Association between childhood rhabdomyosarcoma and maternal history of stillbirths.

A case/control study of childhood rhabdomyosarcoma (RMS) was conducted utilizing information from birth records. Cases among Connecticut residents age 19 and younger diagnosed between 1960 and 1988 were identified from the Connecticut Tumor Registry files. Connecticut birth certificates were located for 103 of the 130 cases identified. A random sample of control birth certificates was frequency-matched to the cases by year of birth, sex, and ethnic origin, with a control:case ratio of 2:1. Information abstracted from birth certificates included birth weight, length of pregnancy, plurality, birth order, mother's prior stillbirths, mother's age, father's age, and father's occupation. Data were analyzed by conditional logistic regression. The major finding was an association between RMS and the mother having had one prior stillbirth or more (odds ratio = 3.7; 95% confidence interval = 1.5 to 8.9). Particularly noteworthy was the observation that 6 mothers of cases, but no mothers of controls, had had 2 or more prior stillbirths. The trend for increasing risk of RMS with increasing number of mother's prior stillbirths was highly significant (p = 0.0004). This association suggests that RMS and a class of stillbirths share a common etiologic factor. This factor may be genetic or may involve in utero exposure to an exogenous or endogenous agent.