate of 25%.37 However, many of these studies allowed patients with double depression to participate, and it is unclear whether similar effects would have been obtained in primary dysthymia (not double depression). In a double-blind treatment study of 416 young adults (310 completers) with primary dysthymia, Thase et al.44 reported intent-to-treat response rates of 64% to imipramine, 59% to sertraline, and 44% to placebo. Although significant, the 15% advantage for sertraline over placebo was not robust. Of note, the magnitude of the advantage for fluoxetine over placebo in our study was 8% in intent-to-treat analyses and 14% in completer analyses.

Overall, it appears that in young adult dysthymic patients, SSRIs show advantages over placebo that may be greater than in elderly dysthymic patients. One possible explanation is that the nature of the young adult and elderly dysthymic populations is inherently different: elderly patients with dysthymic disorder are characterized by late age at onset (more than 80% had onset of dysthymia after age 50 in this study), do not show female preponderance, and have limited comorbid Axis I and II pathology.5'6 In contrast, young adults with dysthymic disorder typically have an early age at onset, are predominantly female, and commonly have comorbid anxiety, alcohol/substance abuse, and personality disorders.45'46

The majority of patients in this study were male, as was the case in a multicenter trial of elderly patients with dysthymic disorder and minor depression.12 Although treatment response was unrelated to gender in this study, there is some evidence that adult men may not respond as well as women to SSRI treatment.47 Interestingly, there is preliminary evidence that low levels of serum testosterone may characterize elderly men with dysthymic disorder.48 Also, the bulk of the patients in our study of geriatric dysthymic disorder had a late age at onset, and cardiovascular disease was common. These features indirectly suggest the possibility that cerebrovascular disease, which has been shown to characterize both late-onset major depression49"52 and minor depression,53 plays an etiologic role in late-onset dysthymic disorder. Also, there is preliminary evidence that patients with late-onset major depression and comorbid cognitive impairment may show poor treatment response and prognosis.54'55 However, we found no association between treatment response and either age at onset or cardiovascular disease, indirectly suggesting that factors reflecting brain pathology did not contribute to the relatively low response rate to fluoxetine.

Study limitations included the specialty mood-disorders clinic site that may limit generalizability to other types of settings and the recruitment of patients primarily by advertising. However, the small fluoxetine-placebo difference is similar to the paroxetine- placebo difference observed in the multicenter primary-care study12 and the fluoxetine-placebo difference in a multicenter geriatric major depression study.20 The findings, therefore, are not inconsistent with the literature.

Conclusions

Fluoxetine showed limited efficacy, with an advantage over placebo, particularly on the CDRS, in elderly patients with dysthymic disorder. In treatment studies of dysthymic disorder, rating scales specific for dysthymia may confer advantages over traditional scales that focus on major depression, and these may need to be included as the primary outcome measure in such trials in the future.

Despite progress in antidepressant therapeutics, no treatment with proven, clinically meaningful efficacy exists for dysthymic disorder in elderly patients. The clinical characteristics of elderly dysthymic patients indicate an entity that is typically distinct from dysthymic disorder in young adults, \and the results of this study suggest that treatments for young adult dysthymic patients cannot be assumed to be effective in elderly dysthymic patients. These findings are indirectly supported by the only other published double-blind, placebo-controlled study of an antidepressant medication in elderly dysthymic patients, which showed marginal differences between paroxetine, problem-solving therapy, and placebo.12 It is possible that geriatric dysthymic disorder is fairly heterogeneous, with some patients being more likely to respond to psychotherapy than to antidepressant medications.

Geriatric dysthymic disorder is a chronic condition that contributes to suffering, disability, and lower quality of life,56 making it particularly important to find an effective treatment. Elderly patients with nonmajor depression appear to have increased cardiac risk and other medical comorbidities as compared with non- depressed elderly patients,57 and successful treatment of dysthymic disorder (and other forms of non-major depression) with psychotropic medications and/or psychotherapy may improve both mental health and medical outcomes.58

Clearly, further research is needed to identify an effective and clinically meaningful treatment for elderly patients with dysthymic disorder. In this research, rating scales specific for dysthymia may confer advantages over traditional depression scales, and appropriate investigative tools such as MRI scan of the brain and neuropsychological testing need to be utilized to identify the factors that may mediate antidepressant treatment response and resistance.

