JOURNAL OFAvailable online at www.sciencedirect.com

PSYCHIATRIC

RESEARCHJournal of Psychiatric Research 42 (2008) 797–801

www.elsevier.com/locate/jpsychires

Remission in depressed outpatients: More thanjust symptom resolution?

Mark Zimmerman *, Joseph B. McGlinchey, Michael A. Posternak, Michael Friedman,Daniela Boerescu, Naureen Attiullah

Department of Psychiatry and Human Behavior, Brown University School of Medicine, Rhode Island Hospital, Bayside Medical Center,

235 Plain Street, Providence, RI 02905, United States

Received 22 December 2006; accepted 26 September 2007

Abstract

Objective: In treatment studies of depression remission is defined according to scores on symptom severity scales. Normalization of func-tioning has often been mentioned as an important component of the definition of remission, though it is not used to identify remittedpatients in studies of treatment efficacy. Conceptually, the return of normal functioning should be as fundamental to the concept ofremission as is symptom resolution because the presence of both symptoms and impaired functioning are core constructs in the diagnosisof mental disorders. In the present report from the Rhode Island Methods to Improve Diagnostic Assessment and Services (MIDAS)project we examined the independent and additive association between level of severity of depressive symptoms and functional impair-ment in predicting depressed patients’ subjective evaluation of their remission status.Methods: Five hundred and fourteen depressed psychiatric outpatients filled out a questionnaire on which they rated the severity of thesymptoms of depression, the level of impairment due to depression, and their quality of life.Results: Symptom severity, functional impairment from depression, and quality of life were significantly and highly intercorrelated, andeach was significantly associated with remission status. The results of a logistic regression analysis indicated that each of the three vari-ables was a significant, independent, predictor of remission status.Discussion: In treatment studies of depression remission is narrowly defined in terms of symptom resolution. Our results support broad-ening the concept of remission beyond symptom levels to include assessments of functioning and quality of life.� 2007 Elsevier Ltd. All rights reserved.

Keywords: Depression; Remission; Symptoms; Psychosocial functioning

1. Introduction

Increasingly, experts in the treatment of depression havesuggested that achieving remission should be viewed as theprimary goal of treatment (Ballenger, 1999; Ferrier, 1999;Nierenberg and Wright, 1999; Rush et al., 1998; Rushand Trivedi, 1995; Stahl, 1999; Thase, 2003). These recom-mendations are based on studies that have consistentlydemonstrated that depressed patients who have responded

0022-3956/$ - see front matter � 2007 Elsevier Ltd. All rights reserved.

doi:10.1016/j.jpsychires.2007.09.004

* Corresponding author.E-mail address: mzimmerman@lifespan.org (M. Zimmerman).

to treatment but failed to achieve symptomatic remissioncontinue to experience more psychosocial impairmentand have a higher likelihood of recurrence of a full depres-sive syndrome (Faravelli et al., 1986; Judd et al., 2000; Juddet al., 1998; Paykel et al., 1995; Simons et al., 1986; Thaseet al., 1992).

In order to establish remission as the goal of treatment itis necessary to have a cogent conceptualization of the con-struct. As Keller (2003) recently noted, ideally remissionwould be defined biologically, based on the normalizationof underlying pathophysiology. However, there are novalid biological state markers of major depressive disorderthat can be used to monitor the progress of the disease. In

798 M. Zimmerman et al. / Journal of Psychiatric Research 42 (2008) 797–801

the absence of such biological tests, we are left withphenomenological definitions of remission. In treatmentstudies, remission is defined according to scores fallingbelow a threshold on symptom severity scales such as theHamilton rating scale for depression (1960) and theMontgomery–Asberg depression rating scale (Montgomeryand Asberg, 1979). Normalization of functioning is oftenmentioned as an important component of the definitionof remission, though it is not used to identify remittedpatients in studies of treatment efficacy. Instead, psychoso-cial functioning is compared in patients who have and havenot achieved symptomatic remission (Judd et al., 2000;Miller et al., 1998).

Conceptually, the return of normal functioning shouldbe as fundamental to the definition of remission as is symp-tom resolution because the presence of both symptoms andimpaired functioning are core constructs in the diagnosis ofmental disorders (American Psychiatric Association, 1994).It seems logical that each of the defining features of a dis-order should be absent in order to declare that the disorderis in remission.

