Persistence of Sexual Dysfunction Side Effects after Discontinuation of Antidepressant Medications: Emerging Evidence

42 The Open Psychology Journal, 2008, 1, 42-50
1874-3501/08 2008 Bentham Open

Open Access
Persistence of Sexual Dysfunction Side Effects after Discontinuation of
Antidepressant Medications: Emerging Evidence
Audrey S. Bahrick*

University Counseling Service, 3223 Westlawn S., The University of Iowa, Iowa City, Iowa 52242-1100, USA
Abstract: Post-market prevalence studies have found that Selective Serotonin Reuptake Inhibitor (SSRI) and Serotonin-
Norepinephrine Reuptake Inhibitor (SNRI) sexual side effects occur at dramatically higher rates than initially reported in
pre-market trials. Prescribing and practice conventions rest on the untested assumption that individuals who develop sexual
dysfunction secondary to SSRI and SNRI antidepressant medications return fully to their pre-medication sexual functioning
baseline shortly after discontinuing treatment. Most individuals probably do return to their previous level of sexual
functioning, however recent case reports, consumer-provided Internet-based information, incidental research findings, and
empirical evidence of persistent post SSRI sexual benefits in the premature ejaculation literature suggest that for some individuals,
SSRI and SNRI-emergent sexual side effects persist indefinitely after discontinuing the medications. The literature
poorly captures the full spectrum of SSRI/SNRI sexual side effects, and a lack of systematic follow-up in the sexual
side effects research precludes detection of post SSRI/SNRI sexual dysfunction, leaving the formal knowledge base inadequate
and even inaccurate, raising informed consent issues, and leaving clinicians vulnerable to practicing in ways that
may be hurtful to patients in spite of their best efforts to inform themselves.
Key Words: SSRI, iatrogenesis, persistent sexual side effects, sexual dysfunction, genital anesthesia.

INTRODUCTION
The present article discusses issues and evidence related
to sexual side effects of selective serotonin reuptake inhibitors
(SSRIs) including fluoxetine (Prozac), paroxetine
(Paxil), sertraline (Zoloft), citalopram (Celexa), escitalopram
(Lexapro), and fluvoxamine (Luvox), and the serotoninnorepinephrine
reuptake inhibitor (SNRI) venlafaxine (Effexor)
1. Sexual side effects are known to occur as a result of
drug treatment, but their prevalence is underestimated in the
product literature, and the research literature poorly captures
the quality and scope of the sexual side effects and how distressing
they may be to patients.
The SSRIs and SNRIs are approved and considered first
line treatments for depressive disorders, generalized anxiety
disorder, panic disorder, social phobia, obsessive-compulsive
disorder, bulimia, premenstrual dysphoric disorder, and post
traumatic stress disorder. Increasingly, they are prescribed
off-label [1] to treat conditions such as peri-menopausal and
post-menopausal hot flashes, chronic fatigue syndrome,
chronic pain syndromes, premature ejaculation, and paraphilias.
In the latter two conditions, the sexual side effects are
intended as the primary and desired effects.
Recent estimates are that on average, 43% of patients
seen by psychologists take medication adjunctively to psychotherapy
[2], and that one in eight adult American has
*Address correspondence to this author at the University Counseling Service,
3223 Westlawn S., The University of Iowa, Iowa City, Iowa 52242-
1100, USA; Tel: 319 335-7294; Fax: 319 335-7298;
E-mail: audrey-bahrick@uiowa.edu
1 Sexual side effects of the SNRI duloxetine (Cymbalta) are less well-studied.
taken an SSRI or SNRI over the last ten years [3]. In the present
author’s practice setting at a large university counseling
service, nearly fifty percent of the generally high-functioning
and sexually active or sexually motivated young adult clientele
are taking psychotropic medications, usually SSRIs or
SNRIs [M. Harris, Clinical Director, University of Iowa
Counseling Service personal communication, Dec.5, 2007].
There is a growing consensus that psychologists need to be
knowledgeable about psychotropic medication effects and
side effects as a competency issue and a standard of care [4].