References

1. Blazer D, Hughes DC, George LK: The epidemiology of depression in an elderly community population. Gerontologist 1987; 27:281-287

2. Carta MG, Carpiniello B, Kovess V, et al: Lifetime prevalence of major depression and dysthymia: results of a community survey in Sardinia. Eur Neuropsychopharmacol 1995; (supp!5):103-107

3. Fichter MM, Brace ML, Schroppel H, et al: Cognitive impairment and depression in the oldest old in a German and in U.S. communities. Eur Arch Psychiatry Clin Ncurosci 1995; 245:319-325

4. Stefansson JG, Lindal E, Bjornsson JK, et al: Lifetime prevalence of specific mental disorders among people born in Iceland in 1931. Acta Psychiatr Scand 1991; 84:142-149

5. Devanand DP, Nobler MS, Singer T, et al: Is dysthymia a different disorder in the elderly? Am J Psychiatry 1994; 151:1592- 1599

6. Devanand DP, Turret N, Moody BJ, et al: Personality disorders in elderly patients with dysthymic disorder. Am J Geriatr Psychiatry 2000; 8:188-195

7. Abrams RC, Spielman IA, Alexopoulos GS, et al: Personality disorder symptoms and functioning in elderly depressed patients. Am J Geriatr Psychiatry 1998; 6:24-30

8. Reynolds CF III, Frank E, Kupfer DJ, et al: Treatment outcome in recurrent major depression: a post-hoc comparison of elderly ("young-old") and midlife patients. Am J Psychiatry 1996; 153:1317- 1324

9. Kocsis JH, Frances AJ, Voss C, et al: Imipramine treatment for chronic depression. Arch Gen Psychiatry 1988; 45:253-257

10. Kocsis JH, Frances AJ, Voss C, et al: Imipramine and social- vocational adjustment in chronic depression. AmJ Psychiatry 1988; 145:997-999

11. Ravindran Ay Bialik RJ, Lapierre YD: Therapeutic efficacy of specific serotonin reuptake inhibitors (SSRIs) in dysthymia. Can J Psychiatry 1994; 39:21-26

12. Williams JW Jr, Barrett J, Oxman T, et al: Treatment of dysthymia and minor depression in primary care: a randomized controlled trial in older adults. JAMA 2000; 284:1519-1526

13. Nobler MS, Devanand DP, Kim MK, et al: Fluoxetine treatment of dysthymia in the elderly. J Clin Psychiatry 1996; 57:254-256

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

15. Schulz KF, Grimes DA: Unequal group sizes in randomized trials: guarding against guessing. Lancet 2002; 359:966-970

16. Mason BJ, Kocsis JH, Leon AC, etal: Measurement of severity and treatment response in dysthymia may have been limited by the structure and format of existing rating instruments. Psychiatr Ann 1993; 23:625-631

17. Fawcett J, Epstein P, Fiester SJ, et al: Clinical management- imipramine/placebo administration manual. NIMH Treatment of Depression Collaborative Research Program. Psychopharmacol Bull 1987; 23:309-324

18. EndicottJ, Nee J, Harrison W, etal: Quality of Life Enjoyment and Satisfaction Questionnaire. Pychopharmacol Bull 1993; 29:321326

19. Montgomery SA, Kasper S: Comparison of compliance between serotonin reuptake inhibitors and tricyclic antideprcssants: a meta- analysis. Int CHn Psychopharmacol 1995; 9:33-40

20. Tollefson GD, Holman SL: Analysis of the Hamilton Depression Rating Scale factors from a double-blind, placebo-controlled trial of fluoxetine in geriatric major depression. Int Clin Psychopharmacol 1993; 8:253-259

21. Schneider L, Clary C, Finkel S, et al: An 8-week multicenter, parallel-group, double-blind, placebo-controlled study of sertraline in elderly outpatients with major depression. Am J Psychiatry 2003; 160:1277-1285

22. Schatzberg A, Cantillon M: Antidepressant early response and remission with venlafaxine or fluoxetine in depressed geriatric outpatients. Presented at the European College of Neuropsychopharmacology Meeting, Nice, 2000

23. Gottfries CG: Scandinavian experience with citalopram in the elderly. Int Clin Psychopharmacol 1996; 11:41-44

24. Nyth AL, Gottfries CG, Lyby K, et al: A controlled multicenter clinical study of citalopram and placebo in elderly depressed patients and placebo in elderly depressed patients with and without concomitant dementia. Acta Psychiatr Scand 1992; 86:138- 145

