Table 1.
Characteristics of included studies evaluating trazodone for primary and secondary insomnia
| Study (Year) | Population Features (N) | Study Design | Trazodone Dosage mg/d (N) | Duration | Sleep Measures (Subj/Obj) | Conclusions | Strengths and Limitations |
|---|---|---|---|---|---|---|---|
| Morin et al14 (2016) | Primary insomnia Age ≥21 (N=224) | Randomized controlled trial 2-tx stages | 50–150 mg/d (N=not indicated) | 6 weeks | -sleep diaries -PSG -(Subj/Obj) | Ongoing | Strengths: Dual-site RCT; large target sample size; remitters will be followed for 12 months; use of both clinically relevant primary outcomes; inclusion of insomnia patients with and without psychiatric comorbidity. Limitations: N/A |
| Roth et Roth et al15 (2011) | Primary insomnia Age range 18–65 (N=16) | Randomized double-blind, placebo-controlled | 50mg/d (N=16) | 3 weeks | -PSG (Obj) | Trazodone is efficacious for sleep maintenance but may be associated with motor and cognitive impairments | Strengths: Strict inclusion of primary insomniacs; use of polysomnography to confirm diagnosis. Limitations: Small sample; of the 63 individuals who gave informed consent, 47 did not complete the study; no a priori justification of sample size. |
| Zavesicka et al16 (2008) | Primary insomnia Mean age 46 (N=20) | Randomized comparative trial CBT vs. CBT+Traz | 100mg/d (N=10) | 8 weeks | self-reported -PSG (Subj/Obj) | CBT monotherapy & combined with trazodone are effective for short-term management of chronic primary insomnia | Strengths: Randomized design; comparative clinical trial to CBT. Limitations: Single-site study; effects of trazodone were not compared with placebo; small treatment groups |
| Wichniak et al17 (2007) | Primary insomnia Mean age 46 (N=20) | Randomized comparative trial CBT vs. CBT+Traz | 25–150 mg/d (N=28) | 3 months | -LSEQ (Subj) | Trazodone improves sleep quality and daytime functioning | Strengths: Three-month study; inclusion of multiple patient-reported sleep scales. Limitations: Small sample size. |
| Walsh et al18 (1998) | Primary insomnia Age range 21–65 (N=306) | Randomized, double-blind, placebo-controlled | 50mg/d (N=100) | 2 weeks | Sleep Question (Subj) | Trazodone improves subjective sleep latency, sleep duration, & quality | Strengths: Double-blind, placebo-controlled RCT; large sample size; robust exclusion criteria. Limitations: Only single dose of each drug was included; no inclusion of PSG measures. |
| Giannaccini20 (2016) | Secondary insomnia: Mood disorder Middle-age (N=17) | Pilot study | 10–20 mg/d (N=17) | 1 month | HAM-D (Subj) | Trazodone would rebalance sleep and mood by interacting with | Strengths: Initial pilot study to investigate the involvement of melatonin system in low-dose efficacy of the typical antidepressant, trazodone on insomnia patients with mood disorders; inclusion of medication serum levels. Limitations: Small sample size. |
| Eraslan et al21 (2014) | Secondary insomnia: post-menopausal Mean age 51 (N=83) | Open and randomly assigned to trazodone or zopiclone | 50–100m g/d (N=28) | 4 weeks | HAM-D (Subj) | Both trazodone & zopiclone improved sleep quality and sexual dysfunction | Strengths: Homogenous cohort of post-menopausal women; exclusion of women taking psychotropic drugs known to cause sleep problems. Limitations: Mild attrition; small sample size; did not control for extraneous variables (e.g., hormone levels). |
| Camargos et al22 (2014) | Secondary insomnia: Alzheimer Age >60(N=36) | Randomized, double-blind placebo-controlled | 50mg/d (N=15) | 2 weeks | Actigraphy (Obj) | 50mg was safe and effective in the treatment of insomnia | Strengths: First double-blind, placebo-controlled study of trazodone in patients with Alzheimer’s disease. Limitations: No a priori power calculations; small sample size; no use of PSG data; problems with daily diary recordings. |
