Sleep Medicine 34 (2017) 126e133

Contents lists available at ScienceDirect

Sleep Medicine
journal homepage: www.elsevier.com/locate/sleep

Review Article

Treatment of insomnia with tricyclic antidepressants: a meta-analysis

of polysomnographic randomized controlled trials
Yang Liu a, b, Xiaomin Xu a, b, Meixue Dong a, b, Shiyu Jia a, b, Youdong Wei a, b, *
a
Department of Neurology, The First Afliated Hospital of Chongqing Medical University, Chongqing, China
b
Chongqing Key Laboratory of Neurobiology, Chongqing, China

a r t i c l e i n f o a b s t r a c t

Article history: Background: Insomnia represents a signicant public health burden worldwide. Antidepressants have
Received 21 October 2016 often been the insomnia treatment of choice in recent decades. Some tricyclic antidepressants (TCAs)
Received in revised form have been shown to improve sleep efciency.
8 March 2017
Objective: Assess the efcacy and safety of TCAs for the treatment of insomnia using a meta-analysis of
Accepted 8 March 2017
Available online 27 March 2017
randomized control trials (RCTs).
Methods: Relevant studies were identied in electronic databases such as PubMed, Cochrane, Embase,
and Web of Science, up until July 2016. We included all polysomnographic (PSG) RCTs using TCAs to treat
Keywords:
Meta-analysis
insomnia. The primary outcome measure was the total sleep time (TST), although other polysomno-
Tricycle antidepressants graphic measures were also investigated. Next-day somnolence and dropout rates were also assessed.
Insomnia Results: The meta-analysis included nine RCTs. TCAs signicantly improved TST compared with placebo
(SMD 0.61, 95% CI 0.50e0.71, P < 0.00001). Participants receiving TCAs were not more likely to drop
out than those receiving a placebo because of adverse side effects (1.71% vs 1.19%, RR 1.37, 95% CI 0.67
e2.80, P 0.39) or any other reason (7.08% vs 8.20%; RR 0.86, 95% CI 0.60e1.23, P 0.42). However,
the incidence of somnolence was higher in participants receiving TCAs (6.06% vs. 3.21%; RR 1.82, 95%
CI 1.10e3.00, P 0.02).
Conclusions: Based on our limited data analysis with two medications at particular doses (most studies
included extremely low doxepin), we assert that TCAs can be an effective pharmacological treatment for
insomnia. TCAs were found to improve sleep outcome measures, with the notable exception of an 82%
increase in somnolence. Overall TCAs have very problematic and dangerous side effects, while TCAs were
not found to increase the dropout rate compared with the placebo.
2017 Elsevier B.V. All rights reserved.

1. Introduction of stress, anxiety, depression, substance abuse, suicide, cardiovas-

Insomnia can be dened as a difculty in initiating sleep,
maintaining sleep, or as a struggle with frequent early-morning
awakenings; of note, the sleeping difculties must signicantly
interfere with daily functioning to receive the diagnosis [1].
Furthermore, these difculties must be present for at least three
months to correspond with the current diagnostic criteria of
insomnia [2]. The prevalence of sleeping problems during the past
12 months ranges from 23% to 56% in the United States, Europe, and
Japan [3], whereas the prevalence of insomnia in these same
countries ranges from 9% to 15% [4]. Insomnia can increase the risk
* Corresponding author.
E-mail address: havonewei@163.com (Y. Wei).
cular disease and can decrease immune functioning [5]. Insomnia
has also been reported to negatively affect work productivity and
cognitive abilities [6,7] and therefore represents a signicant public
health burden worldwide.
Because the pathophysiological origins of insomnia remain
unclear, current research is mainly focused on symptom reduction
and minimizing side effects [8]. Previous studies have shown that
insomniacs can benet from easily available pharmacological
treatments [9,10]. Indeed, the use of pharmacological drugs is the
most widely used treatment method to aid patients with their
insomnia symptoms (approximately 50% in the United States and
Western Europe and up to 90% in Japan) [3]. The pharmacological
management of insomnia in clinical practice includes the use of
benzodiazepines, benzodiazepine receptor agonists (BZRA), and
antidepressants [11,12]. However, previous research has also

