Does Exposure to Ambient Air Pollution Affect Gestational Age and Newborn Weight?—A Systematic Review

Current evidence suggests that airborne pollutants have a detrimental effect on fetal growth through the emergence of small for gestational age (SGA) or term low birth weight (TLBW). The study’s objective was to critically evaluate the available literature on the association between environmental pollution and the incidence of SGA or TLBW occurrence. A comprehensive literature search was conducted across Pubmed/MEDLINE, Web of Science, Cochrane Library, EMBASE, and Google Scholar using predefined inclusion and exclusion criteria. The methodology adhered to the PRISMA guidelines. The systematic review protocol was registered in PROSPERO with ID number: CRD42022329624. As a result, 69 selected papers described the influence of environmental pollutants on SGA and TLBW occurrence with an Odds Ratios (ORs) of 1.138 for particulate matter ≤ 10 μm (PM10), 1.338 for particulate matter ≤ 2.5 μm (PM2.5), 1.173 for ozone (O3), 1.287 for sulfur dioxide (SO2), and 1.226 for carbon monoxide (CO). All eight studies analyzed validated that exposure to volatile organic compounds (VOCs) is a risk factor for SGA or TLBW. Pregnant women in the high-risk group of SGA occurrence, i.e., those living in urban areas or close to sources of pollution, are at an increased risk of complications. Understanding the exact exposure time of pregnant women could help improve prenatal care and timely intervention for fetuses with SGA. Nevertheless, the pervasive air pollution underscored in our findings suggests a pressing need for adaptive measures in everyday life to mitigate worldwide environmental pollution.


Introduction
Intra-uterine growth is a crucial indicator reflecting the well-being of the fetus.Therefore, fetal growth abnormalities could arise from various pregnancy-related complications and are directly linked to increased fetal mortality [1].According to the Royal College of Obstetricians and Gynaecologists guidelines, a newborn is considered small for gestational age (SGA) if the birth weight is below the 10th percentile based on customized growth charts [2][3][4].It is estimated that while the majority of hypotrophic infants fall under the The risk of bias was assessed independently by two authors (SF and BG) using the Newcastle-Ottawa scale [21].The third reviewer (JM) resolved any discrepancies in the selection process.Predominantly, the studies included were of moderate to high quality.
Due to the included studies' heterogeneity, it was impossible to perform a quantitative synthesis.Nevertheless, a comprehensive comparison of the included studies is provided in the summary.

Results
The study selection process is comprehensively presented in Figure 1, providing a flow diagram of the evaluation process.Initially, a total of 7932 articles were identified from the database search (Pubmed/MEDLINE = 1778, Web of Science = 2035, Cochrane Library = 138, EMBASE = 674, and Google Scholar = 3307).After removing duplicates, 3064 publications underwent preliminary evaluations based on their titles and abstracts.Just 353 articles were selected for full-text screening, as shown in Figure 1.The study was written according to guidelines, and the PRISMA checklist is published in Supplementary Table S1 [20].
groups differ between the studies.Despite different values used as the cut-off point for inclusion in the exposed group, authors in the studies always compared the exposed group (from the second to the fourth quartile) to the group considered unexposed (values in the first quartile).This distinction between the exposed and unexposed groups based on pollution thresholds allows for comparisons of the impact of air pollution within culturally, geographically, socially, and environmentally similar cohorts.
The risk of bias was assessed independently by two authors (SF and BG) using the Newcastle-Ottawa scale [21].The third reviewer (JM) resolved any discrepancies in the selection process.Predominantly, the studies included were of moderate to high quality.
Due to the included studies' heterogeneity, it was impossible to perform a quantitative synthesis.Nevertheless, a comprehensive comparison of the included studies is provided in the summary.