As part of the Rhode Island Methods to Improve Diag-nostic Assessment and Services (MIDAS) project werecently conducted a survey of the factors depressedpatients considered important in defining remission fromdepression (Zimmerman et al., 2006a; Zimmerman et al.,2006b). We found that patients considered both symptomresolution and functional improvement important in deter-mining remission. In fact, normalization of function wasmore often judged to be the most important factor in deter-mining whether a depressive episode has remitted.

In the present report from the MIDAS project, weexamined the independent and additive associationbetween level of severity of depressive symptoms and func-tional impairment in predicting depressed patients’ subjec-tive evaluation of their remission status. We also examineda third variable, subjectively rated quality of life, as it hasalso been emphasized as an important factor in determin-ing the adequacy of treatment response of medical disor-ders (Kennedy et al., 2001; Thunedborg et al., 1995).

2. Methods

The study was conducted from August 2003 until July2004. Participants were 535 psychiatric outpatients whowere being treated for a DSM-IV major depressiveepisode in the Rhode Island Hospital, Department ofPsychiatry outpatient practice. This private practice grouppredominantly treats individuals with medical insuranceon a fee-for-service basis, and it is distinct from the hospi-tal’s outpatient residency training clinic that predominantlyserves lower income, uninsured, and medical assistancepatients. The sample included 182 (34.0%) men and 353(66.0%) women who ranged in age from 21 to 80 years(M = 44.2, SD = 11.5). The Rhode Island Hospital institu-tional review committee approved the research protocol,and all patients provided informed, written consent.

Patients completed two questionnaires. One of the ques-tionnaires was a symptom measure of depression thatincluded a question regarding functional impairment dueto depression (‘‘Overall, how much have symptoms ofdepression interfered with or caused difficulties in your lifeduring the past week? 0 = not at all; 1 = a little bit; 2 = amoderate amount; 3 = quite a bit; 4 = extremely’’) andquality of life (‘‘How would you rate your overall qualityof life? 0 = very good, my life could hardly be better;1 = pretty good, most things are going well; 2 = the goodand bad parts are about equal; 3 = pretty bad, most thingsare going poorly; 4 = very bad, my life could hardly beworse.’’) (Zimmerman et al., 2004a). The questions weretaken from the Diagnostic Inventory for Depression, a reli-able and valid measure of depressive symptoms, psychoso-cial impairment due to depression, and quality of life(Zimmerman et al., 2004b).

The second questionnaire assessed patients’ opinionregarding the importance of different factors in determin-ing remission from depression. On the front side of thetwo-sided questionnaire the instructions read as follows:

During the past decade, researchers who study the treat-ment of depression have discussed the best method of eval-uating response to treatment. One area of controversy iswhat are the most important factors in determining whohas responded well to treatment. Some experts say thatthe most important thing to look at are the symptoms ofdepression – a person should be considered in remissionwhen the symptoms of depression (such as depressedmood, sleep and appetite changes, fatigue, problems con-centrating, etc.) have gone away. Other experts say thatthe most important thing to look at is how a person is func-tioning, regardless of whether they are still experiencingsome symptoms of depression. Other aspects of remissionhave also been proposed.

The purpose of this brief questionnaire is to learn whatpatients believe are the most important factors in determin-ing whether someone is in remission from their depression.Please rate how important you think each of the followingfactors are in determining whether someone is in remissionfrom depression. After rating the importance of each item,circle the number of the item that you think is the mostimportant factor.

Before completing the questionnaire, please provide thefollowing background information:

The background information requested included gender,age, education, current level of severity of depression asrated on 5-point rating scale (0 = none, 1 = minimal,2 = mild, 3 = moderate, 4 = severe), and current remissionstatus from depression (0 = no, 1 = yes). Thus, the threedependent variables (symptom severity, impairment infunction, and quality of life) were rated on 5-point scalesof severity. The reliability and validity of these single-itemassessments of these constructs is described elsewhere(Zimmerman et al., 2006b).

