| Reference | Study type | Population | Intervention and comparison | Outcomes | Risk of bias [ «Additional comments for included studies....»2 Additional comments] |
|---|---|---|---|---|---|
| RCT=randomized controlled trial; ISI=Insomnia Severity Index; PSQI=Pittsburgh Sleep Quality Index; MRT=mirtazapine; PLA=placebo; SOL=sleep onset latency; WASO=wake after sleep onset; NA=number of awakenings; TST=total sleep time; SE=sleep efficiency; N/A=not available; GRC=global rate of change=a scale of -5 to +5 where greater value indicates greater improvement of the sleep-related problem. | |||||
| «Nguyen PV, Dang-Vu TT, Forest G, et al. Mirtazapin...»1 | RCT, double blind, placebo-controlled, parallel group. | Adults aged 65y and older with chronic insomnia (n=60); Quebec, Canada. | Mirtazapine 7.5 mg (n=30) vs placebo (n=30) at bedtime. (One week baseline – four
weeks treatment – one week follow-up). Sleep diary and actigraphy at baseline, weak 5 and weak 6. |
|
Low. |
| «Bakker MH, Hugtenburg JG, Bet PM, et al. Effective...»2 | RCT, double blind, placebo-controlled, three parallel groups, multicentre. | Adults aged 18-85y with insomnia disorder with sleep maintenance problem for whom non-pharmacological treatment not efficient (n=80); patients in general practice; The Netherlands. | Mirtazapine 7.5-15 mg (n=27) vs amitriptyline 10-20 mg (n=26) vs placebo (n=27). Total duration 16 weeks. |
|
Low. |
| Reference | Comments |
|---|---|
| ICSD-3=International Classification of Sleep Disorders, version 3, of the American Association of Sleep Medicine. | |
| «Nguyen PV, Dang-Vu TT, Forest G, et al. Mirtazapin...»1 | This was a double-blind, randomized, placebo-controlled study in a geriatric outpatient clinic of a university hospital in Quebec, Canada. The patients were adults aged 64 years and older with chronic insomnia as defined by ICSD-3. Overall, 231 patients were screened, of whom 60 (45 from the community and 15 from the outpatient clinic) were randomized (30 in the mirtazapine 7.5 mg group and 30 in the placebo group). After one-week baseline, medication was given once at bedtime for 28 days, followed by one-week follow-up. Adverse effects were queried once weekly by telephone. The primary efficacy endpoint was the mean change in ISI score from baseline to 28 days post-treatment. Being a responder (clinically meaningful improvement in insomnia) was defined as a reduction of ≥6 points on ISI during the treatment period. Remission was defined as ISI of <8 points. The primary safety endpoints included any adverse events reported during the trial and all adverse events leading to premature withdrawal. Demographic and clinical findings and safety measures were recorded. Only 75% (45 of 60) completed the trial (19 with mirtazapine and 26 with placebo). Sensitivity analyses were computed using the baseline data carry-forward (BDCF) assumption. Age of the patients at entry was between 64 and 85 years representing the elderly Canadian population. The prevalence of comorbidities was 88% (n=53), with a mean of 3.4±2.7 comorbidities and 57% (n=34) had three or more comorbidities. Based on MADRS scores, 50% (n=30) of patients had no symptoms of depression, 47% (n=28) had mild symptoms and 3% (n=2) had moderate symptoms. There was a possible selection bias. Altogether 40% (12 of 30) in the mirtazapine group had used hypnotics prior to study, whereas the corresponding percentage in the placebo group was 13% (4 of 30). This was not discussed in the publication. We (authors of this review) found that this difference is statistically significant (Fisher's exact test p=0.039) and could explain partly the greater withdrawal due to adverse effects in the mirtazapine group. The medication was given for only 4 weeks, due to which longer-term efficacy and safety remain unknown. Also, the dosage was 7.5 mg, which is higher than the low dose (<7.5 mg). 11 (33%) of the randomized patients in the MRT group withdrew vs. 4 (13%) in the PLA group. |
