Skip navigation
Link to Back issues listing | Back Issue Listing with content Index | Subject Index

More evidence on chronic headache

Antidepressants for chronic headache
Cluster headache
Chronic headache is a big problem for individuals, their professional carers and health services. Reviews examining the efficacy of treatments are welcome. Two reviews, of very different type, inform on chronic headache treated with antidepressants [1], a common condition, and cluster headache [2], much more rare.

Antidepressants for chronic headache

This thorough review from the USA [1] did extensive searching for randomised trials comparing antidepressant prophylaxis with placebo for chronic headache. A range of outcomes were extracted, including global improvement (usually at least 50% reduction in headache burden, itself obtained by multiplying frequency of headaches by their severity), headache burden, and use of analgesics.


There were 38 randomised trials with useful data, predominantly in chronic migraine and chronic tension headache. Average trial size was 50 patients and the average duration was 10 weeks. More recent studies used International Headache Society classifications for headache diagnosis. Tricyclic drugs were represented primarily by amitriptyline (12/19), serotonin blockers by pizotifen (12/18) and SSRIs by femoxetine (3/7). Dose varied (for instance for amitriptyline between about 10 mg and 150 mg daily; and readers should beware that the notation 'qd' in American means once-a-day).

Pooled data for global improvement, headache burden and analgesic use all favoured antidepressant over placebo. For global improvement 31% more patients improved with placebo, with a number needed to treat of 3.2 (95% confidence interval 2.5 to 4.3). With a rate ratio of 2.0, this implies that about 60% of patients using antidepressants have a global improvement approximating to about 50% reduction in headache burden.

There was no observed difference between different classes of drugs. Type of headache, type of antidepressant or duration of treatment did not affect results, nor did several aspects of trial design (parallel group versus crossover), or trial size or trial quality. Adverse events were not described in the review.

Cluster headache

Cluster headache (unilateral, excruciating severe attacks of pain principally in the ocular, frontal and temporal areas recurring in separate bouts with daily or almost daily attacks for weeks to months) is rare. It affects fewer than one person in a thousand in their lifetime. A review [2] looks at all the evidence on cluster headache, including classification, epidemiology, aetiology, pathophysiology and drug and surgical treatment, while another, though not overtly systematic, is comprehensive [3].

It's a fun review, impossible to summarise for Bandolier . The best evidence for drug treatment is for subcutaneous sumatriptan.


Two interesting if different reviews on chronic headache. One question for which the answer remains blurred is when are headaches sufficiently chronic to need prophylactic treatment? Is it one a day, one a week, or one a month? There are two ways of looking at this.

A small proportion of triptan users consume large numbers of tablets, sometimes in excess of one a day. They represent a significant minority of triptan users and triptan use, and clearly need expert help.

Another view is that even one headache a month, that causes significant disability and for which triptans are not effective, should be a trigger for prophylactic therapy.

Whatever, prophylactic treatment of chronic headache is a growth area where significant change can be expected. Watch this space.


  1. GE Tomkins et al. Treatment of chronic headache with antidepressants: a meta-analysis. American Journal of Medicine 2001 111: 54-63.
  2. Zakrzewska. Cluster headache: a review of the literature. British Journal of Oral and Maxillofacial Surgery 2001 39: 103-113.
  3. Dodick DW et al. Cluster headache. Cephalalgia 2000 20: 787-803.
previous or next story in this issue