Electrohysterogram (EHG) signals have emerged as a valuable tool for analyzing uterine activity and predicting pregnancy outcomes. This study investigates key statistical features extracted from EHG signals and examines their significance in differentiating pregnancy cases. A comprehensive statistical analysis, including analysis of variance (ANOVA) and Tukey's post-hoc test, was conducted on twelve extracted features. Results indicate that six features—standard deviation, root mean square (RMS), entropy, spectral entropy, Higuchi fractal dimension (HFD), and approximate entropy (ApEn)—exhibited significant differences among the cases. These findings highlight the potential of these features in distinguishing pregnancy conditions and improving predictive models for preterm birth risk.