Mitchell AJ. Two-week delay in onset of action of antidepressants: new evidence. Br J Psychiatry 188: 105-106

Abstract

Many sources purport that antidepressants have a delayed onset of action, measured in weeks rather than days. Recent data using weekly or daily mood ratings demonstrate that maximum improvement occurs during the first 2 weeks, with some improvement within the first 3 days. Methodological issues may underlie the delayed-onset hypothesis.

Full text

SummarySummary Man y sour ces purportthatMan y sour ces purportthat
an tidep r essants hav e a de l a yed onse tofan tidep r essants hav e a de lay ed onsetof
action, measuredinweeksrather thanaction, measuredinweeksrather than
days.R ecen tdata using week lyor dailyda ys .Rec ent da ta using weekl yor da il y
mood ratings demonstratethat maximummood ratings demonstratethatmaximum
i mpro vemen t occurs d ur i ng the first 2im provemen toccurs du r i ng the fi rst 2
weeks, wi t h some i mpr ovemen t w i t h i nweeks, wi th some im pro vement wi th i n
the fi rst 3 days.Methodologicalissues maythefirst 3 days.Methodologicalissuesmay
underlie the delayed-onset hypothesis.unde rl ie the del a yed-onset h ypo t hesis.
Declaration of interestDeclaration of interest None .None .
The notion that onset of action of anti-The notion that onset of action of anti-
depressants takes several weeks is widelydepressants takes several weeks is widely
quoted in clinical guidelines (Andersonquoted in clinical guidelines (Anderson etet
alal, 2000; National Institute for Clinical, 2000; National Institute for Clinical
Excellence, 2004) and has formed the basisExcellence, 2004) and has formed the basis
for considerable biological and pharma-for considerable biological and pharma-
cological research (Blier, 2003). It is alsocological research (Blier, 2003). It is also
a notion that has been thoroughly trans-a notion that has been thoroughly trans-
lated into clinical practice, where clinicianslated into clinical practice, where clinicians
regularly tell patients that the anti-regularly tell patients that the anti-
depressant is likely to take 2 to 4 weeks todepressant is likely to take 2 to 4 weeks to
start to work (Garfieldstart to work (Garfield et alet al,2004).Yetthis, 2004). Yet this
was not always the accepted view, and newwas not always the accepted view, and new
data suggest that this statement may be atdata suggest that this statement may be at
best misleading and at worst inaccurate.best misleading and at worst inaccurate.
THE DELAYED- ONSE TTHE DELAYED- ONSE T
HYPOTHESISHYPOTHESIS
Conceptually, any drug that has a delayedConceptually, any drug that has a delayed
onset of action would be expected to showonset of action would be expected to show
weak or negligible early efficacy butweak or negligible early efficacy but
superior efficacy in the medium term. Thissuperior efficacy in the medium term. This
was examined in three antidepressant trialswas examined in three antidepressant trials
by a group from Columbia University, Newby a group from Columbia University, New
York; QuitkinYork; Quitkin et alet al (1987) observed that(1987) observed that
patients who achieved a sustainedpatients who achieved a sustained
response tended to do so after severalresponse tended to do so after several
weeks’ delay and, furthermore, those whoweeks’ delay and, furthermore, those who
responded early often failed to continue toresponded early often failed to continue to
full remission. Yet in the past 15 years thisfull remission. Yet in the past 15 years this
observation has been challenged, because aobservation has been challenged, because a
delayed response has not been replicated bydelayed response has not been replicated by
a number of independent groups (Dewa number of independent groups (Dew etet
alal, 2001; Szegedi, 2001; Szegedi et alet al, 2003). Posternak, 2003). Posternak
& Zimmerman (2005) recently conducted& Zimmerman (2005) recently conducted
a meta-analysis of 47 double-blinda meta-analysis of 47 double-blind
placebo-controlled antidepressant trialsplacebo-controlled antidepressant trials
encompassing 5100 patients on activeencompassing 5100 patients on active
medication and 3400 on placebo. Acrossmedication and 3400 on placebo. Across
all studies when outcome was defined asall studies when outcome was defined as
a reduction in Hamilton Rating Scale fora reduction in Hamilton Rating Scale for
Depression (HRSD) score, the authorsDepression (HRSD) score, the authors
showed that 23% of all differences betweenshowed that 23% of all differences between
