Van Howe Robert S, Storms Michelle R
Department of Pediatrics and Human Development, Michigan State University College of Human Medicine, Marquette, Michigan 49855, USA.
Am J Perinatol. 2008 May;25(5):283-9. doi: 10.1055/s-2008-1076602. Epub 2008 Apr 24.
Our purpose was to identify factors that may contribute to hypoglycemia in large for gestational age (LGA) infants and subsets of infants for whom blood glucose screening would be clinically helpful. LGA infants of 36 to 42 weeks of gestational age using the Denver criteria were identified from the 1999 through 2001 birth log at Marquette General Hospital, Marquette, Michigan, a rural referral hospital, for a retrospective chart review. Infants of mothers with diabetes were excluded. The impact of maternal factors, intrauterine factors, and birthweight on blood glucose levels was assessed using marginal mixed models, Poisson regression, and receiver operating characteristic curves. We found 457 infant records documented blood glucose levels among the 727 infant records analyzed. The lowest blood glucose levels occurred in the first 90 minutes of life. Blood glucose levels were positively associated with weeks of gestation and breastfeeding. Birthweight did not predict hypoglycemia in any of the multivariable models constructed. Symptoms consistent with hypoglycemia occurred only in boys and were negatively associated with gestational age. Models using prenatal and perinatal factors to predict hypoglycemia or symptomatic hypoglycemia had little clinical value. In this large community-based sample of LGA infants, we found no evidence that higher birthweight increases the risk for hypoglycemia. The clinical usefulness of routine blood glucose monitoring in healthy LGA infants is not supported by this retrospective analysis and needs to be established by prospective studies.
我们的目的是确定可能导致大于胎龄(LGA)婴儿低血糖的因素,以及确定哪些亚组婴儿进行血糖筛查在临床上会有帮助。我们从密歇根州马奎特市一家农村转诊医院——马奎特综合医院1999年至2001年的出生记录中,根据丹佛标准确定了孕龄为36至42周的LGA婴儿,以进行回顾性病历审查。患有糖尿病母亲的婴儿被排除在外。使用边际混合模型、泊松回归和受试者工作特征曲线评估母亲因素、宫内因素和出生体重对血糖水平的影响。在分析的727份婴儿记录中,我们发现有457份记录了血糖水平。最低血糖水平出现在出生后的前90分钟。血糖水平与孕周和母乳喂养呈正相关。在构建的任何多变量模型中,出生体重均不能预测低血糖。仅在男婴中出现了与低血糖相符的症状,且与孕周呈负相关。使用产前和围产期因素预测低血糖或有症状低血糖的模型几乎没有临床价值。在这个基于社区的大型LGA婴儿样本中,我们没有发现证据表明较高的出生体重会增加低血糖风险。这项回顾性分析不支持对健康LGA婴儿进行常规血糖监测的临床实用性,需要通过前瞻性研究来确定。