Lampl 

et al. The Journal of Headache and Pain

https://doi.org/10.1186/s10194-023-01573-6
(2023) 24:39
The Journal of Headache
and Pain

RESEARCH Open Access

European Headache Federation (EHF) critical

re‑appraisal and meta‑analysis of oral drugs
in migraine prevention—part 1: amitriptyline
Christian Lampl1,2*, Jan Versijpt3†, Faisal Mohammad Amin4, Christina I. Deligianni5, Raquel Gil‑Gouveia6,7,
Tanvir Jassal8, Antoinette MaassenVanDenBrink9†, Raffaele Ornello10†, Jakob Paungarttner2†,
Margarita Sanchez‑del‑Rio11, Uwe Reuter12, Derya Uluduz13, Tessa de Vries9†, Dena Zeraatkar8† and
Simona Sacco10† 

Abstract 
Objective  The aim of this paper is to critically re-appraise the published trials assessing amitriptyline for migraine
prophylaxis.
Methods  We report our methods and results following the Preferred Reporting Items for Systematic Reviews
(PRISMA), by searching MEDLINE, EMBASE, Cochrane CENTRAL, and ClinicalTrials.gov for randomized trials of pharma‑
cologic treatments for migraine prophylaxis. We included randomized trials that compared amitriptyline with pla‑
cebo for migraine prophylaxis in adults. Our outcomes of interest were informed by the Outcome Set for preventive
intervention trials in chronic and episodic migraine (COSMIG) and include the proportion of patients who experi‑
ence a 50% or more reduction in migraine days per month, migraine days per month, and adverse events leading to
discontinuation.
We assessed risk of bias by using a modified Cochrane RoB 2.0 tool and the certainty of evidence by using the GRADE
approach.
Results  Our search yielded 10.826 unique records, of which three trials (n = 622) were eligible for data synthesis and
analysis. We found moderate certainty evidence that amitriptyline increases the proportion of patients who experi‑
ence a 50% or more reduction in monthly migraine days, compared to placebo (relative risk: 1.60 (95% CI 1.17 to 2.19);
absolute risk difference: 165 more per 1,000 (95% CI 47 more to 327 more). We found moderate certainty evidence
that amitriptyline increases the proportion of patients who discontinue due to adverse events compared to placebo
(risk difference: 0.05 (95% CI 0.01 to 0.10); absolute risk difference: 50 more per 1,000 (95% CI 10 more to 100 more).
Conclusions  Our meta-analysis showed that amitriptyline may have a prophylactic role in migraine patients, how‑
ever these results are far from robust. This warrants further large-scale research to evaluate the role of amitriptyline in
migraine prevention.

†
Antoinette MaassenVanDenBrink, Raffaele Ornello, Jakob Paungarttner, Jan
Versijpt, Tessa de Vries, Dena Zeraatkar and Simona Sacco contributed equally
to this work.
*Correspondence:
Christian Lampl
[email protected]
Full list of author information is available at the end of the article

© The Author(s) 2023. Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which
permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the
original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or
other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line
to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory
regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this
licence, visit http://​creat​iveco​mmons.​org/​licen​ses/​by/4.​0/. The Creative Commons Public Domain Dedication waiver (http://​creat​iveco​
mmons.​org/​publi​cdoma​in/​zero/1.​0/) applies to the data made available in this article, unless otherwise stated in a credit line to the data.
Lampl et al. The Journal of Headache and Pain (2023) 24:39 Page 2 of 12

Keywords  Migraine, Amitriptyline, Prophylactic treatment, Meta-analysis

Introduction they also relate to the various adverse effects caused by

Migraine is a highly disabling disease that often requires this drug.
preventive treatment, especially in highly frequent epi- The aim of this paper is to critically re-appraise the
sodic and chronic migraine. Patients need prophylactic published trials assessing amitriptyline for migraine
drugs to reduce the migraine burden, either to decrease prophylaxis. We focus on amitriptyline because, com-
the occurrence of acute attacks and/or the need of anal- pared to other antidepressants, it is the most widely stud-
gesics. All older prophylactic drugs that are used in ied for migraine and thus has the largest evidence base
migraine have been developed for other indications and supporting its efficacy and safety for migraine.
were later found effective in migraine. Tricyclic anti-
depressants (TCAs) were among the first medications Methods
identified as having a preventive benefit for migraine. We report our methods and results following the Pre-
Amitriptyline was discovered in the late 1950s and was ferred Reporting Items for Systematic Reviews (PRISMA)
approved by the U.S. Food and Drug Administration [15].
(FDA) in 1961. The beneficial use of amitriptyline in
migraine was first reported in the late 1960s by Fried- Search strategy (Supplement 1)
man [1] and Mahloudji [2]. Studies of migraine pre- In consultation with an experienced research librarian,
ventive use in the USA show that TCAs are the second we searched MEDLINE, EMBASE, Cochrane CENTRAL,
most prescribed medications for migraine prevention, and ClinicalTrials.gov from inception to August 13, 2022
after topiramate [3]. Amitriptyline is considered as a for randomized trials of pharmacologic treatments for
level B drug for migraine prophylaxis by the American migraine prophylaxis, without language restrictions. We
Headache Society (AHS) and American Academy of supplemented our search by retrieving references of sim-
Neurology (AAN), meaning it is regarded as "probably ilar systematic reviews and meta-analyses [16].
effective” even though it has not been approved by the
FDA for the prophylactic use in migraine [4]. In Europe,
Screening and study eligibility
amitriptyline is considered as a ‘drug of second choice’
[5]. The exact mechanism of action of amitriptyline in Following training and calibration exercises to ensure suf-
migraine prophylaxis is unclear. The neurotransmit- ficient agreement, pairs of reviewers, working indepen-
ter 5-hydroxytryptamine (5-HT, serotonin) is involved dently and in duplicate, reviewed titles and abstracts of
in migraine pathophysiology [6] and the acute antimi- search records and subsequently the full texts of records
graine medication class of triptans targets the 5-HT deemed potentially eligible at the title and abstract
receptor subtypes 5-HT1B/1D/(1F) [7]. TCAs inhibit the screening stage. Reviewers resolved discrepancies by dis-
uptake of 5-HT in the synaptic cleft [8] so it is likely that cussion, or, when necessary, by adjudication with a third
the antimigraine effect of amitriptyline results from its viewer. We included randomized trials that compared
effects on serotonergic transmission. Moreover, inhi- amitriptyline with placebo for migraine prophylaxis in
bition of reuptake of noradrenaline leads to increased adults. We excluded trials that investigated abortive
concentrations of this neurotransmitter in the synaptic rather than prophylactic interventions and trials that ran-
cleft, which could exert antinociceptive effects through domized children or adolescents. We excluded trials that
activation of α2-adrenoreceptors [8, 9]. In addition to randomized fewer than 25 participants as we anticipated
5-HT and noradrenaline reuptake inhibition, TCAs that smaller trials may be unrepresentative and at higher
have multiple other targets, including anticholinergic risk of publication bias [17].
and antihistaminergic effects, they affect sodium, cal-
cium [10] and potassium channels [11], and exert an Data extraction
effect on adrenergic α1-adrenoreceptors, N-methyl-D- Following training and calibration to ensure sufficient
aspartate (NMDA) and opioid receptors [12]. In a rat agreement, pairs of reviewers, working independently
model, amitriptyline was shown to suppress cortical and in duplicate, extracted data from eligible stud-
spreading depression (CSD), which is thought to be ies. Reviewers resolved discrepancies by discussion
the underlying mechanism of migraine aura [13]. These and if necessary, by adjudication with a third party.
many sites of action could potentially contribute to the We extracted trial characteristics, patient characteris-
antimigraine effect of amitriptyline (Fig.  1) [14], but tics, diagnostic criteria, type of migraine, intervention
characteristics, and outcomes of interest at the longest
 Lampl et al. The Journal of Headache and Pain (2023) 24:39 Page 3 of 12

