This article is only available in the PDF format. Download the PDF to view the article, as well as its associated figures and tables.
Dr Lowenfels has demonstrated some interesting ways to make use of general statistical data. We prefer, however, to rely on scientific principles that require comparisons to be conducted between similar groups and examined in similar ways. If he had applied the International Coffee Association's survey methods to determine the coffee-drinking habits of the "case" group, if those habits were unaffected by the disease, and if the data were free from the other biases we discussed, his table might warrant the calculations he has performed.If coffee is indeed a causal culprit, the continuing rise in annual incidence of pancreatic cancer seems strikingly inconsistent with the continuing decline in coffee drinking since 1950. On the other hand, the changing data for annual incidence of pancreatic cancer are probably grossly inaccurate, since many of the cases detected with improvements in modern diagnostic technology would not have been recognized in the
Feinstein AR, Horwitz RI. Coffee and Cancer of the Pancreas-Reply. JAMA. 1982;247(7):980. doi:10.1001/jama.1982.03320320019019