http://dx.doi.org/10.1016/j.sleep.2017.03.007
1389-9457/ 2017 Elsevier B.V. All rights reserved.
 Y. Liu et al. / Sleep Medicine 34 (2017) 126e133
revealed signicant concerns regarding potential drug abuse, as
well as tolerance and dependency issues with benzodiazepine and
BZRA use [13]. Therefore, to minimize these signicant side effects,
clinicians decreased their prescriptions of benzodiazepines by 30%
and increased their prescriptions of antidepressants by 100% from
1987 to 1991 [14]. Consequently, antidepressants were often the
insomnia treatment of choice in previous decades [10,15].
Although antidepressants are not as addictive as benzodiaze-
pines or BZRAs, they still can nonetheless cause various effects
depending on the specic drug used [9]. For instance, trimipramine
can reduce REM sleep and nocturnal wakefulness [16]. Indeed,
some tricyclic antidepressants (TCA), such as amitriptyline, tri-
mipramine, and doxepin, have been shown to improve clinician
sleep ratings, augment sleep duration and reduce the number of
nocturnal disturbances; however, other TCAs, such as imipramine
and desipramine, have not been shown to be as efcient [17].
Nonetheless, the therapeutic benet of several TCAs still re-
mains unknown and consequently merits further study to identify
better treatment options for insomnia. Therefore, the goal of the
present study was to conduct a meta-analysis of all published
randomized control trials (RCTs) that used TCAs as an insomnia
treatment to evaluate the treatment efcacy and the adverse effects
associated with various TCAs. Given that some TCAs, such as
imipramine, can cause some daily somnolence [18], we isolated
somnolence as a key side effect to consider separately from other
side effects in our meta-analysis.
2. Methods
For the present meta-analysis, we followed the Preferred
Reporting Items for Systematic Reviews and Meta-Analyses
(PRISMA) guidelines.
2.1. Data sources and search strategy
Relevant studies were systematically identied through
PubMed, Cochrane, Embase, and Web of Science database searches.
All database searches were performed up to August 2016. The En-
glish keywords TCAs and insomnia were searched across all
databases (additional details about the search strategies are pre-
sented in Supplemental Table 1). We did not use any publication
language restrictions in our searches. Conference summaries and
reference lists were also scanned to avoid omitting relevant trials.
2.2. Study selection
The inclusion of studies in the meta-analysis was based on the
following criteria: 1) RCTs focused on the comparison of TCAs with
a placebo condition for the treatment of insomnia, 2) The study
investigated individuals with a diagnosis of insomnia, 3) The
research used either a parallel or a cross-over study design, 4) The
study were selected for the presence of PSG measures, 5) The study
documented efcacy outcomes.
The following exclusion criteria were used to remove studies
from the meta-analysis: 1) Uncontrolled, non-randomized or quasi-
randomized trials, 2) Insomnia was accompanied by other signi-
cant medical disorders, 3) Participants consisted of healthy adults
and the study was using a model of transient insomnia, 4) The
study did not perform a contrast with a placebo condition, 5) Data
included in the study was incomplete or unavailable.
2.3. Outcome measures
The primary polysomnographic (PSG) outcomes were total
sleep time (TST), latency to persistent sleep (LPS), wake after
127
sleep onset (WASO), and sleep efciency (SE). Additional primary
outcomes included next-day somnolence and dropout rate. Pre-
treatment data and post-treatment data were available in all
trials. We used the group mean change scores to calculate the
effects of treatments, the presence of next-day somnolence and
the dropout rate. Drop-out rate calculations included subjects
who discontinued the study for any reason and for adverse side
effects.
2.4. Data extraction and assessment of risk of bias
Two reviewers independently screened the titles and abstracts
of identied RCTs and obtained the full texts of selected studies to
further judge whether they fullled the inclusion criteria. Data
extracted from the RCTs included study designs, general charac-
teristics, intervention type, PSG-related outcomes, dropout rates
and the proportion of next-day somnolence. The methodological
quality of the articles included in the meta-analysis was assessed
by using the Risk of bias tool developed by The Cochrane
Collaboration [19]. Any disagreement was resolved by reaching
consensus.
2.5. Statistical analysis
The meta-analysis was performed using RevMan5.3 software
(Cochrane Information Management System). Only crossover and
parallel designs were included and all studies were assessed based
on the group mean change of the outcome values. We calculated
the standardized mean difference (SMD) along with the 95% con-
dence interval (CI) for all continuous outcomes, such as the TST,
LPS, WASO and SE. We calculated risk ratios (RR) along with the 95%
CI for dichotomous outcomes, such as the dropout rate. If the
outcome data was not statistically heterogeneous (I2 < 50% or
P > 0.10), the data was assessed with Galbraith plots and the I2
statistic and analyzed with a xed effect model. If the statistical
heterogeneity was signicant, we used a random-effects model to
calculate pooled estimates.
We also performed subgroup analyses based on subject age,
sample size, and TCA type, to investigate whether these variables
modulated treatment efcacy. By removing the crossover trials one
at a time, a sensitivity analysis was conducted to reanalyze the
overall effect size associated with TCAs. Finally, we used the funnel
plot method to examine potential publication biases [20].
3. Results
3.1. Study selection and characteristics
The electronic database searches identied 2916 potential
studies: 799 from Pubmed, 1094 from Cochrane, 106 from Embase,
and 917 from Web of Science. Subsequently, that number was
reduced to 46 trials that were scrutinized in full text based on in-
formation found in the title and abstracts. Nine trials [21e29] were
found to meet the inclusion criteria and were included in the meta-
analysis (Fig. 1). Within these nine studies, those that administered
different doses of the same drug were considered as distinct trials.
The general characteristics of each study are reported in Table 1. The
meta-analysis included 968 subjects that were randomized to
receive either TCAs or placebo. Two different TCAs were investi-
gated: doxepin [21,22,24e26,28,29] and trimipramine [23,27]. The
age of the patients ranged in age from 18 to 93 years and the sample
was mostly composed of women (65% vs 34% for men). The dura-
tion of treatment ranged between two days and 12 weeks. Five
trials [22,24e27] used parallel designs and four [21,23,28,29] used
crossover designs.
128 Y. Liu et al. / Sleep Medicine 34 (2017) 126e133

