Sie sind auf Seite 1von 6

Übersichtsarbeit · Review Article

Verhaltenstherapie 2017;27:265–270 Published online: July 18, 2017

DOI: 10.1159/000460826

The Potential Role of Iatrogenic Comorbidity

in the Interaction between Pharmacotherapy and
Psychotherapy in Anxiety Disorders
Giovanni A. Fava a,b Giada Benasi c Fiammetta Cosci d
Department of Psychology, University of Bologna, Bologna, Italy;
Department of Psychiatry, University at Buffalo, Buffalo, NY, USA;
c Center for Integrative Medicine, Weill Cornell Medical College, New York, NY, USA;
Department of Health Sciences, University of Florence, Firenze, Italy

Keywords Schlüsselwörter
Iatrogenic comorbidity · Psychotherapy · Antidepressant drugs · Iatrogene Komorbidität · Psychotherapie · Antidepressiva ·
Benzodiazepines · Anxiety disorders Benzodiazepine · Angststörungen

Summary Zusammenfassung
The combination of pharmacotherapy and psychotherapy in Die Kombination von Pharmakotherapie mit Psychotherapie bei
the setting of anxiety disorders is often viewed as a potential Angststörungen wird oft als mögliche Ursache einer Verstär-
source of augmentation of clinical effects, with little attention kung klinischer Effekte angesehen, ohne das mögliche Auftre-
paid to the potential occurrence of negative events. In most of ten negativer Ereignisse zu beachten. In den meisten Studien
the studies, however, if benefits ensued, they were modest and waren die Vorteile, die daraus hervorgingen, jedoch nur gering-
likely to fade. Further, 4 high-quality and well-designed individ- fügig und ließen meist wieder nach. Des Weiteren schlagen 4
ual studies suggest that the addition of a benzodiazepine or an qualitativ hochwertige und gut angelegte Einzelstudien vor,
antidepressant to cognitive behavioral treatment of anxiety dis- dass die zusätzliche Gabe eines Benzodiazepins oder eines An-
orders could be detrimental compared to placebo at follow-up. tidepressivums zu einer kognitiven Verhaltenstherapie bei
The aim of this review was to outline a novel hypothesis, which Angststörungen zu einer Verschlechterung im Vergleich zu Pla-
needs to be adequately tested but may shed some new light on cebo beim Follow-up führt. Ziel dieses Übersichtsartikels war
this interaction. Any type of psychotropic drug treatment, par- die Darstellung einer neuen Hypothese, die zwar noch entspre-
ticularly after long-term use, may increase the risk of experienc- chend getestet werden muss, die aber ein neues Licht auf diese
ing additional psychopathological problems that do not neces- Interaktion werfen könnte. Jede Art von Behandlung mit einem
sarily subside with discontinuation of the drug or of modifying psychotropen Medikament kann besonders nach längerfristiger
responsiveness to subsequent treatments. The changes are Einnahme zu einer Erhöhung des Risikos für das Erfahren wei-
persistent and not limited to a short phase, such as in the case terer psychopathologischer Probleme, die nicht notwendiger-
of withdrawal reactions, and cannot be subsumed under the weise mit dem Absetzen des Medikamentes abklingen, oder für
generic rubrics of adverse events or side effects. The term ‘iat- ein verändertes Ansprechen auf weitere Behandlungen führen.
rogenic comorbidity’ refers to unfavorable modifications in the Diese Veränderungen sind dauerhaft und nicht wie im Falle von
course, characteristics, and responsiveness of an illness that Entzugsreaktionen auf eine kurze Phase begrenzt, und sie kön-
may be related to treatments that were administered previous- nen auch nicht unter den Oberbegriffen «unerwünschte Er-
ly. The likelihood of iatrogenic comorbidity needs to be consid- eignisse» oder «Nebeneffekte» zusammengefasst werden. Der
ered in clinical practice: The concurrent use of pharmacothera- Begriff «iatrogene Komorbidität» bezieht sich auf ungünstige
py and psychotherapy may yield advantages in the short term, Modifikationen im Verlauf, in den Eigenschaften und im An-
but its costs at some later point in time may largely outweigh sprechen einer Krankheit, die mit den vorher angewendeten
such benefits. Behandlungsmethoden in Zusammenhang stehen. Die Wahr-
© 2017 S. Karger GmbH, Freiburg scheinlichkeit des Auftretens einer iatrogenen Begleiterkran-
kung muss in der klinischen Praxis berücksichtigt werden: Die
gleichzeitige Anwendung von Pharmakotherapie und Psycho-
therapie kann kurzfristig Fortschritte erzielen, ihre zu einem
späteren Zeitpunkt auftretenden Kosten können diesen Nutzen
jedoch weit überwiegen.

