To the Editor.—
Kendell and Kemp1 conclude that the three studies they report do not support the influenza A virus hypothesis in schizophrenia. We cannot accept their conclusions due to serious inadequacies in their data sources and methods.Kendell and Kemp reason that the severe June and November 1918 and March 1919 influenza epidemics that devastated Scotland should have increased the rate of schizophrenia among those exposed during gestation. To identify people with schizophrenia born in 1918 and 1919, they used a national Scottish register of psychiatric hospitalizations. They inspected their data visually and concluded that there was no increase in the number of people born with schizophrenia in 1918 and 1919. For the period in question, however, the register did not record date of birth; only the patient's report of his or her age was recorded. Year of birth was calculated by subtracting the age reported from the