| Tanimukai et al23 (2013) | Secondary insomnia: Cancer patients Mean age 61 (N=267) | Observational study | 12.5–50 mg/d (N=30) | Not defined | Request of insomnia prescription (Subj) | Trazodone may be effective in treatment of insomnia and nightmares in patients with cancer | Strengths: Homogenous cohort of advanced cancer patients. Limitations: Single site; small sample; 7-day study duration for each participant; lack of objective measures. |
| Khazaie et al19 (2013) | Secondary insomnia: 3rd trim pregnancy (N=67) | Randomized, placebo-controlled | 50mg/d (N=20) | 6 weeks | Actigraphy (Obj) | Trazodone improved sleep quality during pregnancy | Strengths: First RCT to assess the effect of trazodone in pregnant women; homogenous cohort. Limitations: Relatively small sample; no long-term follow-up past 6 weeks; drug serum levels were not measured; only 3 nights of actigraphy monitoring. |
| Doroudgar et al24 (2013) | Secondary insomnia: Psychiatric inpatients age 18–65 (N=64) | Observational study | 102mg (12.5–300mg/d) (N=30) | Up to 2 weeks | Patient interview and sleep log (Subj) | Trazodone was preferred over quetiapine for improvement of total sleep | Strengths: Stringent exclusion criteria. Limitations: Small sample; variable medication regimen across treatment groups; observational study; no objective measures; single-site study. |
| Stein et al26 (2012) | Secondary insomnia: methadone-maintenance Mean age 38 (N=137) | Randomized, double-blind placebo-controlled | 50–150m g/d Self-titrate (N=69) | 6 months | PSQI PSG (Subj/Obj) | Trazodone did not improve subjective or objective sleep disturbance | Strengths: First placebo-controlled RCT for opiate dependent persons; participant follow-up; sufficient power; use of PSG measure. Limitations: Smaller sample size. |
| Camargos et al25 (2011) | Secondary insomnia: Dementia Mean age 79 (N=178) | Retrospective study | 50mg/d (N=34) 100mg/d (N=1) | Not defined | Not defined | Trazodone is effective in treating for insomnia associated with dementia | Strengths: Representative cohort of older adults with dementia; longitudinal follow-up over 1 year of the study, strict exclusion criteria; comprehensive demographic characteristics provided. Limitations: No objective measures; observational study; absence of a placebo-controlled group |
| Galecki et al27 (2010) | Insomnia | Article in Polish | Article in Polish | Article in Polish | Article in Polish | Trazodone can be effective in the treatment of insomnia with small adverse reactions | Strengths: N/A Limitations: N/A |
| Paterson et al28 (2009) | Healthy men Age range 21–34(N=12) | Randomized, double-blind placebo-controlled | 100 mg/d + caffeine (N=12) | PSG EEG (Obj) | Trazodone improved sleep latency & total sleep time | Strengths: Use of PSG measure; double-blind study. Limitations: Small sample size. | |
| Sheehan29 (2009) | Secondary insomnia: Depressed patients (N=412) | Randomized placebo-controlled | 150–375mg/d (N=206) | 8 weeks | HAM-D (Subj) | Trazodone improved middle & late insomnia | Strengths: Large randomized study sample; inclusion of Intent-to-treat analysis. Limitations: No follow-up after 8 weeks of study. |
| Friedmann et al30 (2008) | Secondary insomnia: Alcohol-detox Ages 18–65 (N=173) | Randomized, double-blind, placebo-controlled | 50–150m g/d (N=88) | 12 weeks | PSQI (Subj) | Trazodone was associated with improved sleep quality, but can lead to increased drinking when stopped | Strengths: Double-blind, placebo-controlled RCT; reasonable follow-up; robust statistical analysis for missing data; stringent exclusion criteria. Limitations: Small sample; single site; no objective measures; study attrition. |
