Maternal hypertension and infant growth

Dear Editor,

The Jornal de Pediatria published, in its 91st volume, a very interesting article entitled: “Growth of preterm low birth weight infants until 24 months corrected age: effect of maternal hypertension.”¹ The authors tackled a very important issue in the context of infant and maternal health worldwide, mainly considering the impact of the gestational period on the offspring's health and disease pattern during the life course. However, we would like to highlight some points in order to contribute to this subject.

According to the Task Force on Hypertension in Pregnancy, the hypertensive disorders in pregnancy are classified into: preeclampsia/eclampsia; chronic hypertension; preeclampsia superimposed on chronic hypertension; and gestational hypertension.² In our unpublished systematic review, we analyzed 45 papers (from 2008 to 2015) on the association between hypertensive disorders in pregnancy and offspring's medium- and long-term health outcomes. We found that the high heterogeneity of results among the studies was mainly caused by different classifications of maternal hypertension, and by the quality of adjustments performed by the authors. Thus, in this article, some methodological questions were raised.

Firstly, the authors defined two study groups according to the exposure or not to gestational hypertension syndrome; however, in their study description, it is not clear if women with chronic hypertension were also included in the hypertensive group. It is important to highlight that each hypertensive disorder has a different and complex clinical presentation, with diverse consequences in the offspring. Therefore, it is important that studies address the hypertensive disorders independently (i.e., chronic hypertension vs. gestational hypertension vs. preeclampsia) in their analysis.

Secondly, regarding the sample selection used by the authors, all children included in the study were born preterm (gestational age < 37 weeks) and had low birth weight (LBW; 1500 g to 2499 g). It is known that preterm birth and LBW are abnormal events, and the pathways that lead to these conditions are mostly pathological,³ so normotensive mothers must also have been exposed to adverse conditions during pregnancy. Therefore, by restricting the sample to preterm infants with LBW, the authors render normotensive mothers a greater chance of having these other adverse conditions, compared to the average population. The odds ratios (OR) of 0.47 for inadequate weight and 0.20 for inadequate length at 24 months described by the authors reflect this bias. The protective effect of maternal hypertension on growth was probably a result of unmeasured causes of LBW and preterm birth in the normotensive group. These other unmeasured or unknown disorders may possibly influence child growth as much as maternal hypertension.

Thirdly, to decrease the selection bias, the authors should had controlled the outcome for the causes of preterm birth and LBW, such as intrauterine infection, nutritional disorders, smoking, alcohol and drug consumption, violence, lower socioeconomic status, and other chronic diseases. However, the authors limited their logistic regression model to the following variables: gestational age, gender, and adequacy of birth weight for gestational age; a choice of variables that, besides being insufficient, may be a source of collision bias.

In conclusion, cohort studies have great importance in the scientific investigations, and the one conducted by the authors will contribute to this essential field of research. Nonetheless, the analysis of observational studies is extremely challenging and must be carefully performed.

Conflicts of interest

The authors declare no conflicts of interest.