drug and placebo groups were alreadydrug and placebo groups were already
apparent by week 1, and that 57% ofapparent by week 1, and that 57% of
possible differences had appeared bypossible differences had appeared by
week 2. Close inspection of responders inweek 2. Close inspection of responders in
both arms for the exact timing of theirboth arms for the exact timing of their
improvement revealed no differences inimprovement revealed no differences in
the antidepressantthe antidepressant v.v. placebo response ratesplacebo response rates
during any week of treatment. Examinationduring any week of treatment. Examination
of antidepressant improvement trajectoriesof antidepressant improvement trajectories
in isolation was also valuable. It was foundin isolation was also valuable. It was found
that 60% of improvement that occurredthat 60% of improvement that occurred
across all trials with antidepressants tookacross all trials with antidepressants took
place during the first 2 weeks. To look atplace during the first 2 weeks. To look at
this another way, in 80% of all trials the re-this another way, in 80% of all trials the re-
sponse was greater in weeks 1 and 2 than itsponse was greater in weeks 1 and 2 than it
was either in weeks 3 and 4 or in weeks 5was either in weeks 3 and 4 or in weeks 5
and 6. On the basis of these data there seemsand 6. On the basis of these data there seems
to be little doubt that the largestto be little doubt that the largest
improvement per unit time produced by anti-improvement per unit time produced by anti-
depressants occurs within the first 2 weeks ofdepressants occurs within the first 2 weeks of
treatment. Yet we still do not know exactlytreatment. Yet we still do not know exactly
how early within this period antidepressantshow early within this period antidepressants
begin to work – that is, whether there is anybegin to work – that is, whether there is any
clinicalclinical delay in the onset of action.delay in the onset of action.
MEASURING INITIAL ONSETMEASURING INITIAL ONSET
OF THER APEUTIC ACTIONOF THERAPEUTIC ACTION
If we are to address initial onset of action –If we are to address initial onset of action –
that is, how long it takes a drug tothat is, how long it takes a drug to beginbegin toto
work – we should not be satisfied withwork – we should not be satisfied with
studies that use standard definitions ofstudies that use standard definitions of
response or remission, as neither of theseresponse or remission, as neither of these
accurately gauges first onset. Althoughaccurately gauges first onset. Although
some authors have defined onset of actionsome authors have defined onset of action
as a 20% or 33% reduction in baselineas a 20% or 33% reduction in baseline
severity scores, this approach conflates aseverity scores, this approach conflates a
true continuous variable (degree of suffer-true continuous variable (degree of suffer-
ing from depression) with a pseudo-linearing from depression) with a pseudo-linear
one (Parkerone (Parker et alet al, 1997). In addition, most, 1997). In addition, most
clinical trials have not considered whatclinical trials have not considered what
proportion of improvement is due toproportion of improvement is due to
natural (untreated) remission, because fewnatural (untreated) remission, because few
studies leave cohorts completely untreated.studies leave cohorts completely untreated.
Of the handful of studies that have used aOf the handful of studies that have used a
waiting-list arm (mostly psychotherapywaiting-list arm (mostly psychotherapy
studies), the typical rate ofstudies), the typical rate of spontaneousspontaneous
remission of major depression (many stu-remission of major depression (many stu-
dies allow waiting-list patients to receivedies allow waiting-list patients to receive
naturalistic treatment) is approximatelynaturalistic treatment) is approximately
10% per month, with a mean episode10% per month, with a mean episode
duration of 6 months (Posternak & Miller,duration of 6 months (Posternak & Miller,
2001). This indicates that although both drug2001). This indicates that although both drug
and placebo arms offer substantial benefit,and placebo arms offer substantial benefit,
some early remission would have occurredsome early remission would have occurred
without intervention, particularly if patientswithout intervention, particularly if patients
were recruited at peak depression severity.were recruited at peak depression severity.
