2.5 Sample size calculation
To test the hypothesis that GCF-PLAP concentrations are significantly higher during the first trimester in pregnancies that will subsequently develop PE, the estimated sample size was calculated based on previously published data41 and in our previous observations of differences between GCF-PLAP concentrations in women with PE and normotensive pregnant women (2044 ± 217 and 1880 ± 82 pg/ml; mean ± standard deviation, respectively)99. To test the hypothesis of mean differences in GCF-PLAP between patients affected by PE and heathy controls, a minimum cohort size of 406 pregnant women was calculated based on the following assumptions: a 6.4% prevalence of PE in the entire cohort; a significance level of 5%; a power of 80%; a two-sided test; and a loss-to-follow-up of 5%.