© 2017 S. Karger GmbH, Freiburg Prof. Dr. Giovanni A. Fava

Department of Psychology
Fax +49 761 4 52 07 14 University of Bologna Accessible online at: Viale Berti Pichat 5, 40127 Bologna, Italy
Introduction different forms of anxiety disorders and yielded a detrimental in-
teraction. These studies represent only part of the randomized con-
The combination of pharmacotherapy and psychotherapy in the trolled trials (RCT) that were performed with such a combination
setting of anxiety disorders is often viewed as a potential source of in anxiety disorders. In most of the cases benefits did not ensue or
augmentation of clinical effects, with little attention paid to the po- were modest and faded after treatment was completed [Forand et
tential occurrence of negative events. Forand et al. [2013] have re- al., 2013].
viewed the complex literature that is concerned with combining The first of these studies is a cross-national trial conducted by
medication and psychotherapy in the treatment of major mental Marks et al. [1993] in 2 centers, London and Toronto. In this re-
disorders. What emerges from their analysis is the fact that a com- search, 154 patients with a diagnosis of chronic panic disorder with
bined treatment is not necessarily a better treatment and that there agoraphobia were randomized to receive alprazolam and exposure
is the need of proper assessment tools for evaluating the pros and (AE group), alprazolam and relaxation (AR group), placebo and
cons of each approach. exposure (PE group), or placebo and relaxation (PR group), the lat-
Uhlenhuth et al. [1969] examined the combinations of pharma- ter being a form of psychological placebo. The study was double-
cotherapy and psychotherapy in psychiatric disorders. They out- blind for drug and single-blind for psychological treatment. Dur-
lined 4 models of interaction: (a) addition (the effects of 2 interac- ing a total period of 10 months, the patients had individual outpa-
tions combined equals the sum of their individual effects), (b) po- tient treatment for 8 weeks, taper of medication from week 8 to 16,
tentiation (the effect of 2 interventions combined is greater than and then follow-up to week 43. At the end of the treatment, both
the sum of their individual effects), (c) inhibition (the effect of 2 alprazolam and exposure were effective on non-panic measures,
interventions combined is less than each individual effect), (d) re- with exposure having twice the effect size of alprazolam. Interest-
ciprocation (the effect of the 2 interventions combined equals the ingly, during tapering and follow-up, PE patients maintained and
individual effect of the more potent intervention). Most of the slightly increased their gains, whereas patients receiving alpra-
studies that were reviewed were compatible with the reciprocal zolam (whether in combination with exposure or relaxation) lost
concept of interaction, and only a few, with an additive effect. theirs, with AE being significantly worse than PE. This pattern is
Subsequent research has substantiated these conclusions in well represented by rates of relapse. Indeed, according to an index
mood and anxiety disorders. In some cases, psychotherapy and of global improvement, the percentages of patients who improved
pharmacotherapy have partially additive effects, as was found to be and did not relapse at the end of the study (week 43) were 36% in
the case with the sequential model in depression which consists of AE, 62% in PE, 29% in AR, and 18% in PR.
the use of pharmacotherapy in the acute phase and of psychothera- Subsequently, Barlow et al. [2000] conducted a 5-arm multi-
py in its residual phase [Guidi et al., 2016]. In most cases, the ef- center double-blind trial in which 312 patients affected by panic
fects of the 2 interventions combined were equivalent to the effect disorder with or without mild agoraphobia were treated with cog-
of 1 intervention, particularly when follow-up evaluations were nitive-behavioral therapy (CBT) alone, imipramine and medical
performed [Friedman and Thase, 2009; Forand et al., 2013; management, CBT and imipramine, placebo and medical manage-
Biesheuvel-Leliefeld et al., 2015]. There are, however, some high- ment, or CBT and placebo. The patients were assessed at the end of
quality and well-designed individual studies in anxiety disorders both an acute and a maintenance phase of 3 and 6 months, respec-
that suggested that the addition of a benzodiazepine or an antide- tively, and after 6 months of follow-up. At the end of the treatment,
pressant to cognitive behavioral treatment could be detrimental all active groups were significantly better than the placebo group.
compared to placebo at follow-up [Marks et al., 1993; Barlow et al., However, after treatment discontinuation, only CBT alone and the
2000; Haug et al., 2003; Nordahl et al., 2016]. We will briefly de- combination of CBT and placebo were significantly superior to
scribe these studies and we will then discuss a recent conceptual- placebo, whereas imipramine-treated patients, with or without the
ization, subsumed under the rubric of iatrogenic comorbidity, addition of CBT, were significantly worse than those who received
which needs to be adequately tested but may shed some new light CBT, either alone or in combination with placebo. In particular,
on the interaction between pharmacotherapy and psychotherapy. according to the intention-to-treat analysis of the Panic Disorder
Finally, we will provide some cautionary clinical notes on the con- Severity Scale (PDSS) [Shear et al., 1997], the response rates at fol-
current use of pharmacotherapy in the psychotherapeutic setting low-up were 41% for CBT and placebo, 32.4% for CBT alone, 25%
of anxiety disorders. for CBT and imipramine, 19.7% for imipramine, and 9.1% for
In the study of Haug et al. [2003], 375 patients affected by gen-
Description of Studies eralized social phobia were treated for 24 weeks with sertraline and
exposure (SE group), sertraline and general medical care, placebo
In the literature, we can find 4 randomized placebo-controlled and exposure (PE group), or placebo and general medical care. Ex-
trials in which a combination of medication (benzodiazepine posure was given in 8 sessions for the first 12 weeks of treatment.
[Marks et al., 1993] or antidepressant [Barlow et al., 2000; Haug et The patients were evaluated at post-treatment (week 24) and at
al., 2003; Nordahl et al., 2016]) and cognitive behavioral treatment follow-up (week 52). At post-treatment, only sertraline and the
was compared with each treatment alone in the management of combination of sertraline and exposure were significantly superior