When trying to measure initial onset ofWhen trying to measure initial onset of
therapeutic action, a major methodologicaltherapeutic action, a major methodological
problem is finding an appropriate outcomeproblem is finding an appropriate outcome
measure (Leonmeasure (Leon et alet al, 2001). An ideal, 2001). An ideal
measure should have inherent linearitymeasure should have inherent linearity
across a broad range of psychopathologicalacross a broad range of psychopathological
domains which translates into sensitivity todomains which translates into sensitivity to
early change (Maierearly change (Maier et alet al, 1988). This, 1988). This
hypothetical tool should also be appliedhypothetical tool should also be applied
early and frequently to increase temporalearly and frequently to increase temporal
resolution. The last point requires clarifica-resolution. The last point requires clarifica-
tion. If there is even a possibility that anti-tion. If there is even a possibility that anti-
depressants begin to workdepressants begin to work withinwithin 2 weeks,2 weeks,
how could this be detected if the firsthow could this be detected if the first
measurement takes place 2 weeks aftermeasurement takes place 2 weeks after
starting the drug? The same argument holdsstarting the drug? The same argument holds
for 1 week or even 1 day. So although therefor 1 week or even 1 day. So although there
are definite disadvantages of measuringare definite disadvantages of measuring
mood too often in a long-term study, if amood too often in a long-term study, if a
study is concerned with early onset of ac-study is concerned with early onset of ac-
tion, measurement at 1 week (currently antion, measurement at 1 week (currently an
unusually prompt measurement) is cer-unusually prompt measurement) is cer-
tainly too late. This approach is not new.tainly too late. This approach is not new.
Studies of the response to electroconvulsiveStudies of the response to electroconvulsive
therapy (ECT) often measure mood aftertherapy (ECT) often measure mood after
each application two or even three timeseach application two or even three times
weekly (Husainweekly (Husain et alet al, 2004). As an aside,, 2004). As an aside,
it would be interesting to see how muchit would be interesting to see how much
of the variance in rapidity of onset of actionof the variance in rapidity of onset of action
of ECT compared with pharmacotherapyof ECT compared with pharmacotherapy
or psychotherapy is explained by differ-or psychotherapy is explained by differ-
ences in the frequency of assessment alone.ences in the frequency of assessment alone.
Scales designed for regular daily mood rat-Scales designed for regular daily mood rat-
ings have already appeared in other medicalings have already appeared in other medical
specialties (Petersspecialties (Peters et alet al, 2000). Although, 2000). Although
these have often been in the form of simplethese have often been in the form of simple
visual analogue scales or mood diaries, onevisual analogue scales or mood diaries, one
group has used factor analysis to develop agroup has used factor analysis to develop a
daily mood scale for depression (Parker &daily mood scale for depression (Parker &
Roy, 2003). Daily or even twice dailyRoy, 2003). Daily or even twice daily
manual or electronic mood diaries,manual or electronic mood diaries,
although currently unfashionable, appearalthough currently unfashionable, appear
105105
BRITISH JOURNAL OF PSYCHIATRYBRITISH JOURNAL OF PSYCHIATRY (2006), 188, 105^106(2006), 188, 105^106 EDITORIALEDITORIAL
Two-week delay in onset of actionTwo-w eek del ay in onset of action
of antidepressants: new evidenceof antidepressants: new evidence
ALE X J. MI TCHELLALE X J. MITCHELL

MITCHELLMITCHELL
to be easy to use and reliable (Sherliker &to be easy to use and reliable (Sherliker &
Steptoe, 2000). Subjective self-rating meth-Steptoe, 2000). Subjective self-rating meth-
ods are an equally important and powerfulods are an equally important and powerful
method of assessing first onset, and providemethod of assessing first onset, and provide
different information to classical objectivedifferent information to classical objective
scales when compared head to headscales when compared head to head
(Moller(Moller et alet al, 1996)., 1996).
LARGE-SCALE STUDIESLARGE-SCALE STUDIES
Some clinicians will remain unconvinced bySome clinicians will remain unconvinced by
evidence extracted from multiple smallevidence extracted from multiple small
trials, albeit in the form of a meta-analysistrials, albeit in the form of a meta-analysis
(Posternak & Zimmerman, 2005). Support(Posternak & Zimmerman, 2005). Support
for this argument would be strengthened byfor this argument would be strengthened by
at least one really large study with measuresat least one really large study with measures
providing sufficient resolution in time.providing sufficient resolution in time.
