ORIGINAL ARTICLE

Efficacy of Interpersonal Psychotherapy

for Depressed Adolescents
Laura Mufson, PhD; Myrna M. Weissman, PhD; Donna Moreau, MD; Robin Garfinkel, PhD

Background: Psychotherapy is widely used for de- improvement in overall social functioning, functioning
pressed adolescents, but evidence supporting its effi- with friends, and specific problem-solving skills. In the
cacy is sparse. intent-to-treat sample, 18 (75%) of 24 patients who re-
ceived IPT-A compared with 11 patients (46%) in the con-
Methods: In a controlled, 12-week, clinical trial of In- trol condition met recovery criterion (Hamilton Rating
terpersonal Psychotherapy for Depressed Adolescents Scale for Depression score #6) at week 12.
(IPT-A), 48 clinic-referred adolescents (aged 12-18 years)
who met the criteria for DSM-III-R major depressive dis- Conclusions: These preliminary findings support the
order were randomly assigned to either weekly IPT-A or feasibility, acceptability, and efficacy of 12 weeks of IPT-A
clinical monitoring. Patients were seen biweekly by a in acutely depressed adolescents in reducing depressive
“blind” independent evaluator to assess their symp- symptoms and improving social functioning and inter-
toms, social functioning, and social problem-solving skills. personal problem-solving skills. Because it is a small
Thirty-two of the 48 patients completed the protocol (21 sample consisting largely of Latino, low socioeconomic
IPT-A–assigned patients and 11 patients in the control status adolescents, further studies must be conducted with
group). other adolescent populations to confirm the generaliz-
ability of the findings.
Results: Patients who received IPT-A reported a nota-
bly greater decrease in depressive symptoms and greater Arch Gen Psychiatry. 1999;56:573-579

N
UMEROUS clinical trials therapy for the depressed adolescents.13-17
show the efficacy of drugs Brent et al17 demonstrated the efficacy of
and psychotherapy, indi- 12 to 16 weeks of individual cognitive be-
vidually or combined, for havior therapy (CBT) for depressed ado-
acute or maintenance lescents. The recovery rates were: CBT
treatments with depressed adults.1 Simi- (64.7%); systemic behavioral family therapy
lar data for depressed adolescents are (37.9%); and nonsupportive therapy
sparse. Trials of tricyclic antidepressants (39.4%). Symptom relief was more rapid in
with adolescents have failed to demon- CBT according to both interviewer and pa-
strate efficacy.2-6 These studies have been tient reports. All 3 treatments showed simi-
criticized on methodological grounds, in- lar reductions in suicidality and improve-
cluding small sample size, diagnostic vari- ment in social functioning.17
ability, comorbidity, variable dosing, and This article reports the results of a
insufficient treatment time.7-10 Recently, randomized controlled clinical trial of In-
Emslie et al11 demonstrated the efficacy of terpersonal Psychotherapy for Depressed
8 weeks of treatment with fluoxetine, a se- Adolescents (IPT-A) in comparison to
From the Division of lective serotonin reuptake inhibitor, in clinical monitoring in a sample of clinic-
Clinical–Genetic Epidemiology comparison to placebo. Fifty-six percent referred depressed adolescents. An open
(Drs Mufson and Weissman) of the patients treated with fluoxetine clinical trial18 and 1-year follow-up19 pro-
and the Department of Child
showed an improvement of their condi- vided preliminary support for the use of
Psychiatry (Drs Moreau and
Garfinkel), New York State tion on the Clinical Global Impressions IPT-A with depressed adolescents. This
Psychiatric Institute and Scale (CGI)12 in comparison to 33% who study hypothesis was that at the end of 12
College of Physicians and received placebo. weeks of treatment, patients treated with
Surgeons, Columbia University, Several studies have tested the effi- IPT-A in comparison to the control treat-
New York. cacy of group or individual psycho- ment would show a notably greater de-

ARCH GEN PSYCHIATRY/ VOL 56, JUNE 1999

573

©1999 American Medical Association. All rights reserved.

