REVIEW ARTICLE
Insomnia and Somnolence Associated With
Second-Generation Antidepressants During the Treatment
of Major Depression
A Meta-Analysis
Siegfried Alberti, MD,* Alberto Chiesa, MD,* Costanza Andrisano, MD,* and Alessandro Serretti, MD, PhD*
Background: Sleep reduction or enhancement is frequently observed
with second-generation antidepressant treatments, and they can be benefi-
cial or harmful depending on the symptom profile of each subject. Never-
theless, relatively little attention has been given so far to rank those effects
across compounds. The aim of this meta-analysis is to provide quanti-
tative data about short-term rates of insomnia and somnolence associ-
ated with 14 second-generation antidepressants during the treatment
of major depression.
Methods: A literature search and a search of unpublished documents
were performed. Eligible studies focusing on MD patients treated with
second-generation antidepressants were entered in the analysis. Our pri-
mary outcome measures were insomnia and somnolence rates induced
by antidepressants as compared with those associated with placebo. Sensi-
tivity analyses were carried out as well.
Results: Ten second-generation antidepressants showed higher rates of
insomnia than placebo. The highest incidence was found for bupropion
and desvenlafaxine. Agomelatine was the only antidepressant with a lower
likelihood of inducing insomnia than placebo. Eleven antidepressants were
associated with higher rates of somnolence than placebo. Fluvoxamine and
mirtazapine showed the highest frequency of somnolence. Bupropion in-
duced somnolence to a lower extent than placebo. Sensitivity analyses
showed a degree of variation of those findings.
Discussion: Antidepressants are associated with different insomnia and
somnolence rates, mainly depending on their mechanisms of action.
Despite some limitations, we underscore that the treatment-emergent in-
somnia and/or somnolence are frequent, and they could be used in clinical
practice to face the specific needs of each patient.
Key Words: depression, antidepressants, tolerability, individualized
therapy
(J Clin Psychopharmacol 2015;35: 296303)
M ajor depression (MD) is one of the most relevant health-care
problems. Indeed, it is thought to affect about 121 million
people throughout the world,1 it has an estimated lifetime pre-
valence of 13.2% in the United States,2 and it is associated with
significant disability,3 mortality,4,5 and costs for both affected in-
dividuals and the health-care system.6
From the *Department of Biomedical and Neuromotor Science, University
of Bologna, Bologna; and Department of Clinical and Experimental Medicine
and Pharmacology, University of Messina, Messina, Italy.
Received June 19, 2014; accepted after revision March 4, 2015.
Reprints: Alessandro Serretti, MD, PhD, Department of Biomedical and
Neuromotor Science, University of Bologna, Viale Carlo Pepoli 5, 40123
Bologna, Italy (email: alessandro.serretti@unibo.it).
Supplemental digital contents are available for this article. Direct URL citation
appears in the printed text and is provided in the HTML and PDF versions
of this article on the journals Web site (www.psychopharmacology.com).
Copyright 2015 Wolters Kluwer Health, Inc. All rights reserved.
ISSN: 0271-0749
DOI: 10.1097/JCP.0000000000000329
296 www.psychopharmacology.com Journal of Clinical Psychopharmacology Volume 35, Number 3, June 2015
Copyright 2015 Wolters Kluwer Health, Inc. All rights reserved.
According to current guidelines, the so-called second-
generation antidepressants have played in the last 2 decades an
increasingly important role for the treatment of MD.7,8 Although
there is no complete consensus as to which antidepressants should
be included under the label of second-generation antidepres-
sants, there is general consensus that they broadly include selective
serotonin reuptake inhibitors (SSRIs), serotonin and norepineph-
rine reuptake inhibitors, and other drugs with related mechanisms
of action that usually selectively target neurotransmitter receptors
and reuptake.911 Second-generation antidepressants have been
found to be effective and generally well tolerated.1216 For this rea-
son, they are considered the first-line treatment in clinical settings.
Despite their positive balance of efficacy and side effects,
second-generation antidepressants are often associated with mild-
to-moderate side effects that, although unrelated to risks of mortal-
ity, can undermine compliance with treatment and impair quality
of life. In fact, these side effects including, among others, sex-
ual dysfunction, weight gain, gastrointestinal disorders, and sleep
disorders have been found to represent the main reason for drop-
out during antidepressant treatment, in particular during the first
3 months of treatment.17,18
Of note, recent meta-analyses showed that significant dif-
ferences exist among different antidepressants in terms of, for
instance, the incidence of drug-related sexual dysfunction19 and
weight gain,20 mainly depending on the underpinning pharmaco-
logical mechanisms of each specific drug. On the other hand, it is
less clear so far the extent to which second-generation antidepres-
sants have an impact on sleep aspects in patients treated with these
drugs. Sleep-related issues are of particular importance as they
may influence quality of life of subjects and also the risk of re-
lapse.21 Depression presents with a variable of sleep-related symp-
toms ranging from the most frequent insomnia to somnolence.
Therefore, it is of particular importance to use the most appropri-
ate compound for each specific symptom profile, for example,
a patient with insomnia and/or anxiety will benefit more from
a sleep-inducing antidepressant. In particular, although some re-
views have been published that provide a qualitative comparison
of the modifications on sleep induced by different antidepressants,
for example,22 a comprehensive meta-analysis comparing the
rates of insomnia and daytime somnolence among these antide-
pressants has not been performed thus far.