Results
The study selection process is comprehensively presented in Figure 1, providing a flow diagram of the evaluation process.Initially, a total of 7932 articles were identified from the database search (Pubmed/MEDLINE = 1778, Web of Science = 2035, Cochrane Library = 138, EMBASE = 674, and Google Scholar = 3307).After removing duplicates, 3064 publications underwent preliminary evaluations based on their titles and abstracts.Just 353 articles were selected for full-text screening, as shown in Figure 1.The study was written according to guidelines, and the PRISMA checklist is published in Supplementary Table S1 [20].The systematic review comprised a total of 69 papers .Supplementary Table S2 shows the Newcastle-Ottawa risk bias score of the included studies.
Six studies were conducted in South America: four in Brazil [47,56,59,73], one in Chile [31], and one in Peru [40].Two studies were from the other parts of the world-one from Australia [75] and one from South Africa [24].The selected studies encompassed aggregated data from 20,024,479 pregnant women between 1975 and 2021.The quality of the included studies showed that the majority of the studies were of intermediate to high quality [21].
Proximity to major roads was shown to be a risk factor for SGA and TLBW in four studies [63,69,70,74].
In the reviewed studies, air pollution measurements were based on data from research stations monitoring air quality in the area inhabited by the study cohorts.In all works, exposure assessment consisted of extracting pollution data from national or regional air quality databases.The assessment of individual pollution exposure was determined based on the residence location of a mother relative to the locations of the monitoring sites during a given time window using models such as Distributed Lag Models (DLMs), the General Additive Model (GAM), and the Land Use Regression Model (LUR).This process is called spatial-temporal exposure assessment and is one of the most popular methods used in air quality research.It involves using statistical models to determine the relationship between the level of air pollution and landscape characteristics and land use within a given area.This method can be used to estimate the level of air pollution based on geographic data, such as terrain maps, traffic flow, pollutant emission sources, and other variables.Geocoding the addresses of study participants is also a popular method used in air quality research.It involves assigning geographic coordinates to the addresses of study participants, enabling a comprehensive analysis of the relationship between the level of air pollution and geographical location.
Upon evaluation, the majority of the included studies demonstrated a moderate to high quality, as assessed using the Newcastle-Ottawa Scale [21].All but 12 studies were adjusted for associated variables such as maternal age, BMI, pregnancy, ethnicity, and socioeconomic status.Those without aOR scored lower on the Ottawa-Newcastle scale [43,50,51,63,68,72,74,79,80,82,88,90].The studies were based on retrospective and prospective cohorts compared with adjusted healthy pregnancies, producing high-quality results.Each study in the table has its Newcastle-Ottawa risk bias score (NOS) listed [21].The detailed quality assessment of the included studies is presented in Supplementary Table S2.

PM 2.5 and PM 10 Exposure
There are four critical pollutants that the WHO considers crucial to human health: particulate matter, O 3 , NO x , and SO 2 .Despite the relatively large amount of epidemiological data on the impact of particulate matter, epidemiological data on gaseous pollutants are less abundant, especially regarding nitrogen compounds and sulfur dioxide.
PM 10 and PM 2.5 are atmospheric aerosols smaller than 10 and 2.5 micrometers in diameter, respectively.The toxicity of the pollutants is the result of many factors, including the location of deposition, which is different depending on particle size and reactivity [91].The smaller the particles are, the more they sediment into the lower airways, which allows them to affect the alveolar-capillary barrier directly [13].They trigger cytotoxicity, leading to a local and systemic inflammatory response (via cytokines and mediators) [92].Similar results were shown in a meta-analysis conducted by Liu et al. [93].
PMs trigger pro-inflammatory signals through a Reactive Oxygen Species-dependent mechanism [94].Oxidative stress, characterized by an imbalance between oxidants and antioxidants, can cause cell damage by oxidizing nucleic acids, proteins, and lipids, leading to cell death via apoptosis or necrosis [95].There is much high-quality evidence in the literature from in vivo studies that chronic exposure to PM 2.5 increases serum Interleucin 6 (IL-6), Tumor Necrosis Factor alpha (TNF-α), total cholesterol (TC), and Low-density lipoprotein C (LDL-C) levels, increases the expression of oxidative stress-related genes, causes progression of atherosclerosis, and leads to increased inflammation and redox levels in mice [95].Increasing the antioxidant capacity of exposed cells has been shown to reduce the harmful effects of PM 2.5 and PM 10 [96].
It is speculated that PM 2.5 and smaller particles (<0.1 um), called ultrafine particles (UFP), are able to reach other distant organs via the cardiovascular system [97].The toxic effects of PM 2.5 may be realized directly at the level of the placenta and the developing fetus, which in turn may trigger inflammation and oxidative stress, finally impeding trophoblast invasion, placental vascularisation via anti-angiogenic factors such as the sFlt-1 pathway, and placental dysfunction, a pivotal contributor to SGA [98][99][100].This smaller size may explain the demonstrated significantly higher incidence of SGA for PM 2.5 .
In studies by Fernando Costa Nascimento et al. and Brown et al., a paradoxical protective effect of both PM 2.5 and PM 10 on TLBW was shown [47,53].The possible reason for these negative associations may be the fact that high levels of exposure to air contamination throughout gestation led to miscarriage or stillbirth, which was not included as an outcome in those studies, thereby resulting in a selective survival bias for healthier fetuses.