3.2. Efcacy outcomes

Using a xed effect model that included eight of the reported

RCTs (one RCT included two or three active treatments), an effect
size of 0.61 (95% CI 0.50e0.71, P < 0.00001) was found when
comparing the TST improvement of TCA patients compared with
those who received the placebo (Fig. 2). Subjects receiving TCA
therapy had a signicantly reduced LPS compared with those
receiving a placebo (SMD 0.11, 95% CI 0.21e0.00, P 0.04)
and low heterogeneity (P 0.38; I2 7%) (see Fig. 3). In addition,
using the data from four RCTs and a xed effect model for low
signicant heterogeneity (P 0.21; I2 31%), WASO was also found
to signicantly reduced in TCA patients (SMD 0.51, 95%
CI 0.65 to 0.38, P < 0.00001) (see Fig. 4). TCAs were also
associated with an improved SE (SMD 0.46, 95% CI 0.31e0.61,
P < 0.00001); however, the associated statistical heterogeneity was
found to be not signicant (P 0.50; I2 0%) (see Fig. 5).

3.3. Subgroup analysis

We conducted subgroup analyses on the TST data to examine

Fig. 1. Flowchart of study selection.
the potential inuence of age, TCA type, and sample size. Included
trials in the TCA type subgroup analysis compared the efcacy of
tripramine and doxepin relative to a placebo. TST was larger in
doxepin subgroups (SMD 0.61, 95% CI 0.50e0.71, P < 0.00001)
compared with tripramine subgroups (SMD 0.59, 95%
CI 0.10e1.08, P 0.02) (see Fig. 6a).

Table 1
Demographic and clinical characteristics of the included RCTs.

Study Mean Female Center Insomnia Treatment Intervention vs Dose Duration Study TST pre treated TST post treated
age (%) criteria groups placebo (n) range design
(mg/d)