 PATIENTS AND METHODS Clinical Monitoring

Patients were assigned a therapist and told they would be

PATIENTS seen monthly for 30-minute sessions, with an option for a
second session within the month, to discuss their symp-
English-speaking patients of normal intelligence, aged 12 toms and functioning. The therapists were given a brief
to 18 years, with a legal guardian who could give informed treatment manual instructing them to refrain from advice
consent, and who met criteria for DSM-III-R major depres- giving or skills training, and to use the sessions to review
sive disorder with an intake Hamilton Rating Scale for De- depressive symptoms, school attendance, assess suicidal-
pression (HRSD)20 score of 15 or more were eligible to par- ity, and just listen supportively. The therapist was avail-
ticipate. Patients who were actively suicidal; were in another able by beeper during regular clinic hours. Clinical moni-
treatment for the same condition; had a chronic medical ill- toring was chosen as a comparison treatment in an effort
ness; or met criteria for psychosis, bipolar I or II, conduct to create an ethical wait-list condition and was modeled
disorder, substance abuse disorder, current eating disorder, after a “call-me-if-you-need-me” control condition previ-
and/or obsessive-compulsive disorder were excluded. ously used in an adult IPT study.23
Patients were recruited from the Child Anxiety and Patients in both treatment conditions were told that if
Depression Clinic at Babies Hospital Columbia Presbyte- they felt worse in between sessions that they should contact
rian Medical Center, New York, NY, and the Clinical Re- their therapist for an immediate evaluation. Patients in IPT-A
search Center, New York State Psychiatric Institute, New could have up to 3 extra sessions during the protocol with-
York, between 1993 and 1996. Most patients were self- out necessitating removal from the study, while the control
referred or referred by parents or mental health profes- patients could have 1 extra session a month without being
sionals from school-based clinics. Seventy-nine adoles- removed from the study. If therapists or independent evalu-
cents suspected of being depressed were screened by one ators (IEs) felt the patient needed a change in treatment, a
of us (L.M.) who conducted a clinical interview using the child psychiatrist (D.M.) blind to treatment condition evalu-
HRSD. Of the 57 adolescents eligible to participate, 48 agreed ated the patient. If the patient’s depression had worsened, if
to randomization. Treatment was free. he or she was suicidal and/or if functioning was deteriorat-
A randomization schedule for the first 100 patients was ing (ie, chronic school refusal), the patient was removed from
constructed by drawing 100 random numbers from the uni- the study and referred for other active treatment. Patients also
form distribution on [0,1] using the Statistical Analysis Sys- were removed if they had failed to attend 3 consecutive ap-
tem (SAS Inc, Cary, NC) (seed = 071450). The lowest 5 num- pointments with their therapist and/or IE.
bers within each block of 10 numbers were assigned to
IPT-A; the highest 5 numbers within each block were as- Therapist Training and Treatment Quality
signed to the control condition.
The IPT-A training program was supervised by one of us
TREATMENT (L.M.), an expert IPT-A therapist. Therapists included 2
child psychiatrists, a licensed clinical psychologist, and a
Interpersonal Psychotherapy for Depressed Adolescents masters’ level psychologist with more than 10 years of pre-
vious clinical experience. All therapists participated in di-
Interpersonal Psychotherapy for Depressed Adolescents dactic and clinical practicum. Each therapist treated 2 de-
is a brief, specified psychotherapy originally developed by pressed adolescents for 12 weeks each. Three sessions per
Klerman et al21 for depressed adult outpatients and adapted case were randomly rated by the 2 expert IPT-A raters
for adolescents.22 The adaptation for adolescents (IPT-A) (M.M.W. and D.M.) using the Therapist Strategy Rating
addresses common adolescent developmental issues, eg, Form and the Therapy Process Rating Form, both of which
separation from parents, exploration of authority in rela- were used in the National Collaborative Study for the Treat-
tionship to parents, development of dyadic interpersonal ment of Depression.24
relationships, initial experience with the death of relative All 4 therapists were rated as competent. Patients
or friend, and peer pressure. In addition, a fifth problem treated during therapist training were excluded from the
area of single-parent families was added.22 Patients were seen study sample. All the training and study therapy sessions
weekly for 12 weeks with once weekly additional tele- (for both conditions) were videotaped for treatment ad-
phone contact between therapist and patient during the first herence and integrity. Therapists received weekly super-
4 weeks of treatment. vision by one of us (L.M.) based on the videotapes. The

crease in depressive symptoms and greater improve- COMPLETION RATES AND EARLY TERMINATION
ment in global and social functioning.
CompletionoftreatmentwassignificantlyhigherintheIPT-A
RESULTS (88%) compared with the control condition (46%). Reasons
for noncompletion in the control condition were worsen-
BASELINE CHARACTERISTICS ing of symptoms and functioning as well as noncompliance
(Table 2). Five patients in the control condition were re-
Patients in both treatment groups did not differ on sex, moved from the study at week 2—4 for suicidality and 1 for
mean age, ethnicity, parental education, socioeconomic psychotic features. The rate of attrition for the control con-
status, history of suicide attempt, suicidal ideation, or base- dition was 3 patients at week 6; 3 at week 8, and 2 at week
line diagnoses (Table 1). 10. Patients in IPT-A attended a mean number of 9.8 (75%)

ARCH GEN PSYCHIATRY/ VOL 56, JUNE 1999

574

©1999 American Medical Association. All rights reserved.