266 Verhaltenstherapie 2017;27:265–270 Fava/Benasi/Cosci

to placebo. At follow-up, patients in the PE and in the placebo after their discontinuation. Any type of psychotropic drug treat-
groups had a significant improvement on several scales assessing ment, particularly after long-term use, may increase the risk of ex-
social phobia and daily functioning, whereas patients treated with periencing additional psychopathological problems that do not
sertraline, either alone or in combination with exposure, had a sig- necessarily subside with discontinuation of the drug or of modify-
nificant deterioration as compared to treatment with both PE and ing responsiveness to subsequent treatments [Fava et al., 2014;
placebo. Offidani et al., 2014; Fava et al., 2016]. The term ‘iatrogenic comor-
Most recently, Nordahl et al. [2016] conducted a comparative bidity’ refers to unfavorable modifications in the course, character-
trial in which 102 patients with a primary diagnosis of social anxi- istics, and responsiveness of an illness that may be related to treat-
ety disorder with or without avoidant personality disorder were al- ments administered previously [Fava et al., 2014; Offidani et al.,
located to 4 different conditions: paroxetine and clinical manage- 2014; Fava et al., 2016]. Such vulnerabilities may occur during
ment, cognitive therapy (CT), paroxetine and CT, or placebo and treatment administration and/or manifest themselves after its dis-
clinical management. The study was double-blind for the medica- continuation. The changes can be persistent and not limited to a
tion and placebo groups. Treatment with paroxetine and placebo short phase, such as in the case of withdrawal reactions [Fava et al.,
lasted for 26 weeks, with tapering that commenced at week 23. All 2015a], and cannot be subsumed under the generic rubrics of ad-
patients were assessed at post-treatment (week 12) and at the 12- verse events or side effects.
month follow-up. At post-treatment, the results with CT did not We will illustrate some examples that are concerned with the
differ significantly from those with the combination of CT and use of antidepressant drugs; yet, similar phenomena have been de-
paroxetine. However, while the CT group improved significantly scribed with other types of psychotropic drugs [Fava et al., 2014].
more than the groups treated with both paroxetine and placebo, no The best-known manifestation involves the phenomenon of
differences were observed between the results achieved with the switching into mania or hypomania; this is a well-known compli-
combination of CT and paroxetine and with paroxetine alone. At cation of the use of antidepressant drugs (ADs) in bipolar disorder
follow-up, CT maintained its benefits and was significantly better [Tondo et al., 2010]. The occurrence of mania and other forms of
than the combination of CT plus paroxetine on the Liebowitz So- behavioral activation in unipolar depression may unveil unrecog-
cial Anxiety Scale (LSAS) [Liebowitz, 1987]. The groups treated nized bipolar illness or may be drug induced, since they may also
with paroxetine (alone or in combination with CT) were no longer occur in allegedly unipolar patients. A systematic review and me-
different from those treated with placebo. Furthermore, the recov- ta-analysis concerned with excessive mood elevation and behavio-
ery rates were greater with CT both at post-treatment and at fol- ral activation of children and adolescents disclosed that the rates
low-up. Specifically, the share of patients who achieved recovery at of excessive arousal/activation with antidepressants were very high
the 12-month follow-up was 68% in the CT group, 40% in the both in anxiety (13.8%) and in depression (9.8%), and much lower
combination group, 24% in the paroxetine group, and 4% in the with placebos (5.2% vs. 1.1%, respectively) [Offidani et al., 2013a].
placebo group. Furthermore, in almost half of the pediatric patients who partici-
The results of all these studies are consistent in challenging the pate in antidepressant trials, such reactions occur in the absence of
idea that the combination of pharmacotherapy and psychotherapy a family history of bipolar disorder [Joseph et al., 2009]. Hence,
may improve the outcome in the treatment of anxiety disorders. the risk of developing behavioral activation may occur also with
Even when drugs may be more effective than placebo, and the com- the use of ADs in anxiety disorders, particularly in younger pa-
bination of drugs and cognitive behavioral treatment may have tients, and the symptoms do not necessarily subside upon discon-
some advantages over psychotherapy alone during treatment, only tinuation of the AD [Joseph et al., 2009; Tondo et al., 2010; Offida-
benefits obtained with psychotherapy are maintained and slightly ni et al., 2013a]. Such findings are in line with the contraindica-
improved thereafter. On the contrary, the effect of medications tions that occur with the use of ADs in bipolar disorder [Tondo et
strictly depends on their administration and is lost after treatment al., 2010].
discontinuation. Adding a benzodiazepine or an antidepressant to Further, several side effects of ADs are transient and may dis-
cognitive-behavioral treatment was found to be associated with appear after a few weeks following treatment initiation, but po-
greater rates of relapse at follow-up. This means not only that the tentially serious adverse events may persist or ensue later. They
combination of pharmacotherapy and psychotherapy is ineffective encompass gastrointestinal symptoms (e.g., nausea, diarrhea, gas-
in the long-term treatment of anxiety disorders, but also that its ef- tric bleeding, dyspepsia), hepatotoxicity, weight gain and meta-
fect could be detrimental. These results are rather important if we bolic abnormalities, cardiovascular disturbances (e.g., heart rate,
consider that anxiety disorders are often a disabling and chronic QT interval prolongation, hypertension, orthostatic hypoten-
condition, necessitating a treatment that entails long-term efficacy. sion), genitourinary symptoms (e.g., urinary retention, inconti-
nence), sexual dysfunction, hyponatremia, osteoporosis and risk
of fractures, bleeding, central nervous system (CNS) disturbances
Iatrogenic Comorbidity (e.g., lowering of the seizure threshold, extrapyramidal side ef-
fects, cognitive disturbances), sweating, sleep disturbances, affec-
There are adverse events that are limited to the period of psy- tive disturbances (e.g., apathy, switches, paradoxical effects), oph-
chotropic drug administration and effects that may persist long thalmic manifestations (e.g., glaucoma, cataract), and hyperprol-

