Air Pollution During Pregnancy and Neonatal Outcome: A Review

Birth weight

Most data regarding potential harmful effects of air pollution during pregnancy exist using birth weight as an outcome marker, as it is routinely measured and easily obtainable from hospital charts. Furthermore, it seems to be a good marker for intrauterine development and a good predictor of subsequent morbidity. The overall effect for a single subject (hazard) is rather small; however, on a population level, this may still have significant consequences. Table 1 gives an overview of recent publications, although the comparison of results is hampered by the heterogeneity between studies. As shown in Table 1 in detail, many studies found evidence for lower birth weight at term after higher exposure to air pollution, whereas other studies did not find any association. Apart from random variability, several factors may be responsible for this discrepancy: accuracy and method of exposure assessment, study size, absolute levels of pollutants, and land-use differences between studies. Moreover, how confounding is being addressed may also play a major role in determining possible effects. For example, social class as one of the most relevant confounding factors⁽¹¹⁾ may be differently distributed within polluted areas depending on countries or regions.⁽¹²⁾

To overcome this heterogeneity, international multicenter projects are trying to investigate the effect of air pollution on a large scale, with comparable exposure assessment and considering similar confounding factors in the analysis (ESCAPE,⁽¹³⁾ ICAPPO,⁽¹²⁾ GA²LEN,⁽¹⁴⁾ ENRIECO⁽¹⁵⁾). Within Spain, four cohorts were merged together to investigate a possible effect of exposure to nitrogen dioxide (NO₂) on anthropometric measures. The meta-analysis from these four cohorts, including 2,337 mothers, showed weak evidence for lower birth length: –0.9 (95% CI –0.27; –0.01) mm per increase in 10 μg/m³ of NO₂.⁽¹⁶⁾ In a world-wide project, 13 groups of researchers, already involved in projects regarding the effect of air pollution, formed the International Collaboration on Air Pollution and Pregnancy Outcomes (ICAPPO) and conducted a reanalysis of their data in a consistent manner. The first results, summarized quantitatively using a meta-analysis, provide an estimate of the relative risk of low birth weight (defined as lower than 2,500 g) associated with each 10 μg/m³ increase in mean particulate matter (PM₁₀) concentration. The results of the crude correlation show a weak but significant association: odds ratio (OR) 1.03 (95% CI 1.01; 1.05). Results remained stable after adjustment for confounding: OR 1.02 (95% CI 1.00; 1.03). However, there was significant heterogeneity between studies.⁽¹⁷⁾

Neonatal prematurity

An analysis of the effect of air pollution on term birth weight without considering premature birth can bias the results. If fetuses more susceptible to air pollution encounter premature birth, they are not eligible for studies on the effect on birth weight in term-born babies. This may attenuate a possible effect. Table 2 gives an overview of current evidence of premature birth after prenatal exposure to air pollution. Most of the studies found a positive association between exposure to air pollution and premature birth; however, again comparison between studies is difficult due to heterogeneity.

One of the most recently published studies about the effect of NO₂, PM₁₀, and sulfur dioxide (SO₂) on premature birth was conducted by Zhao et al. in Guagzhou China.⁽¹⁸⁾ The study area shows a rather high pollution level with a NO₂ mean of 61 μg/m³, PM₁₀ mean of 82 μg/m³, and SO₂ mean of 51 μg/m³. They performed time-series analysis and found a consistent cumulative effect of the exposure to pollutants, resulting in a risk ratio (RR) for preterm birth for NO₂ of 1.054 (95% CI 1.008; 1.1), for PM₁₀ of 1.067 (95% CI 1.013; 1.124), and for SO₂ of 1.13 (95% CI 1.05; 1.21) per increased concentration of 100 μg/m³of pollutant.⁽¹⁸⁾ Another time-series analysis performed in Atlanta (United States) did not confirm such a clear correlation: most of the results were not significant; however, NO₂ exposure in the last 6 months of pregnancy was associated with an increased rate of preterm deliveries: RR 1.06 (1.02; 1.09) per increase of 5 ppb NO₂.⁽¹⁹⁾ In a case-control survey nested in a birth cohort, Ritz et al. examined the effect of PM_2.5, NO₂, ozone (O₃), and carbon monoxide (CO) exposure in the first trimester of pregnancy, as well as in the 6 weeks before birth.⁽²⁰⁾ They compared the odds for preterm birth within a subsample where detailed information about other relevant risk factors such as smoke and alcohol exposure was available. They found increased odds of preterm birth after higher exposure to all pollutants for both exposure time spans, e.g., OR for preterm birth of 1.0, 1.15, and 1.29 for PM_2.5 tertile increase (<18.6; 18.6 to 21.4;>21.4 μg/m³).⁽²⁰⁾