Given mental health professionals’ responsibility for client
welfare, when clients are taking or are contemplating taking
SSRIs/SNRIs, psychologists have an obligation to inform
clients of the possibility of sexual side effects, and in collaboration
with clients and prescribing professionals, to consider
the impact of adding a new, medication-related sexual
dysfunction to the client’s condition. Prescribing psychologists
are optimally positioned to monitor and to be responsive
to treatment implications of emergent medication side
effects [5] .
The possibility that treatment-related sexual side effects
may persist in some patients after stopping the medications
seems to be unrecognized among the research and professional
communities, yet is increasingly identified among
Internet communities [e.g. SSRIsex@yahoogroups.com;
http://www.sexual-dysfunction.info/forum/], and described
in a series of recent case reports [6-9]. My immersion in the
formal literature, in my clinical practice, and in consumerbased
Internet information has led me to believe that the existing
knowledge base with regards to SSRI/SNRI sexual
side effects is unintegrated, inadequate, and even inaccurate,
raising concerns about informed consent, and leaving cliniPersistence
of Sexual Dysfunction Side Effects The Open Psychology Journal, 2008, Volume 1 43
cians vulnerable to unknowingly contributing to iatrogenic
harm.
The present article presents available evidence indicating
that SSRI and SNRI-emergent sexual dysfunctions, for an
unknown number of patients, may not resolve upon cessation
of medication. As discussed in Rivas-Vasquez, Rey, Blais, et
al. [5], sorting out the etiology of sexual dysfunction can be
challenging. Evidence considered in the present article is
confined to sexual dysfunctions that were not present at
baseline, and emerged only with drug initiation, maximizing
the probability that the dysfunction is drug-induced. Evidence
presented comes from a variety of sources including
incidental research findings, empirical findings, case reports,
and consumer-reported experiences. In addition, two distinctive
sexual symptoms are highlighted that are rarely reported
in the research literature, but are often identified as among
those symptoms that persist after treatment discontinuation
in both case reports and among members of an Internet
community. It is proposed that these unusual symptoms,
genital anesthesia and anhedonic orgasm/ejaculatory anhedonia
(i.e. diminished or absent genital- tactile sensitivity
and orgasm or ejaculation that is not associated with physical
sensations of pleasure), often missed by our assessment instruments,
may serve as lingering markers of past SSRI or
SNRI exposure [10], further helping to distinguish medication
effects from sexual dysfunctions that may be primarily
of psychological origin.
Searching the data bases of research studies yielded no
published investigations that bear directly on the problem of
persistent sexual side effects after discontinuation of SSRIs/
SNRIs. With few exceptions, the research investigating
SSRI/SNRI sexual side effects has not included follow up
after discontinuation of the medication. However a number
of investigations, not primarily motivated by concern about
long term drug safety, have followed SSRI-emergent sexual
side effects for a six month period after discontinuation of
the medications [11-13]. The researchers did not interpret
their findings as raising concerns about the persistence of
sexual side effects after discontinuation of SSRIs, but such
alternative interpretations of the findings will be discussed.
UNDERESTIMATION, MINIMIZATION, AND GAPS
IN KNOWLEDGE ABOUT SSRI/SNRI SEXUAL SIDE
EFFECTS
Post market research has clearly established that the
SSRIs and SNRIs can affect most every aspect of sexual
functioning at rates significantly higher than the 2-16% rates
reported in pre-market trials and currently listed in the drug
insert literature. Large prospective studies in which baseline
assessment excludes participants with pre-existing sexual
dysfunction have found rates of treatment-emergent sexual
dysfunctions such as decreased libido, delayed orgasm, anorgasmia,
erectile dysfunction, and difficulties with arousal, of
between 36 and 70% [14, 15]. The 2-16% pre-market rates
of SSRI and SNRI-induced sexual side effects are based on
spontaneous reports of individuals in initial trials who had
been on the medications a short time. When individuals are
directly and systematically asked about changes in sexual
functioning via a structured clinical interview or a self-report
inventory, dramatically higher rate information is obtained as
compared with reliance on individuals to spontaneously
als to spontaneously volunteer personally sensitive information
about changes in sexual functioning [15, 16].