Iatrogenic Comorbidity Verhaltenstherapie 2017;27:265–270 267

actinemia [Carvalho et al., 2016]. At times, such adverse events The second manifestation of iatrogenic comorbidity involves
may persist after drug discontinuation, yielding iatrogenic co- persistent post-withdrawal symptoms after discontinuation of
morbidity. For instance, long-term use of antidepressant drugs SSRIs. Chouinard and Chouinard [2015] have recently provided a
such as selective serotonin reuptake inhibitors (SSRIs) may in- classification of symptoms that may ensue upon discontinuation of
duce weight gain, after an initial period characterized by reduced psychotropic drugs, with particular reference to SSRIs and seroto-
appetite, and the increased weight does not necessarily recede nin and noradrenalin reuptake inhibitors (SNRIs). They differenti-
upon AD discontinuation [Carvalho et al., 2016]. It has been sug- ated between withdrawal syndromes (i.e., symptoms occurring
gested that an increase in exposure to antidepressants via a multi- with a decrease or discontinuation of a drug that are new and not
tude of mechanisms may be a driving force for the obesity pan- part of the original illness), rebound syndromes (i.e., a rapid return
demic [Lee et al., 2016]. Similarly, the prevalence of sexual side of the original symptoms at a greater intensity than before treat-
effects can be as high as 50–70% among individuals taking SSRIs, ment), and post-withdrawal or tardive receptor sensitivity disor-
and such effects may persist even after AD discontinuation [Car- ders (i.e., return of the original illness at a greater intensity and/or
valho et al., 2016]. with additional features and/or symptoms related to emerging new
Negative effects may also occur as a result of psychotherapeutic disorders). Withdrawal symptoms following discontinuation of an-
treatment, whether due to technique, patient or therapist variables, tidepressant treatment have been described with any type of AD,
or inappropriate use [Barlow, 2010; Linden, 2013; Scott and Young, but particularly with SSRIs, venlafaxine, and duloxetine [Choui-
2016]. Side effects of psychotherapy are difficult to recognize be- nard and Chouinard, 2015; Fava et al., 2015b]. They have been
cause of the number of variables that are involved, including the generally defined as ‘discontinuation syndromes’, with the aim of
various stages of the psychotherapeutic process [Linden, 2013]. It avoiding any hint to the dependence potential of SSRIs that may
has been underscored how much research on the negative aspects affect marketing, but such definition is misleading since they are
of psychotherapy is insufficient [Scott and Young, 2016]. actually withdrawal reactions. The withdrawal syndrome is charac-
Two other manifestations of iatrogenic comorbidity related to terized by a broad range of somatic symptoms, such as headache,
the use of ADs are, however, less known and may have particular dizziness, fatigue, diminished appetite, sleep disturbances (i.e.,
importance in interpreting the interaction between pharmacother- vivid dreams and insomnia), somnolence, flu-like symptoms, nau-
apy and psychotherapy. sea and vomiting, and genital hypersensitivity. Less common phys-
One is concerned with the fact that, in some cases, long-term ical symptoms include myalgia, Parkinsonism, balance difficulties,
use of ADs may enhance the biochemical vulnerability to depres- and cardiac arrhythmias. Psychological symptoms may ensue as