Infant mortality

A World Health Organization review from 2005 concludes that good evidence exists for an association between air pollution exposure during pregnancy and infant mortality.⁽²¹⁾ A review from Glinianaia et al. found best evidence for postneonatal mortality due to respiratory reasons.⁽²²⁾ Since then, two large studies have been published confirming these results. Woodruff et al. conducted a study with a population of 3,583,495 births including 6,639 postneonatal deaths in 96 counties throughout the United States.⁽²³⁾ They considered the monitoring data of pollutants in the first 2 months after birth as a proxy for chronic exposure, as these data were highly correlated with annual averages, after adjusting for season. Linking the data with birth and death records, they showed that infants exposed to the highest quartile of pollution (>34 μg/m³ PM₁₀ and>18,7 μg/m³ PM_2.5) had elevated odds especially for respiratory mortality compared with the lowest quartile (<23.3 μg/m³ PM₁₀ and<11.7 μg/m³ PM_2.5): for PM₁₀ OR 1.31 (95% CI 1.00; 1.71) and for PM_2.5 OR 1.39 (95% CI 1.04; 1.85).⁽²³⁾ The second study was conducted in South Korea. Son et al. assessed PM₁₀, PM_2.5, and total suspended particles (TSP) during pregnancy and found a hazard ratio for all causes of mortality of 1.44 (95% CI 1.06; 1.97) for 8.9 μg/m³ TSP increase, 1.65 (95% CI 1.18; 2.31) for 6.9 μg/m³ PM₁₀ increase, and 1.53 (95% CI 1.22; 1.9) for 3.15 μg/m³ PM_2.5 increase.⁽²⁴⁾ When they analyzed respiratory causes of mortality, the hazard ratios increased to 3.78 (95% CI 1.18; 12.13) for 8.9 μg/m³ TSP increase, 6.2 (95% CI 1.5; 25.6) for 6.9 μg/m³ PM₁₀ increase, and 3.15 (95% CI 1.26; 7.85) for 3.15 μg/m³ PM_2.5 increase.⁽²⁴⁾

In the framework of an estimation of years of life lost attributable to air pollution, Röösli et al.⁽²⁵⁾ conducted a meta-analysis of five studies on infant mortality.^(23,26–29) They found an overall relative risk of 1.056 (95% CI 1.026; 1.088) per increase of 10 μg/m³ PM₁₀. Applying this risk estimate to the Swiss situation in the year 2000 (77,800 infants aged below 1 year and an average PM₁₀ level of 19.6 μg/m³) resulted in 1,705 (95% CI 915; 2,482) estimated years of life lost due to ambient air pollution.⁽²⁵⁾

Lung development and respiratory effects

Development of airways and the lung begins with the embryonic phase at 4–7 weeks of gestational age, reaches the alveolar phase at around 36 weeks of gestational age, and continues until adolescence. This maturational process takes place over a relatively long time period compared with that of other organs, so potentially harmful factors have a greater opportunity to interfere with the developmental process. Moreover, the repair mechanisms of the developing lung tissue are not as efficient as those of the mature lung.⁽¹⁾ Considering all these aspects, it seems quite plausible that—analogous to tobacco smoke⁽⁶⁾ or postnatal exposure to air pollution⁽³⁰⁾—prenatal exposure to air pollution may also affect lung development.