Accordingly, the aim of the present work is to provide quanti-
tative data about the short-term rates of insomnia and somnolence
associated with second-generation antidepressants during the
treatment of MD. In particular, we investigated the extent to which
14 second-generation antidepressants, including agomelatine,
bupropion, citalopram, desvenlafaxine, duloxetine, escitalopram,
fluoxetine, fluvoxamine, milnacipran, mirtazapine, paroxetine,
reboxetine, sertraline, and venlafaxine impact insomnia and som-
nolence over the short-term period in patients with MD.
Journal of Clinical Psychopharmacology Volume 35, Number 3, June 2015
MATERIALS AND METHODS
Literature Research
A literature research was undertaken using MEDLINE,
ISI Web of Science, the Cochrane database, and references
of retrieved articles (this is not a Cochrane Collaboration Study).
The research included articles written in English and published up
to January 2013. The main key words were agomelatine, bupro-
pion, citalopram, desvenlafaxine, duloxetine, escitalopram, fluox-
etine, fluvoxamine, milnacipran, mirtazapine, paroxetine,
reboxetine, sertraline, and venlafaxine in combination with ma-
jor depression or major depressive disorder. A research of un-
published documents was performed as well. More in detail,
documents were requested to the following pharmaceutical
companies: Eli Lilly, Lundbeck, Organon, Solvay, Pfizer,
GlaxoSmithKline, Bristol-Myers Squibb, and Pierre Fabre, and
to the main Medical Control Agencies, including the Food and
Drug Administration, the European Medicines Agency, the Phar-
maceuticals and Medical Devices Agency and the Therapeutic
Goods Administration.
Selection of Trials
Two reviewers independently searched eligible articles for
inclusion (A.C. and S.A.). Included studies had to do the follow-
ing: focus on patients aged 18 or older with MD according to
the Diagnostic and Statistical Manual for Mental Disorders,
Fourth Edition, Text Revision, criteria, with or without further psy-
chiatric comorbidities; investigate the use of one or more of the
above-mentioned antidepressants; provide dichotomous measures
of drug-related insomnia and/or somnolence rates within the 12th
week of treatment; and have a minimum duration of 6 weeks. In-
somnia and somnolence are broad definitions, which may range
to mild-to-severe symptoms. In the present analysis, we reported
their rates as present in included papers, to which we refer. Usu-
ally, these symptoms have been assessed with specific rating
scales or using clinical criteria.
We also included data of a sample of patients with MD re-
ported in 2 previously published studies.23,24 However, from those
studies, which included also bipolar disorder, we selected only
data of patients with MD from the original datasets and included
these data as if they were derived from a single uncontrolled study.
Excluded were as follows: 1) case reports and case series,
2) review articles and meta-analyses, 3) long-term studies (lasting
more than 12 weeks) that did not provide information about the
emergence of drug-related insomnia and/or somnolence rates in
the short-term period (612 weeks); 4) studies providing data on
patients with and without MD joined together; 5) studies focusing
on depressed patients taking other psychotropic drugs at study en-
try (ie, patients treated with other psychotropic drugs in the days
before the entry in the study who switched to the antidepressant
without a washout period); 6) studies focusing on patients in re-
mission at study entry; 7) studies focusing on the combination
of a pharmacological treatment with a nonpharmacological ther-
apy of any kind; And 8) studies providing data on more antide-
pressants or more psychoactive drugs joined together. A notable
exception was represented by studies investigating the association
between a single antidepressant and anxiolytics (a benzodiazepine
or a benzodiazepine-like drug), although this was considered in
sensitivity analyses. The concomitant treatment with cardiovascu-
lar drugs was allowed only if the authors specified that the admin-
istration of these drugs was stable at study entry and remained
stable during the study period. Studies with different methodolog-
ical designs were included. However, sensitivity analyses were
performed so as to estimate the possible impact of study design,
2015 Wolters Kluwer Health, Inc. All rights reserved.
Copyright 2015 Wolters Kluwer Health, Inc. All rights reserved.
Insomnia, Somnolence, and Antidepressants
sponsorship, and concomitant therapies on the outcomes of inter-
est (see below).
Outcome Measures
Our 2 primary outcome measures were short-term insomnia
and somnolence rates reported during the treatment period with
a second-generation antidepressant in patients with MD. In partic-
ular, whenever it was possible, we referred to side effects emerg-
ing within the first 8 weeks of treatment. If 8-week data were
not available, we focused, in the following order, on the emer-
gence of insomnia and somnolence rates within the 10th or the
12th week of treatment, as reported by the authors of original stud-
ies, or alternatively, in case of studies with a shorter duration,
on the same outcome measures emerged within the first sixth
week of treatment. We included data about insomnia and somno-
lence rates only if these data were described as treatment-related
side effects and labeled with the following words: insomnia,
reduced time of sleep, or sleep difficulty for the insomnia
rates and somnolence drowsiness or sleepiness for the som-
nolence rates. It is worth noting that, whether a study reported
overlapping data for a given side effect (eg, somnolence and drows-
iness), we included that associated with the largest incidence.
Of note, in case of overlapping samples, the decision to
include a given study was based on the following algorithm:
(1) whether no difference existed between two studies in terms
of outcome measures, we considered the study with the largest
sample size, and (2) whether the studies provided different out-
comes, we choose the study whose outcomes were more in line
with the criteria of the present meta-analysis.