Fig. 1  Potential mechanisms of action for the anti-migraine effect of the tricyclic antidepressant amitriptyline. Amitriptyline inhibits the uptake
of serotonin and noradrenaline in the synaptic cleft, and possibly exerts its antimigraine effects by affecting serotonergic transmission or through
antinociceptive effects via activation of the α2 adrenoreceptor [8]. In addition, tricyclic antidepressants affect sodium [14] calcium [10] and
potassium [11] channels, exert an effect on adrenergic α1, NMDA and opioid receptors [12] and suppress cortical spreading depression (CSD), which
could be underlying migraine aura [13]

reported follow-up time at which patients were still using results. Reviewers resolved discrepancies by discussion
the interventions being investigated. Our outcomes of and if necessary, by adjudication with a third party.
interest were informed by the Outcome Set for preven-
tive intervention trials in chronic and episodic migraine Data synthesis and analysis
(COSMIG) and include the proportion of patients who For all outcomes, we performed frequentist random-
experience a 50% or more reduction in migraine days effects meta-analysis using the restricted maximum
per month, migraine days per month, and adverse events likelihood (REML) estimator [21]. We also performed
leading to discontinuation [18]. We prioritized extracting sensitivity analyses using the Paule-Mandel hetero-
monthly migraine days when reported but also extracted geneity estimator. We analyzed 50% or more reduc-
monthly headache days or monthly migraine attacks tion in monthly migraine days as relative risks, monthly
when monthly migraine days were not reported. migraine days as mean differences, and adverse events
leading to discontinuation as risk differences, since we
Risk of bias assessments expected many studies to report no or few events with
Following training and calibration to ensure suffi- placebo. To facilitate interpretation, we report dichoto-
cient agreement, reviewers working independently mous outcomes as number of events per 1,000 patients.
and in duplicate, assessed risk of bias using a modified We summarize heterogeneity using the ­I2 statistic and
Cochrane RoB 2.0 tool [19, 20]. For each trial, we rated interpret an ­I2 value of 0% to 40% as not important, 30%
each outcome as either ‘low risk of bias’, ‘some concerns – to 60% as moderate heterogeneity, 50% to 90% as sub-
probably low risk of bias’, ‘some concerns –probably high stantial heterogeneity, and 75% to 100% considerable het-
risk of bias’, and ‘high risk of bias’ across the following erogeneity [22].
domains: bias arising from the randomization process, We anticipated that the effects of treatments may vary
bias due to departures from the intended intervention, based on risk of bias, baseline monthly migraine days,
bias due to missing outcome data, bias in measurement and the proportion of patients that had previously used
of the outcome, and bias in selection of the reported prophylactic therapy. To test for subgroup effects based
Lampl et al. The Journal of Headache and Pain (2023) 24:39 Page 4 of 12