| Paterson et al31 (2007) | Healthy men & rats Age range 21–34(N=12) | Randomized, double-blind placebo-controlled | 100 mg/d + caffeine (N=12) | 4 weeks | PSG LSEQ (Subj/Obj) | Both zolpidem and trazodone improved subjective sleep latency | Strengths: Established an effective model of onset insomnia; experimental design; use of PSG monitoring. Limitations: Translational model; did not account for sensitivity to caffeine; small sample size. |
| Saletu et al32 (2005) | Secondary insomnia: somatoform pain disorders (N=11) | Sleep laboratory study | 100 mg/d (N=6) | PSG Psychometry (Subj/Obj) | SPD induced changes in subjective & sleep quality which was mitigated by trazodone | Strengths: Use of PSG monitoring; homogenous cohort. Limitations: Small sample size. | |
| Kaynak et al33 (2004) | Secondary insomnia: Depressed women age range 20–50 (N=12) | Double-blind, crossover, placebo-controlled | 100 mg/d (N=6) | 14 days | PSG PSQI (Subj/Obj) | Trazodone is effective in the treatment of antidepressant-associated insomnia | Strengths: Clinician rating scale used; objective measures included; PSG recordings were scored blinded; random assignment to treatment; homogeneous female cohort. Limitations: Small sample size; single dose amount. |
| Schwartz et al34 (2004) | Secondary insomnia: psychiatric inpatients (N=15) | Open and randomly assigned to trazodone or zaleplon | 50–100m g/d (N=15) | As-needed basis | Trazodone may be a better agent to promote longer, deeper subjective quality sleep for psychiatric inpatients with insomnia | Strengths: Random assignment; flexible-dosing schedule; homogenous inpatient cohort. Limitations: Small sample size. | |
| Saletu-Zyhlarz et al35 (2003) | Secondary insomnia: depressed patients all age groups (N=549) | Open-label study | 50–300mg/d controlled-release (N=549) | 6 weeks | HAM-D Self-rating (Subj) | Insomnia was ranked to be the most improved symptom after treatment of trazodone | Strengths: Large sample; multicenter study across 80 outpatient clinics; small attrition rate. Limitations: No inclusion of objective measures. |
| Karam-Hage & Brower36 (2003) | Secondary insomnia: Alcohol-dependent Mean age 44 (N=50) | Open pilot study | (105 =/-57 mg) at bedtime (N=16) | 4–6 weeks | Sleep problems questionnaire (Subj) | Significant sleep improvement was reported during treatment with either trazodone or gabapentin | Strengths: First study to compare gabapentin and trazodone for treating insomnia in alcohol-dependent patients; required 4 weeks of abstinence to initiate participation in the study; included study follow-up. Limitations: Small sample; non-randomization; no placebo control; non-blinded; 2 patients admitted to drinking during the study period. |
| Le Bon et al37 (2003) | Secondary insomnia: Alcohol-dependent (N=16) | Double-blind, placebo-controlled | Titrated up to 200 mg/d (N=8) | 4 weeks | PSG (Objective) | Trazodone can be potentially helpful in the treatment of alcohol post-withdrawal insomnia | Strengths: PSG monitoring; inclusion of secondary measures (HRSD); robust inclusion and exclusion criteria to ensure a homogenous cohort. Limitations: Small sample size. |
| Saletu-Zyhlarz et al38 (2002) | Secondary insomnia: Depressed patients Healthy controls (N=22) | Single-blind, crossover, placebo-controlled | 100mg/d (N=11) | 1 night | -PSG -SASAQS (Subj/Obj) | 100mg trazodone increased total sleep and sleep efficiency, but not sleep latency | Strengths: Objective and subjective measures; placebo-controlled design; homogenous cohort. Limitations: Small sample size; short observation duration; no follow-up; evening and morning blood pressure was higher in depressed patients. |
| Saletu-Zyhlarz et al39 (2001) | Secondary insomnia: Dysthymic patients mean age 50 +/-14 (N=22) | Single-blind, crossover, placebo-controlled | 100mg/d (N=11) | 1 night | -PSG -SASAQS (Subj/Obj) | 100mg trazodone improved objective and subjective sleep and awakening quality (increased slow-wave sleep) | Strengths: PSG monitoring and subjective measures; placebo-controlled design; homogenous cohort. Limitations: Small sample. |
