Childhood lymphatic leukemia: prenatal seasonality and possible association with congenital varicella.

A statistically significant seasonality by month of birth, which differed for children diagnosed as having acute lymphatic leukemia in the under 2, 2-3 and 4-9-year age groups, was observed in urban counties of upstate New York. This suggested that the mothers of patients diagnosed in these three age groups might have been exposed to leukemogenic factors during different trimesters, but with each trimester consisting of the same specific group of months. Since a similar birth-month seasonality was not observed in rural regions, it seemed likely that leukemogenic factors might operate with a greater regularity in urban areas. Using these observations and reported trends for acute leukemia in the United States and New York State (excluding New York City), an effort was made to determine whether varicella, influenza, rubeola or rubella had similar epidemiologic features. Only varicella manifested both the urban-rural differences in seasonality and concomitant variations in time trends that were comparable to reported mortality trends for acute leukemia. Rank correlation coefficients for varicella and lymphatic leukemia incidence rates by month were also statistically significant when leukemia cases diagnosed in the three age groups and born in urban countries, were placed in the month of their appropriate trimesters. A retrospective search of varicella case records identified 63 instances of this viral disease complicating pregnancy. Three children resulting from these pregnancies subsequently developed acute lymphatic leukemia.