on these factors, we performed pairwise meta-regres- authors. Because we were unable to source established
sions comparing results rated at low versus high risk of minimally important differences for migraine from the
bias and trials below versus above the median number of literature, based on consensus of the authors, we con-
monthly migraine days or proportion of patients that had sidered a 15% increase in the proportion of patients
previously used prophylactic therapy. We assessed the who experienced a 50% or more reduction in monthly
credibility of subgroup effects using the ICEMAN tool migraine days, a reduction of 2 monthly migraine days,
[23]. For analyses with 10 or more studies, we planned and a 2% increase in patients who experienced adverse
to test for publication bias by visually inspecting funnel events leading to discontinuation as minimally impor-
plots and Eggers tests [24]. We performed all analyses tant. We report results using GRADE simple language
using the meta and metafor packages in R (version 4.1.2) summaries (i.e., describing high certainty evidence with
[25, 26]. declarative statements, moderate certainty evidence with
‘probably’, low certainty evidence with ‘may’ and very low
Assessment of the certainty (quality) of evidence indicated by ‘very uncertain’) [29].
We assessed the certainty of evidence using the GRADE
approach [27]. For each outcome, we rated certainty Results
of each comparison as either high, moderate, low, or Our search yielded 10.826 unique records, of which five
very low based on risk of bias, inconsistency, indirect- trials were eligible for the narrative description [30–34]
ness, imprecision, and publication bias. We made judge- and three for data synthesis and analysis [31–33]. Figure 2
ments of imprecision using the minimally contextualized presents details about study selection.
approach [28]. The minimally contextualised approach
considers only whether confidence intervals include the Narrative description of amitriptyline
null effect and thus does not consider whether plausible in placebo‑controlled trials
effects, captured by confidence intervals, include both In the first clinical trial, published by Gomersall and
important and trivial effects. To evaluate the certainty Stuart in 1973 [30], amitriptyline (10–60  mg per
of no effect, we used minimally important differences, day) reduced the number of migraine attacks by 42%
sourced from the literature and by consensus from the (p 
< 0.001), in about half of the subjects by  > 50%.

Fig. 2  Selection of studies for the systematic review. Our search yielded a total of 10,826 unique records. After title and abstract screening 1,276
records proved potentially eligible and after full-text review 5 records proved eligible. We excluded records if they did not describe full-text
peer-reviewed reports of randomized trials that compared amitriptyline with placebo for prophylaxis of migraine in adult patients
 Lampl et al. The Journal of Headache and Pain (2023) 24:39 Page 5 of 12

However, only 20 subjects of 26 who initiated did com- with a minimum of two migraine days per month [31,
plete the trial. 33] and one trial recruited patients with a minimum of 4
The Couch and Hassanein (1979) trial used a com- migraine days per month and a maximum of 15 headache
posite migraine score including frequency, severity, and days per month [34]. Table 1 presents the trial character-
duration of attacks as the primary outcome parameter istics and Fig. 3 presents the risk of bias ratings.
for efficacy [31]. This specific score was reduced by more
than 50% in 55% of the amitriptyline-treated patients 50% responder rate
(dose up to 100 mg per day), compared with 34% of the Two trials [33, 34] reported on 50% or more reduction
placebo-treated patients. The therapeutic gain in that in monthly migraine days in 289 patients and one trial
study was 21%. Data on migraine frequency were not [31] reported on 50% responder rate in 100 patients. We
presented, and patients with comorbid depression were performed a sensitivity analysis excluding the trial that
not excluded. reported responder rate. The sensitivity analysis pro-
In another placebo-controlled trial published in 2005 duced results consistent with the main analysis (Fig.  4).
the prophylactic activity of propranolol and amitriptyline Two out of three trials were rated at high risk of bias, due
on frequency, duration and severity of migraine attacks to missing outcome data (Fig.  3). Two of the trials also
was compared in 105 patents. Amitriptyline (25  mg failed to describe methods for allocation concealment.
twice per day) significantly reduced the frequency, dura- We were unable to make confident judgements about
tion and intensity of migraine attacks after treatment of potential for selective reporting due to lack of publicly
45  days [32]. After discontinuation, the rebound effect accessible protocol or registration files for two trials —
was higher than in the propranolol group. likely since these trials were performed/published before
Couch published an analysis of a trial that was per- trial registration practices became common. We found
formed between 1976 and 1979 subsequently in 2011 moderate certainty evidence that amitriptyline probably
[33]. 391 subjects with migraine and chronic daily head- increases the proportion of patients who experience a
ache were included. There was a significant improvement 50% or more reduction in monthly migraine days, com-
in headache frequency for amitriptyline 25 mg over pla- pared to placebo (Table 2; Figs. 4 and 5). The certainty of
cebo at 8 weeks (p 0.018) but not at 12, 16, or 20 weeks. evidence was downgraded by one level due to concerns
There were no significant differences in headache sever- about risk of bias. We anticipated that the effects of ami-
ity or duration between amitriptyline and placebo at any triptyline may be different based on risk of bias (i.e., low
time point during the study. The drop-out rate was 52% vs. high risk of bias), mean monthly migraine days at
at week 20. baseline, and the proportion of patients who reported
Another placebo-controlled trial with 196 patients ran- having previously used prophylactic drugs and had
domized to receive either melatonin as active comparator planned to perform subgroup analyses investigating the
or amitriptyline was published in 2016 [34]. Amitripty- effects of these variables on results. Due to lack of report-
line 25  mg was superior to placebo (p < 0.05) for reduc- ing of mean monthly migraine days at baseline and the
ing migraine days per month after 12  weeks compared proportion of patients who had previously used prophy-
to baseline but not superior to melatonin. Melatonin lactic drugs, we were unable to perform subgroup analy-
was better than amitriptyline for the secondary end- ses addressing these factors. The subgroup analysis based
point (50% responder rate) and was better tolerated than on risk of bias did not suggest that the trial at low risk of
amitriptyline. bias produced results that were different from the trial at
high risk of bias (Fig.  6). A sensitivity analysis using the
Data synthesis and analysis Paule-Mandel heterogeneity estimator yielded results
We included three trials in our quantitative analysis, consistent with the primary analysis (Supplement 2).
including 622 patients [31, 33, 34]. We excluded one trial
from the quantitative analysis since it included only 20 Monthly migraine days
participants [30] and one other trial [32] because it only Only one trial, including 118 patients, reported on the
reported total number of participants and not the num- reduction in monthly migraine days [34]. The trial was
ber of participants in each arm, with migraine attacks rated at low risk of bias (Fig. 3). We found high certainty
and not migraine/headache days as primary outcome evidence that amitriptyline reduces monthly migraine
parameter which precludes analysis. Two of the three days (Table  2). We were unable to perform subgroup
trials were industry-funded and performed in the USA analyses based on risk of bias, mean monthly migraine
[31, 33] and the third trial was funded by a public grant days at baseline, and the proportion of patients who
from Brazil [34]. More than three quarters of patients reported having previously used prophylactic drugs due
were middle-aged women. Two trials recruited patients to too few trials.
 Lampl et al. The Journal of Headache and Pain

