Women in both of these case- control studies were examined at the end of pregnancy or after delivery, which does not convincingly prove an an- tecedent exposure and thus causality. Despite this limitation, theseearlystudiesledtothehypothesisthatperiodontopathic bacteria, primarily Gram-negative anaerobes, may serve as a source for endotoxin and lipopolysaccharides, which then increases local inflammatory mediators including PGE2, and cytokines, and that this increases systemic inflammatory me- diators that can then lead to preterm birth [23].