Three such studies have in fact been pub-Three such studies have in fact been pub-
lished involving 429 (Stassenlished involving 429 (Stassen et alet al, 1996),, 1996),
1277 (Stassen1277 (Stassen et alet al, 1997) and 369 patients, 1997) and 369 patients
(Parker(Parker et alet al, 2000) respectively. The results, 2000) respectively. The results
mirror those of the meta-analysis men-mirror those of the meta-analysis men-
tioned above. The Zurich group used dailytioned above. The Zurich group used daily
depression ratings on the HRSD and Zungdepression ratings on the HRSD and Zung
Self-Rating Depression Scale. They foundSelf-Rating Depression Scale. They found
that regardless of which antidepressant thethat regardless of which antidepressant the
patient was taking, there was a measurablepatient was taking, there was a measurable
early effect on day 1. By day 3, 20% of pa-early effect on day 1. By day 3, 20% of pa-
tients had shown some improvement, andtients had shown some improvement, and
by day 7 50% had improved. Furthermore,by day 7 50% had improved. Furthermore,
90% of those who showed any response90% of those who showed any response
during the first 3 weeks went on to becomeduring the first 3 weeks went on to become
full responders. Drug–placebo differencesfull responders. Drug–placebo differences
(where apparent) could be detected as early(where apparent) could be detected as early
as day 5. In the third study, Parker’s groupas day 5. In the third study, Parker’s group
examined responders and non-respondersexamined responders and non-responders
treated by 27 Australian and New Zealandtreated by 27 Australian and New Zealand
psychiatrists (Parkerpsychiatrists (Parker et alet al, 2000). Patients, 2000). Patients
were requested to complete a self-reportwere requested to complete a self-report
mood rating every third day. All patientsmood rating every third day. All patients
showed a decrease in depression (and anxiety)showed a decrease in depression (and anxiety)
within the first 3 days, but with little furtherwithin the first 3 days, but with little further
improvement in non-responders from days 4improvement in non-responders from days 4
to 6. Again, early improvement within 1 weekto 6. Again, early improvement within 1 week
was a strong predictor of responder status.was a strong predictor of responder status.
Given these findings, it is pertinent toGiven these findings, it is pertinent to
ask why the view that antidepressants haveask why the view that antidepressants have
a delayed onset of action is so commonlya delayed onset of action is so commonly
held. There are two possibilities. The firstheld. There are two possibilities. The first
is imprecision concerning the use of theis imprecision concerning the use of the
term ‘onset of action’ when we really meanterm ‘onset of action’ when we really mean
‘time to substantial remission’. This is‘time to substantial remission’. This is
understandable if we assume that patientsunderstandable if we assume that patients
want to know when they will feel muchwant to know when they will feel much
improved rather than when they will beginimproved rather than when they will begin
to feel a little better. The second is a failureto feel a little better. The second is a failure
to distinguish initial therapeutic benefit,to distinguish initial therapeutic benefit,
which occurs within days of starting anwhich occurs within days of starting an
antidepressant, from the concept of drugantidepressant, from the concept of drug
v.v. placebo separation, which accrues moreplacebo separation, which accrues more
slowly. In the Posternak and Zimmermanslowly. In the Posternak and Zimmerman
meta-analysis of antidepressantmeta-analysis of antidepressant v.v. placeboplacebo
responses, the active arm showed modestresponses, the active arm showed modest
but definite early efficacy, but this wasbut definite early efficacy, but this was
difficult to separate from a significantdifficult to separate from a significant
placebo response until a week or more ofplacebo response until a week or more of
cumulative benefit had occurred. Thesecumulative benefit had occurred. These
limitations are amplified by relying on rela-limitations are amplified by relying on rela-
tively blunt instruments applied infre-tively blunt instruments applied infre-
quently by observers. The implication forquently by observers. The implication for
research is that sensitive daily measuresresearch is that sensitive daily measures
are required to elucidate whether theseare required to elucidate whether these
early patterns of response are dependentearly patterns of response are dependent
on biological correlates of antidepressanton biological correlates of antidepressant
or placebo treatment (Maybergor placebo treatment (Mayberg et alet al, 2002)., 2002).
