Increased inflammatory markers in late pregnancy depression: The role of progesterone and allopregnanolone

Background Previous studies have found relationships between inflammatory markers and perinatal depression, but these have rarely been studied in the context of reproductive hormone levels. Methods Progesterone and allopregnanolone (by ELISA) and cytokines (IL6, IL10, IFN- γ , TNF- α , and GM-CSF, by cytometric bead array) were measured during the third trimester of pregnancy for 38 women with a history of mood disorders. Depression severity was measured with the Edinburgh Postnatal Depression Scale (EPDS). Results In separate linear regression models, IL-6, TNF- α , and GM-CSF measured in the third trimester were significantly correlated with EPDS score (IL-6: β  = 0.17, p  = .024; TNF- α : β  = 0.09, p  = 0.021; GM-CSF: β  = 0.11, p 0.0001 ). In a model controlling for gestational week of blood draw and for the effect of all cytokines on one another, only GM-CSF retained significance ( β  = 0.182, p  = 0.003). In a fully corrected model that also controlled for third trimester levels of progesterone and allopregnanolone, results for GM-CSF were similar ( β = 0.187 , p = 0.002 ), and no other cytokine retained significance. Conclusions This study confirms previous work showing correlations between pro-inflammatory cytokines and depressive symptoms in pregnancy. Our results indicate that some such correlations may be accounted for by the effect of inflammatory markers on one another. Moreover, progesterone and allopregnanolone do not appear to mediate the relationship between a pro-inflammatory cytokine (GM-CSF) and depressive symptoms.