Risk factors for preeclampsia in infertile Chinese women with polycystic ovary syndrome: A prospective cohort study: Methodological issues

Dear Editor‐in‐Chief,

We have read the article by Xia and colleagues1 published in The Journal of Clinical Hypertension in December 2016 enthusiastically. The authors aimed to evaluate preconception risk factors for preeclampsia in women with polycystic ovary syndrome. They conducted a prospective study on 92 infertile Chinese women with polycystic ovary syndrome, in whom 15 were diagnosed with preeclampsia. They found that preconception sex hormone–binding globulin (P=.027) and insulin level at 120 minutes (P=.048) were independently associated with preeclampsia. Although the current study makes valuable contributions to the area, some methodological points need to be considered.

First, the authors constructed a multivariate model in which only independent variables with P<.05 in the univariate model were included in the multivariate model, which is questionable. In the standard model construction, relaxed univariate P values such as P<.2 are suggested to be used.2 In fact, the only independent variables with a large effect are included in the multivariate model when rough univariate P values are used in the variable selection. This means that variables with low effects are missed. Steyerberg named this phenomenon as Testimation bias.3, 4 Hence, Xia and colleagues constructed a multivariate model in which variables with P<.05 were only imported into the multivariate model and we wonder why fasting plasma insulin (P=.131), area of insulin under the curve (P=.137), homeostasis model of assessment—insulin resistance (P=.099), and triglycerides (P=.058) were not included in the model.

Second, the sample size calculation is critical step in epidemiological studies. The minimum required sample size and power for statistical tests is determined in this step.5 The authors compared risk factors between patients with (n=15) and without (n=77) preeclampsia; however, it is not clear whether the minimum power for statistical tests was provided.

Third, at least 10 events per variable should be provided per each variable included in the multivariate model to avoid overparameterization and sparse data bias.6, 7 In the study conducted by Xia and colleagues, only 15 events (preeclampsia) exist, whereas more than five variables were included in the multivariate model. In fact, some significant associations may not be detected in the multivariate constructed by Xia and colleagues, as the overparameterization and sparse data bias attenuates the statistical power.6

A take‐home message for the readers is that Testimation bias should be avoided using the appropriate criterion for the variable selection. On the other hand, appropriate events per variable should be applied to avoid overparameterization.

DISCLOSURES

The authors declare no conflicts of interest.