The release of the stage 1 results from the National Institute of Mental
Health's Treatment of Adolescent Depression Study (TADS) could not have come
at a more appropriate—or highly
Right in the middle of a series of Food and Drug Administration (FDA)
hearings, Congressional inquiries, and widespread debate over the use of
antidepressants to treat youth with depression, the TADS results answer at
least one key question.
"These results should really put to rest the debate over whether
medication is effective in adolescents with major depression," declared
Peter Jensen, M.D., Ruane professor of child and adolescent psychiatry at
Columbia University. "Fluoxetine clearly works in carefully selected,
properly diagnosed teenagers with moderate to severe major
"It was noteworthy that suicide-related ideation and
self-harm-related behaviors dropped dramatically in all groups, along with the
substantial clinical benefits from getting effective treatment with
medication, with or without CBT. Of course, this benefit of medication still
doesn't rule out the possibility or rare but real adverse events in kids on
meds. Even though penicillin can be be lifesaving in certain circumstances,
some people have bad reactions to it, and doctors always need to be cautious
and monitor the patient carefully, regardless of the medication. But we don't
stop using penicillin because of rare adverse events; that would harm many
Jensen, director of the Center for the Advancement of Children's Mental
Health at Columbia, was not a part of the TADS research team.
The TADS group was led by John March, M.D., professor of child and
adolescent psychiatry at Duke University School of Medicine. March led a large
group of researchers and clinicians at 13 academic centers across the country
who studied 439 adolescents with major depression in the multiyear,
multimillion-dollar project funded by NIMH.
The goal was to evaluate the effectiveness in a real-world setting of four
specific treatments: cognitive-behavioral therapy (CBT), the SSRI fluoxetine
(Prozac) alone, or the combination of both CBT and fluoxetine, all compared
with a single control group that received a placebo pill. The CBT arms of the
study involved a "skills-oriented treatment based on the assumption that
depression is either caused by or maintained by depressive thought patterns
and a lack of active, positively reinforcing behavior patterns."
Patients completed 15 CBT sessions of 50 to 60 minutes over the 12 weeks of
the study. Patients on fluoxetine or placebo were monitored during six
20-minute to 30-minute sessions over the 12 weeks. A pharmacotherapist checked
clinical status and medication effects and depending on clinical ratings was
able to adjust doses of fluoxetine or placebo. In addition, patient education
was provided including encouragement on the effectiveness of pharmacotherapy
The study involved multiple stages. Stage 1 lasted 12 weeks and compared
the four treatment groups' acute response. The results of stage 1 were
reported in the August 18 Journal of the American Medical
Association. Further reports from other stages will be reported in the
All 439 patients across the four treatment groups improved from baseline
through six weeks and on to the 12-week assessment that closed stage 1. Two
primary outcomes were chosen: scores on the Children's Depression Rating
Scale-Revised (CDRS-R), at weeks six and 12, compared with baseline; and the
12-week rating on the Clinical Global Impression-Improvement (CGI) scale. A
CGI score of 1 (much improved) or 2 (very much improved) was defined as a"
response" to treatment.
Significantly, response rates varied widely between the four groups. Of the
group receiving CBT/fluoxetine, 71 percent were rated at 12 weeks as
responders. Among those taking fluoxetine alone, 60.6 percent responded. For
those who received CBT alone the response rate was 43.3 percent. And for those
receiving a pill placebo, the response rate was 34.8 percent.
Two things are significant about those numbers, noted Graham Emslie, M.D.,
the Charles E. and Sarah M. Seay Chair in Child Psychiatry and professor of
psychiatry in the Graduate School of Biomedical Sciences at the University of
Texas Southwestern Medical Center at Dallas, who was the principal
investigator for TADS at Southwestern.