sion and worsen its long-term outcome and symptomatic expres- well: agitation, anxiety, panic attacks, dysphoria, confusion, and
sion, decreasing both its likelihood of subsequent response to worsening of mood [Chouinard and Chouinard, 2015; Fava et al.,
pharmacological treatment and the duration of symptom-free pe- 2015b]. Symptoms typically appear within 3 days of stopping anti-
riods [Fava, 2003]. Clinical evidence has been found indicating depressant medication, even if this is done gradually, or initiating a
that, even though ADs are effective in treating depressive episodes, medication taper. They may be easily misinterpreted as a sign of
they are less efficacious in recurrent depression and in preventing impending relapse or a need of keeping the same dosage. Untreat-
relapse. In many cases, antidepressants have been described as in- ed symptoms may be mild and resolve spontaneously in 1–3 weeks;
ducing adverse events such as withdrawal symptoms at discontin- in other cases, they may persist for months or even years, leading
uation, onset of tolerance and resistance phenomena, and switch to what has been defined as persistent post-withdrawal disorder
and cycle acceleration in bipolar patients. Unfavorable long-term [Chouinard and Chouinard, 2015]. This latter form of iatrogenic
outcomes and paradoxical effects (depression inducing and symp- comorbidity has been described with different classes of CNS drugs
tomatic worsening) have also been reported [Fava, 2003]. All these or substances (e.g., protracted insomnia for alcohol and benzodiaz-
phenomena may be explained on the basis of the oppositional epine withdrawal, supersensitivity psychosis from antipsychotic
model of tolerance. Continued drug treatment may recruit pro- drug treatment, dysphoria after cocaine and amphetamine use).
cesses that oppose the initial acute effect of a drug. When the drug Post-withdrawal disorders may resemble rebound syndromes, but
treatment ends, these processes may operate unopposed, at least these disorders persist for at least 6 weeks, in contrast to rebound
for some time, and increase vulnerability to relapse [Fava and symptoms, and may include new illness features [Chouinard and
Offidani, 2011]. In a meta-analysis concerned with the sequential Chouinard, 2015]. Their prevalence is not known at the moment,
use of pharmacotherapy and psychotherapy in depression, pa- due to their very recent definition. The high prevalence of mental
tients randomized to psychotherapy who had their antidepres- disorders in the general population may also be an effect of the
sants tapered and discontinued were significantly less likely to ex- presence of disorders that are a consequence of previous pharma-
perience relapse/recurrence compared to patients with continua- cological treatments [Cosci et al., 2015]: iatrogenic comorbidity
tion of antidepressant medication [Guidi et al., 2016]. In an RCT that would require removal of drug treatment and is then inter-
of mindfulness-based cognitive therapy (MBCT) versus a waiting preted as a justification for new treatment. For instance, much of
list control group [Bakker et al., 2016], the patients who increased the refractoriness to treatment of anxious depression may be actu-
their positive affects the most were the ones who did not take an ally due to secondary post-withdrawal disorders that are secondary
AD. to the use of ADs in anxiety disorders [Fava and Tomba, 2014].