We computed Spearman correlation coefficients amongthe symptom, functioning, and quality of life ratings, and

Table 2Logistic regression evaluating remission status predicted by symptomseverity, impairment and quality of life in 514 depressed outpatients

Predictor Parameterestimate

Oddsratio

95% Confidenceinterval

Symptom severity �1.09*** .34 .24–.47Psychosocial

impairment�0.42* .66 .46–.94

Quality of life �0.62** .54 .34–.85

* Significant at p < .05.** Significant at p < .01.

*** Significant at p < .001.

M. Zimmerman et al. / Journal of Psychiatric Research 42 (2008) 797–801 799

point-biserial correlations between remission status andeach of these three variables. After the univariate analyseswe conducted a multivariate logistic regression analysis todetermine which of the predictor variables were indepen-dently associated with remission status.

3. Results

Twenty-one (3.9%) of the 535 patients did not answerthe question about current remission status leaving a finalsample of 514 patients. There were no demographic differ-ences between the patients who did and did not answer thisquestion. Half of the 514 patients (n = 260; 50.6%) consid-ered themselves to be in remission at the time of the evalu-ation. The mean score on the symptom severity rating was1.9 (SD = 1.2), indicating a mild level of severity. Corre-spondingly, the mean score on the psychosocial impair-ment item was 1.3 (SD = 1.2), indicating, on average,mild levels of impairment, and the mean on the qualityof life item was 1.6 (SD = 0.9).

As expected, symptom severity, functional impairmentfrom depression, and quality of life were significantly andhighly intercorrelated, and each was significantly associ-ated with remission status (Table 1). Compared to patientswho were not in remission, patients who indicated that theywere in remission reported significantly lower symptomseverity (1.2 ± 0.9 vs. 2.7 ± 0.9, t = 18.3, p < .001), lesspsychosocial impairment from depression (0.6 ± 0.8 vs.2.0 ± 1.1, t = 16.0, p < .001), and greater life satisfaction(1.1 ± 0.6 vs. 2.1 ± 0.8, t = 15.4, p < .001).

Table 2 shows the results of the logistic regression anal-ysis. The overall model was significant, and each of thethree variables was a significant, independent, predictorof remission status. The amount of variance accountedfor in predicting remission status increased from 39% basedon symptom severity alone to 56% based on all threevariables.

4. Discussion

The diagnosis of depression is based on the presence ofsymptoms along with functional impairment. Conse-quently, the definition of remission of depressive disorder

Table 1Intercorrelation between measures of symptom severity, psychosocialimpairment, quality of life and remission status in 514 depressedoutpatients in ongoing treatment

Depressionseverity

Psychosocialimpairment

Quality oflife

Remissiona �.63 �.58 �.57Symptom Severity .77 .73Psychosocial

Impairment.79

a 0 = not in remission; 1 = in remission. For the other scales lower scoresindicate less pathological responses (i.e. lower symptom severity, lesspsychosocial impairment, greater life satisfaction).

should be based on the resolution of both symptoms andfunctional impairments. This, however, is not how the fieldhas defined remission. Rather, in treatment studies ofdepression, remission has been defined in symptom termsonly. Specifically, antidepressant efficacy trials have definedremission according to scores falling below a cutoff onsymptom severity scales such as the Hamilton rating scalefor depression and the Montgomery–Asberg depressionrating scale. Thus, a depressed patient whose symptomshave nearly resolved but who has not yet returned towork because of lack of confidence would be consideredto be in remission. On a common sense basis this seemsinappropriate.

We are unaware of any previous studies that have exam-ined whether a multifactorial approach towards definingremission is more valid than the traditional unidimensionalapproach based solely on symptom severity. That is, whileseveral studies have found that symptom-based definitionsof remission predict future relapse, no studies have exam-ined the prognostic validity of remission definitions basedon both symptom status and functioning. It is conceivablethat a multifactorial definition, even though more concep-tually appealing than an exclusively symptom-based defini-tion, would not be more valid than symptom-basedremission definitions because symptom improvementaccounts for such a large portion of the variance in deter-mining remission status that the assessment of the otherdomains does not improve validity.