Table 1  Trial characteristics

Study Registration Funding Country Mean age % Male Duration of % Aura Migraine days per Interventions Number
migraine month at baseline of
(years) patients
(2023) 24:39

Couch 1979 [30] NR Merck Laboratories US NR 16.0 NR NR NR Amitriptyline 100 mg/day or MTD, 100
titrated over 4 weeks, maintained for
4 weeks, oral
Placebo
Couch 2011 [32] NR Merck, Sharp, and US 34.9 19.0 NR NR NR Amitriptyline 100 mg/day or MTD, 391
Dohme Research titrated over 4 weeks, maintained for
Laboratories 12 weeks, oral
Placebo
Gonçalves 2016 [33] NCT 01357031 Fundação de Brazil 36.9 24.6 22.2 16.1 7.3 Amitriptyline 25 mg/day for 12 weeks 131
Amparo a Pesquisa Placebo
de São Paulo
NR not reported, NCT number clinical trial, MTD maximum tolerable dose
Page 6 of 12
 Lampl et al. The Journal of Headache and Pain (2023) 24:39 Page 7 of 12

Fig. 3  Risk of bias ratings. Two out of three trials and one out of two trials were at high risk of bias due to missing outcome data for 50% or more
reduction in monthly migraine days and adverse events leading to discontinuation, respectively. One trial, reporting on monthly migraine days, was
at low risk of bias

Table 2  Amitriptyline compared to placebo for migraine prophylaxis

Patient or population: migraine
Intervention: prophylaxis with amitriptyline
Comparison: placebo
Outcomes № of Certainty of the Relative effect (95% CI) Anticipated absolute ­effectsa
participants evidence (GRADE)
(studies) Risk with placebo Risk difference with
Amitriptyline

50% or more reduction in 389 (3 RCTs) Moderate RR 1.60 (1.17 to 2.19) 275 per 1,000 165 more per 1,000 (47 more
monthly migraine days (downgraded due to risk to 327 more)
of bias)
Monthly migraine days 118 (1 RCT) High - NA MD 1.2 migraine days fewer
(2.1 fewer to 0.3 fewer)
Adverse events leading to 507 (2 RCTs) Moderate RD 0.05 (0.01 to 0.10) 0 per 1,000 50 more per 1,000 (10 more
discontinuation (downgraded due to risk to 100 more)
of bias)
CI confidence interval, MD mean difference, RR risk ratio, RD Risk difference
GRADE Working Group grades of evidence
High certainty: we are very confident that the true effect lies close to that of the estimate of the effect
Moderate certainty: we are moderately confident in the effect estimate: the true effect is likely to be close to the estimate of the effect, but there is a possibility that it
is substantially different
Low certainty: our confidence in the effect estimate is limited: the true effect may be substantially different from the estimate of the effect
Very low certainty: we have very little confidence in the effect estimate: the true effect is likely to be substantially different from the estimate of effect
a
The risk in the intervention group (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention
(and its 95% CI)

Fig. 4  Sensitivity analysis of analysis for 50% or more reduction in monthly migraine days excluding a trial that reported on a 50% reduction in a
migraine score
Lampl et al. The Journal of Headache and Pain (2023) 24:39 Page 8 of 12

Fig. 5  The forest plot shows pooled relative risk and associated confidence intervals comparing 50% or more reduction in monthly migraine days
for amitriptyline versus placebo

Fig. 6  Subgroup analysis comparing results of trials at low vs. high risk of bias for 50% responder rate

Fig. 7  Forest plot showing meta-analysis comparing amitriptyline with placebo for adverse events leading to discontinuation

Adverse events leading to discontinuation Discussion

Two trials, including 507 patients, reported on adverse Amitriptyline is widely used in the prophylactic treat-
events leading to discontinuation [31, 33]. One of the ment of migraine. Our meta-analysis showed that the
two trials was rated at high risk of bias due to missing tricyclic antidepressant amitriptyline may have a prophy-
outcome data [30]. We found moderate certainty evi- lactic role in migraine patients, however, in view of the
dence that amitriptyline probably increases the propor- studies retrieved and included in our meta-analysis, these
tion of patients who discontinue due to adverse events results are far from robust. This warrants further large-
compared to placebo. The certainty of evidence was scale research to evaluate the role of amitriptyline in
downgraded by one level due to risk of bias (Fig. 7). We migraine prevention. As it is in guidelines, it is often used
were unable to perform subgroup analyses based on in the real-life setting. An adequate registry would be
risk of bias, mean monthly migraine days at baseline, able to collect relevant information on its role in migraine
and the proportion of patients who reported having management. The most important adverse effects of ami-
previously used prophylactic drugs due to too few tri- triptyline are drowsiness and anticholinergic symptoms
als. A sensitivity analysis using the Paule-Mandel het- such as dry mouth, constipation, and tachycardia. Weight
erogeneity estimator yielded results consistent with the gain occurs in many patients together with elevated levels
primary analysis (Supplement 1). of leptin, insulin, and C peptide [35], and can be a limiting
Only one trial reported specific adverse events that factor leading to impaired compliance and discontinua-
led to discontinuation, which included rash, hyperten- tion. Occasionally, amitriptyline may provoke glaucoma,
sion, nausea, and numbness of hands and feet [30]. PQ and QT interval prolongation on electrocardiogram
 Lampl et al. The Journal of Headache and Pain (2023) 24:39 Page 9 of 12