O 3 Exposure
O 3 has the most evidence linking it to adverse health effects among gaseous pollutants.It is a pollutant formed by chemical reactions between nitrogen oxides (NO x ) and VOCs in the presence of sunlight.In addition to being a highly reactive molecule capable of inducing oxidative stress, O 3 has been shown to stimulate the synthesis of inflammatory cytokines by alveolar macrophages, such as IL-1ß, IL-6, IL-8, and TNF-α [101].Studies conducted by Huang et al. and Shang et al. established that O 3 exposure was linked to increased term birth weight and a higher incidence of macrosomia [30,36], and the study of Nobles et al. showed a protective effect of exposure to O 3 [43].This phenomenon might be explained by the observation by Beckerman et al. that O 3 concentrations increased with proximity to the expressway, possibly due to O 3 being scrubbed by NO to form NO 2 [102].The negative association with O 3 may suggest a low level of exposure to TRAP, which may result in decreased SGA or TLBW occurrence.

Exposure to Traffic-Related Air Pollutants (TRAPs)
As almost every study shows, inconsistent results could result from exposure to multiple air pollutants.For example, in the study of Gan et al., while SO 2 was found to have a significant impact on the prevalence of SGA, its exposure effects were reported in conjunction with other pollutants [28].Evidence from car emissions studies emphasizes that the combined effect of air pollutants should be recognized as a primary risk factor for SGA [63,69,70,74].While the composition of emissions may vary depending on differences in fuel type between gasoline and diesel vehicles [103], the emissions contain all of the pollutants evaluated in our study (PM 2.5 , PM 10 , O 3 , CO, NO x , SO 2 , VOCs, and more) [104,105].Hence, the studies where only the influence of one pollutant was assessed neglect the combined, synergistic effect of other pollutants.It is also important to note that numerous factors potentially influence placental function and increase SGA risk.Only in some studies was the OR adjusted, and this should be considered when interpreting the findings from those studies.

NO x Exposure
NO x contributes to oxidative stress by generating reactive oxygen species that can overpower the placenta's natural antioxidant barriers, causing cellular and molecular harm.NO x exposure can cause inflammation in placental tissues, leading to functional damage and disrupting the exchange of nutrients and oxygen between the mother and fetus [106].The inflammatory environment in the placenta can cause decreases in placental blood flow by constricting blood vessels and impairing endothelial function [107].Furthermore, a discrepancy in the levels of NO and NO 2 within blood vessels might lead to endothelial dysfunction, negatively impacting placental blood flow.The severity of these adverse outcomes depends on the level and duration of NO x exposure.These alterations could lead to SGA and TLBW [26,28,30,38,43,[48][49][50]57,63,64,68,72,74,81,86].
In the study by Mitku et al., the counterintuitive protective effect of exposure to NO 2 was shown.The protective effect may result from simultaneous exposure to other environmental substances with a stronger protective impact, which conceals the negative effects of NO 2 [24].

CO Exposure
The fetotoxic effect of CO is associated with impaired cellular respiratory function.It irreversibly binds to the hemoproteins (cytochrome a-3 and myoglobin) that carry oxygen in the cell, leading to cellular respiration dysfunction.This results in mitochondrial degradation in CNS and heart cells, which require higher energy levels, cellular damage, and ultimately irreversible tissue damage.It also promotes the formation of oxygen-free radicals [109].At the supracellular level, it prevents hemoglobin from delivering oxygen to tissues.The affinity of CO for hemoglobin is stronger in the fetus compared to children and adults.It is important to remember that fetal damage can occur even if the mother's CO levels are not toxic [110], which could result in the appearance of SGA or TLBW [25,32,36,57,59,64,78,80,81,87,88,90].

