Drinking coffee makes you see ghosts

17th May 2008

All over the media last week were reports of the finding that drinking coffee can lead to people seeing ghosts. Headlines such as:

all had the same theme: that drinking coffee can lead to an increased risk of hallucination and therefore seeing ghosts or sensing dead people around you.

As is usually the case with the reporting of scientific publications, the story reported is often not actually what the research indicates and the headline writers usually extrapolate things so that they become sensationalist! That is certainly the case with this study.

What was the study about?

In what are called "Diathesis-stress models of psychosis", it is proposed that stress plays a role in the development of hallucinatory or schizophrenia-like experiences, which are enhanced by the stress hormone cortisol.

Caffeine has been found to increase the amount of cortisol released in response to stress and so it was proposed that ingestion of caffeine (not necessarily from coffee) would be associated with an increased proneness to psychotic-type experiences.

Previous studies in this area have used participants who have psychoses; however, there is a problem that any medication that they were on could interfere with caffeine. So in this study, the researchers used a 'normal' population of participants as it's known that schizophrenia-like (or schizotypal) experiences occur in non-psychotic people too (the most common occurrence familiar to most people probably being hearing a voice in your ear).

What the researchers were looking for was to see whether:

  1. Caffeine intake was associated with stress levels;
  2. Caffeine intake was associated with hallucinatory experiences and/or persecutory ideation.


The results showed that caffeine consumption was not associated with persecutory ideation, but it was associated with stress and it was associated with hallucinatory experiences. The effect, the size of the association, was found to be small however.

Associations or correlations do not prove causation, and it may be that caffeine causes hallucinatory experiences in people but it may just be that people who have hallucinatory experiences consume more caffeine. Or, it could be something else altogether that causes people to have hallucinatory experiences and to consume more caffeine.

What can we conclude?

From a critical thinking point of view, we may be faced with people claiming that ghosts can be seen as the result of drinking coffee and such like; but this was really a huge extrapolation of the study's findings by the media headline writers and not a conclusion of the study - which, incidentally, had nothing to do with ghosts or the paranormal.

What we can say is that the study's findings support the hypothesis that caffeine consumption may have a small effect on psychosis-like hallucinatory experiences in people who are prone to stress and such experiences in the first place. The effect of caffeine on these experiences, if true, is so small that it probably has little or no meaning outside of a clinical setting.

This was a small, preliminary piece of research that was relevant to its own field, but it's not an answer to seeing ghosts, hearing voices and other, what we might come across as paranormal, experiences.

The problem here is not the actual research or the researchers, but the media's interpretation of it.