268 Verhaltenstherapie 2017;27:265–270 Fava/Benasi/Cosci

Reinterpreting Psychotherapy/Pharmacotherapy of poor outcomes from an index disorder without treatment, re-
Detrimental Effects sponsiveness to the treatment option, and vulnerability to the ad-
verse effects of treatment. A rational approach to treatment takes
Let us consider the psychotherapy/pharmacotherapy combina- into account the balance between potential benefits and adverse ef-
tion that resulted in detrimental effects in the 4 studies we synthe- fects applied to the individual patient [Fava et al., 2015c]. Unfortu-
sized [Marks et al., 1993; Barlow et al., 2000; Haug et al., 2003; Nor- nately, the prescribing clinician is driven by an overestimated con-
dahl et al., 2016] in light of these new insights we have discussed. sideration of potential benefits, paying little attention to the likeli-
All studies involved cognitive-behavioral treatment, albeit with hood of responsiveness and to potential vulnerabilities in relation
somewhat different protocols. In all 4 studies, the detrimental ef- to the adverse effects of treatment [Fava et al., 2015c]. Appraisal of
fects occurred during follow-up and not at the end of the study. adverse effects relies primarily on observational studies and data
This may have 2 different interpretations. Either the process of from routine clinical practice and may not emerge from RCTs, un-
learning during psychotherapy was decreased by the concurrent less these effects occur early in the course of treatment and are spe-
pharmacotherapy (e.g., exposure with the concurrent use of alpra- cifically investigated [Vanderbroucke and Psaty, 2008]. Further,
zolam may entail less enduring effects than with placebo) or iatro- detection of adverse events requires sensitive clinimetric indices
genic comorbidity (to be fully disclosed after drug discontinuation) that depart from conventional assessment strategies geared to
might have ensued. For instance, a patient who suffers from panic demonstrating efficacy against placebo [Fava and Bech, 2016; Bech,
disorder may be successfully treated with a medication but, when 2016; Piolanti et al., 2016]. Finally, the role of financial conflicts of
the drug is tapered and discontinued, may manifest withdrawal interest in shaping the literature appraisal is an increasing source
syndromes that do not subside in a few weeks but persist for of concern [Fava, 2016]. While the dependence potential of benzo-
months or years and/or a rebound syndrome (e.g., panic that is far diazepines is widely acknowledged, the clinician is less likely to be
worse than that characterizing the pretreatment phase) and/or a aware that the same, if not worse, phenomenon may occur with
new syndrome (e.g., depression) which he/she never experienced ADs [Chouinard and Chouinard, 2015; Fava et al., 2015b; Starcevic
before drug treatment. et al., 2015]. Discontinuation syndromes of SSRIs and SNRIs are
SSRIs [Belaise et al., 2014; Chouinard and Chouinard, 2015; actually withdrawal reactions [Chouinard and Chouinard, 2015;
Fava et al., 2015b], which were used in 2 studies, and alprazolam Fava et al., 2015b] and at least as bad as those that may occur with
[Fava, 1988; Cloos et al., 2015; Zito, 2015] have been specifically benzodiazepines. In a systematic review of the direct comparison
linked to iatrogenic comorbidity and addictive properties. In one between ADs and benzodiazepines in anxiety disorders, the superi-
of the studies [Marks et al., 1993], the effects of discontinuing al- ority of benzodiazepines, both in terms of efficacy and side effect
prazolam were specifically investigated [O’Sullivan et al., 1996]. profile, emerged [Offidani et al., 2013b]. This was in sharp contrast
Alprazolam patients deteriorated on anxiety, panic, depression, with meta-analyses and guidelines that were based on such meta-
and phobia, whereas serious withdrawal symptoms did not occur analyses [Cosci et al., 2015; Starcevic et al., 2015].
(probably due to slow tapering). The deterioration in patients Despite the fact that, in real-world clinical practice, most of the
treated with alprazolam persisted up to 6  months post-taper patients with anxiety disorders already assume some form of psy-
[O’Sullivan et al., 1996]. chotropic drug treatment, clinical psychologists have little familiar-
It is thus conceivable that the concurrent use of pharmacothera- ity with psychopharmacology and are substantially unaware of
py and psychotherapy may yield advantages in the short term, but subtle and yet pervasive potential effects of psychotropic drugs in
its costs in terms of iatrogenic comorbidity at some later point in their patients [Guidi and Fava, 2014]. It is a major flaw in their
time largely outweigh such benefits. However, this is simply a hy- clinical training that appears to occur around the world and which
pothesis that needs to be considered among various other potential requires urgent attention.
explanations, such as the possibility that medications may suppress The concept of iatrogenic comorbidity has heuristic value and
the activation of fear-related cognitions and/or may interfere with may alert the clinician to the potential occurrence of detrimental
the consolidation of fear-incongruent learning during exposure effects during treatment of anxiety disorders.
[Forand et al., 2013].