 same therapists treated patients in both the IPT-A and clini- Patients rate themselves for the past 2 weeks and receive 5
cal monitoring condition. scores, 1 for the total of all the domains, and 1 for each do-
main. Use with adolescents has been reported.18,32
ASSESSMENTS Social problem-solving skills were assessed using the
52-Item Social Problem-Solving Inventory–Revised self-
All patients were scheduled to be seen at weeks 0, 2, 4, 6, 8, report measure.33 It comprises 5 scales: positive problem
10, and 12 by an IE who administered the research assess- orientation, negative problem orientation, rational prob-
ments of clinical status blind to the treatment being re- lem solving, impulsive/careless problem-solving style, and
ceived. At week 8, the clinical status of patients in both avoidant coping style based on a model that states that so-
groups were evaluated by a child psychiatrist (D.M.) blind cial problem-solving outcome is largely determined by prob-
to treatment group to determine if treatment should be ter- lem orientation and problem-solving skills.34 It has been
minated. The patients were asked not to tell the IE the demonstrated to have adequate reliability and validity in
treatment they were receiving. If a blind was inadvertently general33,35 and with adolescents.36,37
broken, the patient would be reassigned to a different IE. Suicidality, including ideation, plan, or attempt, was
The main outcomes were diagnosis, symptoms, glo- assessed in the K-SADS-E depression section and suicide
bal and social functioning, and problem-solving skills. The screen section.
Diagnostic Interview Schedule for Children Version 2.3 The K-SADS-E interviews were administered by a
(DISC 2.3) interview25 and the Schedule for Affective Dis- trained nurse clinician and audiotaped versions were co-
orders and Schizophrenia for School-Aged Children rated by another trained interviewer to assess reliability in
(K-SADS-E)26 anxiety and depression sections were ad- 12 cases. There was 82% agreement between the 2 raters
ministered to make a full DSM-III-R diagnosis at week 0 for the major depressive disorder diagnosis according to
prior to study enrollment. The structured interviews, clini- the K-SADS-E. The IEs rated 13 audiotapes of each other’s
cal interview note, and baseline ratings on the HRSD, Beck clinical interviews to establish interrater reliability for the
Depression Inventory (BDI),27 and the Children’s Global HRSD (intraclass correlation coefficient = 0.95) and C-
Assessment Scale (C-GAS)28 were given to an indepen- GAS (intraclass correlation coefficient = 0.84).
dent child psychiatrist to make a best-estimate diagnosis
for each patient at baseline and termination. DATA ANALYSES
Depression was assessed using both a clinician rating
scale, the HRSD, and a self-report measure, the BDI. A score The comparability of the patients in the 2 treatment groups
of 15 or greater on the BDI was used to signify a moderate was examined for demographics. Sex differences and ethnic-
to severe depression, and 9 or less is viewed as recovered ity were examined using the x2 test. Mean age and parental
from a major depressive episode. The BDI is reliable in as- education were analyzed using an unpaired t test. Between
sessing depression in adolescents.29 Internal consistency us- group differences at baseline and at week 12 or termination
ing this sample was excellent (a = .89). The 24-item struc- were calculated using unpaired t tests for outcome measures
tured HRSD, reliable in adolescents,30 was used to assess including the 52-Item Social Problem-Solving Inventory–
changes in depressive symptoms during the course of treat- Revised and the C-GAS. Analyses were conducted for both
ment. A score of 15 or greater was used to indicate a major an intent to treat sample and completer sample. The intent
depression, and a score of less than 6 was used to indicate to treat sample included all subjects who were enrolled in treat-
recovered from a major depressive episode. ment (N = 48). The completer analysis included all subjects
Global functioning was assessed using the C-GAS, a who had completed the 12 week protocol (n = 32). The over-
clinician-rated instrument modified from the adult GAS by all efficacy of treatment was assessed by conducting an analy-
Shaffer et al28 for use with children, and completed by the sis of covariance (ANCOVA) controlling for pretreatment
IEs at evaluation week and week 12 or termination. scores when such scores were expected to affect outcomes,
Using the CGI form, clinicians rated the patient on cur- on all the major outcome measures by treatment condition
rent severity of mental illness and current level of improve- at termination except Social Problem-Solving Inventory due
ment. to missing data on this latter form. Clinical recovery was de-
Change in social functioning over the course of therapy fined as a score of 6 or less on the HRSD, and/or a score of 9
was assessed using the Social Adjustment Scale–Self- or less on the BDI. Data were analyzed for the percentage of
report version (SAS-SR)31 for adolescents, a brief self- patients who reached these cutoff scores at the point of ter-
report instrument that contains 23 questions that fall into mination using x2 analyses on an intent to treat sample. The
4 major categories: school, friends, family, and dating. a level was set at .05 for data analyses.

