Abstract
Background
Fine particulate matter (PM2.5) is recognized as a significant risk factor for adverse health effects, however, its association with perinatal complications and neonatal outcomes remains incompletely understood. Elucidation of this relationship is critical for enhancing perinatal healthcare strategies.
Methods
This research employed a retrospective analysis of patient data from the First People’s Hospital of Shangqiu and Henan Children’s Hospital, covering the period from February 1, 2018, to December 30, 2022. The study cohort consisted of 13,828 patients who underwent either vaginal or cesarean delivery, patients were categorized based on their delivery method (vaginal or cesarean) and PM2.5 exposure levels (< 50 µg/m3, 50–100 µg/m3, > 100 µg/m3). The study compared perinatal complications and neonatal outcomes among these groups.
Results
PM2.5 exposure was associated with a significant rise in gestational hypertension across both delivery modes. In the cesarean delivery group, higher PM2.5 exposure was associated with an increased incidence of oligohydramnios (p = 0.006). Furthermore, lower birth weights were consistently observed within groups subject to higher PM2.5 exposure, irrespective of the mode of delivery. Multivariate logistic regression analysis revealed a significant correlation between elevated PM2.5 exposure levels and an increased risk of gestational hypertension (p < 0.001). Linear analysis demonstrated a negative correlation between elevated PM2.5 exposure levels and neonatal birth weight (p < 0.001). The adjusted Generalized Additive Model (GAM) suggested a non-linear pattern, indicating potential thresholds between PM2.5 exposure and the measured outcomes.
Conclusions
The exposure level of PM2.5 is significantly correlated with an increased incidence of gestational hypertension and a decrease in birth weight, it is essential for obstetricians to incorporate air quality considerations into the framework of prenatal care.
Keywords: PM2.5 exposure, Perinatal complications, Neonatal outcomes, Gestational hypertension, Birth weight
Introduction
Ambient air pollution, especially fine particulate matter (PM2.5), has garnered significant attention in recent decades for its potential implications on health [1]. Numerous studies have established a link between PM2.5 exposure and a range of adverse health outcomes, including those affecting respiratory and cardiovascular systems [2, 3]. Particularly, research has demonstrated that the anatomical and physiological alterations experienced by women during pregnancy, along with the heightened vulnerability of the developing fetus to external influences, elevate the risks for both mother and fetus [4–6], thereby prompting increased apprehension regarding the impact of PM2.5 exposure on pregnancy outcomes and infant health.
Previous studies have suggested a potential association between PM2.5 exposure and pregnancy complications. You et al. demonstrated that high PM2.5 exposure during pregnancy may increase the risk of preterm birth (PTB) and gestational diabetes mellitus (GDM). After adjusting for confounding variables, their study found that for every 5 µg/m³ increase in PM2.5 exposure during pregnancy, the risk of preterm birth increased (p = 0.019), and the risk of GDM throughout pregnancy also increased (p = 0.029) [2]. Pan et al. reported that exposure to high levels of PM2.5 and O3 elevates the risk of GDM [7]. However, conflicting findings from Coogan et al. suggested no definitive association between PM2.5 exposure and either GDM or gestational hypertension [8]. These discrepancies may stem from methodological limitations including restricted small sample sizes and single-center study designs. Furthermore, neonatal outcomes in light of exposure to PM2.5 during pregnancy have recently become a point of emphasis. Koo et al. found that ongoing exposure to high PM2.5 concentrations during pregnancy impacts the incidence of neonatal surgical anomalies [9]. Nevertheless, current evidence remains insufficient regarding PM2.5’s association with critical neonatal parameters including PTB, macrosomia, and birth weight, with no conclusive consensus established on pollution-pregnancy complication relationships.
Therefore, this study aims to fill this gap by conducting a retrospective analysis of patient data from two hospitals in China using statistical models such as generalized additive models (GAM), linear regression analysis, and multivariate logistic regression analysis. To exclude the potential confounding effects of delivery modes on the research results, we stratified patients based on the different delivery modes they received, aiming to more accurately clarify the relationship between PM2.5 exposure and perinatal complications as well as neonatal outcomes. This will provide a deeper understanding of the relationship between air pollution and pregnancy outcomes. The insights derived from this study are vital for shaping pregnancy care practices and for developing targeted interventions to lessen the harmful effects of air pollution on maternal and neonatal health.
