Journal ofNeurology, Neurosurgery, and Psychiatry 1996;61:285-290 285 J Neurol Neurosurg Psychiatry: first published as 10.1136/jnnp.61.3.285 on 1 September 1996. Downloaded from A non-selective (amitriptyline), but not a selective (citalopram), serotonin reuptake inhibitor is effective in the prophylactic treatment of chronic tension-type headache Lars Bendtsen, Rigmor Jensen, Jes Olesen Abstract cemed, the most important type of headache.' Objectives-Although the tricyclic anti- Yet remarkably little is known about its patho- depressant amitriptyline is extensively physiology and the treatment available is lim- used in the prophylactic treatment of ited. The only established prophylactic chronic tension-type headache, only few treatment is the tricyclic antidepressant studies have investigated the efficacy of amitriptymine, a non-selective serotonin reup- this treatment and the results are contra- take inhibitor. The efficacy of this treatment dictory. In addition, the new selective has, however, only been investigated in a few serotonin reuptake inhibiting antidepres- placebo controlled studies,3-6 which have sants, which are widely used in depres- reported conflicting results. The scientific sup- sion and of potential value in pain port for the widespread use of amitriptyline in management, have never been investi- the treatment of chronic tension-type gated in a placebo controlled study often- headache, therefore, is scant. sion-type headache. The aim was to The mechanism of action of amitriptyline in evaluate the efficacy of amitriptyline and chronic tension-type headache, as well as in of the selective serotonin reuptake other non-depressive chronic pain states, is inhibitor citalopram in chronic tension- largely unknown, but it is assumed that the type headache. blockage of serotonin reuptake in the CNS Methods-Forty non-depressed patients plays an essential part in its analgesic effect.7-'0 with chronic tension type headache were The selective serotonin reuptake inhibitor included in a 32 week, double blind, citalopram, which has antidepressant proper- placebo controlled, threeway crossover ties comparable with the tricyclic drugs but a study. far better side effect profile," might therefore Results-Thirty four patients completed be of value in the treatment of chronic tension the trial. Amitriptyline reduced area type headache. In addition, a comparison of under the headache curve by 30% com- amitriptyline and citalopram could provide pared with placebo (P = 0.002), whereas information on the mechanism of action of citalopram had no significant effect (P = these drugs in chronic pain. The aim of the 0.68). Explanatory analyses showed that present study was to evaluate the prophylactic http://jnnp.bmj.com/ amitriptyline significantly reduced the effect of amitriptyline and citalopram in duration of headache (P = 0.01), chronic tension-type headache. headache frequency (P = 0.01), and intake of analgesics (P = 0.02) but not headache intensity (P = 0.12). Materials and methods Conclusion-Although amitriptyline did PATIENTS not eliminate the headache, it provided a Forty patients with chronic tension type clinically important reduction of headache diagnosed according to the criteria on September 29, 2021 by guest. Protected copyright. headache in the majority of otherwise of the International Headache Society'2 were treatment resistant patients. The differ- recruited from the outpatient headache clinic ential effect of amitriptyline and citalo- at Glostrup University Hospital, Copenhagen, pram indicates that mechanisms other Denmark. Seven patients had coexisting infre- than inhibition of serotonin reuptake are quent migraine ( < one day a month) whereas Department of Neurology, Glostrup involved in the analgesic effect of the tri- 33 never had migraine. The patients under- Hospital, University of cyclic antidepressants. Amitriptyline, but went a general and a neurological examina- Copenhagen, not citalopram, is valuable in the prophy- tion, including 12 channel ECG and Glostrup, Denmark lactic treatment of chronic tension type laboratory screening, and completed a diag- L Bendtsen R Jensen headache. nostic headache diary'3 during a four week run J Olesen in period. Table 1 gives detailed clinical infor- Correspondence to: (7 Neurol Neurosurg Psychiatry 1996;61:285-290) mation. Dr Lars Bendtsen, Department of Neurology, The inclusion criteria were a diagnosis of Glostrup Hospital, chronic tension type headache and age University of Copenhagen, Keywords: antidepressants; pain; tension type head- DK 2600 Glostrup, ache between 18 and 65 years. Women of child- Denmark. bearing potential had to use adequate contra- Received 2 January 1996 ceptive measures throughout the study. The and in revised form 18 March 1996 Tension type headache is the most common exclusion criteria were previous participation Accepted 22 March 1996 and, as far as socioeconomic impact is con- in a clinical trial, migraine more than one day a 286 Bendtsen, _Jensen, Olesen J Neurol Neurosurg Psychiatry: first published as 