Data Extraction and Analysis
Data were independently extracted from original reports by
2 reviewers (A.C. and S.A.). Quality of included articles was inde-
pendently assessed by the 2 reviewers using validated quality
scales: the Jadad Scale25 for controlled studies and a modified ver-
sion of the New Ottawa Scale26 for uncontrolled studies for the re-
maining studies. For controlled trials, a score of 3 or higher was
considered to be indicative of high quality,25 whereas for uncon-
trolled ones, a score of 4 or higher was considered as indicative
of high quality.26 All disagreements were resolved through dis-
cussion and, in case of persistent disagreement, with the involve-
ment of a third reviewer (A.S.). The agreement rates between the
2 main reviewers were high (94%). After the involvement of the
third reviewer, no final disagreement was observed.
All data were referred to the intention-to-treat populations,
when available. Otherwise, the completers sample was analyzed.
In addition, whether one study included 2 (or more) treatment
arms of patients treated with the same antidepressant at different
dosages, we separately considered each treatment arm as if they
were 2 (or more) independent studies.
Data extracted from the included studies were entered into
the Cochrane Collaboration Review Management Software
(RevMan version 5.1; Cochrane Collaboration, Oxford, UK) and
analyzed by RevMan analysis 1.01. For insomnia and somnolence
rates, odds ratios and their 95% confidence intervals were calcu-
lated, using a random effect model, which takes into account
possible differences in the implementation of intervention and
characteristics of participants included. Heterogeneity across the
studies was assessed by the 2 and 2 statistics and by visual in-
spection of the results. The 2 statistic P < 0.05 and the 2 statistic
higher than 50% were taken to be suggestive of heterogeneity.
Moreover, for studies that did not include a placebo control
group, we calculated the weighted mean of the placebo samples
for the considered variable from placebo-controlled studies that
www.psychopharmacology.com 297
Alberti et al Journal of Clinical Psychopharmacology Volume 35, Number 3, June 2015
investigated the same variable and applied it to studies not includ-
ing a placebo group according to a previously reported methodol-
ogy.20,27 In such cases, the sample size of the virtual placebo
group was considered as equal to the number of patients treated
with antidepressants in the specific studies that did not include a
placebo group to limit the estimate inflation.
We also performed additional sensitivity analyses focus-
ing on the following: 1) double-blind, randomized placebo-
controlled studies, 2) antidepressants monotherapy studies, and
3) nonsponsored studies (studies were considered as sponsored
if they were funded, also in part, by a pharmaceutical company
or if at least one of the authors was member of a pharmaceutical
company) to investigate whether the incidence of insomnia and
somnolence varied as a function of these variables. to be included
in a sensitivity analysis; each drug was required to be investigated
in at least 2 studies providing data about the specific outcome
measure of interest.
RESULTS
Search Result
The original research identified about 7000 published pa-
pers. About 5220 studies were not considered for inclusion be-
cause they were review, meta-analyses, case reports, case series,
and studies conducted on animal models. After the inclusion
and the exclusion criteria were applied to the remaining 1780 stud-
ies, 1535 studies were excluded and 245 studies could be included
in the present meta-analysis.
In addition, 3 of 8 pharmaceutical companies, as well as 3 of
4 Medical Control Agencies, replied to us. Appropriate docu-
ments providing information about one or both our outcomes
of interest were provided by one pharmaceutical company and
by 2 Medical Control Agencies. The inclusion and the exclusion
criteria were also applied to these documents: 345 studies were
excluded, and 30 could be included in our meta-analysis.
Overall, we used the data of 276 trials (4 concerning
agomelatine, 21 bupropion, 24 citalopram, 10 desvenlafaxine, 32
duloxetine, 26 escitalopram, 60 fluoxetine, 8 fluvoxamine, 10
milnacipran, 23 mirtazapine, 44 paroxetine, 8 reboxetine, 37
sertraline, and 52 venlafaxine). Among included studies, 223
studies used a randomized controlled, 7 studies used a non-
randomized controlled design, and 46 studies used an uncon-
trolled design. A summary of included studies is shown
in Supplementary Table 1, Supplemental Digital Content 1,
http://links.lww.com/JCP/A282.
Primary Outcome Measures
Ten of 14 antidepressants showed significantly higher rates
of short-term insomnia than placebo, with the following increas-
ing order of incidence: escitalopram, duloxetine, venlafaxine, parox-
etine, reboxetine, fluoxetine, fluvoxamine, sertraline, desvenlafaxine,
and bupropion. On the other hand, agomelatine was the only anti-
depressant that had a lower likelihood of inducing insomnia than
placebo. Mirtazapine, milnacipran, and citalopram did not differ
from placebo in terms of insomnia rates (Fig. 1).
Pertaining to the short-term somnolence rates, 11 antide-
pressants were associated with a higher likelihood of inducing
this side effect than placebo, with the following increasing
order of incidence: escitalopram, citalopram, fluoxetine, sertraline,
venlafaxine, duloxetine, desvenlafaxine, paroxetine, reboxetine,
mirtazapine, and fluvoxamine. The only drug that was associated
with a significantly lower risk of inducing somnolence than pla-
cebo was bupropion. Milnacipran and agomelatine did not differ
from placebo in terms of somnolence rates (Fig. 1).
298 www.psychopharmacology.com
Copyright 2015 Wolters Kluwer Health, Inc. All rights reserved.
Sensitivity Analyses
Short-Term Insomnia Rates
When data derived from randomized, double-blind, placebo-
controlled studies were separately analyzed, the main results of
the general analysis were overall confirmed. However, citalopram
was associated with a higher likelihood of inducing insomnia,
whereas fluvoxamine and sertraline were associated with insom-
nia to a lower extent than that they were in the general analysis.