Methods
Study design and participants
This retrospective cohort study was conducted at the First People’s Hospital of Shangqiu and the Henan Children’s Hospital. The study population comprised pregnant women who delivered between February 1, 2018, and December 30, 2022. Patient data were extracted from hospital medical records, including demographic information, medical history, method of delivery (vaginal or cesarean section), newborn characteristics, and perinatal outcomes. Three exposure categories were established based on recorded PM2.5 levels: <50 µg/m3, 50–100 µg/m3, > 100 µg/m3. The PM2.5 concentration stratification was established by referencing the World Health Organization’s (WHO) multilevel Interim Targets (ITs), current national standards in China, and sample size considerations. The inclusion criteria were: singleton pregnancy, gestational age of at least 28 weeks at the time of delivery, and complete medical records encompassing detailed delivery and postnatal outcomes. Exclusion criteria included a history of chronic diseases such as hypertension, diabetes mellitus, metabolic disorders and pelvic tuberculosis, which could affect pregnancy outcomes, patients with incomplete clinical data.
Data collection
The PM2.5 exposure levels in mothers during pregnancy were estimated from records provided by local environmental monitoring stations. To ensure precision in our analysis, we utilized the widely recognized online coordinate identification system (https://lbs.amap.com/tools/picker?spm), to obtain the geographic coordinates (latitude and longitude) of both air quality monitoring stations and the residential addresses of participants. This platform is known for its high positioning accuracy, with an average precision of 10–50 m in urban settings. Daily average PM2.5 concentrations (µg/m³) were sourced from ground-based monitoring stations overseen by the China National Environmental Monitoring Center (CNEMC). The measurement of PM2.5 concentrations was carried out using a beta-attenuation monitor (BAM-1020). The limit of detection (LOD) and limit of quantification (LOQ) for this instrument were established at 0.8 µg/m³ and 2.4 µg/m³, respectively, in accordance with the specifications outlined in the instrument’s technical manual. These thresholds are in alignment with the Chinese National Ambient Air Quality Standards, ensuring the reliability and compliance of our data. To evaluate the accuracy of our spatial interpolation methods, we employed a leave-one-out cross-validation (LOOCV) technique. This involved iteratively excluding each monitoring station from the analysis and comparing the predicted PM2.5 concentrations against the observed values. The performance of our model was robust, with a root mean square error (RMSE) of 3.9 µg/m³ and a mean R² value of 0.72, indicating a high level of predictive accuracy and reliability. The daily PM2.5 exposure for each patient was calculated using the inverse distance weighting (IDW) method in R software. IDW is a spatial interpolation method used to predict the value of an unknown point based on the weighted average of known values from surrounding sample points, an approach previously validated in epidemiological studies investigating ambient air pollution and health. The study protocol was consistent with the ethical principles of the Declaration of Helsinki and received approval from the Ethics Committee of The First People’s Hospital of Shangqiu (No: SYY20220112). As a retrospective study, the Ethical Committee of the First People’s Hospital of Shangqiu waived the requirement of informed consent.
Outcome measures
The primary outcomes were the incidence of perinatal complications and various neonatal outcomes. Perinatal complications encompassed conditions such as gestational diabetes, gestational hypertension, intrahepatic cholestasis, placenta previa, premature rupture of membranes, polyhydramnios, oligohydramnios, and postpartum hemorrhage. Neonatal outcomes included gestational age, birth weight, rates of preterm delivery, incidence of macrosomia, neonatal Apgar scores ≤ 7 at 1 min and 5 min, and the incidence of birth defects. All outcomes were defined by standard clinical criteria.