Current American Psychiatric Association treatment
guidelines do not recognize clinically significant differences
in efficacy among the various medications [17]. The choice
to prescribe one SSRI or SNRI over another is often based
on the known side effect profile, with the reported impact on
sexual functioning often a deciding factor [18]. While sexual
side effects are probably the most-researched of all the SSRI
and SNRI side effects, the foci, motivated by competing
drug companies’ efforts to gain a market share, have been
heavily on: determining the comparative prevalence rates of
a variety of sexual side effects; determining strategies for
managing the sexual side effects: demonstrating the efficacy
of one medication to serve as an antidote to the sexual dysfunction
caused by another; and also on the ejaculationdelaying
benefits of the various SSRIs when used as an offlabel
treatment for premature ejaculation (i.e. see Waldinger
[19] for a review).
Despite the substantial body of literature devoted to
comparative sexual side effects prevalence rates and antidotes,
there is little indication of clinically significant differences
among the medications with regards to rates of associated
sexual side effects [20] and no strategy or antidote has
proven reliably effective for alleviating the SSRI or SNRIinduced
sexual side effects. While the consistent impression
given by the literature is that sexual side effects may be easily
resolvable in consultation with the prescriber, the management
of antidepressant sexual side effects, according to
Balon [16], should currently be seen as an art and not a science.
An estimated 5 to 10% of individuals may experience a
diminution of the SSRI or SNRI emergent sexual side effects
over time as they remain on the medication [21], but for the
vast majority, the sexual side effects are intractable and will
continue for at least as long as they take the medication [14,
15].
The research questions related to SSRI/SNRI sexual side
effects that are of most interest to industry sponsors are not
the same questions that are of most interest and urgency
from a public health perspective. As a result, important gaps
in knowledge remain. These gaps include, but are not limited
to: 1) Information as to which sexual side effects are most
characteristic of the SSRIs/SNRIs; 2) The degree to which
side effects remain stable or change over the course of longer
term treatment; 3) When and whether individuals who develop
sexual side effects secondary to the medications return
fully to their pre-medication sexual functioning baseline; 4)
What impact the medications may have on fertility and reproductive
health, and; 5) In what ways the medications may
affect the developing sexuality of adolescents and children.
Gaps 1, 2, and 3 will be explored more fully in this article.
While the potential impacts of SSRIs and SNRIs on fertility
and reproductive health, and on adolescents’ developing
sexuality are beyond the scope of this article, there are
reasons to be concerned: Tanrikut and Schlegel [22] recently
found SSRI usage in two men to be related to impaired semen
parameters, which apparently improved when the men
discontinued the medications. And Scharko [23] notes that
while sexual side effects may be expected to occur in adolescents
at rates similar to those found in adults, his review of
the adolescent SSRI literature found a nearly complete lack
44 The Open Psychology Journal, 2008, Volume 1 Audrey S. Bahrick
of information or research related to SSRI sexual side effects
in adolescents, and no developmentally appropriate instruments
for assessing the medications’ impact on adolescent
sexual functioning.
The assumption of a sexual-functioning return to baseline
shortly after cessation of the medications is deeply embedded
in the literature, as well as in the conventional approach
to practice and prescribing [10]. Yet no primary source was
found to substantiate the resolution of sexual side effects: no
study was found that systematically and intentionally followed
the course of SSRI or SNRI-induced sexual dysfunction
after discontinuation of the medications for the drugsafety-
related purpose of determining when and to what degree
the sexual side effects resolve. A search of the literature
resulted in identifying only a single study in which fluoxetine-
emergent sexual side effects were reported to have resolved
within one to three weeks of cessation of medication
[24]. The author does not specify when or by what means the
follow-up information was obtained or how many of 54 patients
in this two-year, naturalistic, private-practice based
study had discontinued the medication and were available for
follow up. Treatment-emergent sexual side effects probably
do resolve for most individuals after discontinuing the medications,
yet since sexual side effects might persist after treatment
cessation for some individuals, researchers have an
obligation to design studies so as to assess this likelihood. In
the vast majority of studies of SSRI sexual side effects, no
post-treatment follow-up is included in the research protocol,
and/or participants are still actively taking the medication at
study endpoints.