Disclosure Statement
Clinical Implications
No funding was received by the authors to carry out the present review.
G.A.F. and G.B. have no conflict of interest to declare. F.C. received in 2010 an
Richardson and Doster [2014] suggested the consideration of 3 award from Pfizer in the framework of the Global Research Awards for Nico-
dimensions in the process of evidence-based decision: baseline risk tine Dependence (GRAND).

Iatrogenic Comorbidity Verhaltenstherapie 2017;27:265–270 269

Bakker JM, Lieverse R, Gescwhind N, et al: The two sided Fava GA, Cosci F, Tomba E: Iatrogenic comorbidity in men- O’Sullivan GH, Swinson R, Kuch K, et al: Alprazolam with-
face of antidepressants. Psychother Psychosom 2016; tal health. Psychother Psychosom 2015a;84(suppl 1):22. drawal symptoms in agoraphobia with panic disorder:
85: 180–182. Fava GA, Gatti A, Belaise C, et al: Withdrawal symptoms observations from a controlled Anglo-Canadian study.
Barlow DH: Negative effects from psychological treat- after selective serotonin reuptake inhibitor discon- J Psychopharmacol 1996; 10: 101–109.
ments. Am Psychol 2010; 65: 13–20. tinuation: a systematic review. Psychother Psychosom Offidani E, Fava GA, Sonino N: Iatrogenic comorbidity in
Barlow DH, Gorman JM, Shear MK, et al: Cognitive be- 2015b;84: 72–81. childhood and adolescence. CNS Drugs 2014; 28: 769–
havioral therapy, imipramine and their combination Fava GA, Guidi J, Rafanelli C, et al: The clinical inadequacy 774.
for panic disorder. JAMA 2000; 283: 2529–2536. of evidence-based medicine and the need for a concep- Offidani E, Fava GA, Tomba E, et al: Excessive mood eleva-
Bech P: Clinimetric dilemmas in outcome scales for mental tual framework based on clinical judgment. Psychother tion and behavioral activation with antidepressant
disorders. Psychother Psychosom 2016; 85: 323–326. Psychosom 2015c;84: 1–3. treatment of juvenile depressive and anxiety disorders.
Belaise C, Gatti A, Chouinard V-A, et al: Persistent post- Fava GA, Tossani E, Bech P, et al: Emerging clinical trends Psychother Psychosom 2013a;82: 132–141.
withdrawal disorder induced by paroxetine, a selective and perspectives on comorbid patterns of mental dis- Offidani E, Guidi J, Tomba E, et al: Efficacy and tolerability
serotonin reuptake inhibitor, and treated with specific orders in research. Int J Methods Psychiatr Res 2014; of benzodiazepines versus antidepressants in anxiety
cognitive behavioral therapy. Psychother Psychosom 23: 92–101. disorders: a systematic review and meta-analysis. Psy-
2014; 83: 247–248. Forand NR, de Rubeis RJ, Amsterdam JD: Combining chother Psychosom 2013b;82: 355–362.
Biesheuvel-Leliefeld KE, Kok GD, Bockting CL, et al: Ef- medication and psychotherapy in the treatment of Piolanti A, Offidani E, Guidi J, et al: Use of the Psycho
fectiveness of psychological interventions in prevent- major mental disorders; in Lambert MJ (ed): Bergin Social Index (PSI), a sensitive tool in research and
ing recurrence of depressive disorder: meta-analysis and Garfield’s Handbook of Psychotherapy and Be- practice. Psychother Psychosom 2016; 85: 337–345.
and meta-regression. J Affect Dis 2015; 174: 400–410. havior Change, ed 6. New York, John Wiley & Sons, Richardson WS, Doster LM: Comorbidity and multimor-
Carvalho AF, Sharma MS, Brunoni AR, et al: The safety, 2013, pp 735–774. bidity need to be placed in the context of a framework
tolerability and risks associated with the use of newer Friedman ES, Thase ME: Combining cognitive-behavioral of risk, responsiveness, and vulnerability. J Clin Epide-
generation antidepressant drugs. Psychother Psycho- therapy with medication; in Gabbard GO (ed): Text- miol 2014; 67: 244–246.
som 2016; 85: 270–288. book of Psychotherapeutic Treatments. Washington, Scott J, Young AH: Psychotherapies should be assessed for
Chouinard G, Chouinard V-A: New classification of selec- American Psychiatric Press, 2009, pp 263–285. both benefit and harm. Br J Psychiatry 2016; 208: 208–