Hajak (1996) 41.3 30% Single DSM IV and the Doxepin vs 10 vs 10 25 3 weeks Crossover 456.5 vs 402.9 462 vs 345.5
center International Placebo
Classication of
Sleep Disorders
Hajak (2000) 47 77% Multiple DSM-IV and the Doxepin vs 24 vs 23 25e50 4 weeks Parallel 378.88 vs 430.25 vs 408.2
centers International Placebo 398.95
Classication of
Sleep Disorders
(ICSD)
Hohagen (1994) 49.4 47% Single DSM-III-R Trimipramine 19 vs 15 50e200 4 weeks Crossover 357.2 vs 366.2 422 vs 382.8
center vs Placebo
Krystal (2010) 71.4 65% Multiple DSM-IV-TR Doxepin 3 mg 82 vs 777 vs 81 3 vs 1 12 weeks Parallel 326.9 vs 322.4 373.7 vs 360.5 vs
centers vs Doxepin vs 320.6 343.7
1 mg vs Placebo
Krystal (2011) 44.5 73% Multiple DSM-IV-TR Doxepin 6 mg 76 vs 75 vs 76 6 vs 3 35 days Parallel 380.3 vs 380.3 417.1 vs 408.5 vs
centers vs Doxepin vs 380.2 385
3 mg vs Placebo
Lankford (2012) NR 65% Multiple DSM-IV-TR Doxepin vs 130 vs 125 6 4 weeks Parallel NR
centers Placebo
Riemann (2002) NR 38% Multiple DSM-III-R Trimipramine 22 vs 21 25e200 28 days Parallel 348.92 vs 371 406.13 vs 380.44
centers vs Placebo
Roth (2007) 42.4 70% Multiple DSM-IV Doxepin 6 mg 67 vs 67 vs 67 6 vs 3 vs 1 2 days Crossover 339.5 vs 339.5 418.4 vs 415.4 vs
centers vs Doxepin vs 67 vs 339.5 vs 407.5 vs 389.6
3 mg vs 339.5
Doxepin 1 mg
vs Placebo
Scharf (2008) 71 61% Multiple DSM-IV Doxepin 6 mg 76 vs 76 vs 76 6 vs 3 vs 1 2 days Crossover 340.8 vs 340.8 398.4 vs 390.6 vs
centers vs Doxepin vs 76 vs 340.8 vs 377.4 vs 360.7
3 mg vs 340.8
Doxepin 1 mg
vs Placebo

NR, not reported.

 Y. Liu et al. / Sleep Medicine 34 (2017) 126e133 129

Fig. 2. Efcacy of tricyclic antidepressants in increasing total sleep time. Forest plot depicting the change in total sleep time for TCAs compared with placebo.

Fig. 3. Efcacy of tricyclic antidepressants in reducing latency to persistent sleep. Forest plot depicting the reduction of sleep latency for TCAs compared with placebo.

Fig. 4. Efcacy of tricyclic antidepressants in reducing wake after sleep onset. Forest plot depicting the reduction of wake after sleep onset for TCAs compared with placebo.

Fig. 5. Efcacy of tricyclic antidepressants in increasing sleep efciency. Forest plot depicting the change in sleep efciency for TCAs compared with placebo.

In the age subgroup analysis, TST increase was greater for
populations above 65 years of age (n > 65) (SMD 0.58, 95%
CI 0.44e0.73, P < 0.00001) compared with those under the age of
65 (SMD 0.62, 95% CI 0.46e0.77, P < 0.00001) (see Fig. 6b).
The effect size of the ve large study samples (n > 50) was
smaller (SMD 0.60, 95% CI 0.49e0.70, P < 0.00001) than that of
the three small study samples (n < 50) (SMD 0.76, 95%
CI 0.33e1.19, P 0.0006) (see Fig. 6c).
No statistically signicant differences were found when
comparing subgroups, indicating that age, TCA type, and sample
size did not signicantly affect TST.
3.4. Acceptability outcomes
We assessed tolerability outcomes with the data from seven
studies; no signicant differences were found between TCA and
130 Y. Liu et al. / Sleep Medicine 34 (2017) 126e133

Fig. 6. (a) Subgroup analysis comparing the effect of TCA type. (b) Subgroup analysis comparing the effect of age. (c) Subgroup analysis comparing the effect of sample size.

