(British Medical Journal) - The research team at the University of Manchester's Chronic Disease Epidemiology Unit monitored the rates and causes of death among 6569 adults up to the age of 85 for eight years. All of them had taken part in two pain surveys in 1991-2. They were asked if and where they had felt body pain in the preceding month, whether they smoked, and what their levels of psychological distress were.

Just under half of respondents had pain in one area of their bodies and just over a third reported no pain at all. A further 15 per cent complained of pain all over their bodies. And this group included more women and more older people than the other two groups.

Over the eight years of monitoring, 654 people died. Heart disease accounted for four out of 10 of the deaths, followed by cancer, which killed almost one in three. Most of the remaining deaths were attributable to respiratory disease.

Death rates were not only higher among people with widespread and regional pain, but they were also much more likely to die of cancer. Regional pain sufferers were significantly more likely to have died of cancer, while those with widespread pain were twice as likely to have done so. The risk remained even after accounting for a diagnosis of cancer at the time of the survey and other influential factors, such as age, sex, and smoking. The number of accidental, suicidal, or violent deaths was small at 2 per cent, but people with widespread pain were five times as likely to die this way as people who reported no pain.

The authors suggest that perhaps the same factors underlying heightened pain perception may also be involved in an increased risk of cancer, or that widespread pain may shorten survival in people who go on to develop the disease.

An accompanying commentary speculates on the possible reasons for the "unexplained but potentially important finding," one of which relates to undiagnosed cancer.