In the BILD study, a healthy birth cohort in Switzerland,⁽³¹⁾ where air pollution is relatively low (mean PM₁₀ 22 μg/m³), we performed lung function measurements at 5 weeks of postnatal age.^(32,33) Air pollution exposure was assessed during pregnancy and stratified per trimester. Adjusted results showed an increase of minute ventilation in newborns of 24.9 mL/min (95% CI 9.3; 40.5) per 1 μg/m³ increased exposure to PM₁₀ during pregnancy.⁽³⁴⁾ This suggests that prenatal exposure to air pollution may affect lung development. This finding may be one explanation for the above-mentioned increased risk for infant mortality due to respiratory causes. Exposure to air pollution during pregnancy seems to affect lung function not only in infants, but also in older children. In Poland, measurements performed at 5 years of age showed reduced forced vital capacity (FVC) and forced expiratory volume in the first second (FEV₁) if higher prenatal PM_2.5 exposure occurred.⁽³⁵⁾ Another study conducted in California on asthmatic children aimed to investigate the influence of prenatal and early childhood exposure to air pollution on lung function at 6–11 years of age. The results showed consistent evidence for an association between prenatal exposure and FVC, peak expiratory flow (PEF), and FEV₁.⁽³⁶⁾ Impaired lung function should correspond to susceptibility to airway infection and increased respiratory symptoms. Studies on prenatal exposure to NO₂ and respiratory symptoms during the first year of life,⁽³⁷⁾ prenatal exposure to O₃ and NO₂ and occurrence of apnea,⁽³⁸⁾ prenatal polycyclic aromatic hydrocarbon exposure and frequency and duration of respiratory symptoms over the first year of life,⁽³⁹⁾ and prenatal exposure to fine particle matter and severity of respiratory symptoms within different wheezing phenotypes⁽⁴⁰⁾ support this hypothesis. Air pollution affects not only lung function and respiratory symptoms, but also the development of asthma. In a case-control study including 37,401 subjects in Canada, Clark et al. investigated the incidence of asthma up to the age of 3–4 years in relation to individual exposure to air pollution during gestational time and the first year of life.⁽⁴¹⁾ They found a positive association: a 10 μg/m³ increase in NO₂ was associated with an OR of 1.12 (95% CI 1.07; 1.17) for asthma incidence.⁽⁴¹⁾

Immune system

Environmental exposure occurring in early life influences the pattern of immune maturation and resulting immune response.⁽⁴²⁾ Exposure to air pollution may influence immune programming and, together with a genetic predisposition, may cause immunological diseases such as asthma.⁽⁴³⁾ Although knowledge of changes in the immune system of neonates after maternal smoke exposure during pregnancy has existed for some time,^(44–46) data on systemic inflammation in the mother and alteration in the neonatal immune system upon maternal exposure to outdoor air pollution have only recently been published. In one study from Pennsylvania, authors measured blood concentration of C-reactive protein (CRP) in 1,696 women before the 22^nd week of gestation and estimated exposure during the previous 20 days to air pollution from maternal ZIP codes. Among the nonsmokers, the odds ratio to have an increased CRP (>8 ng/ml) was 1.47 (95% CI 1.06; 2.02) for 10 μg/m³ PM₁₀ increase and 1.55 (1.15; 2.11) for 5.2 μg/m³ PM_2.5 increase.⁽⁴⁷⁾ In the BILD study, healthy term-born infants exposed to moderate levels of PM₁₀ showed an attenuated expression of the cytokines interleukin (IL)-10 and IL-1β in cord blood after higher prenatal exposure with different peak effects, depending on stages of pregnancy and cytokines measured.⁽⁴⁸⁾ Lymphocyte proportions were investigated in three studies of one larger project assessing the effect of air pollution on fetal and childhood development in the Czech Republic. The first study found an association between exposure to polycyclic aromatic hydrocarbons (PAH) and PM_2.5 and higher natural killer (NK) cell fractions in cord blood.⁽⁴⁹⁾ Furthermore, they found a decreased percentage of T lymphocytes and an increase of B lymphocytes in cord blood when the exposure occurred during the 14 days before birth.⁽⁵⁰⁾ Finally, they investigated the effect on lymphocyte distribution in cord blood for each month of exposure during pregnancy separately. Using this time period–based approach, they were able to show that exposure during early gestation was associated with increased percentages of T lymphocytes (CD3⁺ and CD4⁺ cells) and reduced percentages of B lymphocytes (CD19⁺) and NK cells, whereas exposure during late gestation was associated with reduced percentages of T lymphocytes and increased percentages of B lymphocytes and NK cells.⁽⁵¹⁾ Using a comparable approach, this group was also able to show an association of PAH exposure during gestational months 4–7 with elevated concentrations of cord-blood IgE, with robust results after adjusting for several confounders.⁽⁵²⁾