Furthermore, fluvoxamine was the only antidepressant that
showed no significant difference from placebo in terms of insom-
nia rates (Fig. 2).
In addition, when we focused only on studies providing data
about antidepressant monotherapy, duloxetine proved to be the
antidepressant associated with the highest rates of insomnia
(Supplementary Figure 1, Supplemental Digital Content 2,
http://links.lww.com/JCP/A283). Finally, when we considered
only nonsponsored studies, we found that fluoxetine was the drug as-
sociated with the highest risk of insomnia, whereas fluvoxamine was
the drug associated with the lowest risk (Supplementary Figure 2,
Supplemental Digital Content 3, http://links.lww.com/JCP/A284).
Short-Term Somnolence Rates
When we focused only on double-blind, randomized placebo-
controlled studies, the highest rates of somnolence were found for
mirtazapine, followed by fluovoxamine. Another remarkable dif-
ference from the general analysis was the significantly higher
incidence of somnolence associated with citalopram (Fig. 2).
The remaining results were consistent with those observed in
the primary analysis.
When we specifically focused on antidepressants mono-
therapy studies, we found that bupropion was no longer asso-
ciated with a lower rate of somnolence than placebo and did
not differ significantly from placebo (Supplementary Figure 1,
Supplemental Digital Content 2, http://links.lww.com/JCP/A283).
Finally, when data derived from nonsponsored studies were
separately analyzed, mirtazapine showed a higher association with
somnolence than fluvoxamine and showed the highest somnolence
rate among all included antidepressants (Supplementary Figure 2,
Supplemental Digital Content 3, http://links.lww.com/JCP/A284).
Details of all analyses are reported in Supplementary Table 2,
Supplemental Digital Content 4, http://links.lww.com/JCP/A285.
DISCUSSION
The aim of the present work was to quantify and compare for
the first time the short-term rates of insomnia and somnolence
associated with 14 second-generation antidepressants during MD
treatment. Our analysis showed several important findings. First,
bupropion, desvenlafaxine, duloxetine, escitalopram, fluoxetine,
fluvoxamine, paroxetine, reboxetine, sertraline, and venlafaxine
were associated with significantly higher rates of insomnia as com-
pared with placebo. These findings can be considered consistent
with current knowledge of the mechanisms of action of these
drugs. Indeed, all these drugs directly affect brain levels of neuro-
transmitters involved in the regulation of the sleep-wake cycle. In
particular, bupropion is the only antidepressant that selectively
inhibits the reuptake of dopamine and noradrenaline. More in
detail, it has been well documented that the dopaminergic stimula-
tion in the substantia nigra and ventral tegmental area is involved
in arousal.28,29 Furthermore, norepinephrine has been found to
promote waking through the stimulation of 1-adrenoreceptors,
as well as through the activation of other wake-promoting systems,
and the inhibition of systems involved in sleep stimulation.30
2015 Wolters Kluwer Health, Inc. All rights reserved.
Journal of Clinical Psychopharmacology Volume 35, Number 3, June 2015
FIGURE 1. General analysis. CIs are reported for each drug.
These mechanisms of action could explain our observation that
bupropion is the antidepressant associated with the highest rates
of insomnia. It is also remarkable that the strong association be-
tween bupropion and insomnia is consistent with findings re-
ported by previous reviews.3133 Desvenlafaxine was the second
drug associated with insomnia following bupropion. The pharma-
codynamic properties of this drug could partially explain this re-
sult: indeed, both the noradrenergic and the serotonergic activity
of this drug have a reverse relationship with sleep duration,30,34,35
possibly leading to the usually higher risk of insomnia observed
in association with serotonin and norepinephrine reuptake inhibi-
tors in comparison with SSRIs. However, this does not explain
the reason behind our observation that desvenlafaxine could be
more likely to induce insomnia as compared with its racemic
form, venlafaxine. One could hypothesize that this finding could
be the result of the dosage of this drug used in included studies
or pharmacokinetics issues, although we were not able to test this
hypothesis.3639
2015 Wolters Kluwer Health, Inc. All rights reserved.
Copyright 2015 Wolters Kluwer Health, Inc. All rights reserved.
Insomnia, Somnolence, and Antidepressants
Pertaining to sertraline, the high incidence of insomnia asso-
ciated with this drug could be the consequence of its mechanism
of action. Indeed, sertraline has a weak dopaminergic action in the
striatum and the nucleus accumbens,40 thus promoting arousal.
Another important finding of our work is that agomelatine
proved to be the only second-generation antidepressant associated
with a lower likelihood of inducing insomnia compared with pla-
cebo. This seems to be linked to its peculiar action of inhibiting
serotonin 5-HT(2C) receptors and stimulating melatonin M1 and
M2 receptors.4144 However, this effect may not be related to the
overall antidepressant effect.
It is also worth noting that escitalopram and citalopram
were associated with lower rates of insomnia in comparison
with the remaining SSRIs. As to escitalopram, this could be
linked to the highest selectivity and peculiar mechanism of ac-
tion on the serotonergic transporter compared with the other
SSRIs.45 Pertaining to citalopram, our finding could be related
to its weak antagonist action on histamine1 receptor that has
www.psychopharmacology.com 299
Alberti et al Journal of Clinical Psychopharmacology Volume 35, Number 3, June 2015

FIGURE 2. Sensitivity analysis I: Double-blind, randomized placebo-controlled studies. CIs are reported for each drug.