| Warner et al40 (2001) | Secondary insomnia: PTSD patients (N=74) | Survey on usefulness of trazodone | dose range 50–200 mg/d (N=74) | Empirically designed question (Subj) | Trazodone was effective in patients with chronic PTSD, insomnia, and frequent nightmares | Strengths: Homogeneous cohort. Limitations: No control group; moderate levels of participant discontinuation due to priapism or daytime sedation; no objective measures; single institution. | |
| Mashiko et al41 (1999) | Secondary insomnia: Depressed patients (N=75) | Dose-finding study; randomized non-controlled | 50, 75, or 100mg/d (N=75) | 6 weeks | Self-rating for sleep (Subj) | Trazodone at dosage 50–100mg/d improved sleep disorders and most effective at 100mg/day | Strengths: First dose-finding study on trazodone that excluded concomitant use of hypnotics during depressive state; multi-dose comparison groups. Limitations: Limited demographic characteristics provided. |
| Haffmans & Vos42 (1999) | Secondary insomnia: Depressed patients with brofaromine-induced insomnia (N=17) | Randomized, double-blind, crossover, placebo-controlled | 50mg/d (N=7) | 1 week | PSG HAM-D (Subj/Obj) | Low doses of trazodone may be safe and effective in the treatment of MAO-I induced insomnia | Strengths: Double-blind, placebo-controlled design with random assignment; exclusion of participants on other psychoactive medications. Limitations: Small sample with large variability in sleep parameters. |
| Yamadera et al43 (1999) | Healthy men (N=12) | Non-randomized, non-controlled | 50–100 mg/d (N=12) | 2 nights | PSG (Obj) | Trazodone increased slow-wave sleep | Strengths: PSG monitoring. Limitations: Small sample; non-randomized single-blind study method. |
| Arriaga et al44 (1997) | Secondary insomnia: Major depression (N=9) | Single-blind | 50–250 mg/d (N=9) | 5 weeks | EEG (Obj) | Trazodone improved objective insomnia and increased amounts of slow-wave sleep | Strengths: PSG monitoring; longer duration study with 2-week placebo run-in period. Limitations: Small sample size; single-blind study method. |
| Nierenberg et al45 (1994) | Secondary insomnia: Depressed fluoxetine bupropion-induced insomnia (N=17) | Randomized, double-blind, crossover, placebo-controlled | 50–100 mg/d (N=17) | 6.5 days (mean) | PSQI (Subj) | Trazodone is an effective hypnotic for patients with antidepressant-associated insomnia | Strengths: Double-blind crossover trial; homogenous cohort. Limitations: Small sample size; no objective measures. |
| Parrino et al46 (1994) | Secondary insomnia: Dysthymic patients middle aged (N=6) | Non-randomized, non-controlled, single-blind | 50–100 mg/d (N=6) | 6 weeks | PSG VAS (Subj/Obj) | No significant sleep initiation and maintenance, while significant increase of slow-wave sleep and reductions of stage 2 | Strengths: PSG monitoring; homogenous cohort; 6-week phase schedule with increasing dosing schedule. Limitations: Small sample size. |
| Van Bemmel et al47 (1992) | Secondary insomnia: Depressed outpatients (N=8) | Non-randomized, non-controlled, single-blind | 300–400 mg/d (N=8) | 5 weeks | PSG (Obj) | Trazodone did not influence slow-wave sleep but did suppress REM sleep | Strengths: PSG monitoring. Limitations: Small sample size and non-randomized design. |
| Scharf & Sachais48 (1990) | Secondary insomnia: Depressed patients (N=6) | Non-randomized, non-controlled, single-blind | 150–400 mg/d (N=6) | 8 weeks | PSG (Obj) | Trazodone significantly improved symptoms of depression and polysomnographic sleep architecture | Strengths: Double-blind, crossover study design; PSG monitoring; 4 dosing schedules. Limitations: Small sample size. |
| Ware & Pittard49 (1990) | Healthy men (N=6) | Double-blind, crossover, placebo-controlled | 50–200 mg/d (N=6) | 4 nights | PSG (Obj) | Trazodone significantly increased time spent in deep sleep without affecting normal sleep architecture | Strengths: Double-blind, crossover study design; PSG monitoring; 4 dosing schedules. Limitations: Small sample size. |