Miscellaneous outcomes

The harmful effect of air pollution on the fetus may potentially involve not only lung or immune development, but any organ or system. Evidence of effects on congenital anomalies is limited and controversial. Data from North England did not show any association of black smoke and SO₂ with any congenital malformation,⁽⁵³⁾ whereas a later study from the same area investigated the exposure to CO and NO₂ and found an association with ventricular septal defects and cardiac septal malformations, but no consistent results after exposure to O₃ and PM₁₀.⁽⁵⁴⁾ Another study from the United States found just one significant correlation out of 60 investigated: PM₁₀ and patent ductus arteriosus.⁽⁵⁵⁾ A recent meta-analysis of 10 studies (including two of the above) shows clear evidence of an increased relative risk (between 3% and 20% per unit of exposure) for coarctation of the aorta and tetralogy of Fallot after NO₂ and SO₂ exposure, for atrial septal defects after PM₁₀ exposure, and for cleft lip with or without cleft palate after O₃ exposure.⁽⁵⁶⁾

Evidence of an effect on neurodevelopment and behavior is sparse, but seems quite consistent. Data from a case-control study in California (United States) suggests an association between residential proximity to freeways (<300 m) and cases of autism.⁽⁵⁷⁾ The INMA (Infancia y Medio Ambiente) project from Spain found an association between prenatal exposure to NO₂ and benzene, and cognitive development in the second year of life, strongest in infants who were not breast-fed and whose mothers had a lower fruit and vegetable intake.⁽⁵⁸⁾ This finding supports the hypothesis that the adverse effects of air pollution act through oxidative pathways.

Methodological considerations

To fully appreciate possible associations between air pollution during pregnancy and health outcomes, some important issues regarding exposure assessment need to be considered.

The precision of estimating individual exposure affects the power of the study and, therefore, the detected effect of air pollution on health. The most common approach is to assess the residential exposure at the mother's address, with a lower or higher resolution: district, postal code, or street number. This approach assumes that in the considered area the pollution level is homogeneous, that the subject did not move during pregnancy, and that the mothers stay most of the time at home, or that the exposure at the work place is comparable to that at home. Aguilera et al. collected data on time-activity patterns of pregnant women and analyzed the correlation between air pollution exposure during pregnancy and birth weight.⁽⁵⁹⁾ Limiting the analysis to a subgroup of women, who spent less than 2 hr per day outside home, the association between aromatic hydrocarbons [benzene, toluene, ethylbenzene, xylene (BTEX)] and birth weight changed from –7.6 g (95% CI –54.9; 39.8) to –76.6 g (95% CI –146.3; –7.0) per increase of 4 μg/m³ BTEX, indicating that effects are underestimated due to nondifferential exposure misclassification, and the true, stronger association may only be seen in studies with more exact exposure assessment.⁽⁵⁹⁾

