From the Department of Epidemiology, Biostatistics and Occupational Health, McGill University, Montreal, QC, Canada.
Child Health Evaluative Sciences, The Hospital for Sick Children Research Institute, Toronto, ON, Canada.
Epidemiology. 2023 Mar 1;34(2):247-258. doi: 10.1097/EDE.0000000000001574. Epub 2022 Dec 1.
Evidence on the effects of in utero exposure to maternal diabetes on cerebral palsy (CP) in offspring is limited. We aimed to examine the effects of pregestational (PGDM) and gestational diabetes (GDM) separately on CP risk and the mediating role of increased fetal size.
In a population-based study, we included all live births in Ontario, Canada, between 2002 and 2017 followed up through 2018 (n = 2,110,177). Using administrative health data, we estimated crude and adjusted associations between PGDM or GDM and CP using Cox proportional hazards models to account for unequal follow-up in children. For the mediation analysis, we used marginal structural models to estimate the controlled direct effect of PGDM (and GDM) on the risk of CP not mediated by large-for-gestational age (LGA).
During the study period, 5,317 children were diagnosed with CP (187 exposed to PGDM and 171 exposed to GDM). Children of mothers with PGDM showed an increased risk (hazard ratio [HR]: 1.84 [95% confidence interval (CI): 1.59, 2.14]) after adjusting for maternal sociodemographic and clinical factors. We found no associations between GDM and CP (adjusted HR: 0.91 [0.77, 1.06]). Our mediation analysis estimated that LGA explained 14% of the PDGM-CP association.
In this population-based birth cohort study, maternal pregestational diabetes was associated with increased risk of CP, and the increased risk was not substantially mediated by the increased fetal size.
关于母体糖尿病对后代脑瘫（CP）影响的证据有限。我们旨在分别研究孕前（PGDM）和妊娠期糖尿病（GDM）对 CP 风险的影响，以及胎儿大小增加的中介作用。
在一项基于人群的研究中，我们纳入了 2002 年至 2017 年期间在加拿大安大略省的所有活产儿，并随访至 2018 年（n=2110177）。我们使用行政健康数据，通过 Cox 比例风险模型估计 PGDM 或 GDM 与 CP 之间的粗关联和调整关联，以考虑到儿童之间随访时间的不平等。对于中介分析，我们使用边缘结构模型来估计 PGDM（和 GDM）对非由大于胎龄儿（LGA）介导的 CP 风险的控制直接效应。
在研究期间，有 5317 名儿童被诊断为 CP（187 名暴露于 PGDM，171 名暴露于 GDM）。在调整了母亲的社会人口统计学和临床因素后，PGDM 母亲的孩子发生 CP 的风险增加（风险比[HR]：1.84[95%置信区间（CI）：1.59，2.14]）。我们没有发现 GDM 与 CP 之间的关联（调整 HR：0.91[0.77，1.06]）。我们的中介分析估计，LGA 解释了 14%的 PDGM-CP 关联。
在这项基于人群的出生队列研究中，母体孕前糖尿病与 CP 风险增加相关，而这种风险增加并没有被胎儿大小增加显著中介。