tive serotonin reuptake inhibitor withdrawal. Psy- Guidi J, Fava GA: Emerging trends in clinical psychology. 209.
chother Psychosom 2015; 84: 63–71. Riv Psichiatr 2014; 49: 227. Shear MK, Brown TA, Barlow DH, et al: Multicenter Col-
Cloos JM, Bocquet V, Rolland-Portal I, et al: Hypnotics Guidi J, Tomba E, Fava GA: The sequential integration of laborative Panic Disorder Severity Scale. Am J Psychia-
and triazolobenzodiazepines – best predictors of high- pharmacotherapy and psychotherapy in the treatment try 1997; 154: 1571–1575.
dose benzodiazepine use: results from the Luxembourg of major depressive disorder: a meta-analysis of the Starcevic V, Barkoulias V, Viswasam K, et al: Inconsistent
National Health Insurance Registry. Psychother Psy- sequential model and a critical review of the literature. portrayal of medication dependence, withdrawal and
chosom 2015; 84: 273–283. Am J Psychiatry 2016; 173: 128–137. discontinuation symptoms in treatment guidelines for
Cosci F, Guidi J, Balon R, et al: Clinical methodology mat- Haug TT, Blomhoff S, Hellstrom K, et al: Exposure therapy anxiety disorders. Psychother Psychosom 2015; 84:
ters in epidemiology: not all benzodiazepines are the and sertraline in social phobia. Br J Psychiatry 2003; 379–380.
same. Psychother Psychosom 2015; 84: 262–264. 182: 312–318. Tondo L, Vázquez G, Baldessarini RJ: Mania associated
Fava GA: Fading of therapeutic effects of alprazolam in Joseph MF, Youngstrom EA, Soares JC: Antidepressant- with antidepressant treatment: comprehensive meta-
agoraphobia. Case reports. Prog Neuropsychopharma- coincident mania in children and adolescents treated analytic review. Acta Psychiatr Scand 2010; 121: 404–
col Biol Psychiatry 1988; 12: 109–112. with selective serotonin reuptake inhibitors. Future 414.
Fava GA: Can long term treatment with antidepressant Neurol 2009; 4: 87–102. Uhlenhuth EH, Lipman RS, Covi L: Combined pharmaco-
drugs worsen the course of depression? J Clin Psychia- Lee SH, Paz-Filho G, Mastronardi C, et al: Is increased an- therapy and psychotherapy. J Nerv Ment Dis 1969;148:
try 2003; 64: 123–133. tidepressant exposure a contributory factor to the obe- 52–64.
Fava GA: The hidden costs of financial conflicts of interest sity pandemic? Transl Psychiatry 2016; 6:e759. Vanderbroucke JP, Psaty BM: Benefits and risks of drug
in medicine. Psychother Psychosom 2016; 85: 65–70. Liebowitz MR: Social phobia. Mod Probl Pharmacopsy- treatments. How to combine the best evidence on
Fava GA, Bech P: The concept of euthymia. Psychother chiatry 1987; 22: 141–173. benefits with the best data about adverse effects. JAMA
Psychosom 2016; 85: 1–5. Linden M: How to define, find and classify side effects in 2008; 300: 2417–2419.
Fava GA, Offidani E: The mechanisms of tolerance in anti- psychotherapy. Clin Psychol Psychother 2013; 20: 286– Zito JM: High-dose benzodiazepine use among long-term
depressant action. Progr Neuropsychopharmacol Biol 296. users: when will we ever learn? Psychother Psychosom
Psychiatry 2011; 35: 1593–1602. Marks IM, Swinson RP, Basoglu M, et al: Alprazolam and 2015; 84: 259–261.
Fava GA, Tomba E: Treatment of comorbid anxiety disor- exposure alone and combined in panic disorder with
ders and depression; in Emmelkamp PMG, Ehring T agoraphobia. Br J Psychiatry 1993; 162: 776–787.
(eds): The Wiley Handbook of Anxiety Disorders. Nordahl HM, Vogel PA, Morken G, et al: Paroxetine cog-
Chichester, Wiley-Blackwell, 2014, pp 1165–1182. nitive therapy and their combination in the treatment
Fava GA, Cosci F, Offidani E, et al: Behavioral toxicity re- of social anxiety disorder with or without avoidant
visited. Iatrogenic comorbidity in psychiatric evalua- personality disorder. Psychother Psychosom 2016; 85:
tion and treatment. J Clin Psychopharmacol 2016; 36: 346–356.

270 Verhaltenstherapie 2017;27:265–270 Fava/Benasi/Cosci