(ECG), as well as benign prostate hypertrophy. Ami- is regarded as "probably effective [39]. In the 2009 revised
triptyline is metabolized by cytochrome P450 (CYP) European guidelines on the drug treatment of migraine,
isoenzymes, particularly CYP2D6, which is responsible amitriptyline is recommended as drug of ­2nd choice for
for multiple interactions [36]. So far, three placebo-con- migraine prophylaxis [5]. Besides these recommenda-
trolled trials found amitriptyline significantly better than tions there is still a need for further clinical trials in indi-
placebo at reducing a headache index or frequency, but viduals of all ages, since it is still based on old trials with
the magnitude of effect, albeit significant as compared to small numbers of participants, different treatment end-
placebo is limited. Furthermore, the trial by Couch pub- points and old regulatory approval standards.
lished in 2011 with patient enrollment initiation between Nowadays, based on standards from Food and Drug
1977 and 1979 showed that amitriptyline was superior to Administration (FDA) and European Medicines Agency
placebo in migraine prophylaxis at 8 weeks but, because (EMA), drugs do not get approved without at least two
of a robust placebo response, not at subsequent time well designed positive placebo-controlled trials (https://​
points. Therefore, this study must be rated as negative. www.​ f da.​ g ov/​ d rugs/​ d evel​ o pment-​ a ppro​ v al-​ p roce​ s s-​
There are many limitations in the described trials, that drugs). Some of the trials considered in this review had
have to be raised and critically analyzed. Some of them limited sample size, which leaves the findings unclear
are listed below: (i) baseline observation period: was this for several outcome measures. Length of follow-up was
prospective or historically driven? Was baseline attack often too short (mean length, 12  weeks; recommended,
frequency measured by a standardized questionnaire 24 weeks), and the clinical outcomes measured (scales or
or not? If not, then this is extremely susceptible to bias. indices) often did not have a well-established rationale
(ii) blinding: how was blinding performed and main- and were not prespecified. The appropriateness of sta-
tained, especially during the titration phase? Can there tistical analyses was a frequent matter of concern, par-
be unblinding, e.g. due to side effects that can be quite ticularly considering multiple treatment comparisons,
pronounced at the high doses of amitriptyline used? (iii) repeated measurements over time, and questionable sub-
analysis: was the analysis of the primary endpoint pro- group analyses.
spectively determined or was there the possibility of a Another heterogeneity is the fact that some of the pre-
retrospective interpretation and selection of only the so- sented studies examined migraine preventive efficacy
called positive endpoints? (iv) outcomes: what was the only in those patients without concomitant depression,
primary endpoint? Was it and the time of assessment whereas others allowed concurrent depression. In the
predetermined or was the most positive endpoint only past few years, the association between migraine and
selected at variable time points after the trial? (v) dropout depression has been described in both clinic- and com-
rate: how were the results adjusted for dropouts? How munity-based populations [40]. Many researchers main-
were dropouts handled? (vi) one of the trials was con- tain that chronic migraine pain can induce a reactive
ducted in the 1970s, but not published until 2011 [33]. depression that becomes more evident the more chronic
This is highly unusual, and raises questions on the solid- the pain is. To explain a development from migraine to
ity of the data, unless one could study the original raw depression, it has been hypothesized that unpredictable
data. (vii) how where different types of headaches diag- attacks of severe pain might lead to anxiety and depres-
nosed and discriminated? Amitriptyline is effective in sion. However, in longitudinal studies, the evidence sup-
tension-type headache, and many patients have a combi- ports a bidirectional relationship between migraine and
nation of both tension-type headache and migraine [37], depression, with each disorder increasing the risk of the
which complicates effect assessment and interpretation if other [41, 42]. In such cases, amitriptyline may provide
the inclusion and end-point definition are too vague and more benefit than other drugs. However, this approach
include both headache types. is not successful in all migraine patients, and finding a
Taken together, the quality of the studies included in means of identifying patients who are likely to respond
the current meta-analysis is questionable. Nevertheless, to amitriptyline should be a high-priority research goal.
one guideline recommends amitriptyline as first line The strengths of the current review include a com-
agent with a dose range between 30 and 150  mg with a prehensive search strategy and rigorous assessment
medium to high efficacy and mild or infrequent side of the certainty of evidence using the latest GRADE
effects [38]. According to the 2012 published guidelines guidance [27]. We also focus on outcomes relevant to
for preventing episodic migraine (defined as headaches patients, informed by an established core outcome set.
that occur fewer than 15 times per month) established by We assessed the certainty of evidence using the GRADE
the American Headache Society (AHS) and the Ameri- approach [27]. While the GRADE framework presents a
can Academy of Neurology (AAN), amitriptyline is a comprehensive framework for considering all factors that
level B medication for migraine prophylaxis, meaning it may bear on the certainty of evidence, its application is
Lampl et al. The Journal of Headache and Pain (2023) 24:39 Page 10 of 12

ultimately subjective, and others may come to different was designed using Servier Medical Art by Servier (smart.servier.com), licensed
under a Creative Commons Attribution 3.0 Unported License.
conclusions about the certainty of evidence. Our review
does not provide any information on function, disability, Authors’ contributions
or quality of life, though these outcomes are likely to cor- CL, JV and AMvdB initiated the review drafting and revision of the article. CL,
JV, RO, JP, AMvdB, MSdR, UR, DU, RG and SS contributed with a critical review
relate with monthly migraine days, responder rates and of the article. Statistical analyses were performed by DZ and TJ. JP, AMvdB, TdV
adverse events. and DZ prepared figures and tables. All authors read and approved the final
manuscript.

