Maternal exposure to benzene and toluene and preterm birth. A longitudinal study

ABSTRACT BACKGROUND: Exposure to air pollutants has several effects on human health, including during pregnancy. OBJECTIVE: To identify whether exposure to benzene and toluene among pregnant women contributes to preterm delivery. DESIGN AND SETTING: Longitudinal study using data on newborns from mothers living in São José dos Campos (SP) in 2016, who had been exposed to benzene and toluene. METHODS: A logistic regression model with three hierarchical levels was constructed using maternal variables relating to newborns, and using benzene and toluene concentrations in quartiles. Occurrences of cesarean births, twins or malformations were excluded. Maternal exposure windows of 5, 10, 15, 30, 60 and 90 days prior to delivery were considered. RESULTS: Out of the 9,562 live births, 3,671 newborns were included and 343 newborns were born at less than 37 weeks of gestation (9.3%). The average birth weight was 3,167.2 g. Exposure to benzene and toluene was significantly associated (P = 0.04) with preterm delivery in the five-day window. There was no association in any of the other exposure windows. CONCLUSIONS: It was possible to identify that maternal exposure to benzene and toluene has an acute effect on preterm delivery.


INTRODUCTION
Preterm birth is defined as birth before 37 full weeks of gestation and is associated with higher neonatal morbidity and mortality, 1 along with diseases resulting from prematurity. The effects from preterm birth may extend into adulthood. 2 In a multicenter study conducted in Brazil, the prematurity rate was estimated to be 11.7%. However, the rate reported through the national information system for live births (SINASC) was 7.1%. 3 Many factors have been associated with preterm delivery, including pregnancies at ages of below 19 and above 34 years, low maternal schooling level, low number of prenatal visits, twin pregnancy, shortened cervix, premature rupture of the amniotic membrane, inflammation and recent exposure to environmental pollutants. 4,5 Carbon monoxide (CO), ozone (O 3 ), nitrogen dioxide (NO 2 ) and sulfur dioxide (SO 2 ) are the gaseous pollutants that have been most studied. 6 Particulate matter with an aerodynamic diameter of less than 10 μm (PM 10 ) has also been investigated. This is composed of a mixture of solid and liquid particles that adsorb polycyclic aromatic hydrocarbons and ions such as sulphates, nitrates and metals. 7 Besides these, reports have also shown that associations with benzene and toluene exist. These substances are emitted mainly by the petrochemical industry, but also come from the vehicular fleet. It has not been defined whether their effects occur in an acute or chronic manner, or what the mechanism of action for these effects might be. However, it is known that the action of air pollutants is associated with the pro-oxidant effects of lipids and proteins, along with formation of free radicals, which generate oxidative stress and inflammation. [8][9][10][11]

OBJECTIVE
The objectives of this study were to identify the possible effects of maternal exposure to benzene and toluene at the onset of preterm delivery and to evaluate whether these are acute or chronic. and the time at which prenatal care began, which was categorized as favorable when this was no later than the third month or as unfavorable when this was from the fourth month onwards.
The proximal level included the sum of concentrations of the pollutants recorded in the 5, 10, 15, 30, 60 and 90-day windows prior to childbirth.
The pollutants benzene and toluene were analyzed in terms of daily averages, in μg/m 3 . For analysis purposes, the sums of the concentrations in the 5, 10, 15, 30, 60 and 90-day windows before childbirth were considered. The aim of this procedure was to assess whether the cumulative effects of these pollutants on the onset of preterm delivery occurred in an acute or a chronic manner. Bivariate analysis was performed on the variables within the three levels, separately. At the distal level, variables with a P-value ≤ 0.20 were included in the multivariate analysis and variables that presented P-value < 0.05 at this level were maintained at the next phase of the analysis. For the statistical analysis, we used the Epi Info software, version 7.2. The significance level used in the analyses was alpha = 5%.
Because this study used data that are available online with public access, and it was impossible to identify the subjects analyzed, the study was not submitted to a research ethics committee. The results from the bivariate analysis on the distal and intermediate levels showed that only the number of prenatal care consultations was significant (P-value < 0.01). These results are shown in  The χ 2 trend for the accumulated five-day periods is shown in Table 4. It shows that both benzene and toluene had significant cumulative effects.

DISCUSSION
This study identified an association between preterm birth and maternal exposure to benzene and toluene five days before delivery, after adjusting for the number of consultations attended by the pregnant woman during prenatal care. Thus, a possible acute effect was found.
It is known that exposure to air pollutants is associated with several respiratory and cardiovascular diseases, as well as with preterm birth. 8 Few studies in the literature have correlated exposure to benzene and toluene with premature delivery. In Brazil, to the best of our knowledge, this is the first study on this association.
The analysis on distal and intermediate factors (  5 However, the factor of maternal age was not found to be associated with preterm birth, contrary to the findings from a Chinese study. In that study, two age groups were formed and the group older than 30 years presented higher risk of prematurity. 14 There was also no association with maternal marital status, since women living without partners were not found to be at higher risk of preterm birth. This was discordant with the findings from an American study, which showed that those living without a partner presented significantly higher risk (relative risk, RR 10.67%; 95% CI: 10.23-11.12), in relation to those with a partner (RR = 7.21%; 95% CI: 6.92-7.50). 15 Regarding schooling level as a risk factor for low birth weight, our findings were concordant with those from a Brazilian study carried out in the city of Rio de Janeiro, where no association between low schooling level and low birth weight (OR = 0.93) was found. 16 However, the results from the present study were contrary    to the findings from a study conducted in Spain (OR = 2.05) 17 and to those from a Brazilian study conducted in the city of São José do Rio Preto (P < 0.01), which found strong associations between maternal schooling level and low birth weight. 18 Other comparisons with the Brazilian literature are very difficult, given the scarcity of studies on the effect of maternal exposure to air pollutants and occurrence of premature delivery. In a single study conducted by Lima et al., 19 a significant association between premature delivery and exposure to PM 10 was found on the day of delivery (lag 0), the day before delivery (lag 1) and three days prior to delivery. However, a different approach was used in their study.
Analysis on acute exposure ( we found an association with maternal exposure for five days. 10 On the other hand, there was no significant association between maternal exposure to benzene and toluene over other exposure windows such as 30, 60 and 90 days. The χ 2 trend analysis on the five-day windows, with adjustment for the number of consultations, showed that the higher the concentrations of benzene and toluene were, the greater the chance of birth of a preterm fetus was. This highlights the possible cumulative effect of exposure. One possible mechanism for the onset of preterm birth caused by exposure to air pollutants is thought to relate to release of cytokines and reactive oxygen species, thereby causing oxidative stress. 20 In the case of benzene and toluene, there are no studies that explain the mechanism of action of these two pollutants in triggering preterm birth. Moreover, there are no World Health Organization (WHO) guidelines for acceptable values for benzene or toluene, unlike in relation to other air pollutants. 6 This study had certain limitations. The mothers' socioeconomic conditions, home address, living conditions, preexisting diseases and smoking status were not available through SINASC.
The concentrations of the air pollutants were considered homogeneous throughout the city, which might not reflect the reality.
Moreover, the pregnant women had been free to circulate during their pregnancies. Another possible limitation of the present study may have been its non-inclusion of other air pollutants. However, a previous analysis had not shown any correlation between these air pollutants and the outcome.

CONCLUSIONS
Despite the limitations of the present study, it was possible to identify the deleterious effect of maternal exposure to benzene and toluene on the preterm delivery in a medium-sized city in Brazil, and to ascertain that this was an acute effect. The results presented here may be useful for public policy implementation, with the aim of reducing these concentrations.