To the best of our knowledge, this study is the first to evaluate the association of preconception exposure to traffic indicators and air pollution with the glucose concentration obtained in OGTT of healthy pregnant women in a middle income country. The main advantage of our study is the use OGTT results as a sensitive test for glucose homeostasis evaluations. We found that exposure to PM1, PM2.5 and PM10 was positively associated with FBG concentration, demonstrating that the levels of these pollutants might increase the risk of glucose intolerance. Moreover, TSL-100 m was positively associated with FBG and 1-h post-load glucose concentrations. Furthermore, DHMR was negatively associated with FBG and 1-h post-load glucose concentrations.
4.1. Available evidence
Given that this is the first study looking at the association between exposure to PM1, PM10 and traffic indicators with FBG, 1-h and 2-h post-load glucose concentrations obtained in OGTT as indicators of glucose homeostasis in healthy pregnant women; we could not compare or finding regarding these pollutants with results of previous studies. However, a study by Lu et al. 2017 on the 3859 subjects aged over 30 years found that higher FBG, 1-h, 2-h and 3-h glucose concentrations in pregnant women who lived in areas with higher PM2.5 level . A part of this study (i.e., significant positive association between PM2.5 exposures and FBG concentration) is in line with our findings; while, the associations of 1-h and 2-h glucose were inconsistent with our findings. Another study by Fleisch et al. 2014 on 2093 women found that second-trimester PM2.5 exposure was associated with IGT occurrence but not GDM . Choe et al. 2019 reported that PM2.5 exposure in 2nd trimester was associated with GDM development . A population-based retrospective cohort study by Shen et al. 2017, found higher PM2.5 exposure during 12 week preconception period as well as first two trimesters of pregnancy was significantly associated with increase in the risk of GDM in pregnant women . Moreover, the relationship between air pollution exposures and glucose homeostasis in non-pregnant healthy adults has been reported in previous studies. Peng et al. 2016 found that PM2.5 exposure was significantly associated with increase in FBG concentration in non-diabetic subjects . A study by Riant et al. 2018 on 2895 participants aged 40–65 years in France reported that PM10exposure was associated with higher FBG and HbA1c . A study by Chen et al. 2016, reported that short term exposure the PM10 (four days) was associated with higher FBG concentration as well as IFG occurrence . A systematic review and meta-analysis by Elshahidi et al. 2019 found that higher PM2.5 and PM10 exposures were associated with GDM development . These reports are in line with our findings.
We found, FBG and 1-h post-load glucose concentrations were positively associated with TSL-100 m and negatively associated with DHMR. There is limited evidence that investigated the relationship between traffic-related air and noise pollution with glucose tolerance during pregnancy [24, 30, 31]. Hooven et al. 2009 investigated the association between residential proximity to traffic and outcome of glucose homeostasis during pregnancy and reported there was no significant association between traffic indicators and GDM occurrence . Pedersen et al. 2017 examine the association of exposure to air and noise pollution in pregnant women and reported that there was no significant association of exposure to both pollutants and GDM development . In contrast, Malmqvist et al. 2013 in a study based on birth registry data of 81,000 pregnant women in Sweden, found that exposure to traffic indicators was significantly associated with GDM development . In our study, we found a positive correlation between TSL-100 m with PMs as well as a negative correlation between DHMR and PMs concentrations. Previous studies have shown that higher street length was significantly correlated with higher levels of PMs, especially in the smaller buffer sizes (e.g., 100 m) [20, 32–34]. These results could be explained our findings on the significant association of traffic indicators and glucose intolerance.
4.2. Biological plausibility
Although the precise mechanisms of the effect of traffic indicators and air pollution exposure on glucose tolerance are not fully understood, a number of potential mechanisms have been proposed. It has been shown that inhaled PMs into respiratory tract can pass through the alveolar cell and affect metabolism in extrapulmonary organs, e.g., liver [34, 35]. Similarly, non-water-soluble PMs with an aerodynamic diameter of ≤ 0.1 µm could alter glucose metabolism by entering the target cells [20, 36]. Another potential mechanism could be inhaled PMs that activate immunity cells, resulting in cytokines release [37, 38]. Some of these cytokines change glucose metabolism and hence glucose concentration in circulation . Besides, inhaled PMs could induce an autonomic nervous system imbalance, which directly affected insulin sensitivity [40, 41]. Moreover, previous studies suggested that exposure to air pollution induces oxidative stress and adipose tissue inflammation, which disrupts insulin signaling and results in insulin resistance [42, 43]. Insulin resistance could in turn increase FBG, 1-hr, and 2-h glucose concentrations. Moreover, exposure to air pollution may also affect the methylation of genes related to glucose metabolism. The change in methylation patterns affects glucose concentration by altering peripheral insulin sensitivity during pregnancy [44, 45]. Finally, changes in glucose homeostasis in healthy pregnant women might be due to the metabolic induction change in the hypothalamus [46, 47]. Our results of the associations between traffic indicators as well as PMs exposures and glucose intolerance could be explained by one or all of the above mechanisms.
4.3. Strength and limitation
The advantage of our study included use novel markers, access to full residential address histories and detailed information on exposures. Moreover, we studied the preconception exposure to air pollution as well as traffic indicators and glucose homeostasis during pregnancy, which no considered in previous studies. Furthermore, this study is the first report of LMICs about air pollution exposure and glucose intolerance in pregnant women.
Our study has limitations as well. The sample size of our study was relatively small. We measured PMs exposure using the LUR models, and we did not measure individual exposure to PMs during and before pregnancy. Diet can also affect blood glucose concentrations during pregnancy, which was not assessed in our study. Furthermore, we did not evaluate the level of maternal stress that may affect blood glucose levels. These limitations should be considered in future studies.