The exposure to coarse particles (PM_10–2.5) and the consequent effect are not necessarily comparable between different areas. Particulate matter is measured as total mass, but despite similar gravimetric values dramatically different particle composition may arise, depending on the source. Particles measured, e.g., at the seaside, are derived mostly from natural sources (sand and ocean droplets), whereas urban airborne particulates derive mainly from vehicle resuspension from the road and abrasion processes from wheels and brakes. These last highly correlate with combustion-derived ultrafine particles that share the same source. Biological effects could also be exhibited through redox activity and content of organic and elemental carbon and PAH, which are all higher in ultrafine particles.⁽⁶⁰⁾ Beside the composition, the surface of all particles also needs to be considered. The smaller the particles, the smaller the mass and the larger the relative surface, which may be relevant for the biological activity. Most of the available epidemiological studies on air pollution measured PM₁₀ or PM_2.5 (e.g., Table 1). Thus, depending on the design of a study or the source of particulate matter in the study area, PM₁₀ levels are or are not a surrogate measure for exposure to ultrafine particles.^(60–62) However, this is difficult to judge just from the publication without additional information.

Another open question is the shape of the association between exposure and effect. Most studies found or assumed a linear correlation between air pollution and birth outcome, although many biological processes are under feedback control and, when related to continuous predictors, may not behave linearly.⁽⁶³⁾ For the INMA study, Ballester et al. performed cubic smoothing splines to describe nonlinear correlations for birth weight and length.⁽⁶⁴⁾ Reduction in birth weight in that study was evident between 35 and 55 μg/m³ NO₂, and in birth length above 40 μg/m³ NO₂.⁽⁶⁴⁾ This kind of analysis can be useful for policy makers in order to evaluate the merit of setting their limits of pollutants. In the above example, reducing NO₂ levels from 60 to 45 μg/m³ would lead to small effects, whereas benefit would derive from reducing NO₂ down to 35 μg/m³. These cutoff levels may, of course, vary between different locations and for different outcomes.

As the development of the fetus is a very dynamic process, the effect of exposure can differ in the different periods, suggesting changing windows of vulnerability. These timing effects are not surprising and have been shown in several of the studies cited above.^{(48,50,51,65)} The susceptibility to air pollution and the effect from increased exposure depend on both the stage of pregnancy and the developmental process assessed. The pathophysiological basis of these timing effects relates to the mechanisms by which air pollutants impact upon the developing fetus. The first period of pregnancy is crucial for the correct establishment of a fully functional vascularity of the placenta. Factors acting during this time period may alter the implantation of trophoblasts and lead to subsequent chronic placental insufficiency.⁽⁶⁶⁾ The last trimester, on the other hand, is characterized by the fastest somatic growth. At the last workshop on “methodologies in epidemiological studies of air pollution and birth outcome,” the identification of these windows of vulnerability was regarded as one of the main challenges for future research.⁽⁶⁷⁾ The high intercorrelation of exposures during those time windows and throughout the whole period of pregnancy, however, makes those timing effects difficult to disentangle.

General considerations

As shown above, effects of different pollutants have been largely investigated in recent years in epidemiological studies. However, the ways by which particles actually are transported beyond the lungs in humans and the exact mechanisms of their effects on fetuses are still unclear to a large extent. Although animal and in vitro studies suggest it is very likely that small amounts of ultrafine particles can be transported to other organs, it is not clear if this translocation of particles, or rather the larger fraction, which deposits in the upper respiratory tract,⁽⁶⁸⁾ is responsible for health effects in humans. Despite this uncertainty, we review current knowledge on toxicokinetics and toxicodynamics of pollutants in the following paragraphs and try to relate it to possible mechanisms in health effects. We give higher weight to ultrafine compound rather than coarse particles, as better data for ultrafine particles are available despite the fact that most epidemiological studies mentioned further above use coarse particles to estimate pollution exposure.

