Abstract
Preterm deliveries are a major multifactorial public health problem in French Guiana. Desert dust episodes have been associated with preterm delivery in Guadeloupe, a territory with similarities to French Guiana. We thus tried to replicate this finding in the context of French Guiana. A retrospective ecological cohort study combined daily PM10 concentration measurements during pregnancy and term at delivery extracted from French Guiana's computerized pregnancy delivery registry. Daily PM10 concentrations during the course of pregnancy were analyzed as mean concentrations and as the proportion of intense dust episodes (≥55 μg PM(10)/m(3)). These exposure variables were studied in relation to the outcome of preterm delivery. Overall, 3,321 pregnant women with complete daily PM10 measurements were included, of whom 374 (11.26%) delivered prematurely. Among preterm deliveries, 168 (44.9%) were spontaneous deliveries and 206 (55.1%) were induced. Rank-sum tests showed that, for spontaneous and induced spontaneous deliveries, both mean PM10 concentrations and proportions of intense desert dust episodes were significantly greater among preterm births than among term births. Although the proportion of intense desert dust episodes during pregnancy was significantly associated with spontaneous preterm deliveries, the relation was U-shaped, with an adjusted odds ratio (AOR) = 2 (95%CI = 1.2-3.1) for lowest values relative to median values and AOR = 5.4 (95%CI = 3.2-8.9) for the highest values relative to median values. Similarly, the proportion of intense desert dust episodes during pregnancy was also significantly associated with induced preterm deliveries in a U-shaped manner (AOR = 2.7 (95%CI = 1.6-4.5) for the lowest relative to median values and AOR = 6.8 (95%CI = 3.9-11.9) for the highest relative to median values). Although in our study the relation between PM10 concentrations appeared non-linear, the highest mean concentrations and intense desert dust episodes were indeed associated with both spontaneous and induced preterm delivery.