The BMJ recently ran a vote on bmj.com to identify the “top 10 non-diseases.”1 Some critics thought it an absurd exercise,2 but our primary aim was to illustrate the slipperiness of the notion of disease. We wanted to prompt a debate on what is and what is not a disease and draw attention to the increasing tendency to classify people's problems as diseases.

In 1979 the BMJ published a study that did something similar.3 Non-medical academics, medical academics, general practitioners, and secondary school students were invited to say whether 38 terms did or did not refer to a disease. Almost 100% thought that malaria and tuberculosis were diseases, but less than 20% thought the following to be diseases: lead poisoning, carbon monoxide poisoning, senility, hangover, fractured skull, heatstroke, tennis elbow, colour blindness, malnutrition, barbiturate overdose, drowning, or starvation (figure). People were split 50:50 over whether hypertension, acne vulgaris, or gall stones were diseases. The doctors were more likely to view the terms as referring to diseases. The authors of this study included Guy Scadding, who spent much of his life spelling out to doctors that no general agreement exists on how to define a disease.