| Jacobsen50 (1990) | Secondary insomnia: Major depression-MAOI-induced insomnia (N=48) | Open trials | 50–75 mg/d (N=48) | Low-dose trazodone may be safe and effective for treating MAOI-induced and other insomnias | Strengths: Homogenous depressed cohort treated with MAOIs. Limitations: Small sample size. | ||
| Botros et al55 (1989) | Secondary insomnia: Depressed patients (N=20) | Randomized, double-blind, parallel-group | 50mg/d (N=10) | 3 weeks | Sleep questionnaires | Trazodone improved sleep quality | Strengths: Randomized, double-blind design; compared efficacy of trazodone with amitriptyline; robust exclusion criteria; investigated the effect of trazodone on cognition and psychomotor functioning. Limitations: Small sample size. |
| Davey51 (1988) | Secondary insomnia: Depressed patients (N=183) | Randomized, double-blind, parallel-group | 50mg/3x/d (N=87) 150mg/d (N=95) | 6 weeks | LSEQ (Subj) | 150mg trazodone produced greater benefit to sleep and quality of sleep during first few weeks | Strengths: Multicenter, double-blind, randomized trial; larger sample; efficacy study of 2 dosing regimens; Limitations: No objective measures. |
| Moon & Davey52 (1988) | Secondary insomnia: Depressed outpatients (N=39) | Randomized, double-blind, parallel-group | 150mg/d (N=19) | 6 weeks | HAM-D LSEQ (Subj) | 150mg trazodone improved ease of getting to sleep and quality of sleep | Strengths: Double-blind, randomized trial; homogeneous depressed cohort; Limitations: Small sample. |
| Mouret et al53 (1988) | Secondary insomnia: Depressed inpatients (N=10) | Non-randomized, non-controlled | 400–600m g/d (N=10) | 5 weeks | PSG (Obj) | Trazodone would be indicated for depressed individuals with insomnia | Strengths: Homogeneous depressed inpatient cohort; PSG monitoring; plasma levels of trazodone monitored. Limitations: Small sample size. |
| Blacker et al54 (1988) | Secondary insomnia: Depressed outpatients (N=227) | Randomized, double-blind, parallel-group | 150mg/d | 6 weeks | LSEQ (Subj) | Trazodone improved sleep quality and ease of getting to sleep and was associated with lower incidence of dry mouth and drowsiness | Strengths: Multicenter, double-blind, randomized trial; larger sample; compared efficacy of trazodone to mianserin, dothiepin, and amitriptyline. Limitations: No objective measures. |
| Ather et al56 (1985) | Secondary insomnia: Elderly depressed (N=149) | Randomized, double-blind, parallel-group | 100–300 mg/d (N=51) | 6 weeks | VAS (Subj) | Trazodone improved subjective reports of sleep | Strengths: Randomized, double-blind design; homogenous elderly cohort. Limitations: Small sample size. |
| Wheatley57 (1984) | Secondary insomnia: depressed patients (N=146) | Randomized, controlled trial | 100–150 mg/d (N=146) | 6 weeks | HAM-D VAS (Subj) | Trazodone administered as a single or twice daily dosage improved sleep | Strengths: Larger sample size; randomized, controlled trial. Limitations: No objective measures. |
| Montgomery58 (1983) | Self-reported “poor sleepers” mean age 61 (N=9) | Non-randomized, non-controlled | 150mg/d (N=9) | 3 weeks | PSG VAS (Subj/Obj) | Trazodone improved subjective sleep ratings during the first 2 weeks | Strengths: Objective measures; initial 2-week placebo dosing schedule. Limitations: Small sample size; non-randomized. |
Abbreviations: PSG = polysomnographic, Obj = objective, Sub = subjective, CBT = cognitive behavioral therapy, VAS = visual analogue scale, HAM-D = Hamilton Depression Scale, PSQI = Pittsburgh Sleep Quality Index, EEG = electroencephalogram, LSEQ = Leeds Sleep Evaluation Questionnaire, NPI = neuropsychiatric inventory, SPD = somatoform pain disorder, SASAQS = Self-assessment of Sleep & Awakening Quality Scale