been associated to the higher rates of sedation usually associ-
ated with this drug.46
In the sensitivity analysis focusing on double-blind random-
ized placebo controlled studies, results were mainly similar but
probably closer to what is observed in clinical practice regarding
the rank of compounds.
When we focused on monotherapy studies only, the results
of the general analysis were largely confirmed. However,
duloxetine and fluoxetine were associated with higher rates
of insomnia than those observed in the general analysis. A
possible explanation for these findings could be related to
the notion that the concomitant use of benzodiazepines and
benzodiazepine-like drugs could interfere with the result ob-
served of the general analysis. As an example, it has been well
documented that fluoxetine, beside its serotonergic action, in-
creases also dopamine and norepinephrine levels in prefrontal
cortex4749 to higher levels than other SSRIs, thus being more
likely to induce insomnia.4852
Finally, also in the sensitivity analysis focusing on non-
sponsored studies, the overall results were confirmed, although
with some differences.
Pertaining to somnolence, we found that the following de-
creasing order of incidence: fluvoxamine, mirtazapine, reboxetine,
paroxetine, desvenlafaxine, duloxetine, venlafaxine, sertraline,
citalopram, fluoxetine, and escitalopram. As to fluvoxamine, this
association was already reported by the review of Wagner et al.
that classified somnolence as the second most frequent side effect
of fluvoxamine over the short-term period.53 This effect could be
probably related to the inhibition of melatonin degradation asso-
ciated with this drug.54,55 Nevertheless, it is worth noting that a
reduction of melatonin levels has been associated with depressed
mood and sleep disorders.56,57 Therefore, the influence on
melatonin metabolism could be useful for the treatment of
MD by improving both mood and insomnia associated with this
disease.58,59 Moreover, fluvoxamine confidence intervals were
very large, indicating much heterogeneity across studies, this

300 www.psychopharmacology.com 2015 Wolters Kluwer Health, Inc. All rights reserved.

Copyright 2015 Wolters Kluwer Health, Inc. All rights reserved.

Journal of Clinical Psychopharmacology Volume 35, Number 3, June 2015
was the case also for insomnia. In fact, in RCT studies and par-
ticularly in nonsponsored studies, fluvoxamine had the lowest
insomnia rates compared to the other compounds. Therefore,
fluvoxamine may be beneficial in depressed subjects presenting
insomnia but with attention on the individual reaction, which may
vary considerably.
Pertaining to mirtazapine, the main reason that could explain
the association between this drug and high levels of somnolence
could be its high affinity for histamine H1 receptors with a conse-
quent antihistaminic effect,60 thus resulting in a higher somno-
lence rates than several SSRIs.61
Reboxetine was the third drug associated with somnolence
after mirtazapine. This finding seems at odds with the mechanism
of action of reboxetine but could be due to a rebound effect of noc-
turnal insomnia, given its REM sleep disruption. However, we un-
derscore the dearth of studies reporting the incidence of this side
effect during treatment with reboxetine. As a consequence, this re-
sult should be considered with caution.
A further relevant finding was that bupropion was the only
drug associated with a lower incidence of this side effect than pla-
cebo. This result is in agreement with the aforementioned bio-
chemical activity of the drug and with its more activating effects.
The results of the sensitivity analysis focusing on double-
blind randomized placebo controlled studies confirmed the over-
all results of the general analysis. However, it is worth noting that,
unlike the general analysis, mirtazapine showed a higher likeli-
hood of inducing insomnia than fluvoxamine. This result seems
to be consistent with the aforementioned pharmacodynamic
properties, especially if one considers that the inhibition of hista-
minergic receptors could be more relevant in inducing a sedative
state than fluvoxamine's melatonergic action. Furthermore, a re-
cent study demonstrated that fluvoxamine has no significant
effect on histaminergic system, whereas mirtazapine has a high
affinity for H1 receptors,62 thus explaining the higher impact of
mirtazapine on weight gain. When we focused on antidepressants
monotherapy studies only and on nonsponsored studies, overall
findings were mainly confirmed.
The reviewed findings are noteworthy and could have impor-
tant clinical implications. Indeed, patients with MD may present
various forms of sleep disturbances, ranging from insomnia to
hypersomnia. So it is important to carefully consider each antide-
pressants effect on sleep-wake rhythm as a criterion for choosing
the most appropriate treatment on the basis of the individual
symptom profile.
However, several limitations should be considered before firm
conclusions are drawn from the results of the present meta-analysis.
First, we included both controlled and uncontrolled studies
as well as both double-blind, single-blind, and open trials, al-
though double-blind randomized controlled studies showed
superiority to other study designs and are usually preferred for
meta-analytic purposes.63,64 Note, however, that we performed a
sensitivity analysis focusing on randomized, double-blind placebo
controlled studies, finding overall results that largely overlap
those observed in the general analysis, although with interesting
differences. However it has been suggested that outpatients may
show a reduced response (and compliance) in outpatient settings;
this is another possible bias in the present analysis.
Second, for studies that did not include a placebo control group,
we calculated a virtual placebo. Nevertheless, it is worth mentioning
that in previous meta-analyses, an analysis aimed at comparing the re-
sults observed in placebo-controlled trials separately analyzed with
those derived including virtual placebo samples found no significant
differences between the 2 inclusion criteria.20
Third, we decided to evaluate the outcomes of our interest
with dichotomous measures, although some studies reported
2015 Wolters Kluwer Health, Inc. All rights reserved.
Copyright 2015 Wolters Kluwer Health, Inc. All rights reserved.