The SSRI/SNRI sexual side effects literature generally
reflects the perspective that the sexual side effects pose a
threat to treatment compliance [i.e. 30, 37, 43]. Indeed, it is
believed that up to two-thirds of patients discontinue a
guidelines-recommended course of antidepressants due to
adverse events, particularly sexual dysfunction [25]. Less
frequently mentioned, or only perfunctorily treated is the
responsibility of prescribing professionals to address the
interaction of a new or worsening sexual dysfunction with
the condition being treated, and the impact on general quality
of life in the collaborative negotiation of treatment alternatives.
The failure of the SSRI-research literature to fully engage
in addressing the psychological costs of medicationinduced
sexual dysfunction suggests that clinicians may
sometimes fail to invite fully collaborative interactions in
response to patients’ concern about SSRI-emergent sexual
dysfunctions. Instead, they may engage in something more
akin to what Higgens et al. [26] describe as a “compliance
monologue”[p. 441] motivated by, among other things, the
professional’s own discomfort with initiating an in-depth
conversation about sexual functioning, bolstered by an a
priori allegiance to a pharmacological model of treatment.
In response to individuals who express concern about
treatment-emergent sexual side effects, clinicians are advised
to reassure patients that the side effects are “medically benign”
[27 p. 23] and that “ all data suggest return of sexual
functioning to baseline once the medication is stopped” [28
p. 106]. The conventional wisdom seems to dismiss even the
possibility that sexual side effects may continue after stopping
the medications by defining true drug-induced sexual
dysfunction as, “that which occurs after the agent is started
or the dosage is increased, is not partner-specific, is not lifelong
or recurrent, and resolves when drug therapy is discontinued”
(italics added) [29, p. 825]; or, that which “ is not
better explained by physical illness or stress, whose onset is
with drug initiation or dose increase, is present in all sexual
situations, reappears with reintroduction of the drug, and
dissipates with drug discontinuation or dose reduction” (italics
added) [30, p. 1489].
The assumption of benignness, along with such intuitively
attractive, yet empirically untested definitions of druginduced
sexual side effects, contributes to the systematic
failure to include post-treatment follow-up in our research
designs. These assumptions also underscore the reasons that
clinicians, and even patients themselves may fail to consider
past SSRI or SNRI medication use as a possible cause for
otherwise unexplained persistent sexual dysfunctions. Current
definitions are unsatisfactory, as they do not allow for
the possibility that sexual side effects that begin on medication
and persist after medication discontinuation could be
medication-related.
GENITAL ANESTHESIA AND PLEASURELESS ORGASM:
UNCOMMON OR CHARACTERISTIC OF
SSRI/SNRI SEXUAL SIDE EFFECTS?
There are indications that some SSRI/SNRI sexual side
effects thought to be rare are actually common [10]. The
most frequently documented sexual side effects are diminished
libido, unspecified problems with arousal, and delayed
orgasm or anorgasmia. Delayed ejaculation or orgasm, and
anorgasmia have been those symptoms that the literature
links most clearly and most frequently to SSRI treatment, vs.
to depression itself [16]. However the symptoms of genital
anesthesia and pleasureless orgasm, outside the range of
common experience and appearing to often occur together,
are frequently reported among men and women in Internet
communities, in an accumulating case reports literature [6-9,
31-36], and in one research investigation [37].
These counterintuitive symptoms can be described as a
preservation of aspects of normal physical sexual functioning,
but with a loss of the concurrent capacity to experience
sexual pleasure or arousal. Genital anesthesia has been described
in Internet forums and in case reports as genitals that
are numb or nearly numb, which may respond to stimulation
by erection or lubrication, but without attendant subjective
feelings of arousal. Pleasureless orgasm or ejaculatory anhedonia
can be described as orgasm or ejaculation that is preceded
by little sense of building arousal, is identifiable by
rhythmic perineal muscle contractions in women or by ejaculation
in men, but is experienced as pleasureless or nearly so.
These qualitatively different symptoms are not easily classifiable
in regard to any specific category of the sexual response
cycle, and as Bahrick has noted, elude capture in
most prevalence studies [10]. Decreased orgasmic intensity
or ejaculatory anhedonia are not mentioned in recent reviews
of disorders of ejaculation and orgasm in men [38], or in
women [39], and SSRI/SNRI-related decreased genital sensitivity
or genital anesthesia, when mentioned, are reported to
be uncommon [6, 14, 15]. However it is more accurate to say
that the symptoms are uncommonly assessed.