Funding
Conclusions This paper was funded by the European Headache Federation (EHF).
Our systematic review and meta-analysis suggest ami-
triptyline to increase the proportion of patients who Declarations
experience a 50% or more reduction in monthly migraine
days. In fact, duration of treatment in the available stud- Competing interests
Jakob Paungarttner, Tessa de Vries and Dena Zeraatkar report no conflict of
ies was rather limited, whereas in real-life treatment is interest.
required for a longer period thus making tolerability Christian Lampl has received consulting fees and honoraria for lectures/
more compelling. While amitriptyline may remain the presentations from AbbVie/Allergan, Eli Lilly, Lundbeck, Novartis, Pfizer and
Teva. CL participated in clinical trials as the principal investigator for Eli Lilly.
first drug of choice in some patients who, for reason of Intellectual Christian Lampl is president of the European Headache Federation
comorbidities, may particularly benefit from its effect, and associate editor for The Journal of Headache and Pain.
there are no scientific data that can support to include Jan Versijpt received personal fees and nonfinancial support from Teva, per‑
sonal fees from Novartis and Lundbeck, and grants and nonfinancial support
it among the options to be mandatorily considered as from Allergan/Abbvie. Jan Versijpt serves as a member of the Board of Direc‑
first-line treatments for migraine prevention. We want tors in the European Headache Federation.
to reinforce that, drugs approved and recommended for Faisal Mohammad Amin has received honoraria for delivering lectures and/or
participation in advisory boards for Pfizer, Teva, Lundbeck, Novartis and Eli Lilly.
migraine prevention, must be supported by studies that Faisal Mohammad Amin serves as associate editor for the journals Headache
adopt a high standard in terms of design and reporting. Medicine, Acta Neurologica Scandinavica, Frontiers in Neurology and Frontiers
In Pain Research. Faisal Mohammad Amin serves as president of the Danish
Headache Society and a member of the Board of Directors in European
Abbreviations Headache Federation.
TCAs Tricyclic antidepressants Christina I Deligianni has received HIS research fellowship grant, scholarship
FDA Food and Drug Administration from Hellenic society of Neurology. Christina I Deligianni serves as a member
AHS American Headache Society of the Board of Directors in the European Headache Federation.
AAN American Academy of Neurology Raquel Gil-Gouveia reports honoraria for lectures/presentations from AbbVie/
5-HT 5-Hydroxytryptamine Allergan, Eli Lily, Lundbeck, Novartis, Teva, Organon, Pfizer; participated in clini‑
NMDA N-methyl-D-aspartate cal trials as the principal investigator for AMGEN, Novartis, Lundbeck. research
CSD Cortical spreading depression grants from Novartis. Raquel Gil-Gouveia serves as a member of the Board of
PRISMA Preferred Reporting Items for Systematic Reviews Directors in European Headache Federation.
COSMIG Outcome Set for preventive intervention trials in chronic and Antoinette MaassenVanDenBrink received honoraria, research and/or travel
episodic migraine grants from Allergan/Abbvie, Amgen/Novartis, Eli Lilly, Satsuma and Teva as
REML Restricted maximum likelihood principal investigator. Intellectual Antoinette MaassenVanDenBrink is vice-
AAFP American Academy of Family Physicians president of the European Headache Federation and associate editor for The
ACP American College of Physicians Journal of Headache and Pain.
ASIM American Society of Internal Medicine Raffaele Ornello reports personal fees from Eli Lilly, Novartis, and Teva, and
EMA European Medicines Agency non-financial support from Allergan/AbbVie, Novartis, and Teva; he is member
ECG Electrocardiogram of the Junior Editorial Board of “The Journal of Headache and Pain” and is Asso‑
CYP Cytochrome P ciate Editor in Headache and Neurogenic Pain for “Frontiers in Neurology”.
NR Not reported Margarita Sanchez-del-Rio has received consulting fees and honoraria for
NCT Number clinical trail lectures/presentations from Eli Lily, Lundbeck, Novartis, Teva and Pfizer. Intel‑
MTD Maximum tolerable dose lectual as Secretary of the European Headache Federation, Review Editor on
the Editorial Board of Headache and Neurogenic Pain (specialty section of
Frontiers in Neurology). Margarita Sanchez-del-Rio serves as a member of the
Supplementary Information Board of Directors in the European Headache Federation.
The online version contains supplementary material available at https://​doi.​ Uwe Reuter has received consulting fees, research grants or fees for presenta‑
org/​10.​1186/​s10194-​023-​01573-6. tions from Abbie, Allegan, Amgen, Lilly, Lundbeck, Medscape, Novartis, Pfizer,
StreaMedup, Teva and the German Government (BMBF). Uwe Reuter is treas‑
urer of the European Headache Federation.
Additional file 1. 
Derya Uluduz receives honoraria from Allergan-Abbvie, TEVA; consulting
Additional file 2.  fees and honoraria for lectures/presentations from Eli Lily, Novartis, Allergan/
Abbvie and Neutec participated in clinical trials as the sub-investigator for
AMGEN, Novartis. Derya Uluduz serves as a member of the Board of Directors
Acknowledgements in the European Headache Federation.
The authors gratefully acknowledge the time and effort of Prof. Michel Ferrari, Simona Sacco reports personal fees as speaker or advisor for Abbott, Allergan-
Leiden University Medical Center, The Netherlands for his critical review of the Abbvie, AstraZeneca, Eli Lilly, Lundbeck, Novartis, NovoNordisk, Pfizer, Teva;
manuscript and his comments. We also thank Sarah Kirsh, Muizz Hussain, and research grants from Novartis and Uriach; fees for CME/education from
Jordan Pepper for their valuable contribution in the statistical analysis. Figure 1 Medscape and Neurodiem Ology Medical Education; Intellectual as president
 Lampl et al. The Journal of Headache and Pain (2023) 24:39 Page 11 of 12