The implication for clinical practice is thatThe implication for clinical practice is that
current evidence suggests it would be morecurrent evidence suggests it would be more
accurate to say to patients that in 90% ofaccurate to say to patients that in 90% of
cases substantial improvement occurs with-cases substantial improvement occurs with-
in the first 2 weeks, but that benefit con-in the first 2 weeks, but that benefit con-
tinues to build up over several weeks. (Intinues to build up over several weeks. (In
the meta-analysis by Posternak & Zimmer-the meta-analysis by Posternak & Zimmer-
man (2005), 60 out of 66 study cohorts onman (2005), 60 out of 66 study cohorts on
active medication showed a reduction inactive medication showed a reduction in
HRSD score of 50% or more within 2 weeks.)HRSD score of 50% or more within 2 weeks.)
In those who have shown no response by 2In those who have shown no response by 2
weeks there appears to be a law of dimin-weeks there appears to be a law of dimin-
ishing returns, which suggests that it mayishing returns, which suggests that it may
be pertinent to re-examine another com-be pertinent to re-examine another com-
monly quoted recommendation – that anmonly quoted recommendation – that an
antidepressant trial must be at least 6 to 8antidepressant trial must be at least 6 to 8
weeks before switching drugs.weeks before switching drugs.
REFE RENCESREFERENCES
Anderson, I. M., Nutt, D. J. & Deakin, J. F.W.Anderson,I.M.,Nutt,D.J.&Deakin,J.F.W. (2000)(2000)
Evidence-based guidelines for treating depressiveEvidence-based guidelines for treating depressive
disorders with antidepressants: a revision of the1993disorders with antidepressants: a revision of the 1993
British Association for Psychopharmacology guidelines.British Association fo r Psychophar mac ology guide lines.
Journal of PsychopharmacologyJournal of Psychopharmacology,, 1414, 3^20., 3^20.
Blier, P.Blier, P. (2003)(2003) The pharmacology of putative early-The pharmacology of putative early-
onset antidepressant strategies.onset antidepressant strategies. EuropeanEuropean
NeuropsychopharmacologyNeuropsychopharmacology,, 1313, 57^66., 57^66.
Dew, M. A., Reynolds, C. F., Mulsant B.,Dew, M. A., Reynolds, C. F., Mulsant B., et alet al (2001)(2001)
Initial recovery patterns may predict which maintenanceInitial recovery patterns may predict which maintenance
therapies for depression will keep older adults well.the rapi es for depr ession wi ll keep ol de r adults we ll.
Journal of Affective DisordersJournal of Affective Disorders,, 6565,155 ^166.,155^166.
Garfield, S., Francis, S. A. & Smith, F. J.Garfield, S., Francis, S. A. & Smith, F. J. (2004)(2004)
Building concordant relationships with patients startingBuild i ng c oncor dant re lat i onsh i ps wit h pat i en ts starti ng
antidepressant medication.antidepressant medication. Patient Education andPatient Education and
Counsel ingCounseling,, 5555, 241^246., 241^24 6.
Husain, M. M., Rush, A. J., Fink, M.,Husain, M. M., Rush, A. J., Fink, M., et alet al (2004)(2004)
Speed of response and remiss ion in major depressi veSpeed of response and remi ss ion i n major depressive
disorder with acute electroconvulsive therapy (ECT): adisorder with acute electroconvulsive therapy (ECT): a
Consortium for Research in ECT (CORE) report.Consortium for Research in ECT (CORE) report.
Journal of Clinical PsychiatryJournal of Clinical Psychiatry,, 6565,485^491., 4 85^491.
Leon, A. C., Blier, P., Culpepper, L.,Leon, A. C., Blier, P., Culpepper, L., et alet al (2001)(2001) AnAn
ideal trial to test differential onset of antidepressantideal trial to test differential onset of antidepressant
effect.effect. Journal of Clinical PsychiatryJournal of Clinical Psychiatry,, 6262,34^36.,34^36.