placebo use (7.08% vs 8.20%; RR 0.86, 95% CI 0.60e1.23,
P 0.42) (see Fig. 7a). Dropout rates specically related to adverse
side effects were also not statistically signicant between TCA and
placebo use (1.71% vs 1.19%; RR 1.37, 95% CI 0.67e2.80, P 0.39)
(see Fig. 7b). The incidence of next-day somnolence was signi-
cantly different between TCA (6.06%) and placebo (3.21%) groups
(RR 1.82, 95% CI 1.10e3.00, P 0.02) (see Supplemental Fig. 1).
3.5. Sensitivity analysis
Sensitivity analyses, which excluded studies with crossover
designs, conrmed the positive effect of TCAs on TST improvement
(SMD 0.52; 95% CI 0.36e0.67, P < 0.00001) and conrmed that
TCAs did not decrease study heterogeneity (I2 0%; P 0.73)
(Supplemental Fig. 2).
 Y. Liu et al. / Sleep Medicine 34 (2017) 126e133
Fig. 6. Continued
3.6. Quality assessment and publication bias
The quality of the included studies was assessed by the
Cochrane risk-of-bias method (Supplemental Fig. 3a and 3b). Do-
mains with a high risk of bias were reported in only two studies.
Sequence generation and allocation concealment were not clearly
described in most of the studies. Only half of the trials specied that
the investigators were blind to the group allocation. Most studies
provided complete outcome data and showed non-selective
outcome reporting.
In addition, the inverted funnel plots derived from these studies
were found to be approximately symmetrical (see Supplemental
Fig. 4a and 4b).
4. Discussion
Our analyses showed that the use of TCAs was associated with
TST and SE improvements and with WASO and LPS reductions.
The effect of TCAs on TST was not modulated by subject age,
sample size, or TCA type. Furthermore, sensitivity analyses, which
did not include crossover study designs, revealed a relatively
consistent TST effect for patients with insomnia. TCAs were found
to be a relatively safe and acceptable treatment option; the side
effect-associated dropout rates were mildly higher than for pla-
cebo, whereas the dropout rates associated with all other causes
were not different from the rates observed for placebo. However,
the pooled outcomes of the intake trials were not statistically
signicant. The ndings of some of the individual studies
included in our meta-analysis have showed associations of dox-
epin use with headaches, dizziness, dry mouth, diarrhea, and
upper respiratory tract infections, among other symptoms. No
clinically meaningful changes were observed regarding ECG data,
other vital sign measurements, body weight, or physical exami-
nation ndings. Overall, doxepin was well-tolerated.
131
Trimipramine was also well-tolerated, although with the excep-
tion of more frequent occurrences of mouth dryness. Trimipr-
amine was also associated with less frequent restlessness
symptoms compared with doxepin. Somnolence was found to be
statistically elevated 82% by the use of TCAs, no other side effects
were analyzed. As a whole, the efcacy and safety of TCAs were
found to be acceptable and the PSG sleeping outcome measures
were found to be improved after TCA treatment compared with
the placebo. Although we searched for randomized controlled
trials of all tricyclic antidepressants, only studies of doxepin and
trimipramine were retrieved. Our analyses were performed based
on results. It is not clear that these two are representative of the
entire class, and such a broad conclusion is limited.
The results from Yeung's systematic review [12] showed that a
low dose doxepin treatment for only 1e2 nights is sufcient for
improving sleep quality and was shown to be safe. Our results are
also in agreement with a previous study [30], which shows that
adverse effects and study withdrawals did not signicantly differ
between the insomnia patients receiving TCAs and those
receiving a placebo. However, sedating antidepressants are not
always considered as rst-line pharmacological treatment op-
tions, potentially because of the currently limited available evi-
dence demonstrating their efcacy [10,12]. For instance, Buscemi
et al. found that evidence supporting the effectiveness of anti-
depressants is limited and suggested that, in fact, they pose a risk
of harm [11]. However, given the recent review suggesting that
the doxepin and trazodone can provide effective treatment for
insomnia, we conducted a meta-analysis of available studies to