of 13 sessions including the parent session while the those No significant differences were noted in depression (HRSD
inthecontrolconditionattendedameannumberof2.8(56%) and BDI) scores at week 0 between groups. An ANCOVA
of 5 sessions. Patients in IPT-A attended a mean number of controlling for baseline levels of depression showed that
5.8 (83%) of 7 independent evaluations in comparison to the IPT-A group reported fewer depressive symptoms than
the patients in the control group who attended a mean num- the control group at week 12 (HRSD, P,.02; BDI, P,.05)
ber of 4.2 (60.0%) of 7 independent evaluations. (Table 3).
Thirty-two of 48 patients completed the protocol and
TREATMENT OUTCOME were included in a completer analysis (Table 3). The
ANCOVAs for the completer sample similarly demon-
An intent-to-treat analysis with last score carried for- strated that the IPT-A group in comparison to the con-
ward was conducted with the HRSD and BDI (Table 3). trol group reported significantly fewer depressive symp-

ARCH GEN PSYCHIATRY/ VOL 56, JUNE 1999

575

©1999 American Medical Association. All rights reserved.

 Table 1. Demographics and Sample Characteristicsa Table 3. Depression Symptoms at Entrance, Week 12,
or Termination by Treatment Group*
Clinical IPT-A–
Monitoring Group Treated Group Clinical IPT-A–
Characteristics (n = 24) (n = 24) Monitoring Treated
Group Group F df P
Mean (SD) age, y 15.9 (1.7) 15.7 (1.4)b
Female 70.8 75.0c Intent to Treat (n = 48)
Hispanic 62.5 79.2d (n = 24) (n = 24)
Living in single-parent home 70.8 66.7e HRSD
Mean No. of parental education, y 11 11f Week 0 18.7 (8.6) 19.2 (7.5) ... ... ...
Public assistance, mother, No. 11 10g Week 12 or 11.8 (8.9) 6.3 (7.7) 6.0 1,45 .02
Comorbid diagnoses termination
Dysthymic disorder 13 29h BDI
Any anxiety disorder 88 88i Week 0 22.8 (10.6) 18.8 (8.5) ... ... ...
Current suicidal ideation 50 35j Week 12 or 12.9 (12.6) 5.9 (8.1) 4.2 1,44 .05
History of suicide attempt 25 30k termination
Completers (n = 32)
a
Numbers may vary due to missing data. IPT-A indicates Interpersonal (n = 11) (n = 21)
Psychotherapy for Depressed Adolescents. All values are expressed as HRSD
percentages unless otherwise indicated.
b Week 0 15.9 (8.0) 18.6 (7.6) ... ... ...
t46 = 0.33, P = .75.
x 1 = 0.11, P = .75.
c 2 Week 12 or 11.5 (9.4) 4.9 (5.7) 7.2 1,28 .01
x 1 = 1.6, P = .56.
d 2 termination
x 1 = 0.35, P = .56.
e 2 BDI
f
t33 = 0.29, P = .77. Week 0 17.2 (11.4) 17.9 (8.2) ... ... ...
x 1 = 0.18, P = .67.
g 2
Week 12 or 9.4 (12.4) 4.4 (5.9) 2.4 1,29 .14
x 1 = 2.02, P = .16.
h 2
termination
i
Fisher exact 2-tailed test, P = .70.
x 1 = 1.11, P = .29.
j 2 Intent-to-Treat Sample (n = 48)
x 1 = 1.17, P = .68.
k 2 SAS-SR at week 12
School 1.9 (0.97) 2.2 (1.3) 1.7 1,40 .20
Friends 2.4 (0.96) 1.9 (0.87) 5.8 1,44 .02
Family 1.9 (0.86) 1.8 (0.83) 1.4 1,44 .24
Table 2. Reasons for Early Termination* Dating 3.6 (1.2) 2.5 (1.0) 5.9 1,43 .02
Total overall 2.2 (0.70) 1.9 (0.63) 7.1 1,44 .01
Clinical IPT-A– functioning
Monitoring Group Treated Group
Characteristics (n = 24) (n = 24) *Numbers may vary due to missing data. IPT-A indicates Interpersonal
Completers, No. (%) 11 (46) 21 (88)† Psychotherapy for Depressed Adolescents; HRSD, Hamilton Rating Scale for
Noncompleters, No. (%) 13 (54) 3 (12) Depression; BDI, Beck Depression Inventory; SAS-SR, Social Adjustment
Scale–Self-report version; and ellipses, not applicable. Values expressed as
Reasons for early termination
mean (SD).
for noncompleters
Patient initiated withdrawal 3 1
Patient removal 10 2
Reason for patient removal
Suicidality 4 2 significant group differences between treatment condi-
Noncompliance 4 0 tions on the C-GAS at week 12.
School refusal 1 0
Psychotic symptoms 1 0 Clinical Global Impressions
*Numbers may vary due to missing data. IPT-A indicates Interpersonal
Psychotherapy for Depressed Adolescents. All values are expressed as the
The intent-to-treat sample was missing CGI data at week
sample of the population under study unless otherwise indicated. 12 on 8 control patients. No significant group differ-
†x21 = 9.38, P = .002. ences were noted in baseline CGI scores. At week 12, an
ANCOVA shows that those patients in the IPT-A group
toms at week 12 on the HRSD (P,.01), but there were were rated as significantly less depressed (mean = 2.4,
no differences between groups at week 12 for the BDI. SD = 1.6) than those in the control group (mean = 4.2,
An examination of the differences in rates of recov- SD = 1.1) (F1,37 = 18.8, P,.001) by the treating clini-
ery according to preset criteria chosen from the Na- cians. Significantly more IPT-A patients were reported
tional Collaborative Study for the Treatment of Depres- as improved than control patients (x21 = 16.7, P,.001,
sion26 (HRSD, #6; BDI, #9) revealed that significantly Fisher exact test, 2 tailed). Patients were categorized as
more IPT-A–assigned patients (75%) than control pa- improved if they were rated as very much, much, or mini-
tients (46%) met recovery criteria on the HRSD (x21 = 4.3, mally improved; no change; or as worse if they were rated
P = .04). At termination, 13 patients (27.1%) still met cri- as minimally, much, or very much worse. Specifically,
teria for major depression; 3 in the IPT-A group and 10 the conditions of 20 (95.5 %) of the 21 IPT-assigned pa-
in the control group. tients were reported better compared with 7 (61.5%) of
No significant differences were noted between treat- the 11 control patients; the conditions of 2 (15.4%) of
ment groups on any measure of suicide plan or attempt the control patients were reported as the same at week
at week 12 as measured on the K-SADS. There were no 12 than baseline; and the condition of 1 (4.5%) of