Maier, W., Philipp, M., Heuser, I.,Maier,W.,Philipp,M.,Heuser,I.,et alet al (19 8 8)(19 8 8)
Improving depression severity assessment. 1 . Reliability ,Improving depression severity assessment. 1 . Reliability,
internal validity and sensitivity to change of threeinternal validity and sensitivity to change of three
observer depression scales.observer depression scales. Jour nal of Psychi atri cJour na l of Psy chiatri c
ResearchResearch,, 2222,3^12., 3^12.
Mayberg, H. S., Silva, J. A., Brannan, S. K.,Mayberg, H. S., Silva, J. A., Brannan, S. K., et alet al
(2002)(2002)
The functional neuroanatomy of the placeboThe functional neuroanatomy of the placebo
effect.effect. American Journal of PsychiatryAmerican Journal of Psychiatry,, 159159 , 7 28^737., 728^737.
Moller,H.J.,Muller,H.& Volz,H.P.Moller,H.J.,Muller,H. &Volz,H.P. (1996)(19 9 6) How t oHow to
assess the onset of antidepressant effect: comparison ofassess the onset of antidepressant effect: comparison of
global ratings and findings based on depression scales.global ratings and findings based on depression scales.
PharmacopsychiatryPharmacopsychiatry,, 2929, 57^62., 57^62.
National Institute for Clinical ExcellenceNational Institute for Clinical Excellence (200 4)(2004)
Depression: Management of Depression in Primary andDepression: Management of Depression in Primary and
Secondary Care.Secondary Car e . Cli nical G uide line 23 . London: N ICE.C lin ical Gu i deline 23. London: N ICE.
www.nice.org.uk/pdf/C G023quickrefgu ide.pdfwww.nice.org.uk/pdf/C G023qui ckrefguide.pdf
Parker, G. & Roy, K.Parker, G. & Roy, K. (2003)(2003) The de velopme nt of a six-The dev e l op me nt of a si x-
item daily self-report measure assessing identifieditem daily self-report measure assessing identified
depressive domai ns .depressive domains. Journal of Affective DisordersJournal of Affective Disorders,, 7373,,
289^294.289^294.
Parker, G., Roussos, J., HadziPavlovic, D.,Parker, G., Roussos, J., HadziPavlovic, D., et alet al (19 9 7 )(19 9 7)
Plum bi ng the depths: some problems in quanti fying de-Plumbing the depths: some problems in quantifying de-
pression severity.pression severity. Journal of Affective DisordersJournal of Affective Disorders,, 4242,49^58.,49^58.
Parker,G.,Roy,K.,Menkes,D.B.,Parker, G., Roy, K., Menkes, D. B., et alet al (2000)(2000) HowHow
long does ittake for antidepressant therapies to act?long does it take for antidepressant therapies to act?
Austra lian and New Zeal and Jour nal of Psychi atryAustra lian and New Zealand Jour nal of Psychiatry,, 3434,,
65^70.65^ 70.
Peters, M. L., Sorbi, M. J., Kruise, D. A.,Peters, M. L., Sorbi, M. J., Kruise, D. A., et alet al (2000)(2000)
Electronic diary assessment of pain, disability andElectronic diary assessment of pain, disability and
psychological adaptation in pat ients di fferi ng in durati onpsychologi cal adaptat ion in patients d ifferi ng in duration
of pain .of pain . PainPain,, 8484,181^192.,181^192.
Posternak, M. A. & Miller, I.Posternak, M. A. & Miller, I. (2001)(2001) Untreated short-Untreated short-
term course of major depression: a meta-analysis ofterm course of major depression: a meta-analysis of
outcomes from stu d ies u si ng wait-list contro l groups.outcomes from studies us ing w a it -list contr o l groups.
Journal of Affective DisordersJournal of Affective Disorders,, 6666,139^146.,139^146.
Posternak, M. A. & Zimmerman, M.Posternak, M. A. & Zimmerman, M. (2005)(2005) Is thereIs t here
a delay in the antidepressant effect? A meta-analysis.a delay in the antidepressant effect? A meta-analysis.
