As mentioned in the introduction to this chapter, anecdotal accounts of dietary initiation of migraine attacks in some individuals have punctuated the scientific literature since time immemorial. Even so, serious doubt exists in the minds of many neurologists as to whether the condition really exists as a clinical entity (e.g. Blau 1992; Sacks 1993) or whether the phenomenon is merely a manifestation of the placebo effect. Socrates said: 'Let no one persuade you to give the drug against headache to him who before has not opened his soul to your treatment' (Plato, Charmides). And even when an experiment appears to be carefully controlled, there may be pitfalls: Strong (2000), pondering on why some dietary migraine patients claim to get headaches from placebos, pinpointed the control gelatine capsule, composed of animal and vegetable protein, as a possible migraine trigger!
Earlier in this chapter, a number of chemical triggers of migraine were listed, including reserpine, fenfluramine, and MCPP, and, in general, their action is accepted without demur by workers in this field. Then there are single chemical substances employed in foodstuffs — aspartame (Lipton et al. 1989; Van Den Eeden et al. 1994) and vanillin (Saint Denis et al. 1996), for example — and few would have a problem in believing that they act as headache triggers in particular subgroups of the migraine population.
It is with the ingestion of complex chemical mixtures, foodstuffs of only partially known composition, that the main problems of plausibility arise. In her original and, indeed, groundbreaking paper, Hanington (1967) was persuaded that the tyramine content of cheese is responsible for the headache that follows its ingestion in some migraineurs. Several subsequent investigators were unable to replicate her finding of a migraine-triggering effect of tyramine capsule administration, and the tyramine story has duly languished. It is not quite dead, however, for Peatfield et al. (1983) did find that intravenous tyramine administration is followed by slight headache in a proportion of subjects, without a blood pressure rise sufficient to account for it. It appears, however, that tyramine ingestion is only associated with headache in patients treated with monoamine oxidase inhibitors, the so-called cheese effect. Even so, the popular press has enthusiastically adopted tyramine as the prime cause of dietary migraine, and tyramine-free diets now abound. Specialized cookery books for migraineurs are still being published, in which tyramine-free diets form an important plank in the author's platform (e.g. Marks and Marks 2000).
For believers in the concept of dietary migraine, the commonest initiators of an attack are alcoholic beverages (Peatfield 1995), and the largest and most notorious subgroup in this category has red wine as its trigger (Peatfield et al. 1984) — although even here there is conflict: on the continent of Europe, white wine is viewed as the major culprit (Relja et al. 1993; Henri 1996; Tournier-Lasserve 1996). Peatfield (1995) holds that cheese/chocolate and red wine sensitivity in particular have closely related mechanisms. It should be noted that there is very little tyramine in red wine (Hannah et al. 1988).
Against this background of uncertainty, it seemed important to try to obtain objective evidence for the existence of the condition and, accordingly, our group carried out two double-blind trials. In the first (Littlewood et al. 1998), we divided a subset of patients who thought their headaches were initiated by red wine into two groups, giving the first a heavily disguised red wine mixture and the second a similarly disguised vodka mixture of identical alcohol content. The subjects claimed to be unable to tell the difference but only those ingesting red wine (9 out of 11) developed headache with associated migraine symptoms. The second trial (Gibb et al. 1991) involved patients who claimed that chocolate set off their migraine episodes. They were again divided into two groups and were fed, respectively, real and mock chocolate of similar taste which they were unable to distinguish between. Compared with the wine experiment, a smaller but still significant proportion of those fed authentic chocolate, but not placebo, developed a headache attack, again with associated symptoms of migraine. What was particularly interesting, as mentioned above (Littlewood et al. 1988, Gibb et al. 1991), was the difference in lag period between ingestion and headache onset: about three hours in the red wine experiment and 22 hours after chocolate. Whatever the biochemical sequence separating the trigger from the event, it seems likely that the difference between migraineurs and normal subjects will turn out to be enhanced sensitivity to an initiating agent.
Again as mentioned above, a recent instalment in the red wine story (Corder et al. 2001) is the presence of specific polyphenol components identified in red wine that are able to inhibit endothelin-1 synthesis. These might be expected to attenuate migraine attacks if the raised level of circulating endothelin during these episodes is relevant to the sequence of clinical signs. It is tempting to speculate, as the authors do, that the inhibitory effect might be responsible for the French paradox, the lower death rate from coronary heart disease in France compared with the UK, despite a comparable dietary intake of saturated fats. It should be noted, however, that the observations of Corder et al. have only been performed in vitro on cultured bovine aortic cells. Before the data can be extrapolated to humans, we need to known considerably more about the in vivo effect of the polyphenols in question, their absorption and bioavailability. We should remember the case of resveratrol, a highly vaunted compound of high in vitro promise, which failed to achieve therapeutic fulfilment in vivo (Goldberg and Soleas 2002).
It may well be, of course, that alcohol itself, rather than red wine specifically, is responsible for the French paradox (Klatsky 2002). And then again, in a recent new departure, it seems likely that light to moderate alcohol ingestion is associated with a lower risk of dementia in individuals aged 55 years or over (Ruitenberg et al. 2002). It is gratifying that a recreational drug which, in moderation, is capable of providing substantial social benefit and support may also improve our actual physical health on a long-term basis.