ARCH GEN PSYCHIATRY/ VOL 56, JUNE 1999

576

©1999 American Medical Association. All rights reserved.

IPT-A–assigned patients was reported worse vs 3 (23.1 of 56.5 reported by Brent et al.17 Not all adolescent de-
%) of the control patients. pression studies used the HRSD as an entrance crite-
rion, but all did have baseline BDI scores allowing for base-
Social Adjustment Scale-Self-Report line comparison of self-reported depression severity. Brent
et al17 report a mean (±SD) baseline BDI score of 24.2 ±
In an ANCOVA controlling for baseline scores, the IPT- 8 and Emslie et al11 report a baseline mean (±SD) Chil-
A–treated patients reported significantly better function- dren’s Depression Inventory/BDI of approximately 15.5
ing in comparison to control patients for their overall level ± 11. This is consistent with our mean(±SD) baseline BDI
of functioning (F1,44 = 7.1, P = .01), functioning with their score of 20 ± 10. Patients in all 3 studies reported high
friends (F1,44 = 5.8, P = .02), and functioning in dating re- comorbidity with anxiety disorders and disruptive dis-
lationships (F1,43 = 5.9, P = .02). No significant differ- orders when measured. Brent et al17 report a remission
ences were noted between groups for the domains of rate of 64.7% for CBT in comparison to a rate of 75% for
school and family (Table 3). IPT-A–treated patients. Their remission criteria were more
stringent in that they had to have an absence of major
52-Item Social Problem-Solving depressive disorder for 3 consecutive weeks while this
Inventory–Revised study required a one time termination rating. There-
fore, the rates are not totally comparable. Still, it does ap-
Adolescents in the IPT-A–assigned group showed sig- pear that our sample is quite comparable to the recent
nificantly better skills at week 12 on positive problem- medication and psychosocial treatment studies11,17 for de-
solving orientation (t21 = −2.4, P,.05) and rational prob- pression severity and level of functioning.
lem-solving (t21 = −2.4, P,.05). No significant group The study results must be seen in light of the fol-
differences were noted at baseline on any subscales and lowing limitations and need to be considered prelimi-
no significant differences were noted between groups on nary: (1) a small sample (only 24 patients in each treat-
the general subscales of negative-problem orientation, im- ment condition), (2) substantial attrition from the control
pulsivity/carelessness, and avoidance style. Among the condition, and (3) the use of self-report measures of so-
subscales of the rational problem-solving scale, the IPT- cial functioning. While a significant outcome in and of
A–assigned patients in comparison to control patients re- itself, the attrition hampered some of the analyses
ported better functioning in generation of alternatives because of missing data.
(t21 = −2.8, P,.01), and solution implementation and veri- Another possible limitation is the nature of the
fication (t21 = −2.9, P,.01). control condition. The control condition had only one
mandated 30-minute session per month vs the IPT-A
COMMENT condition having four 45-minute mandated sessions per
month. The control condition could include a second
Data from the initial IPT-A–treated study show that IPT-A monthly session if desired. There was a difference in thera-
compared with clinical monitoring is acceptable (low rate pist time and contact that may have translated into thera-
of attrition and high rate of attendance) and efficacious pist expectations for greater patient improvement in the
for depressive symptom reduction and social function- IPT-A condition which could bias the results. However,
ing improvement. The attrition rates suggest that al- most outcomes were self-reports or IE reports, so the out-
most weekly review of symptoms and functioning is not comes should be independent of the bias.
as therapeutic as IPT-A for many depressed adolescents. Much consideration went into the selection of the
According to both clinician and self-report instru- control condition following Rush’s guidelines38 for de-
ments, patients who received IPT-A treatment reported veloping new psychosocial treatments. A wait-list or no
a significantly greater decrease in depressive symptoms treatment condition was deemed unethical by the insti-
and a significantly greater improvement in social func- tutional review board and this clinical monitoring con-
tioning overall and with friends and dating relation- dition had a history of use in an adult IPT study.23 At the
ships, a primary focus of adolescent life. The rate of re- time of the study design, there were no published stud-
covery on the HRSD was significantly better for those ies on the efficacy of any other individual psycho-
patients receiving IPT-A treatment. They also reported therapy for depressed adolescents nor of any psycho-
improvement in the ability to think of alternative solu- pharmacological agent, thus the choices of tested
tions to problems, to try them out, and then use them comparison therapies were few. While the condition does
adaptively. However, due to missing data, these latter have its limitations, the results still demonstrate as in phase
analyses must be viewed as preliminary. The results must 2 drug studies, that the active treatment (IPT-A) was bet-
be viewed in the context of the study limitations and the ter than no treatment or minimal treatment (placebo) for
most recent clinical trials with depressed adolescents. adolescent depression.
The severity of depression and suicidality in these Lastly, there is limited generalizability of the find-
patients was comparable to those patients included in re- ings due to the effects of the inclusion and exclusion cri-
cent studies. In the study by Brent et al17 the mean per- teria. Patients in the study were selected on the basis of
centage of adolescents with a history of suicide attempt having major depression, but excluded if they had psy-
was 23.4% vs our mean rate of 27.5%. Emslie et al11 did chotic features, conduct disorder, substance abuse, or ac-
not report rates of suicidality. Our mean baseline C- tive eating disorder or obsessive-compulsive disorder.
GAS score was 52 which is not significantly different from Therefore, it was a relatively “clean” sample with mostly
the mean of 48 reported by Emslie et al11 and the mean comorbid anxiety disorders. The sample also is largely