Statistical analysis
Statistical analysis was performed with SPSS version 26.0 (IBM, Chicago, USA) and R (version 4.3.2). Patients were stratified into subgroups according to their PM2.5 exposure levels. The Shapiro-Wilk test was employed to assess data normality within these subgroups. Continuous variables were expressed as means (± standard deviation), and categorical variables were presented as frequencies (%). For comparisons of continuous variables among the groups, a one-way ANOVA was used, while comparisons of categorical variables were conducted with Chi-square tests, and Kruskal-Wallis tests for post hoc comparisons. A p-value < 0.05 indicated statistical significance. Logistic and linear regression analyses sought to identify potential associations between PM2.5 exposure levels and the defined outcomes. These models included adjustments for confounders such as delivery mode.
Additionally, a GAM was employed to explore potential non-linear associations and to identify possible cutoff points. GAM is a statistical model that extends the generalized linear model by allowing for non - linear relationships between the response variable and the predictor variables. It represents the relationship between the mean of the response variable and the predictors as a sum of smooth functions of the predictors, providing a flexible framework for modeling complex data patterns. Statistical significance was determined using p < 0.05. For regression analyses, the estimated beta coefficient (β) and 95% confidence intervals (CI) were presented.
Result
The study conducted a retrospective analysis of patient data from the First People’s Hospital of Shangqiu and Henan Children’s Hospital over a period spanning February 1, 2018, to December 30, 2022. A total of 13,828 patients were included in the study, with 7,888 in the vaginal delivery group and 5,940 in the cesarean section group. These groups were further categorized according to PM2.5 exposure concentrations: <50 µg/m3, 50–100 µg/m3, and > 100 µg/m3. In the analysis of women who had vaginal deliveries, baseline characteristics such as age, infertility duration, number of pregnancies, educational level, BMI, and smoking status showed no significant differences across different PM2.5 exposure levels (Table 1).
Table 1.
Comparison of baseline parameters of different concentrations of PM2.5
| Delivery mode | Vaginal delivery (n = 7888) | Cesarean delivery (n = 5940) | ||||||
|---|---|---|---|---|---|---|---|---|
| PM2.5 Group | < 50 µg/m3 (n = 3624)1 | 50–100 µg/m3 (n = 3047)1 | > 100 µg/m3 (n = 1217)1 | p-value2 | < 50 µg/m3 (n = 2672)1 | 50–100 µg/m3 (n = 2326)1 | > 100 µg/m3 (n = 942)1 | p-value2 |
| Age (years) | 0.185 | 0.255 | ||||||
| < 30 | 932 (25.72%) | 769 (25.24%) | 289 (23.75%) | 618 (23.13%) | 576 (24.76%) | 232 (24.63%) | ||
| 30–35 | 1,176 (32.45%) | 940 (30.85%) | 363 (29.83%) | 726 (27.17%) | 609 (26.18%) | 260 (27.60%) | ||
| 36–40 | 761 (21.00%) | 668 (21.92%) | 286 (23.50%) | 667 (24.96%) | 623 (26.78%) | 241 (25.58%) | ||
| > 40 | 755 (20.83%) | 670 (21.99%) | 279 (22.93%) | 661 (24.74%) | 518 (22.27%) | 209 (22.19%) | ||
| Infertility years (years) | 4.00 [2.00, 6.00] | 3.00 [2.00, 6.00] | 4.00 [2.00, 7.00] | 0.076 | 3.00 [2.00, 6.00] | 4.00 [2.00, 6.00] | 3.00 [2.00, 6.00] | 0.437 |