| «Bakker MH, Hugtenburg JG, Bet PM, et al. Effective...»2 | This was a double-blind, randomized, placebo-controlled, multicentre study in general practice setting in the Netherlands. Patients were adults aged 18 to 85 years with insomnia disorder and sleep maintenance problems, for whom non-pharmacological treatment was insufficient, and randomized to mirtazapine (7.5–15 mg/day), amitriptyline (10–20 mg/day), or placebo for 16 weeks, with the optional double-dose regimen for weeks 2–14. ISI scores were assessed at baseline and at 6, 12, 20 and 52 weeks. At baseline ISI score was 19.0 (SD=3.9) on average. The primary outcome was the ISI total score at 6 weeks that was clinically relevant and meaningful with improvement (>7 points lower than baseline) or recovery (total score of ≤10 points). The study was conducted during the COVID-19 pandemic, which complicated the recruitment, and instead of the original target (n=156), only 80 patients were randomized and 78 started treatments. At week 3, 89% (69 of 78) received a follow-up prescription from their general practitioner. Double dose was prescribed to 26%, 42% and 68% in the mirtazapine, amitriptyline and placebo group, respectively. Only 77% (60 of 78) completed the 16-week treatment, where treatment was considered as per protocol for 64% (50 of 78). Despite the planned number of 52 patients per study arm (n=156) was not obtained (ISI at 6 weeks), the type II error was still <10%. |
Results
| Reference | Number of patients (I/C) | Duration of study | Outcomes I (intention to treat) | Outcomes C (Placebo) | Statistical significance P-value or EMD (CI) |
|---|---|---|---|---|---|
| Level of evidence: moderate evidence for effect on insomnia severity up to 6 weeks of treatment. The quality of evidence is downgraded due to small sample sizes, big numbers of withdrawals in the mirtazapine group in study 1, and possible selection bias. The quality of evidence is upgraded due to large effect size for mirtazapine in study 2. Also, the effects remained after sensitivity analyses in study 1. I=intervention; C=comparison; CI=confidence interval; ISI=Insomnia Severity Index, MRT=mirtazapine; AMI=amitriptyline; PLA=placebo; EMD=estimated mean difference; d=Cohen's d value (d=0.5: medium effect; d≥0.8: large effect); NNT=number needed to treat to benefit. |
|||||
| «Nguyen PV, Dang-Vu TT, Forest G, et al. Mirtazapin...»1 | Randomized: 30/30 Completed study 19/26 |
Total length 6 weeks. One week baseline, four weeks treatment, one week follow-up. | Mean change in ISI: -6.5 (95% CI: -8.3 to -4.8) Responders: 50% (10/20) Remission: 40% (8/20) |
Mean change in ISI: -2.9 (-4.4 to -1.4) Responders: 21.4% (6/28) Remission: 3.6% (1/28) |
P=0.003 P=0.038 P=0.001 |
| «Bakker MH, Hugtenburg JG, Bet PM, et al. Effective...»2 | Randomized: Mirtazapine (n=27), amitriptyline (n=26), placebo (n=27) |
16 weeks of treatment. Follow-up until 52 weeks. Main outcomes estimated at end of week 6. | Mean score of ISI (6 wks, MRT): 10.2 (5.6), n=25 Mean score of ISI (6 wks, AMI): 12.7 (5.1), n=25 Improvement on MRT: 52% (13/25) Improvement on AMI: 40% (10/25) Recovery on MRT: 56% (14/25) Recovery on AMI: 36% (9/25) |
Mean score of ISI (6 wks, PLA): 15.8 (6.2), n=22 Improvement on PLA: 13.6% (3/22) Recovery on PLA: 14.3% (3/21) |
EMD: -6.0 (-9.0 to -3.0) d= 0.9 -3.4 (-6.3 to -0.4) d=0.6 P=0.006 NNT= 2.6 (1.6 to 7.1) P=0.044 NNT= 3.8 (2.0 to 41.7) P=0.004 NNT= 2.4 (1.5 to 5.8) P=0.095 NNT= 4.6 (2.2 to ¥) |
Comment: In the 16-week study, from 12 weeks onwards there was no statistically significant differences in ISI scores, and the secondary outcomes included many measures of daytime functioning «Bakker MH, Hugtenburg JG, Bet PM, et al. Effective...»2.