elect European Stroke Organisation, second vice president of the European 16. Jackson JL, Shimeall W, Sessums L, Dezee KJ, Becher D, Diemer M et al
Headache Federation, specialty chief editor in Headache and Neurogenic Pain (2010) Tricyclic antidepressants and headaches: systematic review and
for Frontiers in Neurology, associate editor for The Journal of Headache and meta-analysis. BMJ 341:c5222
Pain, assistant editor for Stroke. 17. Dickersin K, Min YI (1993) Publication bias: the problem that won’t go
away. Ann N Y Acad Sci 703:135–46. discussion 46-8.
Author details 18. Haywood K, Potter R, Froud R, Pearce G, Box B, Muldoon L et al (2021)
1
 Department of Neurology and Stroke Unit, Konventhospital Barmherzige Core outcome set for preventive intervention trials in chronic and
Brüder Linz, Linz, Austria. 2 Headache Medical Center Linz, Linz, Austria. episodic migraine (COSMIG): an international, consensus-derived and
3
 Department of Neurology, Vrije Universiteit Brussel (VUB), Universitair multistakeholder initiative. BMJ Open 11(11):e043242
Ziekenhuis Brussel (UZ Brussel), Brussels, Belgium. 4 Danish Headache Center, 19. Pitre T, Mah J, Helmeczi W, Khalid MF, Cui S, Zhang M et al (2022) Medical
Department of Neurology, Rigshospitalet Glostrup, University of Copenhagen, treatments for idiopathic pulmonary fibrosis: a systematic review and
Copenhagen, Denmark. 5 Department of Neurology, Athens Naval Hospital, network meta-analysis. Thorax 77(12):1243–1250
Athens, Greece. 6 Hospital da Luz Headache Center, Neurology Department, 20. Siemieniuk RA, Bartoszko JJ, Ge L, Zeraatkar D, Izcovich A, Kum E et al
Hospital da Luz Lisboa, Lisbon, Portugal. 7 Center for Interdisciplinary Research (2020) Drug treatments for covid-19: living systematic review and net‑
in Health, Universidade Católica Portuguesa, Lisbon, Portugal. 8 Depart‑ work meta-analysis. BMJ 370:m2980
ment of Anesthesia and Department of Health Research Methods, Evidence 21. Veroniki AA, Jackson D, Viechtbauer W, Bender R, Bowden J, Knapp G,
and Impact, McMaster University, Hamilton, Canada. 9 Department of Internal Kuss O, Higgins JP, Langan D, Salanti G (2016) Methods to estimate the
Medicine, Erasmus MC Medical Center, Rotterdam, The Netherlands. 10 Depart‑ between-study variance and its uncertainty in meta-analysis. Res Synth
ment of Biotechnological and Applied Clinical Sciences, University of L’Aquila, Methods. 7(1):55–79. https://​doi.​org/​10.​1002/​jrsm.​1164. Epub 2015 Sep 2.
L’Aquila, Italy. 11 Department of Neurology, Clinica Universidad de Navarra, PMID: 26332144; PMCID: PMC4950030.
Madrid, Spain. 12 Department of Neurology, Charité Universitätsmedizin Berlin, 22. Higgins JPTJ, Chandler J, Cumpston M, Li T, Page MJ, Welch VA (2019)
Berlin, Germany. 13 Department of Neurology Istanbul Cerrahpasa Medical Cochrane Handbook for Systematic Reviews of Interventions, 2nd edn.
Faculty, Istanbul, Turkey. Wiley, Chichester
23. Schandelmaier S, Briel M, Varadhan R, Schmid CH, Devasenapathy N,
Received: 9 March 2023 Accepted: 3 April 2023 Hayward RA et al (2020) Development of the Instrument to assess the
Credibility of Effect Modification Analyses (ICEMAN) in randomized
controlled trials and meta-analyses. Can Med Assoc J 192(32):E901
24. Egger M, Davey Smith G, Schneider M, Minder C (1997) Bias in meta-
analysis detected by a simple, graphical test. BMJ 315(7109):629–634
References 25. Balduzzi S, Rücker G, Schwarzer G (2019) How to perform a meta-analysis
1. Friedman AP (1968) The migraine syndrome. Bull NY Acad Med 44:45–62 with R: a practical tutorial. Evid Based Ment Health 22(4):153–160
2. Mahloudji M (1969) Prevention of migraine. BMJ 1:182–183 26 Viechtbauer W (2010) Conducting meta-analyses in R with the metafor
3. Yaldo AZ, Wertz DA, Rupnow MFT, Quimbo RM (2008) Persistence with package. J Stat Softw. 36(3):1–48. https://​doi.​org/​10.​18637/​jss.​v036.​i03
migraine prophylactic treatment and acute migraine medication utiliza‑ 27. Guyatt GH, Oxman AD, Vist GE, Kunz R, Falck-Ytter Y, Alonso-Coello P et al
tion in the managed care setting. Clin Ther 30:2452–2460 (2008) GRADE: an emerging consensus on rating quality of evidence and
4. Loder E, Burch R, Rizzoli P (2012) The 2012 AHS/AAN guidelines for strength of recommendations. BMJ 336(7650):924–926
prevention of episodic migraine: a summary and comparison with other 28. Zeng L, Brignardello-Petersen R, Hultcrantz M, Siemieniuk RAC, Santesso
recent clinical practice guidelines. Headache 52(6):930–945 N, Traversy G et al (2021) GRADE guidelines 32: GRADE offers guidance on
5. Evers S, Afra J, Frse A, Goadsby PJ, Linde M, May A, Sándor PS (2009) EFNS choosing targets of GRADE certainty of evidence ratings. J Clin Epidemiol
guideline on the drug treatment of migraine–revised report of an EFNS 137:163–175
task force. Eur J Neurol 16(9):968–981 29. Santesso N, Glenton C, Dahm P, Garner P, Akl EA, Alper B et al (2020)
6. Ferrari MD, Saxena PR (1993) On serotonin and migraine: a clinical and GRADE guidelines informative statements to communicate the findings
pharmacological review. Cephalalgia 13:151–165 of systematic reviews of interventions. J Clin Epidemiol 119:126–135
7. Deen M, Christensen CE, Hougaard A, Hansen HD, Knudsen GM, Ashina 30. Gomersall JD, Stuart A (1973) Amitriptyline in migraine prophylaxis.
M (2017) Serotonergic mechanisms in the migraine brain - a systematic Changes in pattern of attacks during a controlled clinical trial. J Neurol
review. Cephalalgia 37(3):251–264 Neurosurg Psych 36:684–690
8. Gray AM, Pache DM, Sewell RD (1999) Do alpha2-adrenoceptors play an 31. Couch JR, Hassanein RS (1979) Amitriptyline in migraine prophylaxis. Arch
integral role in the antinociceptive mechanism of action of antidepres‑ Neurol 36:695–699
sant compounds? Eur J Pharmacol 378(2):161–168 32. Rafieian-Kopaei M, Mehvari J, Shirzadeh H (2005) Different profile of
9. Obata H (2017) Analgesic Mechanisms of Antidepressants for Neuro‑ propranolol and amitriptyline effects on migraine prophylaxis. Int J Phar‑
pathic Pain. Int J Mol Sci. 18(11):2483 macol 1:366–368
10. Joshi PG, Singh A, Ravichandra B (1999) High concentrations of tricyclic 33. Couch JR (2011) Amitriptyline in the prophylactic treatment of migraine
antidepressants increase intracellular Ca2+ in cultured neural cells. and chronic daily headache. Headache 51(1):33–51
Neurochem Res 24(3):391–398 34. Gonçalves AL, Martini Ferreira A, Ribeiro RT, Zukerman E, Cipolla-Neto
11. Galeotti N, Ghelardini C, Bartolini A (2001) Involvement of potassium J, Peres MF (2016) Randomised clinical trial comparing melatonin 3
channels in amitriptyline and clomipramine analgesia. Neuropharmacol‑ mg, amitriptyline 25 mg and placebo for migraine prevention. J Neurol
ogy 40(1):5–84 Neurosurg Psychiatry 87(10):1127–1132
12. Dharmshaktu P, Taya V, Kalra BS (2012) Efficacy of antidepressants as 35. Berilgen MS, Bulut S, Gonen M, Tekatas A, Dag E, Mungen B (2005)
analgesics: a review. J Clin Pharmacol 52(1):6–17 Comparison of the effects of amitriptyline and flunarizine on weight gain
13. Ayata C, Jin H, Kudo C, Dalkara T, Moskowitz MA (2006) Suppression and serum leptin, C peptide and insulin levels when used as migraine
of cortical spreading depression in migraine prophylaxis. Ann Neurol preventive treatment. Cephalalgia 25(11):1048–1053
59(4):652–661 36. MifsudBuhagiar L, Casha M, Grech A, SerracinoInglott A, LaFerla G (2022)
14. Liang J, Liu X, Pan M, Dai W, Dong Z, Wang X, Liu R, Zheng J, Yu S The interplay between pharmacogenetics, concomitant drugs and blood
(2014) Blockade of Nav1.8 currents in nociceptive trigeminal neurons levels of amitriptyline and its main metabolites. Per Med 19(2):113–123
contributes to anti-trigeminovascular nociceptive effect of amitriptyline. 37. Rasmussen BK, Jensen R, Schroll M, Olesen J (1992) Interrelations
Neuromol Med 16(2):308–321 between migraine and tension-type headache in the general population.
15. Page MJ, McKenzie JE, Bossuyt PM, Boutron I, Hoffmann TC, Mulrow CD Arch Neurol 49(9):914–918
et al (2021) The PRISMA 2020 statement: an updated guideline for report‑ 38. Snow V, Weiss K, Wall EM, Mottur-Pilson C (2002) Pharmacologic manage‑
ing systematic reviews. BMJ 372:n71 ment of acute attacks of migraine and prevention of migraine headache.
Ann Int Med 137:840–849
Lampl et al. The Journal of Headache and Pain (2023) 24:39 Page 12 of 12

