Maternal Hemoglobin Concentration during Pregnancy and Risk of Infant Leukemia.

Abstract

Little is known about the etiology of childhood leukemia. Elevated birth weight (> 4000 g) is gaining acceptance as a risk factor, particularly for children diagnosed < 2 years of age, but the underlying biological or physiological explanation for this association has yet to be defined. At least three recent studies have demonstrated a positive association with childhood leukemia and maternal anemia during pregnancy. Physiologic anemia of pregnancy, which results from normal plasma volume expansion, is thought to produce optimal blood viscosity for the low flow area of the placenta and is associated with the highest average birth weights in the offspring. We therefore hypothesized that maternal physiologic anemia of pregnancy would have a positive association with infant leukemia. To test this hypothesis, we utilized data from a case-control study conducted by the Children's Oncology Group (COG) that consisted of 240 incident cases of acute infant leukemia (< 12 months of age) diagnosed between 1996 and 2002 and 255 controls obtained by random digit dialing. Maternal blood counts were obtained by medical record abstraction. Information regarding the patients' diagnoses and cytogenetic/molecular abnormalities were obtained from COG institutions. Physiologic anemia of pregnancy was defined as a hemoglobin concentration of < 11 g/dL in the second or third trimester of pregnancy (WHO classification). Anemia earlier in pregnancy is more frequently due to nutritional deficiencies and can be associated with poor fetal outcomes and thus was excluded. Maternal hemoglobin values from the second or third trimester were available for 178 case and 180 control mothers. Data were analyzed using unconditional logistic regression. We found no evidence that physiologic anemia of pregnancy was associated with birth weight of > 4000 g (odds ratio (OR) 1.05, 95% confidence interval (CI)=0.55–1.99) Overall, the OR for physiologic anemia of pregnancy and risk of infant leukemia was 0.85 (95%CI=0.53–1.37). The association remained non-significant when cases were divided into acute myeloid leukemia (AML) (n=63) and acute lymphocytic leukemia (ALL) (n=115) with ORs of 0.64 (95%CI=0.32–1.30) and 0.98 (95%CI=0.58–1.66), respectively. The association was also not significant when cases were stratified by MLL gene abnormality status (OR=0.77; 95%CI=0.42–1.40 and OR=0.88; 95%CI=0.45–1.73, for cases with (n=92) and without (n=62) an MLL gene abnormality, respectively). These associations remained non-significant when the data were adjusted for birth weight, maternal race, and birth order. More stringent criteria for anemia also did not modify the results. We found no evidence for an increased risk of infant leukemia in the offspring of mothers with physiologic anemia of pregnancy. This work was supported by NIH R01 CA79940 and the Children's Cancer Research Fund.