| Reference | Number of patients (I/C) | Duration of treatment | Difference BL vs. end mean (SD) I |
Difference BL vs. end mean (SD) C |
Statistical significance P-value or EMD (CI) |
|---|---|---|---|---|---|
| Level of evidence: low evidence for effect on sleep quality. The quality of evidence is downgraded because PSQI is not a good measure of the severity of insomnia, but a general measure of sleep disturbances. |
|||||
| «Nguyen PV, Dang-Vu TT, Forest G, et al. Mirtazapin...»1 | 27/30 | 4 weeks | 12.3 (3.3) vs. 7.2 (2.8) | 11.4 (3.1) vs. 7.8 (2.9) | P=0.32 |
| «Bakker MH, Hugtenburg JG, Bet PM, et al. Effective...»2 | Mirtazapine (n=25), amitriptyline (n=26), placebo (n=27) | Analyses at 6 weeks | TST (h; PSQI) on MRT: 4.8 (1.1) vs. 6.4 (1.0), n=25 TST (h; PSQI) on AMI: 5.0 (1.2) vs. 6.0 (1.6), n=25 |
TST (h; PSQI) on PLA: 4.7 (1.1) vs. 5.5 (1.1), n=19 | P=0.004 0.9 (0.2 to 1.6) (MRT vs. PLA) P=0.227 0.4 (-0.3 to 1.0) (AMI vs. PLA) |
| Reference | Number of patients (I/C) | Duration of treatment | Difference BL vs. end mean (SD) I |
Difference BL vs. end mean (SD) C |
Statistical significance P-value or EMD (CI) |
|---|---|---|---|---|---|
| Level of evidence: low evidence for effect on SOL. The quality of evidence is downgraded due to study limitations including inconsistency, indirectness, imprecision, and effect of plausible residual confounding. I=intervention; C=comparison; CI=confidence interval; MD=Estimated mean difference. |
|||||
| «Nguyen PV, Dang-Vu TT, Forest G, et al. Mirtazapin...»1 | 27/30 | 4 weeks | 43.9 (35.1) / 35.0 (30.9) | 53.6 (50.4) / 42.2 (32.1) | P=0.46; NS |
| «Bakker MH, Hugtenburg JG, Bet PM, et al. Effective...»2 | Mirtazapine (n=27), amitriptyline (n=26), placebo (n=27) | Analyses at 6 weeks | MRT: 42.3 (43.8) vs. 32.6 (28.9), n=22 AMI: 50.0 (29.9) vs. 61.0 (61.7), n=19 |
PLA: 88.0 (74.7) vs. 67.2 (61.0), n=21 | MRT vs. PLA: NS -16.4 (-38.2 to 15.3) AMI vs. PLA: NS 13.8 (-19.1 to 46.7) |
| Reference | Number of patients (I/C) | Duration of study | Difference BL vs. end mean (SD) I |
Difference BL vs. end mean (SD) C |
Statistical significance P-value or EMD (CI) |
|---|---|---|---|---|---|
| Level of evidence: low evidence for effect on WASO. The quality of evidence is downgraded due to study limitations including inconsistency, indirectness, imprecision, and effect of plausible residual confounding. I= intervention; C=comparison; CI=confidence interval; BL=baseline; SD=standard deviation; MD= estimated mean difference; NS=not significant. |
|||||
| «Nguyen PV, Dang-Vu TT, Forest G, et al. Mirtazapin...»1 | 27/30 | 4 weeks | 89.9 (48.0) / 42.1 (24.6) | 84.7 (52.6) / 82.2 (74.9) | 0.021 |