ARCH GEN PSYCHIATRY/ VOL 56, JUNE 1999

577

©1999 American Medical Association. All rights reserved.

female and Latino and includes more late than early ado- REFERENCES
lescents. Thus, the results cannot be viewed as general-
izable to the larger adolescent population with major de-
1. Weissman MM, Markowitz JC. Interpersonal psychotherapy: current status. Arch
pression and other comorbid disorders (particularly Gen Psychiatry. 1994;51:599-606.
substance abuse), but rather are specific to late adoles- 2. Kutcher S, Boulos C, Ward B, Marton P, Simeon J, Ferguson HB, Szaliai J, Katic
cents with depression and comorbid anxiety disorders. M, Roberts N, Bubois C, Reed K. Response to desipramine treatment in adoles-
Treatment modifications may need to be made for younger cent depression: a fixed dose, placebo-controlled trial. J Am Acad Child Adolesc
adolescents and males, although the male patients in this Psychiatry. 1994;33:686-694.
3. Geller B, Cooper TB, Graham DL, Marstellar FA, Bryant DM. Double-blind, pla-
study were no less active participants in the IPT-A treat- cebo controlled study of nortriptyline in depressed adolescents using using a
ment than the female patients. “fixed plasma level” design. Psychopharmacol Bull. 1990;26:85-90.
Our response rate in the control condition for the 4. Birmaher B, Waterman GS, Ryan ND, Perel J, McNabb J, Balach L, Beaudry MB,
intent-to-treat sample (46%) is consistent with or greater Nasr FN, Karambelkar J, Elterich G, Quintana H, Williamson DE, Rao U. Ran-
domized, controlled trial of amitriptyline versus placebo for adolescents with “treat-
than placebo response rates in other studies39 (35%3;
ment-resistant” major depression. J Am Acad Child Adolesc Psychiatry. 1998;
68%40). It may be that those patients who were able to 37:527-535.
maintain themselves in the control treatment were those 5. Klein RG, Mannuzza S, Koplewicz HS, Tancer NK, Shah M, Liang V, Davies M.
patients who go on to spontaneous remission. How- Adolescent depression: controlled desipramine treatment and atypical features.
ever, our attrition rate (54%) for the control condition Depress Anxiety. 1998;7:15-31.
6. Kye CH, Waterman GS, Ryan ND, Birmaher B, Williamson DE, Iyengar S, Dachilee
is somewhat higher than that found in other medication S. A randomized controlled trial of amitriptyline in the acute treatment of adoles-
studies.11 Patients in placebo medication conditions may cent major depression. J Am Acad Child Adolesc Psychiatry. 1996;35:1139-44.
believe they are getting active medication which could 7. Campbell M, Spencer EK. Psychopharmacology in child and adolescent psychia-
be a stronger motivator for attendance and compliance try: a review of the past five years. J Am Acad Child Adolesc Psychiatry. 1988;
with the treatment protocol, resulting in a lower attri- 27:269-279.
8. Connors CK. Methodology of antidepressant drug trials for treating depression
tion rate. in adolescents. J Child Adolesc Psychopharmacol. 1992;2:11-22.
No significant differences were noted between groups 9. Jensen PS, Ryan ND, Prien R. Psychopharmacology of child and adolescent ma-
on the HRSD and BDI before week 12, although the trend jor depression: present status and future directions. J Child Adolesc Psycho-
was in the direction of greater improvement in the IPT-A pharmacol. 1992;2:31-45.
10. Strober M. The pharmacotherapy of depressive illness in adolescents:, III: diag-
condition. While medication studies are generally shorter nostic and conceptual issues in studies of tricyclic antidepressants. J Child Ado-
(approximately 8 weeks), these results are consistent with lesc Psychopharmacol. 1992;2:23-29.
the adult studies showing that psychosocial treatment ef- 11. Emslie GJ, Rush AJ, Weinberg WA, Kowatch RA, Huges CW, Carmody T, Rin-
fects usually take longer to emerge.24 Close monitoring telmann J. A double-blind randomized, placebo-controlled trial of fluoxetine in
