To the Editor Kaiser et al1 reported association between transient newborn hypoglycemia and test proficiency in an annual birth cohort at a university hospital in Arkansas. We have some concerns regarding the generalization of the association.
We disagree that a study implemented at a university hospital is population based. Rather, the facility-based approach may concentrate a relevantly large proportion of high-risk populations. The major characteristics of the study population are largely different than those of the general population in the United States: 81.5% of families had Medicaid or no insurance, 74.8% of mothers had an educational level below high school (as the authors mentioned), and the incidence of the overall preterm labor was as high as 31.5% (440 of 1395). Both preterm labor and poor economic conditions are closely related to neonate hypoglycemia and school performance.2 The skewed distribution of those crucial risk factors in the study population may lead to incorrect estimates of the incidence of transient newborn hypoglycemia and the association with school performance in the general population, despite a comparatively large sample size and consideration of several socioeconomic determinants.
Tobe RG. Association Between Transient Newborn Hypoglycemia and Test Proficiency. JAMA Pediatr. 2016;170(3):297. doi:10.1001/jamapediatrics.2015.4073