VOC Exposure
Various potential processes have been proposed, including the impact of VOCs on developing fetuses, its influence on blood viscosity, and its effect on placental perfusion efficiency on the maternal side.Polycyclic aromatic hydrocarbons (PAHs) are believed to have a direct impact on fetal development and DNA transcription [29,111].These air pollutants can impact maternal well-being by affecting the cardiovascular system and causing metabolic alterations.Consequently, there is a reduction in blood supply to the placenta, resulting in a higher occurrence of SGA [29,32,38,45,83,85,86,89]. Nevertheless, there was insufficient data to compare VOC substances in different populations, as each study analyzed a different molecule.Therefore, more studies are needed to compare these substances to other pollutants, such as PM 2.5 , PM 10 , SO 2 , O 3 , or NO x .

Exposure at a Particular Time of Pregnancy
Another important conclusion from the study is the association regarding exposure time.There appear to be specific windows during which the fetus is especially vulnerable or resistant to harmful substances, including air pollutants.Some studies have shown a positive association between SGA and exposure to harmful substances at any time during the pregnancy [25,26,30,37,76].Other studies show a more significant influence in the first trimester, which is vital for organogenesis [26,34,40,57,63,64,72,76]. Nevertheless, it is believed that the influence in the first trimester has a binary effect on the pregnancy, either leading to a miscarriage or not leaving the pregnancy unaffected by any adverse consequences during pregnancy [26], as was shown by Bai et al. and Liu et al. in their meta-analyses assessing pregnancy outcomes including miscarriage [112,113].The influence in the second and third trimesters may better reflect the influence of air pollutants on SGA as organogenesis is almost complete, and the fetus mainly grows during this period [25,31,33,37,43,49,59,64,72,76,81,87,88].Compounds such as PM 10 or PM 2.5 can enter the bloodstream, accumulate in the fetal circulation, and cause oxidative stress.Chronic inflammation during pregnancy prevents the developing fetus from effectively utilizing nutrients to build reserves of adipose tissue, which is most intense in the third trimester.
Another possible cause for SGA after exposure to pollutants in the second and third trimesters could be individually and collaboratively mediated by increases in maternal blood pressure and hemoglobin levels caused by PM 2.5 , PM 10 , or CO.Hence, monitoring and controlling the mother's blood pressure and hemoglobin levels during prenatal care may lower the risk of SGA through gestational exposure to PM 2.5 [33].An additional analysis of fetal growth restriction, especially its early and late variants, could explain the influence of air pollutants during the second and third trimesters of pregnancy.The discrepancies observed in the analyzed studies regarding the timing and effect on the incidence of SGA after pollution exposure might arise due to exposure to different pollutants, varying concentrations, or socioeconomic differences.These factors should be included in further prospective studies and evaluated using multivariate regression models.

Clinical Implementation and Further Research Directions
This study emphasizes the pressing concern of air pollution in a broader context, extending beyond just greenhouse gases responsible for global climate change.Air pollutants, including SO 2 , PM 10 , PM 2.5 , O 3 , and NO x compounds, can negatively impact the maternal body and the developing fetus.Research has shown that being exposed to air pollution can alter placental DNA methylation, which may lead to disturbances in placental trophic function.This, in turn, can impede the delivery of nutrients to the developing fetus and obstruct the elimination of metabolic waste products from the fetal system [54].
Comparing all air pollutants, it seems that PM 2.5 has the most influence on SGA and TLBW occurrence, as the OR of these pollutants seems to be most associated with the evaluated pregnancy complications.Nevertheless, further investigation of separate components of PM should be performed to evaluate the substances that affect the fetuses most.This knowledge is relevant as this compound could be eliminated from our ecosystem or minimized in global production, thus improving maternal and fetal outcomes.
The precise levels of heavy metals, a notable component of PM, have not been extensively examined in previous research.Prospective investigations are needed to measure these different components in particulate matter systematically and to comprehend their distinct impacts on air quality and potential perinatal complications, such as SGA or TLBW occurrence.This detailed method could result in a deeper comprehension of the risks linked to air pollution [114].