Insomnia, Somnolence, and Antidepressants
changes in insomnia and somnolence associated with antide-
pressants as continuous measures. However, our choice derived
from the observation that the majority of studies reported in lit-
erature provided dichotomous data. Furthermore, studies applied
different methods for detecting these symptoms, ranging from
side effect rating scales to clinical observation; this may have in-
troduced a bias. Finally, a more detailed analysis including inten-
sity, duration, necessity of a specific treatment, and so on would
add more from a clinical point of view. However, this information
is not available in literature or to a very low extent.
Fourth, we included studies with a small sample size,
which limits their power to detect differences among different
treatment options.
Fifth, our analyses focused only on the acute-phase treatment
of MD and cannot be generalized to periods longer than 12 weeks.
Also, multiple-treatments meta-analysis methods have been
recently used to simulate head-to-head comparisons. However,
criticisms have been raised against those methods; therefore, we
followed a more conservative approach.
Finally, we did not consider the impact of the adminis-
tered dosage in the outcomes of our interest, and we included
groups of patients treated with different dosages for the
same antidepressant.
CONCLUSION
In conclusion, the results of the present meta-analysis indi-
cate that the majority of second-generation antidepressants are as-
sociated to a higher risk than placebo of inducing insomnia and/or
somnolence during the short-term treatment of MD. Knowledge
of the specific sleep modulation properties of each drug is impor-
tant for compliance and may also be beneficial to treat specific de-
pression profiles were insomnia or somnolence are predominant.
Moreover, for an effective individualized treatment, other proper-
ties of the compounds, such as on weight gain or on sexual dys-
function, should be considered.
Desspite the potential clinical usefulness of these findings,
we suggest that data about several antidepressants should be con-
sidered with caution, pointing to the need for larger properly
powered randomized controlled studies aimed at exploring the dif-
ferences between placebo and antidepressants in terms of insom-
nia and somnolence rates.
AUTHOR DISCLOSURE INFORMATION
This study was partially supported by Abbott Laboratories.
The authors have no other relevant affiliations or financial in-
volvement with any organization or entity in conflict with the sub-
ject matter or materials discussed in this manuscript apart from
those disclosed.
Dr Serretti is or has been consultant/speaker for Abbott,
Angelini, AstraZeneca, Clinical Data, Boheringer, Bristol-Myers
Squibb, Eli Lilly, GlaxoSmithKline, Innovapharma, Italfarmaco,
Janssen, Lundbeck, Pfizer, Sanofi, and Servier. Dr Chiesa is
or has been consultant/speaker for Abbott, Innovapharma.
Dr Andrisano and Dr Alberti have no conflict of interest to
disclose.
REFERENCES
1. World Health Organization WHO: The world health report 2001Mental
Health: New Understanding, New Hope 2001. Available from:
http://www.who.int/whr/2001/en/. 2001. Accessed April 14, 2014.
2. Hasin DS, Goodwin RD, Stinson FS, et al. Epidemiology of major
depressive disorder: results from the National Epidemiologic Survey on
Alcoholism and Related Conditions. Arch Gen Psychiatry. 2005;62:
10971106.
www.psychopharmacology.com 301
Alberti et al
3. Lopez AD, Mathers CD, Ezzati M, et al. Global and regional burden of
disease and risk factors, 2001: systematic analysis of population health
data. Lancet. 2006;367:17471757.
4. Blair-West GW, Mellsop GW, Eyeson-Annan ML. Down-rating lifetime
suicide risk in major depression. Acta Psychiatr Scand. 1997;95:259263.
5. Murray CJ, Lopez AD. Alternative projections of mortality and disability
by cause 19902020: Global Burden of Disease Study. Lancet. 1997;349:
14981504.
6. Gensichen J, Petersen JJ, Von Korff M, et al. Cost-effectiveness of
depression case management in small practices. Br J Psychiatry. 2013;202:
441446.
7. American Psychiatric Association Practice Guideline For The Treatment
of Patients With Major Depressive Disorder, 3rd edition. 2010.
http://psychiatryonline.org/pb/assets/raw/sitewide/practice_guidelines/
mdd.pdf. Accessed April 14, 2014.
8. Taylor D, Paton C, Kapur S. The Maudsley Prescribing Guidelines.
10th Edition. London: Informa Health Care; 2009.
9. Gartlehner G, Hansen RA, Morgan LC, et al. Comparative benefits
and harms of second-generation antidepressants for treating major
depressive disorder: an updated meta-analysis. Ann Intern Med.
2011;155:772785.
10. Spina E, Santoro V, DArrigo C. Clinically relevant pharmacokinetic
drug interactions with second-generation antidepressants: an update.
Clin Ther. 2008;30:12061227.
11. Thaler KJ, Morgan LC, Van Noord M, et al. Comparative effectiveness of
second-generation antidepressants for accompanying anxiety, insomnia,
and pain in depressed patients: a systematic review. Depress Anxiety.
2012;29:495505.
12. Cheeta S, Schifano F, Oyefeso A, et al. Antidepressant-related deaths
and antidepressant prescriptions in England and Wales, 19982000.
Br J Psychiatry. 2004;184:4147.
13. Buckley NA, McManus PR. Fatal toxicity of serotoninergic and other
antidepressant drugs: analysis of United Kingdom mortality data.
BMJ. 2002;325:13321333.