Michael and Mayer [32] describe a case of fluoxetineinduced
anesthesia of the vagina and nipples in an initially
Persistence of Sexual Dysfunction Side Effects The Open Psychology Journal, 2008, Volume 1 45
depressed woman who had experienced lowered libido concurrently
with depression. Michael and Mayor linked her
genital and nipple anesthesia to the treatment and not to the
patient’s condition, as the anesthesia persisted even when the
initially depressed patient became euthymic. Deisenhammer
and Trawoger [36] demonstrated with a re-challenge test that
genital anesthesia was indeed caused by treatment with citalopram
in the case of a 36 year old man treated with citalopram
for an adjustment disorder with depressed mood. The
man, reported not to have any preexisting sexual dysfunctions,
experienced penile anesthesia after three days on the
medication. The symptom resolved shortly after stopping the
medication, which he’d continued for three weeks due to its
mood lifting effect. A year later, the man, described as
healthy and with no mood or sexual difficulties, consented to
re-exposure to citalopram. He again experienced penile anesthesia
after several days on citalopram, which again resolved
several days after stopping the medication.
Deisenhammer and Trawoger [36] state that genital anesthesia
is probably associated with all antidepressants that
enhance serotonergic neurotransmission. Clayton and Montejo
[21] point to a possible basis for the symptom, noting
that serotonergic medications may decrease genital sensation
via diminished nitric oxide function. Nitric oxide is integral
to penile and clitoral tumescence. Deisenhammer and Trawoger
[36] call for controlled studies to evaluate the incidence
of SSRI induced genital anesthesia. To the present
author’s knowledge, no such controlled studies have been
performed. Indeed, our literature supports the assumption
that genital anesthesia and ejaculatory anhedonia/pleasureless
orgasm are rare by failing to systematically include the
symptoms in our instruments, and by failing to report them
transparently when they are included [10].
According to Clayton and Montejo [21], The Changes in
Sexual Functioning Questionnaire (CSFQ) [40], and the Arizona
Sexual Experiences Scale (ASEX) [41], are the primary
validated instruments used to assess sexual dysfunction secondary
to psychotropic medication. Neither the CSFQ nor
the ASEX include specific queries related to changes in
genital sensitivity. And neither unambiguously captures
pleasureless orgasm or ejaculatory anhedonia; the CSFQ
fails to do so by scoring decreased orgasmic intensity together
with items assessing timing and frequency of orgasm,
and the ASEX because the inquiry regarding changes in satisfaction
from orgasm is less specific than changes in intensity
of orgasm.
To the present author’s knowledge, the Rush Sexual Inventory
is the only instrument, validated in 2005 [42], for
assessing medication-related sexual changes that includes
specific queries related to both changes in sensitivity of external
genitalia as well as changes in orgasm intensity. However,
as items are answered in a yes/no format, the instrument
captures the presence, but not the severity of the symptoms.
Two studies were found that assessed SSRI-related
sexual side effects using the Rush Sexual Inventory:
Ferguson [43] did not report specific symptom results, and
Zajecka et al. [37] reported only partial results. Zajecka et al.
[37] found that among 42 depressed patients taking a variety
of SSRIs, 28% of women reported treatment-emergent decreased
genital sensitivity and 25% of men reported treatment-
emergent decreased intensity of orgasm, suggesting the
symptoms are not uncommon. While data were collected for
both genders for both symptoms, results were reported for
only one of the symptoms for each gender. The article’s abstract
notes the Rush Sexual Inventory’s utility for follow-up
assessment, however no follow-up data were provided regarding
the resolution of the SSRI-induced sexual side effects.
Genital anesthesia and pleasureless orgasm or ejaculatory
anhedonia should be of special interest to both researchers
and clinicians. The symptoms are unknown in the general
population, are not known to be associated with any conditions
or disorders for which SSRIs or SNRIs are prescribed,
yet seem to be distinctive markers of medication treatment.
When treatment-emergent sexual side effects persist after
discontinuing SSRIs or SNRIs, patients’ reports may be discounted,
disbelieved, ascribed to a relapse of the preventing
problem, or to the emergence of yet a new mental health
problem by professionals to whom individuals turn for help.