Strength and Limitations
To the best of our knowledge, this study is the latest and most comprehensive investigation of the effects of air pollution on the occurrence of SGA or TLBW infants.A particular strength of this review is the wide range of studies included and their number.The included research papers have evaluated the influence of exposure to different air pollutants on SGA or TLBW, with most being of moderate to high quality.The studies included in our analysis were mainly retrospective and did not account for specific molecular quantities in particulate matter, such as heavy metals.This exclusion is significant since typical pollution measurements often do not include such information.The lack of precise data highlights the necessity for a more detailed method of measuring pollution that can accurately quantify individual pollutant concentrations.
Nevertheless, the study has its limitations.One limitation is the absence of a cohort entirely unexposed to air pollution.Unfortunately, such a cohort is impossible to find because pollution-free environments are rare.Therefore, our study compared the effects of pollution between groups experiencing the highest exposure and those with the lowest measured exposure.Such division accurately reflects the living conditions among different populations across continents, considering similar national, geographic, cultural, and social backgrounds.Our outcomes might have been affected by the uneven distribution of the study and control groups, especially in terms of air quality and related factors.The different criteria and possible bias of the various studies may be significant weaknesses.The study group might live in regions with more industrial activity and air pollution; however, they also have better access to advanced healthcare facilities and higher levels of public health awareness.This difference in comparison to the control group, who may live in less polluted areas with restricted access to modern fetal monitoring, could unintentionally impact the reported occurrence of SGA and TLBW because of varying standards of medical care.However, these differences seem to be similar in the included studies.
Another limitation is the inability to monitor individual exposure precisely.The results of the included studies are based on estimating the level of air pollution based on the place of residence during pregnancy.The literature indicates that about one-quarter of pregnant women change their location during pregnancy [67].One significant limitation was that the results were too heterogeneous to perform a meta-analysis.A notable oversight in several studies was that they failed to evaluate the confounding effect of maternal smoking, as smoke also contains several air pollutants that might influence SGA or TLBW rates.Finally, since it was effectively a combination of air pollutants that was assessed and its makeup changes based on location, assessing the influence of each pollutant proved challenging.

Conclusions
Air pollution is a real threat to the health of both the mother and the fetus.It is a global problem that is particularly pronounced in large urban areas.The ongoing industrialization and urbanization of society, especially in developing countries, will lead to the even greater exposure of pregnant women to air pollution, such as PM 2.5 , PM 10 , O 3 , NO 2 , CO, SO 2 , and other less-discovered pollutants.Patients in the high-risk group for SGA living in large cities and residing close to sources of pollution increase their risk of pregnancy complications, postpartum complications, and developmental issues.Relocating to less polluted environments can reduce the risk of SGA and may benefit the mother and the fetus.The nature, magnitude, and timing of pollutant exposure influence pregnancy outcomes.Understanding the detailed mechanisms of the effect of air pollution during pregnancy and identifying the most vulnerable windows during pregnancy require further research.Clarifying the exact exposure-time association will improve SGA prevention, especially in high-risk pregnancies.Using standardized exposure criteria and methods for individual assessments of exposure to air pollutants will improve our understanding of this pressing issue, one of the biggest challenges of our times.
There is a strong need to objectify individual exposure to pollutants, which can be achieved by prospective cohort studies and measurement of personal exposure or blood biomarkers.Systematizing exposures will allow better characterization of the association between air pollution and adverse birth outcomes.
However, it is essential to note that finding a place on Earth unaffected by air pollution is almost impossible.Thus, humans must adapt to escalating environmental pollution or undertake significant steps to stop the pollution of the Earth.

Table 2 .
Characteristics of the included studies on the influence of Particulate Matter ≤ 10 µm (PM 10).

Table 3 .
Characteristics of included studies about the influence of Particulate.Matter ≤ 2.5 µm (PM 2.5).

Table 4 .
Characteristics of the included studies about the influence of nitrogen substances (Nitrogen oxides (NO x ): nitric oxide-NO and nitrogen dioxide-NO 2 ).

Table 5 .
Characteristics of the included studies about the influence of ozone (O 3 ).

Table 6 .
Characteristics of the included studies about the influence of sulfur dioxide (SO 2 ), carbon monoxide (CO), and volatile organic compounds (VOCs).