25. Georgotas A, McCue RE, Friedman E, et al: Response of depressive symptoms to nortriptyline, phenelzine, and placebo. Br J Psychiatry 1987; 151:102-106

26. Reynolds CF III, Frank E, Perel JM, et al: Nortriptyline and interpersonal psychotherapy as maintenance therapies for recurrent major depression. JAMA 1999; 281:39-45

27. Oslin DW, Streim JE, Katz IR, et al: Heuristic comparison of sertraline with nortriptyline for the treatment of depression in frail elderly patients. AmJ Geriatr Psychiatry 2000; 8:141-149

28. Mulsant BH, Pollock BG, Nebes R, et al: A twelve-week, doubleblind, randomized comparison of nortriptyline and paroxetine in older depressed inpatients and outpatients. Am J Geriatr Psychiatry 2001; 9:406-414

29. Finkel SI, Richter EM, Clary CM: Comparative efficacy and safety of sertraline versus nortriptyline in major depression in patients 70 and older. Int Psychogeriatr 1999; 11:85-99

30. Keller MB, Klein DN, Hirschfeld RMA, et al. Results of the DSMIV mood disorders field trial. AmJ Psychiatry 1995; 152:843849

31. Hellerstein DJ, Batchelder ST, Lee A, et al: Rating dysthymia: an assessment of the construct and content validity off the Cornell Dysthymia Rating Scale. J Affect Disord 2002; 71:85-96

32. Sayer NA, Sackeim H., Coleman E, et al: The relations between observer rating and self-report of depressive symptomatology. Psychol Assess 1993; 3:350-360

33. Beasley CM, Dornseif BE, Bosomworth JC, et al: Fluoxetine and suicide: a meta-analysis of controlled trials of treatment for depression. BrJ Psychiatry 1991; 303:685-692

34. Kocsis JH, Zisook S, Davidson J, et al: Double-blind comparison of sertraline, imipramine, and placebo in the treatment of dysthymia: psychosocial outcomes. AmJ Psychiatry 1997; 154:390395

35. Stewart JW, Quitkin FM, McGrath PJ, et al: Social functioning in chronic depression: effect of 6 weeks of antidepressant treatment. Psychiatry Res 1988; 25:213-222

36. Mazumdar S, Reynolds CF HI, Houck PR, et al: Quality of life in elderly patients with recurrent major depression: a factor analysis of the General Life Functioning Scale. Psychiatry Res 1996; 63:183-190

37. de Lima MS, hotoph M, Wessely S: The efficacy of drug treatments for dysthymia: a systematic review and meta-analysis. Psychol Med 1999; 29:1273-1289

38. Boyer P, LecrubierY, Stalla-Bourdillon A, etal: Amisulpridc versus amineptinc and placebo for the treatment of dysthymia. Neuropsychobiology 1999; 39:25-32

39. Hellerstein DJ, Yanowitch P, Rosenthal J, et al: A randomized, double-blind study of fluoxetine versus placebo in the treatment of dysthymia. AmJ Psychiatry 1993; 150:1169-1175

40. Lecrubier Y, Boyer P, Turjanski S, et al: Amisnlpridc versus imipramine and placebo in dysthymia and major depression. J Affect Disorcl 1997; 43:95-103

41. Smeraldi E: Amisulpridc versus fluoxetine in patients with dysthymia or major depression in partial remission: a double-blind, comparative study. J Affect Disord 1998; 48:47-56

42. Vanelle J, Attar-Levy D, Poirier M, et al: Controlled efficacy study of fluoxetine in dysthymia. BrJ Psychiatry 1997; 170:345-350

43. Versiani M, Amrein R, Stabl M: Moclobemide and imipramine in chronic depression (dysthymia): an international double-blind, placebo-controlled trial: International Collaborative Study Group, hit CHn Psychopharmacol 1997; 12:183-193

44. Thase ME, Eava M, Halbreich U, et al: A placebo-controlled, randomi/cd clinical trial comparing sertraline and imipramine for the treatment of dysthymia. Arch Gen Psychiatry 1996; 53:777784