14. Henry JA, Alexander CA, Sener EK. Relative mortality from overdose
of antidepressants. BMJ. 1995;310:221224.
15. Thanacoody HK, Thomas SH. Tricyclic antidepressant poisoning:
cardiovascular toxicity. Toxicol Rev. 2005;24:205214.
16. Anderson IM. Selective serotonin reuptake inhibitors versus tricyclic
antidepressants: a meta-analysis of efficacy and tolerability. J Affect Disord.
2000;58:1936.
17. Bull SA, Hunkeler EM, Lee JY, et al. Discontinuing or switching selective
serotonin-reuptake inhibitors. Ann Pharmacother. 2002;36:578584.
18. Maddox JC, Levi M, Thompson C. The compliance with antidepressants
in general practice. J Psychopharmacol. 1994;8:4852.
19. Serretti A, Chiesa A. Treatment-emergent sexual dysfunction related to
antidepressants: a meta-analysis. J Clin Psychopharmacol. 2009;29:
259266.
20. Serretti A, Mandelli L. Antidepressants and body weight: a comprehensive
review and meta-analysis. J Clin Psychiatry. 2010;71:12591272.
21. Wirz-Justice A, Benedetti F, Berger M, et al. Chronotherapeutics
(light and wake therapy) in affective disorders. Psychol Med.
2005;35:939944.
22. Mayers AG, Baldwin DS. Antidepressants and their effect on sleep.
Hum Psychopharmacol. 2005;20:533559.
23. Serretti A, Cusin C, Benedetti F, et al. Insomnia improvement during
antidepressant treatment and CLOCK gene polymorphism. Am J Med
Genet B Neuropsychiatr Genet. 2005;137:3639.
24. Serretti A, Olgiati P, Liebman MN, et al. Clinical prediction of
antidepressant response in mood disorders: linear multivariate vs. neural
network models. Psychiatry Res. 2007;152:223231.
302 www.psychopharmacology.com
Copyright 2015 Wolters Kluwer Health, Inc. All rights reserved.
Journal of Clinical Psychopharmacology Volume 35, Number 3, June 2015
25. Jadad AR, Moore RA, Carroll D. Assessing the quality of reports of
randomized clinical trials: is blinding necessary? Controlled Clinical
Trials. 1996;17:112.
26. Wells GA, Shea B, OConnell D. The Newcastle-Ottawa Scale (NOS) for
assessing the quality of nonrandomised studies in meta-analysis. 2005;
Available from: http://www.ohri.ca/programs/clinical_epidemiology/
oxford.asp. Accessed April 14, 2014.
27. Durrleman S, Chaikin P. The use of putative placebo in active control trials:
two applications in a regulatory setting. Stat Med. 2003;22:941952.
28. Jones BE. From waking to sleeping: neuronal and chemical substrates.
Trends Pharmacol Sci. 2005;26:578586.
29. Smith Y, Kieval JZ. Anatomy of the dopamine system in the basal ganglia.
Trends Neurosci. 2000;23:S2833.
30. Osaka T, Matsumura H. Noradrenaline inhibits preoptic sleep-active
neurons through alpha 2-receptors in the rat. Neurosci Res. 1995;21:
323330.
31. Dhillon S, Yang LP, Curran MP. Bupropion: a review of its use in the
management of major depressive disorder. Drugs. 2008;68:653689.
32. Stahl SM, Pradko JF, Haight BR, et al. A review of the neuropharmacology
of bupropion, a dual norepinephrine and dopamine reuptake inhibitor.
Prim Care Companion J Clin Psychiatry. 2004;6:159166.
33. Richelson E. Synaptic effects of antidepressants. J Clin Psychopharmacol.
1996;16:1S7S; discussion 7S9S.
34. Penalva RG, Lancel M, Flachskamm C, et al. Effect of sleep and sleep
deprivation on serotonergic neurotransmission in the hippocampus:
a combined in vivo microdialysis/EEG study in rats. Eur J Neurosci.
2003;17:18961906.
35. Bjorvatn B, Gronli J, Hamre F, et al. Effects of sleep deprivation on
extracellular serotonin in hippocampus and frontal cortex of the rat.
Neuroscience. 2002;113:323330.
36. Clayton AH, Kornstein SG, Rosas G, et al. An integrated analysis of the
safety and tolerability of desvenlafaxine compared with placebo in the
treatment of major depressive disorder. CNS Spectr. 2009;14:183195.
37. Nichols AI, Focht K, Jiang Q, et al. Pharmacokinetics of venlafaxine
extended release 75 mg and desvenlafaxine 50 mg in healthy CYP2D6
extensive and poor metabolizers: a randomized, open-label, two-period,
parallel-group, crossover study. Clin Drug Investig. 2011;31:155167.
38. Preskorn SH, Nichols AI, Paul J, et al. Effect of desvenlafaxine on the
cytochrome P450 2D6 enzyme system. J Psychiatr Pract. 2008;14:
368378.
39. McAlpine DE, Biernacka JM, Mrazek DA, et al. Effect of cytochrome
P450 enzyme polymorphisms on pharmacokinetics of venlafaxine.
Ther Drug Monit. 2011;33:1420.
40. Kitaichi Y, Inoue T, Nakagawa S, et al. Sertraline increases extracellular
levels not only of serotonin, but also of dopamine in the nucleus accumbens
and striatum of rats. Eur J Pharmacol. 2010;647:9096.
41. Zupancic M, Guilleminault C. Agomelatine: a preliminary review of a
new antidepressant. CNS Drugs. 2006;20:981992.
