Should caffeine be a made a class B drug?

The week before last witnessed the first decent media drug panic we’ve had in a while, although unusually this one was not about cannabis, ecstasy, or cocaine. It followed the publication of a paper (from Durham University’s department of psychology) titled ‘Caffeine, stress, and proneness to psychosis-like experiences: A preliminary investigation’ in a fairly obscure Journal called Personality and Individual Difference. The papers key finding was that heavy caffeine users were slightly more prone to hallucinations, but not to feeling persecuted.

Somewhat strangely the story received massive cross-media coverage, in almost all the UK daily’s, news agencies and major media outlets across the world. A google-news search finds 285 media hits ranging from Fox news (Lots ... and Lots of Coffee Linked to Hallucinations), and the Times of India (coffee can lead to hallucinations), to The Australian (Drink coffee, see dead people) and even Reuters with the ridiculous invent-a-quote headline 'Coffeeholics wake the dead'. But as so often with drug panics the actual research findings are rather underwhelming when compared to many of the shock headlines of the stories covering them.

Some dead people spotted outside Starbucks yesterday*

As deconstructed in the Guardian’s unfailingly brilliant Bad Science Column by Ben Goldacre (and the excellent NHS 'behind the headlines' page), the paper's findings, in actuality, showed a statistically significant but very weak correlation between high caffeine intake and proneness to hallucinate (different from actually hallucinating). Moreover The findings were based on weak observational data from a self selecting sample of students (who did an online questionnaire of often strange and ambiguously worded questions), confounding variables were not considered, and much of the media coverage quoted the finding that 7 cups of coffee a day is associated with a three times higher prevalence of hallucinations, a fact that emerged from the Durham University press release but is not supported by the actual research findings. There was also nothing in the paper about coffee making you see ghosts or dead people – a headline that evidently emerged by a process of Chinese whispers between the published paper (which mentions sensing people who were not there), the departmental press release, news agency coverage, and lazy headline writers.

So here, in the finest tradition of drug stories in the media, was a piece of not especially good research being very badly reported; a rather dull and ambiguous research finding sexed up by the headline writers to either fill up space on a quiet news day or shift units, depending on how you want to see it. You may well ask; So What? Nothing new here really.

But this rather sorry and un-illuminating tale does raise some more interesting issues when viewed alongside the longer running but not completely dissimilar cannabis panic. They are alike in key respects – journalistic hype, poor science reporting, the theme of young people in peril, drugs and mental health (specifically psychosis) and so on. But there are also some dramatic differences. The fact that the caffeine story revolved around the discovery that high levels of consumption could lead to psychotic-like symptoms notably failed to provoke any comment from any of the legions of politicians and opinion writers who pounced on every new research paper that mentioned ‘cannabis’ and ‘psychosis’ in the same sentence, before pontificating ad-nauseum about ‘sending out the wrong message’, ‘a mental health time bomb’, 'lethal skunk' (perhaps responsible for all the dead people seen by the coffee drinkers?) and all the rest.

Where was the Home Secretary telling us how the coffee young people are drinking today, the Starbucks' triple shot mega-chinos, are nothing like the innocent cups of Maxwell House that their parents used to drink in the 60s? Where was Melanie Phillips rallying against the moral depravity of Government-sanctioned caffeine-crazed child-zombies prowling our streets? Where was the Independent on Sunday reversing its support for caramel lattes because of the ten-fold increase in ghost sightings? Caffeine panic just doesn’t seem to fire up our legions of drug warriors. There’s no political mileage in it and of course, most of them spend much of the day drinking it. Even more than alcohol, its one drug that it’s OK to consume. A nice, safe drug. Our drug. Not really a drug at all.

Further highlighting this peculiar conceptual dis-juncture that exists in the minds of the media, public and politicians alike, between illegal and legal drugs, some of the media coverage was unintentionally hilarious; like the interview on the Radio 1 website with a woman, who after drinking several bottles of diet cola, says that: "I can act weird and I talk weird, as if you are high on drugs". And notable in the media coverage was the almost complete absence of the word 'psychosis', even though it was in the title of the research paper being reported on, and psychotic symptoms were the paper's primary subject, and indeed primary findings. Did the uncomfortable echoes of the cannabis/psychosis media narrative lead to the word being unconsciously deleted from the coverage?.

We shouldn’t forget that the key stated motivation for the recent reclassification of cannabis from C back to B was the potent new strains of ‘skunk’ cannabis and its effects on mental health , specifically the danger of inducing psychosis (the widely misunderstood and rarely explained word, endlessly repeated in the media). Compared to the caffeine story there was certainly a wide array of political forces marshaling behind the skunk/psychosis panic; the tabloids and opposition politicians trying to score points by hyping the panic to make the Government look irresponsible, and a new Prime Minister wanting to assert his moral credentials with a few judicious policy U-turns on third-rail type issues. There is more and arguably better research too, both on increasing potency and mental health risks of cannabis, but in the froth of hype, panic, moral grandstanding and political posturing, any grasp on reality or objective science and harm assessments was lost (chronicled on this blog in almost gruesome detail over the last couple of years).

Another lesson then, from this latest caffeine mini-panic, is that the media clearly doesn’t even need the sorts of partisan politics and culture-wars tedium that swilled around the cannabis debate to fall back on bad science reporting of drug stories. They will evidently do that completely unprompted. Understanding of and reporting on statistics in the media is almost universally poor. Combine this fact with the need to produce sexy drug-shock type headlines (usually produced by generalist, crime or home affairs reporters with no science or stats background) and you have a recipe for a seemingly endless tide of poor drug coverage. It may give bloggers something to sound off about, but in policy terms the impacts of this phenomenon are almost all bad.

By way of epilogue consider some of the recent coverage of caffeine in the Daily Mail, the long time standard bearer for hysterical bad-science drug reporting. A Mail story from the 26th of January is headlined: ‘Coffee may raise child cancer risk: New evidence that caffeine could damage babies'' DNA'. With this effort the Mail threatens to become a self-parodying laughing stock, so luminous is its idiocy. Behold; they have managed to conjure a drug shock headline of classic DNA-melting drug-baby vintage, reporting on scary ‘new evidence’ from research that is not only unpublished, but has yet to even begin.

Is this the beginning of a Daily Mail push for caffeine to be classified as a Class B drug? Are they groping around for something new to be outraged about now the cannabis 'problem' has been solved with an additional three years on prison sentences? Probably not, but I almost hope it is – it would expose the contradictions and hypocrisy in the media's drugs reporting.

See also:
the caffeine wikipedia page
NHS caffeine advice

*photo thanks to ioerror