39. Silberstein SD, Holland S, Freitag F, Dodick DW, Argoff C, Ashman E (2012)
Quality Standards Subcommittee of the American Academy of Neurology
and the American Headache Society. Evidence based guideline update:
pharmacological treatment for episodci migrainbe prevention in adults.
Neurology 78(17):1337–134
40. Minen MT, Begasse De Dhaem O, Kroon Van Diest A, Powers S, Schwedt
TJ, Lipton R, Silbersweig D (2016) Migraine and its psychiatric comorbidi‑
ties. J Neurol Neurosurg Psychiatry 87:741–749
41. Lampl C, Huber G, Adl J, Lutrhingshausen G, Franz G, Marecek S, Weber J,
Baumhackl U, Mueller T (2009) Two different doses of amitriptyline ER in
the prophylaxis of migraine: long-term results and predictive factors. Eur J
Neurol 16:943–948
42. Haythornthwaite JA (1993) Migraine headaches and psychopathology:
future directions. J Psychiatr Res 27:183–186

Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in pub‑
lished maps and institutional affiliations.

Ready to submit your research ? Choose BMC and benefit from:

• fast, convenient online submission

• thorough peer review by experienced researchers in your field
• rapid publication on acceptance
• support for research data, including large and complex data types
• gold Open Access which fosters wider collaboration and increased citations
• maximum visibility for your research: over 100M website views per year

At BMC, research is always in progress.

Learn more biomedcentral.com/submissions

Journal of Headache & Pain is a copyright of Springer, 2023. All Rights Reserved.