Abstract
Mechanistic pathways linking maternal polyunsaturated fatty acid (PUFA) status with gestational length are poorly delineated. This study examined whether inflammation and sleep quality serve as mediators, focusing on the antiinflammatory ω-3 docosahexaenoic acid (DHA; 22:6n3) and proinflammatory ω-6 arachidonic acid (AA; 20:4n6). Pregnant women (n = 135) provided a blood sample and completed the Pittsburgh Sleep Quality Index (PSQI) at 20-27 weeks gestation. Red blood cell (RBC) fatty acid levels were determined by gas chromatography and serum inflammatory markers [interleukin (IL)-6, IL-8, tumor necrosis factor-α, IL-1β, and C-reactive protein] by electrochemiluminescence using high sensitivity kits. Both higher serum IL-8 (95% CI = 0.10,3.84) and poor sleep (95% CI = 0.03,0.28) served as significant mediators linking lower DHA:AA ratios with shorter gestation. Further, a serial mediation model moving from the DHA:AA ratio → sleep → IL-8 → length of gestation was statistically significant (95% CI = 0.02, 0.79). These relationships remained after adjusting for depressive symptoms, age, BMI, income, race, and smoking. No interactions with race were observed in relation to length of gestation as a continuous variable. However, a significant interaction between race and the DHA:AA ratio in predicting preterm birth was observed (p = 0.049); among African Americans only, odds of preterm birth decreased as DHA:AA increased (p = 0.048). These data support a role for both inflammatory pathways and sleep quality in linking less optimal RBC PUFA status with shorter gestation in African American and European American women and suggest that African-Americans have greater risk for preterm birth in the context of a low DHA:AA ratio.