Toxicokinetics of pollutants

The first step to consider is the route of the pollutant from the environment into the body. Possible points of entry are the skin, the gastrointestinal tract, and the respiratory tract. From the anatomical point of view, the small intestine and the lungs have a wider surface (200 m² and 140 m², respectively) than the skin (2 m²), and the lungs have the thinnest barrier (<1 μm at the alveolar segments compared with 20–25 μm of the small intestine and 30–50 μm of the skin).⁽⁶⁹⁾ In the lung, surfactant plays a major role for the clearance of particles and their interaction with airway epithelial cells.⁽⁷⁰⁾ Particles with a diameter>1 μm are usually cleared by mucociliary clearance procedures and do not pass the air–blood barrier.⁽⁷¹⁾ Smaller particles are, however, able to pass through the airway epithelium.⁽⁷²⁾ Those processes depend on the size and the properties of the surface and involve passive diffusion, if the ultrafine particle is uncharged, or caveolae-mediated pinocytosis.⁽⁷³⁾ This endocytosis pathway depends not only on the characteristics of the particles, but also on the cell types and their state of differentiation.⁽⁷⁴⁾ In vivo and in vitro studies have shown that another pathway exists by which particles pass the epithelial barrier, which is mediated by cells of the immune system. Macrophages and dendritic cells are able to take up particulates at the apical side of the epithelium^(75,76) and transport them via the epithelial barrier to the interstitial tissue.⁽⁷⁷⁾ Macrophages and dendritic cells express tight junction proteins and interact with each other by a transepithelial cellular network, allowing the particle to translocate through the epithelium.⁽⁷⁸⁾ Dendritic cells containing the particles are then activated to migrate to the draining lymph nodes and into the bloodstream, where they activate further processes by cellular mediators such as cytokines.^(75,79,80)

The passage of ultrafine particles from the bloodstream through the placenta may depend primarily on the stage of pregnancy. In the first months, the thickness of the placental barrier reaches up to 20 μm, there is no perfusion from maternal blood, and nutrients reach the embryo only by passive diffusion. Subsequently, the fetal capillaries increase along with maternal blood supply in the maternal–fetal interface until the 10^th week of gestation, when the full fetomaternal circulation is completed. As pregnancy progresses, the placental barrier gets thinner, the blood perfusion improves from both sides, and the exchange of nutrients and other elements increases.⁽⁸¹⁾ The membrane of the syncytiotrophoblast has similar properties and selective transporters as the blood–brain barrier, the blood–testis barrier, and the blood–retina barrier.⁽⁸²⁾ Efficacy of those barriers is strongly influenced by inflammatory processes and—best known for the blood–brain barrier—efficacy of the barrier decreases if inflammation occurs. The same mechanism may partly explain enhanced effects of air pollution in diseased subjects, and this could happen at the placental barrier as well.

Very small particles, such as nanoparticles, are well known to cross the placental barrier easily. Wick et al. investigated the placental passage of nanoparticles using an ex vivo human placental perfusion model and fluorescently labeled polystyrene (PS) beads with diameters of 50, 80, 240, and 500 nm.⁽⁸³⁾ They showed that PS beads up to a diameter of 240 nm were able to cross the placental barrier with likely transport routes of passive diffusion and clathrin- or caveolin-mediated endocytosis.⁽⁸³⁾

