Sir,

Szegda et al. (2017) examined depression during pregnancy and its association with two birth outcomes: preterm birth and small-for-gestational age (SGA). Their study is important because, although the sample was small (n = 1262), only 2.5% (n = 31) of the women used antidepressant drugs during pregnancy. The sample therefore offered a good opportunity to examine the effect of untreated depression on pregnancy outcomes. However, the authors did not examine the use of other psychotropic medication, nor did they include medical illness and treatments among the covariates.

Furthermore, there are two important limitations to the analysis that was reported:

  1. 1.

    Preterm birth and SGA were examined as categorical variables. The authors could have increased the power in their analyses by considering these as continuous variables; that is, as actual gestational age at birth and as age-corrected birth weight.

  2. 2.

    A large number of analyses were performed. Additionally, in some places the same data were analyzed in different ways; for example, early pregnancy depression ratings were examined as a continuous variable, as a categorical variable for at least probable minor depression, and as a categorical variable for at least probable major depression. No correction for multiple hypothesis testing was applied. Therefore, the significant findings in the analyses could have been false positives (Type 1 errors).

The statistically significant findings reported in this study may therefore be considered tentative, at best.