| «Bakker MH, Hugtenburg JG, Bet PM, et al. Effective...»2 | Mirtazapine (n=27), amitriptyline (n=26), placebo (n=27) | Analyses at 6 weeks |
MRT: 151.9 (58.8) / 98.1 (59.0); n=22 AMI: 118.0 (63.1) / 85.5 (51.1); n=19 |
PLA: 121.3 (72.2) / 96.5 (57.2); n=21 | MRT vs. PLA: NS -25.0 (-54.4 to 4.4) NS AMI vs. PLA: -14.4 (-44.9 to 16.1) |
| Reference | Number of patients (I/C) | Duration of study | Difference BL vs. end mean (SD) I |
Difference BL vs. end mean (SD) C |
Statistical significance P-value or EMD (CI) |
|---|---|---|---|---|---|
| Level of evidence: moderate evidence for effect on TST. The quality of evidence is downgraded due to imprecision. I=intervention; C=comparison; CI=confidence interval; BL=baseline; SD=standard deviation; NS=not significant. |
|||||
| «Nguyen PV, Dang-Vu TT, Forest G, et al. Mirtazapin...»1 | 27/30 | 4 weeks | 354.8 (75.4) / 425 (54) | 336.5 (83.7) / 362 (77) | P=0.006 |
| «Bakker MH, Hugtenburg JG, Bet PM, et al. Effective...»2 | Randomized: Mirtazapine (n=27), amitriptyline (n=26), placebo (n=27) |
Analyses at 6 weeks | MRT: 5.2 (1.2) / 6.9 (1.1), n=22 AMI: 5.7 (1.0) / 6.6 (1.4), n=19 |
PLA: 5.4 (1.2) / 5.9 (1.8), n=21 | MRT vs. PLA: P<0.001 1.2 (0.6 to 1.8) AMI vs. PLA: NS 0.5 (-0.1 to 1.0) |
| Reference | Number of patients (I/C) | Duration of study | Difference BL vs. 4 weeks mean (SD) I | Difference BL vs. 4 weeks mean (SD) C | P-value |
|---|---|---|---|---|---|
| Level of evidence: low. The quality of evidence is downgraded due to a small number of participants. Actigraphy was used only in one study. I=intervention; C=comparison; CI=confidence interval. |
|||||
| «Nguyen PV, Dang-Vu TT, Forest G, et al. Mirtazapin...»1 | 27/30 | 4 weeks | 373.7 (67.0) / 430 (58) P=0.002 |
395.8 (58.9) / 394 (0.10) NS |
P=0.07 |
| «Bakker MH, Hugtenburg JG, Bet PM, et al. Effective...»2 | N/D | - | - | - | - |
| Reference | Number of patients (I/C) | Duration of study | Difference BL vs. end mean (SD) I |
Difference BL vs. end mean (SD) C |
Statistical significance P-value or EMD (CI) |
|---|---|---|---|---|---|
| Level of evidence: moderate evidence for effect on sleep efficiency. The quality of evidence is downgraded due to study limitations including inconsistency, indirectness, imprecision, and effect of plausible residual confounding. I=intervention; C=comparison; CI=confidence interval. |
|||||
| «Nguyen PV, Dang-Vu TT, Forest G, et al. Mirtazapin...»1 | 27/30 | 4 weeks | 0.68 (0.15) / 0.81 (0.08)/ | 0.66 (0.14) / 0.70 (0.16) | P=0.003 |
| «Bakker MH, Hugtenburg JG, Bet PM, et al. Effective...»2 | Randomized: Mirtazapine (n=27), amitriptyline (n=26), placebo (n=27) |
