Spending more on turning patients into people.

deriving an average respectively of 44% and 32% of calories from fat. Furthermore, although careful steps were taken to validate the dietary questionnaire, the inevitable imprecision of data derived from questionnaires was evident from the information provided. The best evidence of any causal association would be provided by an intervention study. The United States National Cancer Institute will shortly decide whether to proceed with a large randomised trial to test the hypothesis that restricting dietary fat can reduce the incidence of breast cancer.27 The trial would include 32 000 women aged 45 to 69 with risk factors for breast cancer. A pilot study has shown that a substantial reduction in fat intake (from an average of 39% of calories to 21%) can be achieved and sustained for at least one year. Critics of the proposed trial, which would last a decade and cost around $100m, argue that the resources might be spent more effectively on observational studies or laboratory research. But the organisers believe that this may be the only way to settle the long standing debate about fat and breast cancer. D C G SKEGG Professor of Preventive and Social Medicine and Director of the Hugh Adam Cancer Epidemiology Unit, University of Otago, Dunedin, New Zealand