Toxicodynamics of pollutants

In this section, hypothetical ways of action and interaction of the particle matters within the organism, including the compound's effect on processes at the organ, cellular, and molecular levels, are reviewed. An understanding of the bio-physico-chemical interactions at the air–blood barrier in the mother and the placental barrier helps to explain the heterogeneous effects in different subjects exposed to similar pollutants. The nano–bio interaction depends on (1) the particle surface, (2) the suspending medium or biological fluid, and (3) the solid–liquid interface's contact zone with biological substrate. These three components dynamically interact with each other, i.e., they change their characteristics as they come into contact. In a given medium, some of the main characteristics of a particle relevant for the surface properties include the chemical composition, the shape, the surface crystallinity and roughness, and the lipo- or hydrophilicity. Other characteristics, such as charge, state of aggregation, dispersion, dissolution, biodegradability, hydration, and valence, may change depending on the properties of the suspending media.⁽⁶¹⁾ In biological fluids or interstitium, pH, ionic strength, salts, and the type and concentration of large organic molecules such as proteins depend also on cell secretion, local or systemic inflammation, and microbiological flora. The particles are cleared from macrophages, depending on the type and amount of proteins on the coat, as well as processes such as opsonization, attachment of immunoglobulins and Fcγ, and complement receptor–mediated phagocytosis.⁽⁷⁴⁾ The particles that resist degradation may lead further to inflammation, or stimulate fibroblast-mediated collagen production resulting in fibrosis. The ultrafine matters that escape phagocytosis may interact with other cells not specialized to recognize and process foreign substances.⁽⁶¹⁾

For example, in chronic atopic lung inflammation, the cells secrete cytokines, immunoglobulins, and degradation products that modify the pH of the suspending medium and protein and glycidic contents. The physiological and pathological flora of the lung may both be influenced by and influence the local environment.⁽⁸⁴⁾ The nanoparticles reaching the bronchial wall and alveolar segments experience coating and charge modification according to the surrounding medium, which determines whether they are bioavailable and may participate in biocompatible or bioadverse interactions.

Considering a nonchanging local environment, what are the effects caused by the particles on the biological target? The most common result is the production of reactive oxygen species (ROS) with subsequent oxidant injury and proinflammatory effects. Evidence on mitochondrial damage with impaired ATP production is also quite consistent.⁽⁶⁰⁾ The combustion-derived ultrafine matters contain a particle component that carries the chemical compounds, many of which are semivolatile organic substances, through which the combined toxic effect is enhanced. The particle transports the organic component through the lung epithelium, the particle surface triggers the oxidative reaction, and the chemical compound enhances subcellular damage. Other cytotoxic mechanisms involve disruption of membrane integrity and transport processes, lysosomal damage, protein unfolding, disruption of the conformation, aggregation, and fibrillation (amyloid fiber) as well as DNA damage.^(61,85,86) Epigenetics play a central role in the hypothesis of permanent phenotypical changes due to prenatal environment exposure. The previously described mechanisms may also result in histone modifications, ATP-dependent chromatin remodeling complexes, noncoding RNA, and DNA methylation, all of which affect protein transcription. During the embryonic and fetal period, an enhanced or reduced protein expression may permanently affect the development of immature cells, organs, and systems comparably to genetic mutations.⁽⁴⁾ For example, epigenetic mechanisms regulate the differentiation of naive T cells in T_H1, T_H2, and Treg. These mechanisms may also lead to transgenerational effects of pollution without involving direct DNA changes. A recent review details the evidence on humans, animals, and in vitro, supporting the hypothesis of the effect of air pollutants (PM and diesel exhaust particles) on epigenetic modifications.⁽⁸⁷⁾

The interaction of ultrafine particles with membrane proteins may limit the activation of receptors for placental growth factors, resulting in reduced placental size and subsequent impaired supply of nutrients and oxygen.^(88–92) Chronic exposure to air pollution may thus lead to a subclinical pulmonary and even systemic inflammation due to oxidative injuries⁽⁹³⁾ and innate immunity stimulation.⁽⁹⁴⁾ Increased blood coagulation and viscosity, in addition to endothelial and vascular function, represent further possible mechanisms of ultrafine matters' toxicity. These may also lead to altered placental perfusion.^(92,95–97) The effect on heart-rate variability due to reduction of autonomic system activity⁽⁹⁸⁾ may affect regulation of blood pressure and consequently change hemodynamic properties of the placenta.