Analyses at 6 weeks | MRT: 62.0 (13.8) / 76.1 (12.1), n=22 AMI: 67.4 (11.6) / 73.4 (13.8), n=19 |
PLA: 61.8 (15.8) / 69.7 (17.2), n=21 | MRT vs. PLA: P=0.014 8.3 (1.6 to 14.9) AMI vs. PLA: NS 0.2 (-6.7 to 7.1) |
| Reference n Duration |
Occurrence n (%) or mean±SD; I |
Occurrence n (%) or mean±SD; C |
P-value |
|---|---|---|---|
| Level of evidence: moderate evidence for adverse effect of drowsiness and vivid dreams. Downgraded due to imprecision. I= intervention; C=comparison (placebo); CI=confidence interval. |
|||
| «Nguyen PV, Dang-Vu TT, Forest G, et al. Mirtazapin...»1 n=27 4 weeks |
≥1 AE 23 (85) Drowsiness 19 (70) Dry mouth 12 (44) Increased appetite 8 (30) Constipation 7 (26) Abnormal dreams 8 (30) Flu-like syndrome 11 (41) Dizziness 8 (30) Tinnitus 3 (11) Photosensitivity 0 (0) Anxiety 0 (0) Falls 0 (0) Weight gain 0.79±1.46 kg |
22 (73) 15 (50) 12 (40) 10 (33) 10 (33) 5 (17) 3 (10) 4 (13) 6 (20) 4 (13) 2 (7) 0 (0) 0.14±1.60 kg |
NS NS NS NS NS NS P=0.012 NS NS NS NS NS NS |
| «Bakker MH, Hugtenburg JG, Bet PM, et al. Effective...»2 MRT, n=25 AMI, n=24 PLA, n=21 4 weeks |
Mirtazapine: (moderate / severe) Drowsiness morning 8 (32) / 4 (16) Drowsiness afternoon 3 (12) / 2 (8) Dry mouth 4 (16) / 1 (4) Vivid dreams 7 (28) / 2 (8) Increased appetite 4 (16) / 3 (12) Weight gain 3 (12) / 4 (16) Constipation 2 (8) / 2 (8) Orthostatic hypotension 1 (4) / 0 (0) Vertigo 0 (0) / 0 (0) Blurred vision 2 (8) / 0 (0) Sexual problems 1 (4) / 1 (4) Headache 1 (4) / 1 (4) Amitriptyline: (moderate/ severe) Drowsiness morning 4 (16.7) / 2 (8.3) Drowsiness afternoon 0 (0) / 0 (0) Dry mouth 6 (25) / 1 (4.2) Vivid dreams 2 (8.3) / 3 /12.5) Increased appetite 2 (8.3) / 0 (0) Weight gain 2 (8.3) / 1 (4.2) Constipation 0 (0) / 1 (4.2) Orthostatic hypotension 0 (0) / 0 (0) Vertigo 0 (0) / 0 (0) Blurred vision 2 (8.3) / 0 (0) Sexual problems 0 (0) / 0 (0) Headache 0 (0) / 0 (0) |
PLA (mod/ severe) 1 (4.8) / 1 (4.8) 3 (14.3) / 1 (4.8) 3 (14.3) / 0 (0) 0 (0) / 2 (9.5) 1 (4.8) / 1 (4.8) 2 (9.5) / 0 (0) 0 (0) / 0 (0) 0 (0) / 2 (9.5) 1 (4.8) / 1 (4.8) 0 (0) / 1 (4.8) 0 (0) / 0 (0) 2 (9.5) / 0 (0) |
P=0.014 NS NS P=0.024 NS NS NS NS NS NS NS NS NS P=0.040 NS NS NS NS NS NS NS NS NS NS NS |
Comment: In the intention-to-treat analysis, there was statistically significant decrease in daytime fatigue (Flinders Fatigue Scale) with low-dose mirtazapine «Bakker MH, Hugtenburg JG, Bet PM, et al. Effective...»2. General fatigue and physical fatigue worsened significantly with amitriptyline at week 20, but not with mirtazapine. Both mirtazapine and amitriptyline as compared to placebo ameliorated work and social adjustment significantly at week 6 but not anymore beyond.