properly ascertain the efcacy and safety of TCAs for the treat-
ment of insomnia. To our knowledge, no other systematic reviews
or meta-analyses on TCAs had previously investigated these
issues.
Our meta-analysis has several advantages over previously
available research ndings. First, we not only focused on comparing
132 Y. Liu et al. / Sleep Medicine 34 (2017) 126e133
Fig. 7. (a) Comparison of withdrawal rates between tricyclic antidepressants and placebo. (b) Comparison of withdrawal rates owing to adverse events between tricyclic anti-
depressants and placebo.
the efcacy and safety of TCA treatments but also separately
analyzed somnolence from other adverse events. Second, we
improved the research design employed by Yeung et al. [12] by
including two TCA types and not just doxepin, and by excluding
subjects for whom insomnia was comorbid with another disease.
There are also several limitations of the present study that should
be highlighted. First, almost all of the studies in our meta-analysis
had only evaluated the effects of short-term TCA treatment and,
therefore, may not be applicable to prolonged treatment regimens.
Future studies are needed to establish the long-term safety and
efcacy of TCAs for the treatment of chronic insomnia. Second, we
only analyzed objective ndings and did not consider patient
subjective sleep reports. Indeed, Morin [31] indicated that different
dimensions of insomnia should be assessed by different modalities
because subjective reports and objective ndings are poorly
correlated in insomnia research. For instance, patients tend to
overestimate LPS and underestimate TST [31]. More studies are
needed to investigate TCA effects on patient subjective outcomes.
Third, we generally assessed the efcacy of a drug by measuring
effect sizes, which differ from clinically signicant treatment effects
and from quality of life improvements.
5. Conclusions
The present study based on a limited data analysis with only
two TCAs (doxepin and trimipramine) at particular doses (most of
the studies were on low-dose doxepin formulations) shows that
TCAs can improve objective PSG outcomes compared with pla-
cebo and that these outcomes are not modulated by subject age,
sample size or TCA type. TCAs were also found to be relatively
safe with acceptable side effects, with the notable exception of an
82% increase in somnolence reported in TCA patients relative to
placebo patients. However, TCAs may have problematic and
dangerous side effects but these were not found to increase the
dropout rate compared with the placebo. Taken together, the
present ndings argue that TCAs can be an effective alternative
pharmacological treatment for insomnia. Future studies, however,
are needed to properly assess the efcacy and safety of TCA at
different doses for long-term use to improve the generalizability
of our conclusions.
Disclosures
The authors declare that they have no conicts of interest.
Acknowledgements
This work was supported by the National Key Clinical Specialties
Construction Program of China, as well as Natural Science Foun-
dation Project of Chongqing Science and Technology Commission
(ID: cstc 2016jcy jA0423).
 Y. Liu et al. / Sleep Medicine 34 (2017) 126e133
Conict of interest
The ICMJE Uniform Disclosure Form for Potential Conicts of
Interest associated with this article can be viewed by clicking on the
following link: http://dx.doi.org/10.1016/j.sleep.2017.03.007.
Appendix A. Supplementary data
Supplementary data related to this article can be found at http://
dx.doi.org/10.1016/j.sleep.2017.03.007.
References
[1] Corine H, Jaap L, Jan BR, et al. Adherence to technology-mediated insomnia
treatment: a meta-analysis, interviews, and focus groups. J Med Internet Res
2015;(9):17.
[2] Kyle SD, Morgan K, Espie CA. Insomnia and health-related quality of life. Sleep
Med Rev 2009;14(1):69e82.
ger D, Poursain B, Neubauer D, et al. An international survey of sleeping
[3] Le
problems in the general population. Curr Med Res Opin 2008;24(1):307e17.
[4] Ohayon MM. Epidemiology of insomnia: what we know and what we still
need to learn. Sleep Med Rev 2002;6(2):97e111.
[5] Taylor DJ, Lichstein KL, Durrence HH, et al. Epidemiology of insomnia,
depression, and anxiety. Sleep 2005;28(11):1457e64.
[6] Lian Y, Xiao J, Liu Y, et al. Associations between insomnia, sleep duration and