45. Klein DN, Norden KA, Ferro T, et al: Thirty-month naturalistic follow-up study of early-onset dysthymic disorder: course, diagnostic stability, and prediction of outcome. J Abnormal Psychol 1998; 107:338-348

46. Klein DN, Riso LP, Donaldson SK, et al: Family study of earlyonset dysthymia. Arch Gen Psychiatry 1995; 52:487-496

47. Kornstein SG, Sehatzberg AF, Thase ME, et al: Gender differences in treatment response to sertraline versus imipramine in chronic depression. AmJ Psychiatry 2000; 157:1445-1452

48. Seidman SN, Araujo AB, Roose SP, et al: Low testosterone levels in elderly men with dysthymic disorder. Am J Psychiatry 2002; 159:456-459

49. Alex\opoulos GS, Meyers BS, Young RC, et al: "Vascular depression" hypothesis. Arch Gen Psychiatry 1997; 54:915-922

50. Krishnan KR, Hays JC, George LK, et al: Six-month outcomes (or MRI-related vascular depression. Depress Anxiety 1998; 8.142- 146

51. Lesser IM, Boone KB, Mchfinger CM, et al: Cognition and whitematter hyperintensitics in older depressed patients. Am J Psychiatry 1996; 153:1280-1287

52. Sackcim HA, Lisanby SH, Nobler MS, et al: MRI hyperintcnsitles and the vascular origins of late-life depression, in Advances in Psychiatry. Edited by Andradrc C. New York, Oxford University Press, 2000, pp 73-116

53. Kumar A, Jin Z, Bilker Wi et ai: Late-onset minor and major depression: early evidence for common neuroanatomlcal substrates detected by using MRI. Proc Nati Acad Sei USA 1998; 95:76547658

54. Alexopoulos GS, Meyers BS, Young RC, et al: Executive dysfunction and risk for relapse and recurrence of geriatric depression. Arch Gen Psychiatry 2000; 57:285-290

55. Hickie I, Scott E, Wilhelm K, et al: Subcortical hypcrintensities on magnetic resonance imaging in patients with severe depression: a longitudinal evaluation. Biol Psychiatry 1997; 42:367-374

56. Pulska T, Pahkala K, Laippala P, et al: Survival of elderly Finns suffering from dysthymic disorder: a community study. Soc Psychiatry Psychiatr Epidemie! 1998; 33:319-325

57. Frasure-Smith N, Lesperance F, Talajic M: Depression following myocardial infarction: impact on 6-month survival. JAMA 1993; 270:1819-1825

58. Shapiro PA, Lesperance F, Frasure-Smith N, et al: An open- label preliminary trial of sertralme for treatment of major depression after acute myocardial infarction (The SADHAT Trial). Am Heart J 1999; 137:1100-1106

O.P. Devanand, M.D., Mitchett S. Nobler, M.D.

Jocelyn Cbeng, B.A., Nancy Turret, M.S.W.

Gregory H. Pelton, M.D., Steven P. Roose, M.D.

Harold A. Sackeini, Ph.D

Received July 9, 2003; revised August 29, 2003; accepted February 14, 2004. From the Late-Life Depression Clinic and the Department ofliiological Psychiatry, New York State Psychiatric Institute, and the College of Physicians and Surgeons of Columbia University, New York, NY. Send correspondence and reprint requests to O.P. Devanand, M.D., New York State Psychiatric Institute, 1051 Riverside Drive, Unit 126, New York, NY 10032. e-mail: dpd3@columbia.edu

2005 American Association for Geriatric Psychiatry

This work was supported by grants MH50513 and MH55716 from the National Institute of Mental Health.

Or. Devanand has received research grant support from Pfizer and Wyeth, and has served as a consultant to Novartis and Janssen. Dr. Roose has received research grant support from Forest Labs, SmithKline Beecham, and Wyeth, has served as a consultant to Forest Labs, SmithKline Beecham, Bristol Myers Squibb, EH Lilly, Organon, and Pharmacia, and has been a speaker for forest Labs and SmithKline Beecham. Dr. Sackeim has received research grant support from Forest Labs and Pfizer, and has served as a consultant to Eli Lilly, Forest Labs, and Pfizer.

Fluoxetine and matched placebo capsules were provided by Eli Lilly and Co.

Copyright American Psychiatric Press, Inc. Jan 2005

Source: American Journal of Geriatric Psychiatry, The

More News in this Category