| Reference | Number of patients (I/C) | Time | Estimated mean difference (CI) between MRT and PLA |
Estimated mean difference (CI) between AMI and PLA |
|---|---|---|---|---|
| Level of evidence:moderate evidence for effect on improvement at week 6 and moderate at week 12. The quality of evidence is downgraded due to a small number of participants. GRC was used only in one study. I=intervention; C=comparison; CI=confidence interval; MRT=mirtazapine; AMI=amitriptyline; PLA=placebo. |
||||
| «Nguyen PV, Dang-Vu TT, Forest G, et al. Mirtazapin...»1 | N/D | - | - | - |
| «Bakker MH, Hugtenburg JG, Bet PM, et al. Effective...»2 | Randomized: Mirtazapine (n=27), amitriptyline (n=26), placebo (n=27) |
Week 6 Week 12 Week 20 |
1.8 (0.6 to 3.0); P=0.003 1.6 (0.4 to 2.8); P=0.009 0.5 (-0.7 to 1.7); P=0.42 |
1.2 (0.0 to 2.4); P=0.050 1.3 (0.1 to 2.5); P=0.033 0.0 (-1.3 to 1.2); P=1.0 |
General Comment
1–7.5 mg mirtazapine in the evening have been used to treat insomnia as off-label prescription. However, we did not find any published randomized placebo-controlled clinical trials with low doses of mirtazapine (<7.5 mg) for symptoms of insomnia nor insomnia disorder. For mirtazapine insomnia is not an indication for use as approved by the national competent authority for regulating pharmaceuticals in Finland, and thus mirtazapine is not a first-line option for prescription in treatment of insomnia.
Instead, in the two studies «Nguyen PV, Dang-Vu TT, Forest G, et al. Mirtazapin...»1, «Bakker MH, Hugtenburg JG, Bet PM, et al. Effective...»2, the dosage of mirtazapine was either 7.5 mg or 15 mg. According to the insomnia treatment guidelines in Finland, the highest dose should not exceed 7.5 mg in treatment of insomnia. Mirtazapine acts as a sedative primarily by blocking histamine-1 (H1) receptors. Mirtazapine is a very strong inhibitor of H1-receptors. It is stronger than doxepin, amitriptyline or other antidepressants and even stronger than diphenhydramine «Richelson E. Pharmacology of antidepressants. Mayo...»3, «Richelson E. Interactions of antidepressants with ...»4. With increasing doses mirtazapine starts to act on other receptors as well, including muscarinic and alpha-1 receptors. This may cause especially anticholinergic adverse effects of mirtazapine compared with the effects induced by its low doses.
Flu-like symptoms and vivid dreams were more common with mirtazapine compared with placebo. Further, morning drowsiness was more common with mirtazapine compared with placebo, about which patients should be warned, and they should avoid driving a vehicle at least during the first days of treatment with mirtazapine. No studies, however, have been published concerning the risk related to driving. In addition, theoretically, low doses of mirtazapine might increase risk of falls.
Mirtazapine increases appetite if doses of 15 mg or more are used «Anttila SA, Leinonen EV. A review of the pharmacol...»5. Anecdotally, lower doses of mirtazapine have been associated with weight gain as well, especially in patients who suffer from obesity, since increased appetite and weight gain may be due to histaminergic antagonism. However, there is no evidence thus far that low doses of mirtazapine (<15 mg) increase weight compared with placebo. It is clinically important that mirtazapine does not seem to worsen sleep apnea, but on the contrary, it may decrease the number of central apneas «Prowting J, Maresh S, Vaughan S, et al. Mirtazapin...»6.
Finally, a clinician needs to pay attention to and consider a finding from a US data from more than 2 million patients from the Veterans Administration database, which demonstrated that mirtazapine, when prescribed for insomnia, has increased suicidality more than other options «Kriikku P, Ojanperä I. Fatal concentrations of ant...»7. Unfortunately, the doses of mirtazapine used were not given, and we may not apply these results directly in Finland. The small tablet sizes of 1 mg, 4.5 mg and 7.5 mg that are used off label in Finland, are not available in the USA. But also in a Finnish study the proportional rate of suicidal deaths related to mirtazapine was second highest after bupropion (8). In that study the indication for use of mirtazapine was not known and it is not known. Also it is not known, which tablet formulations were used in the suicides. However, even that mirtazapine is safe in low doses (≤ 7.5 mg), it seems to be toxic in higher doses. This must be kept in mind when prescribing mirtazapine.