of the adolescent patients shows that the clinical symp- children and adolescents with depression. Arch Gen Psychiatry. 1997;54:1031-
1037.
toms can change rapidly from week to week depending 12. Guy W. ECDEU Assessment Manual for Psychopharmacology (Revised). Rock-
on the current psychosocial stressor suggesting greater ville, Md: National Institute of Mental Health; 1976. DHEW publication ADM 76-
mood reactivity than in adults. 338.
Given the strong, albeit initial findings on IPT-A in 13. Reynolds WM, Coats KI. A comparison of cognitive-behavioral therapy and re-
this study and in a recently completed but unpublished laxation training for the treatment of depression in adolescents. J Consult Clin
Psychol. 1986;44:653-660.
study in Puerto Rico (William Bernal, PhD, written com- 14. Lewinsohn PM, Clarke GN, Hops H, Andrews J. Cognitive-behavioral group treat-
munication, 1998), IPT-A deserves further study and rep- ment of depression in adolescents. Behav Ther. 1990;21:385-401.
lication. In light of recent published studies with de- 15. Kahn JS, Kehle TJ, Jenson WR, Clark E. Comparison of cognitive-behavioral, re-
pressed adolescents, study designs such as IPT-A vs laxation, and self-modeling interventions for depression among middle-school
students. Sch Psychol Rev. 1990;19:196-211.
serotonin reuptake inhibitor or CBT, or IPT-A vs IPT-A
16. Wood A, Harrington R, Moore A. Controlled trial of a brief cognitive-behavioral
plus serotonin reuptake inhibitor are possible future de- intervention in adolescent parents with depressive disorders. J Child Psychol Psy-
signs that seem warranted. chiatry. 1996;37:737-746.
17. Brent DA, Holder D, Kolko D, Birmaher B, Baugher M, Roth C, Iyengar S, Johnson
BA. A clinical psychotherapy trial for adolescent depression comparing cogni-
tive, family, and supportive therapy. Arch Gen Psychiatry. 1997;54:877-885.
Accepted for publication March 8, 1999. 18. Mufson L, Moreau D, Weissman MM. Modification of interpersonal psycho-
This work was supported by an NIMH FIRST therapy with depressed adolescents (IPT-A) phase I and II studies. J Am Acad
Award MH48715 (Dr Mufson) and by Center to Study Child Adolesc Psychiatry. 1994;33:695-705.
Youth Depression, Anxiety, and Suicide grant MH43873- 19. Mufson L, Fairbanks J. Interpersonal Psychotherapy for Depressed Adoles-
cents: a one-year naturalistic follow-up study. J Am Acad Child Adolesc Psy-
09, from the National Institute of Mental Health, Bethesda, chiatry. 1996;35:1145-1155.
Md. 20. Hamilton M. A rating scale for depression. J Neurol Neurosurg Psychiatry. 1960;
We thank the therapist and independent evaluators— 56-62.
Trish Gallagher, PhD, Nechama Sorscher, PhD, Janet Fair- 21. Klerman GL, Weissman MM, Rounsaville BJ, Chevron ES. Interpersonal Psy-
banks, MD, Bruce Waslick, MD, Lisa Kentgen, PhD, Laura chotherapy of Depression. New York, NY: Basic Books; 1984.
22. Mufson L, Moreau D, Weissman MM, Klerman G. Interpersonal Psychotherapy
Whitman, MD, and Jacquie Martin, RN, MA. We greatly ap- for Depressed Adolescents. New York, NY: Guildford Press; 1993.
preciate the helpful review of the manuscript by Priya Wick- 23. DiMascio A, Klerman GL, Weissman MM, Prusoff BA, Neu C, Moore P. A control
ramaratne, PhD, and also want to thank William Jaffee and group for psychotherapy research in acute depression: one solution to ethical
Masako Mori, MA, for their assistance in different phases and methodological issues. J Psychiatr Res. 1979;15:189-197.
of the research. 24. Elkin I, Shea MT, Watkins JT, Imber SD, Sotsky SM, Colllins JF, et al. National In-
stitute of Mental Health Treatment of Depression Collaborative Research Program:
Reprints: Laura Mufson, PhD, Department of Psychia- general effectiveness of treatments. Arch Gen Psychiatry. 1989;45:971-983.
try–Clinical Genetic Epidemiology, Columbia University, 25. Shaffer D, Fisher P, Dulcan MK, Davies M. The NIMH Diagnostic Interview Sched-
722 W 168th St, Unit 14, New York, NY 10032. ule for Children Version 2.3 (DISC 2.3): description, acceptability, prevalence rates,