In the present study, we examined the associationbetween symptom severity, psychosocial impairment, andquality of life and depressed patients’ self-rated evaluationof remission status. We found that symptom severity hadthe highest correlation with remission status, though thecorrelations between remission status and psychosocialimpairment and quality of life were nearly as high. Mostimportantly, the results of the regression analysis indicatedthat each of the three factors was independently associatedwith remission status. This is the first study to provideempirical evidence for broadening the concept of remissionbeyond symptom status.

It will be important to extend the findings of the presentstudy to a study of relapse. One of the principle goals of adefinition of remission is to predict future morbidity, and avalid definition of remission from depression should

800 M. Zimmerman et al. / Journal of Psychiatric Research 42 (2008) 797–801

subdivide treatment responders into groups who are athigher and lower risk of relapse. This is analogous tohow treatment goals for hypertension and hypercholester-olemia were derived (i.e. the prediction of future adversehealth events) (Expert Panel on Detection Evaluation andTreatment of High Blood Cholesterol in Adults, 2001;Joint National Committee on Prevention Detection Evalu-ation and Treatment of High Blood Pressure, 1997). Cur-rent symptom-based definitions of remission, such asthose based on the Hamilton rating scale for depressionand Montgomery–Asberg depression rating scale, alreadyaccomplish this goal (Paykel et al., 1995; Thase et al.,1992; Van London et al., 1998). It is possible that amultifactorial definition of remission will be no better inpredicting future relapse than symptom-based remissiondefinitions because symptom status accounts for such alarge portion of the variance in determining relapse thatthe assessment of the other variables does not improveprognostic ability. This remains an empirical question.

Some limitations of the present study should be noted.The study was conducted in a single outpatient practicein which the majority of the patients was white, female,and had health insurance. The generalizability of theresults to samples with different demographic characteris-tics needs to be demonstrated.

The assessment of remission status was based on a singleyes–no question on a paper-and-pencil questionnaire. It ispossible that some patients did not understand the meaningof the term remission, though the term was defined in theinstructions of the questionnaire. Moreover, expected dif-ferences in symptom severity, psychosocial functioning,and quality of life were found between patients who didand did not indicate that they were in remission.

The assessments of symptom severity, functioning, andquality of life were also based on a single question. Else-where we demonstrated the reliability and validity of thesesingle-item assessments (Zimmerman et al., 2006b). None-theless, perhaps the results would be different if based onmore extensive assessments of these constructs. For exam-ple, it is possible that a more detailed assessment of symp-tom severity would be incrementally more valid than thesingle-item assessment, thereby accounting for more ofthe variance in remission status and eliminating an inde-pendent association between remission status and psycho-social functioning and quality of life. Future studies ofthis issue should include more comprehensive assessmentsof these constructs.

Conflict of interest

None.

Contributors

Author Zimmerman designed the study and wrote themanuscript. Author McGlinchey managed and conductedthe statistical analyses. Authors Zimmerman, Posternak,

Friedman, Boerescu, and Attiullah collected the data. Allauthors contributed to and have approved the finalmanuscript.

Role of the funding source

There was no funding source for this research.

References

American Psychiatric Association. Diagnostic and statistical manual ofmental disorders. 4th ed. Washington, DC: American PsychiatricAssociation; 1994.

Ballenger J. Clinical guidelines for establishing remission in patientswith depression and anxiety. Journal of Clinical Psychiatry1999;60(suppl. 22):29–34.

Expert Panel on Detection Evaluation and Treatment of High BloodCholesterol in Adults. Executive summary of the third report of theNational Cholesterol Education Program (NCEP) expert panel ondetection, evaluation, and treatment of high blood cholesterol in adults(Adult Treatment Panel III). Journal of the American MedicalAssociation 2001;285:2486–2497.

Faravelli C, Ambonetti A, Pallanti S, Pazzagli A. Depressive relapses andincomplete recovery from index episode. American Journal of Psychi-atry 1986;143:888–91.

Ferrier I. Treatment of major depression: is improvement enough? Journalof Clinical Psychiatry 1999;60(suppl. 6):10–4.

Hamilton M. A rating scale for depression. Journal of Neurology,Neurosurgery and Psychiatry 1960;23:56–62.

Joint National Committee on Prevention Detection Evaluation andTreatment of High Blood Pressure. The sixth report of the JointNational Committee on prevention, detection, evaluation, and treat-ment of high blood pressure. Archives of Internal Medicine1997;157:2413–2446.