poor work ability. J Psychosom Res 2015;78(1):45e51.
[7] Altena E, Werf YDVD, Strijers RLM, et al. Sleep loss affects vigilance: effects of
chronic insomnia and sleep therapy. J Sleep Res 2008;17(3):335e43.
[8] Nowell PD, Mazumdar S, Buysse DJ, et al. Benzodiazepines and zolpidem for chronic
insomnia: a meta-analysis of treatment efcacy. Jama 1997;278(24):2170e7.
[9] Winkler A, Auer C, Doering BK, et al. Drug treatment of primary insomnia: a
meta-analysis of polysomnographic randomized controlled trials. Cns Drugs
2014;28(9):799e816.
[10] Schutterodin S, Broch L, Buysse D, et al. Clinical guideline for the evaluation
and management of chronic insomnia in adults. J Clin Sleep Med Jcsm Off Publ
Am Acad Sleep Med 2008;4(5):487e504.
[11] Buscemi N, Vandermeer B, Friesen C, et al. The efcacy and safety of drug
treatments for chronic insomnia in adults: a meta-analysis of RCTs. J General
Intern Med 2007;22(9):1335e50.
[12] Yeung WF, Chung KF, Yung KP, et al. Doxepin for insomnia: a systematic re-
view of randomized placebo-controlled trials. Sleep Med Rev 2014;19:75e83.
[13] [No authors listed] NIH State-of-the-Science Conference Statement on mani-
festations and management of chronic insomnia in adults. NIH Consens State
Sci Statements 2005;22(1):373e8.
133
[14] Walsh JK, Engelhardt CL. Trends in the pharmacologic treatment of insomnia.
J Clin Psychiatry 1992;53(Suppl 4).
[15] Wilson SJ, Nutt DJ, Alford C, et al. British Association for Psychopharmacology
consensus statement on evidence-based treatment of insomnia, parasomnias
and circadian rhythm disorders. J Psychopharmacol 2010;24(11):1577e601.
[16] Nicholson AN, Pascoe PA, Turner C. Modulation of sleep by trimipramine in
man. Eur J Clin Pharmacol 1989;37(2):145e50.
[17] Mayers AG, Baldwin DS. Antidepressants and their effect on sleep. Hum
Psychopharmacol Clin Exp 2005;20(8):533e59.
[18] Nicholson AN, Pascoe PA, Stone BM. Histaminergic systems and sleep :
studies in man with HI and H2 antagonists. Neuropharmacology 1985;24(3):
245e50.
[19] Goforth HW. Low-dose doxepin for the treatment of insomnia: emerging data.
Expert Opin Pharmacother 2009;10(10):1649e55.
[20] Hurwitz R, Blackmore R, Hazell P, et al. Tricyclic antidepressants for autism
spectrum disorders (ASD) in children and adolescents. Cochrane Database
Syst Rev 2012;3(3). CD008372-CD008372.
[21] Hajak G, Rodenbeck A, Adler L, et al. Nocturnal melatonin secretion and sleep
after doxepin administration in chronic primary insomnia. Pharmacop-
sychiatry 1996;29(5):187e92.
[22] Hajak G, Rodenbeck A, Voderholzer U, et al. Doxepin in the treatment of
primary insomnia: a placebo-controlled, double-blind, polysomnographic
study. J Clin Psychiatry 2001;62(6):453e63.
[23] Hohagen F, Montero RF, Weiss E, et al. Treatment of primary insomnia with
trimipramine: an alternative to benzodiazepine hypnotics? Eur Arch Psychi-
atry Clin Neurosci 1994;244(2):65e72.
[24] Krystal AD, Durrence HH, Scharf M, et al. Efcacy and safety of doxepin 1 mg
and 3 mg in a 12-week sleep laboratory and outpatient trial of elderly subjects
with chronic primary insomnia. Sleep 2010;33(11):1553e61.
[25] Krystal AD, Lankford A, Durrence HH, et al. Efcacy and safety of doxepin 3
and 6 mg in a 35-day sleep laboratory trial in adults with chronic primary
insomnia. Sleep 2011;34(10):1433e42.
[26] Lankford A, Rogowski R, Essink B. Efcacy and safety of doxepin 6mg in a four-
week outpatient trial of elderly adults with chronic primary insomnia. Sleep
Med 2012;13(2):133e8.
[27] Riemann D, Voderholzer U, Cohrs S, et al. Trimipramine in primary insomnia:
results of a polysomnographic double-blind controlled study. Pharmacop-
sychiatry 2002;35(5):165e74.
[28] Roth T, Rogowski R, Hull S, et al. Efcacy and safety of doxepin 1 mg, 3 mg,
and 6 mg in adults with primary insomnia. Sleep 2007;30(11):1555e61.
[29] Scharf M, Rogowski R, Hull S, et al. Efcacy and safety of doxepin 1 mg, 3 mg, and
6 mg in elderly patients with primary insomnia: a randomized, double-blind,
placebo-controlled crossover study. J Clin Psychiatry 2008;69(10):1557e64.
[30] Wilt TJ, Macdonald R, Brasure M, et al. Pharmacologic treatment of insomnia
disorder: an evidence report for a clinical practice guideline by the American
College of Physicians. Ann Intern Med 2016.
[31] Morin CM. Measuring outcomes in randomized clinical trials of insomnia
treatments. Sleep Med Rev 2003;7(3):263e79.