Thus, Bahrick has suggested that among all the sexual side
effects that may emerge during treatment or persist after cessation,
genital anesthesia and ejaculatory anhedonia may
provide the most compelling links to the treatment rather
than the conditions being treated [10].
EVIDENCE OF PERSISTENT SEXUAL SIDE EFFECTS
AFTER DISCONTINUATION OF SSRI/SNRI
MEDICATIONS
An Incidental Finding
Montejo et al. [11] appear to have found incidental evidence
of SSRI-induced sexual side effects continuing after
cessation of the medication. The study’s aim was to assess
the impact of changing to another medication in patients
whose depressive symptoms had successfully remitted with a
variety of SSRIs, but who had developed treatment-emergent
sexual dysfunction. Patients were switched either to amineptine
(n=47), an atypical tricyclic antidepressant that is no
longer available, or to paroxetine (n=38). A third group of
depressed patients was treated with amineptine only (n=26)
and had no prior SSRI treatment. All three groups were followed
with multiple assessments over a six month period.
Montejo et al. [11] report for those patients taking amineptine
only, amineptine was not a cause of secondary sexual
dysfunction, and also effectively treated depressive symptoms.
In the group switched to amineptine, the incidence of
sexual dysfunction dropped from 100% to 55% over the six
month period, while depressive symptoms remained in remission.
In the group switched to paroxetine, sexual dysfunction
decreased only slightly from 100% to 89.7% after
six months. The authors interpret these findings to support
with high confidence (p<.001) the conclusion that amineptine is an effective antidepressant that is able to significantly improve the SSRI-caused sexual dysfunction, yet maintain the efficacy of the antidepressant treatment used before. An alternative interpretation is that for 55% of individuals switched to a medication that successfully treated depressive symptoms and was not a cause of secondary sexual dysfunction, the initial SSRI-induced sexual dysfunction persisted for at least six months after discontinuing the SSRI [10]. 46 The Open Psychology Journal, 2008, Volume 1 Audrey S. Bahrick Empirical Evidence: Persistent Sexual Benefits after Discontinuation of SSRIs The well-established ejaculation-delaying effects of all of the SSRIs has led to their ubiquitous off-label use as a treatment for premature ejaculation (PE), and motivated industry to seek approval of SSRIs for a new indication reported to affect approximately 30% of men globally across all age groups [44] and to affect up to 70% of men at some point in their sexual lives [45]. The large emerging literature related to SSRIs as a treatment for PE asks which SSRIs may have the longest or most potent ejaculation-delaying effect; whether chronic or on demand administration of SSRIs is preferable and in what dose, whether newer shorter-acting SSRIs may prove more efficacious than those already on the market, and whether the medications may provide lasting benefits beyond treatment discontinuation. With current inthe- field practice moving well ahead of demonstrated longterm safety, a recent survey of urologists indicated SSRI treatment is the most common first-line approach to managing premature ejaculation [46]. Two recent studies found robust evidence for sustained post-treatment effects of SSRIs on ejaculation latency [12, 13]. Safarinejad and Hosseini [12] evaluated citalopram as a treatment for PE in a prospective, double-blind, placebo controlled, fixed dose (20 mg.), randomized study of fifty-eight non-depressed, psychologically and physically healthy men whose only sexual complaint was PE. Premature ejaculation was defined as intravaginal ejaculatory latency time (IVELT) of less than two minutes in more than 90% of coitus. The impact on sexual functioning of the citalopram (n=29) and placebo (n=29) treatments was assessed every two weeks during the twelve-week treatment period, and in three and six month follow-ups after cessation of treatment. Safety and efficacy measures included: changes in IVELT which subjects recorded with a stop watch; frequency of intercourse, adverse drug events which subjects recorded in an apparently open format diary, and the use of a subset of six out of fifteen questions from the International Index of Erectile Function (IIEF) reflecting a domain of “intercourse satisfaction”. The authors report significantly improved ejaculation latency (IVELT) scores (p<.001) for the citalopram group vs. the placebo group beginning at week one and continuing throughout the 12-week treatment phase, and improved intercourse frequency as well as satisfaction in the citalopram group over the placebo group (p<0.05) n="77)" n="70)">