ARCH GEN PSYCHIATRY/ VOL 56, JUNE 1999

578

©1999 American Medical Association. All rights reserved.

 and performance in the MECA study. J Am Acad Child Adolesc Psychiatry. 1996; lem-Solving Inventory (Social Problem-Solving Inventory): psychological as-
35:865-877. sessment. J Consult Clin Psychology. 1990;2:156-163.
26. Chambers W, Puig-Antich J, Hirsch M, Paez P, Ambrosini PJ, Tabrizi MA, Davies 34. D’Zurilla TJ. Problem-Solving Therapy: A Social Competence Approach to Clini-
M. The assessment of affective disorders in children and adolescents by semi- cal Intervention. New York, NY: Springer-Verlag NY Inc; 1986.
structured interview. Arch Gen Psychiatry. 1985;42:696-702. 35. D’Zurilla TJ, Sheedy CF. The relation between social problem-solving ability and
27. Beck AT, Ward CH, Mendelson M, Mock J, Erbaugh J. An inventory for measur- subsequent level of psychological distress in college students. J Pers Soc Psy-
ing depression. Arch Gen Psychiatry. 1961;4:561-571. chol. 1991;61:841-846.
28. Shaffer D, Gould MS, Brasic J, Ambrosini P, Fisher P, Bird H, Aluwahlia S. A Chil- 36. Sadowsky C, Moore LA, Kelly ML. Psychometric properties of the Social Prob-
dren’s Global Assessment Scale (C-GAS). Arch Gen Psychiatry. 1983;40:1228- lem Solving Inventory (Social Problem-Solving Inventory) with normal and emo-
1231. tionally disturbed adolescents. J Abnorm Child Psychol. 1994;22:487-500.
29. Strober M, Green J, Carlson C. Utility of the Beck Depression Inventory with 37. Sadowski C, Kelley ML. Social problem-solving in suicidal adolescents. J Con-
psychiatrically hospitalized adolescents. J Consult Clin Psychol. 1981;49:482- sult Clin Psychology. 1993;61:121-127.
483. 38. Rush AJ. A phase II study of cognitive therapy of depression. In: Williams JB,
30. Robbins DR, Alessi NE, Colfer MV. Treatment of adolescents with a major de- Spitzer RL, eds. Psychotherapy Research: Where Are We and Where Should We
pression: implications of the DST and the melancholic clinical subtype. J Affect Go? New York, NY: Guilford Press; 1984:216-235.
Disord. 1983;40:629-636. 39. Simeon J, Dinicola V, Ferguson H, Copping W. Adolescent depression: a placebo-
31. Weissman MM, Bothwell M. Assessment of social adjustment by patient self- controlled fluoxetine treatment study and follow-up. Prog Neuropsychopharma-
report. Arch Gen Psychiatry. 1976;33:1111-1115. col Biol Psychiatry. 1990;14:791-795.
32. Garber J, Kriss MR, Koch M, Lindholm L. Recurrent depression in adolescents: 40. Puig-Antich J, Perel JM, Lupatikin W, Chambers WJ,, Tabrizi MA, King J, Goetz
a follow-up study. J Am Acad Child Adolesc Psychiatry. 1988;27:49-54. R, Davies M, Stiller RL. Imipramine in prepubertal major depressive disorders.
33. D’Zurilla TJ, Nezu AM. Development and preliminary evaluation of the Social Prob- Arch Gen Psychiatry. 1987;44:81-89.

ARCH GEN PSYCHIATRY/ VOL 56, JUNE 1999

579

©1999 American Medical Association. All rights reserved.