Judd LL, Akiskal HS, Maser JD, Zeller PJ, Endicott J, Coryell W, et al.Major depressive disorder: a prospective study of residual subthresh-old depressive symptoms as predictor of rapid relapse. Journal ofAffective Disorders 1998;50:97–108.

Judd L, Akiskal H, Zeller P, Paulus M, Leon A, Maser J, et al.Psychosocial disability during the long-term course of unipolar majordepressive disorder. Archives of General Psychiatry 2000;57:375–80.

Keller M. Past, present, and future directions for defining optimaltreatment outcome in depression. Journal of the American MedicalAssociation 2003;289:3152–60.

Kennedy S, Eisfeld B, Cooke R. Quality of life: an important dimension inassessing the treatment of depression? Journal of Psychiatry &Neuroscience 2001;26 suppl.:S23–8.

Miller I, Keitner G, Schatzberg A, Klein D, Thase M, Rush A, et al. Thetreatment of chronic depression, part 3: psychosocial functioningbefore and after treatment with sertraline or imipramine. Journal ofClinical Psychiatry 1998;59:608–19.

Montgomery SA, Asberg M. A new depression scale designed to besensitive to change. British Journal of Psychiatry 1979;134:382–9.

Nierenberg A, Wright E. Evolution of remission as the new standard intreatment of depression. Journal of Clinical Psychiatry 1999;60:7–11.

Paykel ES, Ramana R, Cooper Z, Hayhurst H, Kerr J, Barocka A.Residual symptoms after partial remission: an important outcome indepression. Psychological Medicine 1995;25:1171–80.

Rush A, Trivedi M. Treating depression to remission. Psychiatric Annals1995;25:704–9.

Rush A, Crismon M, Toprac M, Trivedi M, Rago W, Shon S, et al.Consensus guidelines in the treatment of major depressive disorder.Journal of Clinical Psychiatry 1998;59(suppl 20):73–84.

Simons A, Murphy G, Levine J, Wetzel R. Cognitive therapy andpharmacotherapy for depression. Archives of General Psychiatry1986;43:43–8.

M. Zimmerman et al. / Journal of Psychiatric Research 42 (2008) 797–801 801

Stahl S. Why settle for sliver, when you can go for gold? Response vs.recovery as the goal of antidepressant therapy. Journal of ClinicalPsychiatry 1999;60:213–4.

Thase M. Evaluating antidepressant therapies: remission as the optimaloutcome. Journal of Clinical Psychiatry 2003;64(suppl. 13):18–25.

Thase ME, Simons AD, McGeary J, Cahalane JF, Hughes C, Harden T,et al. Relapse after cognitive behavior therapy of depression: potentialimplications for longer courses of treatment. American Journal ofPsychiatry 1992;149:1046–52.

Thunedborg K, Black C, Bech P. Beyond the Hamilton depression scoresin long-term treatment of manic-melancholic patients: prediction ofrecurrence of depression by quality of life measurements. Psychother-apy and Psychosomatics 1995;64:131–40.

Van London L, Molenaar RP, Goekoop JG, Zwinderman AH, Rooij-mans HGM. Three- to 5-year prospective follow-up of outcome inmajor depression. Psychological Medicine 1998;28:731–5.

Zimmerman M, Posternak M, Chelminski I. Using a self-report depres-sion scale to identify remission in depressed outpatients. AmericanJournal of Psychiatry 2004a;161:1911–3.

Zimmerman M, Sheeran T, Young D. The diagnostic inventory fordepression: a self-report scale to diagnose DSM-IV for majordepressive disorder. Journal of Clinical Psychology 2004b;60:87–110.

Zimmerman M, McGlinchey J, Posternak M, Friedman M, Attiullah N,Boerescu D. How should remission from depression be defined? Thedepressed patient’s perspective. American Journal of Psychiatry2006a;163:148–50.

Zimmerman M, Ruggero C, Chelminski I, Young D, Posternak M,Friedman M, et al. Developing brief measures for use in clinicalpractice: the reliability and validity of single-item self-report measuresof depression symptom severity, psychosocial impairment due todepression and quality of life. Journal of Clinical Psychiatry2006b;67:1536–41.