42. Racagni G, Riva MA, Popoli M. The interaction between the internal
clock and antidepressant efficacy. Int Clin Psychopharmacol.
2007;22(Suppl 2):S9S14.
43. Quera-Salva MA, Lemoine P, Guilleminault C. Impact of the novel
antidepressant agomelatine on disturbed sleep-wake cycles in depressed
patients. Hum Psychopharmacol. 2010;25:222229.
44. Lemoine P, Guilleminault C, Alvarez E. Improvement in subjective sleep
in major depressive disorder with a novel antidepressant, agomelatine:
randomized, double-blind comparison with venlafaxine. J Clin Psychiatry.
2007;68:17231732.
45. Owens MJ, Knight DL, Nemeroff CB. Second-generation SSRIs: human
monoamine transporter binding profile of escitalopram and R-fluoxetine.
Biol Psychiatry. 2001;50:345350.
2015 Wolters Kluwer Health, Inc. All rights reserved.
Journal of Clinical Psychopharmacology Volume 35, Number 3, June 2015
46. Sanchez C, Gruca P, Bien E, et al. R-citalopram counteracts the effect of
escitalopram in a rat conditioned fear stress model of anxiety. Pharmacol
Biochem Behav. 2003;75:903907.
47. Jordan S, Kramer GL, Zukas PK, et al. In vivo biogenic amine efflux
in medial prefrontal cortex with imipramine, fluoxetine, and fluvoxamine.
Synapse. 1994;18:294297.
48. Pozzi L, Invernizzi R, Garavaglia C, et al. Fluoxetine increases extracellular
dopamine in the prefrontal cortex by a mechanism not dependent on
serotonin: a comparison with citalopram. J Neurochem. 1999;73:
10511057.
49. Tanda G, Carboni E, Frau R, et al. Increase of extracellular dopamine in the
prefrontal cortex: a trait of drugs with antidepressant potential?
Psychopharmacology (Berl). 1994;115:285288.
50. Bymaster FP, Zhang W, Carter PA, et al. Fluoxetine, but not other selective
serotonin uptake inhibitors, increases norepinephrine and dopamine
extracellular levels in prefrontal cortex. Psychopharmacology (Berl).
2002;160:353361.
51. Clark RN, Ashby CR Jr, Dewey SL, et al. Effect of acute and chronic
fluoxetine on extracellular dopamine levels in the caudate-putamen and
nucleus accumbens of rat. Synapse. 1996;23:125131.
52. Ichikawa J, Meltzer HY. Effect of antidepressants on striatal and
accumbens extracellular dopamine levels. Eur J Pharmacol. 1995;
281:255261.
53. Wagner W, Plekkenpol B, Gray TE, et al. Review of fluvoxamine safety
database. Drugs. 1992;43(Suppl 2):4853; discussion 5354.
54. Hartter S, Wang X, Weigmann H, et al. Differential effects of fluvoxamine
and other antidepressants on the biotransformation of melatonin.
J Clin Psychopharmacol. 2001;21:167174.
55. von Bahr C, Ursing C, Yasui N, et al. Fluvoxamine but not citalopram
increases serum melatonin in healthy subjects an indication that
2015 Wolters Kluwer Health, Inc. All rights reserved.
Copyright 2015 Wolters Kluwer Health, Inc. All rights reserved.
Insomnia, Somnolence, and Antidepressants
cytochrome P450 CYP1A2 and CYP2C19 hydroxylate melatonin.
Eur J Clin Pharmacol. 2000;56:123127.
56. Mendlewicz J, Linkowski P, Branchey L, et al. Abnormal 24 hour pattern
of melatonin secretion in depression. Lancet. 1979;2:1362.
57. Rahman SA, Marcu S, Kayumov L, et al. Altered sleep architecture
and higher incidence of subsyndromal depression in low endogenous
melatonin secretors. Eur Arch Psychiatry Clin Neurosci. 2010;260:
327335.
58. Sunami E, Usuda K, Nishiyama Y, et al. A preliminary study of
fluvoxamine maleate on depressive state and serum melatonin levels
in patients after cerebral infarction. Intern Med. 2012;51:11871193.
59. Dalery J, Honig A. Fluvoxamine versus fluoxetine in major depressive
episode: a double-blind randomised comparison. Hum Psychopharmacol.
2003;18:379384.
60. de Boer TH, Maura G, Raiteri M, et al. Neurochemical and autonomic
pharmacological profiles of the 6-aza-analogue of mianserin, Org 3770
and its enantiomers. Neuropharmacology. 1988;27:399408.
61. Behnke K, Sogaard J, Martin S, et al. Mirtazapine orally disintegrating
tablet versus sertraline: a prospective onset of action study. J Clin
Psychopharmacol. 2003;23:358364.
62. Sato H, Ito C, Tashiro M, et al. Histamine H receptor occupancy by the
new-generation antidepressants fluvoxamine and mirtazapine: a positron
emission tomography study in healthy volunteers. Psychopharmacology
(Berl). 2013;230:227234.
63. Schulz KF, Chalmers I, Hayes RJ, et al. Empirical evidence of bias.
Dimensions of methodological quality associated with estimates of
treatment effects in controlled trials. JAMA. 1995;273:408412.
64. Colditz GA, Miller JN, Mosteller F. How study design affects outcomes
in comparisons of therapy. I Medical Stat Med. 1989;8:441454.
www.psychopharmacology.com 303