Journal of Clinical PsychiatryJournal of Clinical Psychiatry,, 6666,14 8^158.,148^158.
Quitkin, F. M., Rabkin, J. D., Markowitz, J. M.,Quitkin , F. M., Rabkin, J. D., Markowitz, J. M., et alet al
(19 87 )(19 8 7)
Use of pattern analysis to identify true drugUse of patt ern analysis to identify t rue drug
response: a rep licat i on.response: a repl ication. Archives of General PsychiatryArchives of General Psychiatry,,
4444, 259^264., 259^264.
Sherliker, L. & Steptoe, A.Sherliker, L. & Steptoe, A. (2000)(2000) Coping with newCoping w ith new
treatments for cancer: a feasibility study of daily diarytreatments for cancer: a feasibility study of daily diary
measures.measures. Patient Education and CounselingPati ent Education and Counsel i ng,, 4040,11^19.,11^19.
Stassen, H. H., Angst, J. & DeliniStula, A.Stassen, H. H., Angst, J. & DeliniStula, A. (1996)(19 9 6)
Delayed onset of action of antidepressant drugs? SurveyDe l a yed onset of acti on of ant i dep r essant dru gs? Survey
of results of Z urich meta-analyses.of results of Zurich meta-analyses. PharmacopsychiatryPharmacopsychiatry,,
2929, 87^96., 87^96.
Stassen, H. H., Angst, J. & DeliniStula, A.Stassen, H. H., Angst, J. & DeliniStula, A. (1997)(19 9 7 )
Delayed onset of action of antidepressant drugs? SurveyDe l a yed onset of acti on of ant i dep r essant dru gs? Survey
of recent resul ts.of recent resul ts. European PsychiatryEur opean Psych i atry,, 1212,166^176.,166^176.
Szegedi, A., Muller, M. J., Anghelescu, I.,Szegedi, A., Muller, M. J., Anghelescu, I., et alet al (2003)(2003)
Early improvement under mirtazapine and paroxetineEarly improvement under mirtazapine and paroxetine
predicts later stable response and remission with highpredicts later stable response and remission with high
sensitivity in patients with major depression.sensitivity in patients with major depression. Jou rnal ofJournal of
Clinical PsychiatryClinical Psychiatry ,, 6464, 413^420., 413^420.
106106
AL EX J. MITCH ELL, MR CPsy ch , Departmen t of L iaison Psy ch iatry, L e icester General Hospital , Lei cester, U KALEX J. MIT CHELL, MRCPsych, Department of Liaison Psychiatry, Leicester General Hospital, Leicester , U K
Correspondence: Dr Alex J.Mitchell,Departmentof Liaison Psychiatry,Brandon Unit,Leicester GeneralCorrespondence:Dr Alex J.Mitchell,D ep artmentof Liaison Psychiatry,Brandon Unit,Leicester General
Hospital,Leicester LE5 4PW,UK.E-mail: alex.mitchellHospital,Leicester LE5 4PW,UK.E-mail: alex.mitchell@@leicspart.nhs.ukleicspart.nhs.uk
(First received 31March 2005, final revision 1June 2005, accepted14 June 200 5)(First received 31March 2005, final revision 1 June 2005, accepted 14 June 2005)

10.1192/bjp.bp.105.011692Access the most recent version at DOI:
2006, 188:105-106.BJP
ALEX J. MITCHELL
evidence
Two-week delay in onset of action of antidepressants: new
References
http://bjp.rcpsych.org/content/188/2/105#BIBL
This article cites 20 articles, 2 of which you can access for free at:
permissions
Reprints/
permissions@rcpsych.ac.ukwrite to
To obtain reprints or permission to reproduce material from this paper, please
to this article at
You can respond
/letters/submit/bjprcpsych;188/2/105
from
Downloaded
The Royal College of PsychiatristsPublished by
on December 31, 2015http://bjp.rcpsych.org/
http://bjp.rcpsych.org/site/subscriptions/
go to: The British Journal of PsychiatryTo subscribe to