10.1136/jnnp.61.3.285 on 1 September 1996. Downloaded from Table 1 Clinical characteristics (Cipramil®), and placebo tablets. In the first Patients who week of treatment with amitriptyline the Patients included completed the study patients received a daily dose of one 25 mg No of patients 40 34 amitriptyline tablet and one placebo tablet, in Sex (women/men) 25/15 22/12 the second week they received two 25 mg Age (y) 40-0 (18-60) 40 7 (18-60) Area under the headache curve 982 (145-3331) 973 (145-3331) amitriptyline tablets, and in weeks 3-8 they Headache duration (hours/four weeks) 217 (27-487) 220 (27-487) received one 25 mg and one 50 mg amitripty- Headache intensity 4-2 (1 8-7 3) 4 1 (1-8-7-3) Headache frequency (days/four weeks) 24-5 (16-28) 24 7 (16-28) line tablet corresponding to a daily dose of 75 Analgesics (doses/four weeks) 41 8 (0-106) 41 3 (0-93) mg amitriptyline. During the eight weeks of Hamilton depression score 3 3 (2-11) 3 5 (2-11) Years with headache 12 2 (1-36) 11-7 (1-36) treatment with citalopram the patients Frequency of migraine (days/year (n = 7)) 7 6 (2-12) 7-6 (2-12) received a daily dose of one citalopram tablet to a Values are means (range). There were no significant differences in any of the clinical characteris- and one placebo tablet corresponding tics between the 34 patients who completed the study and the six drop outs (P = 0-32-0 98). daily dose of 20 mg citalopram. During the eight weeks of treatment with placebo and during the wash out periods the patients received a daily dose of two placebo tablets. month, serious somatic or psychiatric diseases The patients thus received two tablets daily including depression (Hamilton depression during all 28 weeks of treatment. All tablets score'4 >s 17), misuse of simple analgesics were of identical look and taste, and the (corresponding to more than 2 g aspirin a patients were told to take the tablets two to day), regular intake of opiates or benzodi- three hours before bedtime. azepines, and previous treatment with antide- pressants. All patients gave written informed RECORDING OF EFFICACY VARIABLES consent to participate in the study, which was Throughout the study the patients kept a approved by the regional ethics committee. headache diary with recordings of intensity The patients were informed that the study and duration of headache, intake of analgesics, included placebo periods, but no further infor- and side effects. Intensity was recorded on an mation about the study design was given. 11 point scale (0-10), in which 0 indicated the headache free condition, 5 indicated a moder- STUDY DESIGN AND MEDICATION ate headache, and 10 indicated the worst The study was designed as a double blind, headache imaginable. Localisation and quality placebo controlled, three way crossover trial. of the headache, whether the headache was After a four week run in period, the patients aggravated by physical activity, and presence were randomly allocated to one of the six pos- or absence of nausea, photophobia, and sible treatment sequences (fig 1). Randomisa- phonophobia were also recorded. tion was done in blocks of six patients. Each of the three drugs was given for eight weeks and CLINICAL VISITS the treatment periods were separated by two- Follow up visits were performed at four-week week wash out periods. intervals (fig 1). At each visit, the headache The study medication was tablets contain- diary was checked, medication supplies were ing 25 mg or 50 mg amitriptyline (Saroten®), handed over, side effects reported by the tablets containing 20 mg citalopram patients were recorded, and compliance was http://jnnp.bmj.com/ Citalopram Placebo Figure 1 Study design of a the 32 week, double blind, placebo controlled, three Amitriptyline ----X------- way crossover trial. Each Placebo Citalopram patient was randomly .. ******************** D allocated to one of the six treatment sequences a-f. Clinical visits are indicated on September 29, 2021 by guest. Protected copyright. by *. Placebo Amitriptyline c Citalopram 1- Amitriptyline Placebo Amttn********** * *******ta*opram d Amitriptyline Citalopram e l Placebo Citalopram Amitriptyline ----------s.J f Run-in Period 1 Wash out Period 2 Wash out Period 3 * * * * * * * I I m 0 4 12 14 22 24 -32I Time (weeks) Serotonin reuptake inhibitors and tension-type headache 287 J Neurol Neurosurg Psychiatry: first published as 10.1136/jnnp.61.3.285 on 1 September 1996. Downloaded from Figure 2 Area under the Whitney two tailed test.'5 A possible headache curve (duration x intensity) in 34 patients time/period effect was tested for by comparing with chronic tension type a) the difference in AUC in periods when drug A headache during eight was followed by drug B with the negative dif- weeks of treatment with C.) ference in AUC in periods when B was amitriptyline (circles), drug citalopram (triangles), and C.) followed by drug A.'5 Comparison of clinical placebo (squares). 'a characteristics between patients who com- Asterisks indicate * pleted the study and drop outs was done by differences a) significant U test.