"It's unusual, first of all to see a placebo effect actually so
relatively low," Emslie said, adding, "it's also nice to then see
such a good degree of separation between the active treatments and
Emslie believes the results are at least in part attributable to the
strength of the protocol's patient-selection criteria. At the time of entrance
to the study, patients had to have been ill for at least six weeks without
improvement. Another three weeks on average passed during assessment and
patients being randomly assigned to their treatment group. After nine weeks of
consistent depressive illness, the researchers thought it unlikely that
patients would respond strongly to placebo.
Compared with placebo, CBT/fluoxetine was the strongest therapeutic
approach, with improvement on the CDRS-R scores of patients on the combination
being statistically significant. CBT/fluoxetine was also statistically
significantly superior to either fluoxetine alone or CBT alone. Fluoxetine
alone was superior to CBT alone as well; however, CBT was not statistically
significantly different from placebo.
When the researchers examined patients' scores on the Suicidal Ideation
Questionnaire (SIQ) -Jr. High School Version, an interesting trend appeared.
At baseline, 29 percent of the total population in the study met criteria for
clinically significant suicidal thinking. All four groups' scores on the SIQ
improved (see table at left), with the CBT/fluoxetine group again showing the
most improvement. However, the group receiving CBT alone had the second
greatest improvement, followed by the fluoxetine group.
Both Jensen and Emslie pointed out that the data on harm-related and
suicide-related adverse events during stage 1 indicate that the events were
clustered in the groups that received fluoxetine, either alone or with CBT.
The group receiving fluoxetine alone, however, was the highest in both the
harm-related and suicide-related categories of serious events.
Patients receiving fluoxetine alone had the highest risk (calculated as an
odds-ratio) of experiencing a harm-related event, compared with those
receiving placebo. Patients in the CBT/fluoxetine group were less likely than
those taking fluoxetine alone, but more likely than those receiving placebo to
experience a harm-related event. Compared to placebo, those in the CBT alone
group were slightly less likely to harm themselves or others compared to
However, because the actual numbers of patients in any of the four groups
experiencing a harm-related event were small, the confidence intervals for the
calculated odds ratios were quite wide, and none of the odds ratios
statistically significant. Only when the researchers combined all patients
taking an SSRI (those in the fluoxetine group plus those in the CBT/fluoxetine
group) and compared those with all patients not getting any drug, did the odds
ratio become significant. Those taking an SSRI were more likely to experience
a harm-related event than those taking no drug (odds ratio 2.19, 95 percent
confidence interval 1.03 to 4.62). In addition, when the researchers pooled
all patients receiving CBT (with or without fluoxetine) and compared them with
all patients not receiving CBT (placebo plus fluoxetine groups), again those
receiving CBT appeared to be less likely to experience a harm-related event,
though this result also was not statistically significant.
A similar pattern emerged with the same comparisons for suicide-related
events; however, none of the suicide-related event odds ratios reached
"CBT appeared to be exerting some sort of protective effect, whereas
fluoxetine appeared to be associated with harm-related and suicide-related
events," Emslie pointed out. Seven of the 439 patients attempted suicide
during the trial, six of which were in either the fluoxetine-plus-CBT or the
fluoxetine-alone group, while one was in the placebo group. "There
certainly looks like there is something there, but what exactly it is, I don't
think we understand yet. But fluoxetine is probably not alone. We need to
figure out what other medications might display this same pattern.
Interestingly, looking even closer," Emslie continued, "there was
a stepwise increase in suicide-related events along with increasing response.
It could be that the old theory of improved energy and activation actually is
tied to increased risk. But it is still only a theory."
What is clear is that all patients must be monitored closely, both Jensen
and Emslie stressed. "You can't just write a prescription and send them
on their way. That is clearly not the best practice," Jensen said."
Careful diagnosis plus careful assessment then should always lead to
careful selection of medication—preferably along with some
therapy—and close monitoring."
An abstract of the study, "Fluoxetine, Cognitive-Behavior
Therapy and Their Combination for Adolescents With Depression" is posted