Headache and coffee: what the research says

Coffee ranks high on the headache triggers list. The ingredient that gives you the buzz – caffeine – is one of the best-loved psychostimulants (uppers) ever. As a pick-me-up at work or just to hang out, almost everyone loves their coffee. For most people, life wouldn’t be the same without it. Scandinavian countries are Europe’s biggest coffee drinkers, with Norway, Sweden and Finland taking the lead. Germany doesn’t do too badly either, notching up 4.8 kg of coffee per head of population per year. That’s just under 2 cups a day. So based on the statistics, coffee easily outranks beer as the beverage of choice in Germany.

The problem with caffeine withdrawal

Is coffee good for you? How does caffeine affect our bodily functions and well-being? Experts have been disagreeing on this seemingly since forever. While there is evidence that caffeine can help relieve pain as well as being an effective stimulant, the research outcomes around this issue are not at all consistent. To find out what some of the studies say, read on.

A review by Robert Shapiro says that some regular coffee drinkers who abruptly stop using caffeine may experience a withdrawal syndrome. Typical symptoms include severe, sometimes migraine-like headaches, fatigue, dizziness, muscle pain and cognitive deficits. Caffeine is mildly analgesic (pain-relieving), so drinking coffee on a regular basis is a bit like prolonged use of a mild painkiller. This has led to the theory that caffeine withdrawal headache might be similar in origin to medication-overuse headache. “Coffee withdrawal headache” even got its own listing in the International Classification of Headache Disorders. Right from the earliest studies, the condition was linked with a wildly varied set of symptoms. Only about half of the research subjects showed any susceptibility to the phenomenon at all. Other studies indicated that there may be a genetic component.

Other research finds the exact opposite

A review by Jana Sawynok based at Dalhousie University in Halifax, Canada shows that the results of the available research on the effects of caffeine are widely different and downright contradictory in places. Depending on the study you pick, caffeine relieves pain or caffeine triggers headache and migraine attacks. It is hard to understand why the conclusions are so very different. The author of the review has no explanation either. The human metabolism is so complicated with so many different factors involved that the results of each study only apply to that particular setting. And due to largely unknown genetic confounders, the patient population is probably not normally distributed either. All this makes the studies almost impossible to compare.

The effect of caffeine has been studied in migraineurs, too. Lee et al. based in Seoul, South Korea, divided about 100 people with migraine into two groups. One group was allowed to continue taking caffeine as usual. The other group abstained from caffeine. Both groups continued treating any migraine attacks with their usual medicines – triptans in most cases. The main result, the authors say, was that abstaining from caffeine significantly improved response to therapy. Drug treatment was less effective in the group that continued drinking coffee, they say. Looking for a physiological basis for the observed phenomenon, the authors turned to medication overuse headache (see above). Giving up caffeine – the scientists suggest – may have had a “detoxing” effect similar to stopping medication. There is no hard evidence for any such correlation, of course.

A Norwegian study highlights how confusing and contradictory studies in this field can be. Ottar Sjaastad, a neurologist at Trondheim University Hospital, divided 1700 coffee drinkers (vetted for headache-relevant existing conditions) into three groups which he called "light," "moderate," and "heavy” users based on whether they drank 1-2, 3-4, or 5 or more cups of coffee per day. The overall average was 5 cups. All participants were instructed to stop drinking coffee. They were then asked about any health problems that had occurred. Astonishingly, only seven people in this large subject population (just 0.4%) reported a reaction that met the definition for withdrawal headache.

This surprising outcome naturally required some explanation and raised new questions. One suggestion was that the relatively high and prolonged coffee consumption in the study population might have produced a habituation effect that would have mitigated or prevented the consequences of withdrawal. This striking result prompted the authors of the study to conduct a self-experiment where they increased their personal coffee consumption to more than two liters daily over a period of several weeks, then went cold turkey. None of them got a headache.

Clearly this was quasi anecdotal evidence based on personal observation and not sufficiently powered to disprove the data of other studies. That said, Sjaastad does make a couple of very valid points for research in this area. Withdrawal headache as defined in the international classification system is not easy to diagnose. The necessarily highly subjective accounts provided by the subjects make objective assessment more difficult. The authors believe this could help explain the remarkable degree of variation in existing studies, with reported diagnosis rates ranging from 20% to over 84%. Additional reasons for the variance would be the major differences in study design and methods and the fact that diagnosis was not always by a neurologist.

The science is unclear

This brief excursion through the literature shows two things: 1) there are lots of studies into the coffee-headache connection; 2) the results are bewilderingly contradictory. So what’s going on? Are all the studies just snapshots of unrepresentative populations? Is "caffeine withdrawal headache" not an actual thing, as Sjaastad somewhat provocatively suggests ("The idea that caffeine-withdrawal headache is a ubiquitous, severe headache of considerable social and health importance can, with reasonable certainty, be toned down")?

Looking at the existing data, clearly there is a need to simplify and standardize study design and methods to establish at least a rudimentary degree of comparability between studies. Otherwise, the search for wisdom in this area risks coming dangerously close to “reading the tea leaves” (or, in this case, the coffee grounds...)