| Number of pregnancy (n) | 0.129 | 0.157 | ||||||
| ≤ 1 | 2,198 (60.65%) | 1,903 (62.45%) | 773 (63.52%) | 1,705 (63.81%) | 1,522 (65.43%) | 584 (62.00%) | ||
| > 2 | 1,426 (39.35%) | 1,144 (37.55%) | 444 (36.48%) | 967 (36.19%) | 804 (34.57%) | 358 (38.00%) | ||
| Education (%) | 0.091 | 0.278 | ||||||
| Hish school and lower | 2,095 (57.81%) | 1,687 (55.37%) | 694 (57.03%) | 1,477 (55.28%) | 1,218 (52.36%) | 504 (53.50%) | ||
| College graduate | 1,363 (37.61%) | 1,226 (40.24%) | 482 (39.61%) | 1,096 (41.02%) | 1,005 (43.21%) | 401 (42.57%) | ||
| Post-graduate | 166 (4.58%) | 134 (4.40%) | 41 (3.37%) | 99 (3.71%) | 103 (4.43%) | 37 (3.93%) | ||
| BMI kg/m2 | 0.719 | 0.313 | ||||||
| < 23.9 | 192 (5.30%) | 182 (5.97%) | 77 (6.33%) | 126 (4.72%) | 111 (4.77%) | 62 (6.58%) | ||
| 23.9–29.9 | 2,184 (60.26%) | 1,846 (60.58%) | 728 (59.82%) | 1,629 (60.97%) | 1,428 (61.39%) | 581 (61.68%) | ||
| 29.9–35 | 1,174 (32.40%) | 954 (31.31%) | 391 (32.13%) | 842 (31.51%) | 727 (31.26%) | 277 (29.41%) | ||
| > 35 | 74 (2.04%) | 65 (2.13%) | 21 (1.73%) | 75 (2.81%) | 60 (2.58%) | 22 (2.34%) | ||
| Smoking (%) | 0.324 | 0.902 | ||||||
| Yes | 346 (9.55%) | 270 (8.86%) | 106 (8.71%) | 302 (11.30%) | 281 (12.08%) | 113 (12.00%) | ||
| No | 3,278 (90.45%) | 2,777 (91.14%) | 1,111 (91.29%) | 2,370 (88.70%) | 2,045 (87.92%) | 829 (88.00%) | ||
1 Data are shown as Median [IQR]; or n (%)
2 Kruskal-Wallis rank sum test; or Pearson’s Chi-squared test
BMI, body mass index; aP < 0.05, vs. PM2.5 < 50 µg/m3; bP < 0.05, vs. PM2.5 50–75 µg/m3;
Among the vaginal delivery cohort, those exposed to higher concentrations of PM2.5 (> 100 µg/m3) had a significantly higher incidence of gestational hypertension compared to those exposed to lower concentrations (< 50 µg/m3) (p = 0.045). A similar trend was observed among patients who underwent cesarean delivery, with those exposed to > 100 µg/m3 of PM2.5 having a significantly higher incidence of gestational hypertension compared to both the < 50 µg/m3 and 50–100 µg/m3 exposure groups (p < 0.001). Additionally, among patients who underwent cesarean delivery, those exposed to higher concentrations of PM2.5 had a significantly higher incidence of oligohydramnios (p = 0.006). No significant differences were observed in other perinatal complications, including gestational diabetes, intrahepatic cholestasis, placenta previa, premature rupture of membranes, polyhydramnios, and postpartum hemorrhage, across the PM2.5 exposure groups in both vaginal and cesarean delivery patients (Table 2).
Table 2.
Comparison of perinatal complications of different concentrations of PM2.5
| Delivery mode | Vaginal delivery (n = 7888) | Cesarean delivery (n = 5940) | ||||||
|---|---|---|---|---|---|---|---|---|
| PM2.5 Group | < 50 µg/m3 (n = 3624)1 | 50–100 µg/m3 (n = 3047)1 |
> 100 µg/m3 (n = 1217)1 |
p-value2 | < 50 µg/m3 (n = 2672)1 |
50–100 µg/m3 (n = 2326)1 |
> 100 µg/m3 (n = 942)1 |
p-value2 |
| Gestational diabetes (%) | 0.415 | 0.400 | ||||||
| Yes | 413 (11.40%) | 375 (12.31%) | 152 (12.49%) | 461 (17.25%) | 414 (17.79%) | 181 (19.21%) | ||
| No | 3211 (88.60%) | 2672 (87.69%) | 1065 (87.51%) | 2211 (82.75%) | 1912 (82.21%) | 761 (80.79%) | ||
| Gestational hypertension (%) | 0.045 | < 0.001 | ||||||
| Yes | 286 (7.89%) | 258 (8.47%) | 119 (9.78%)a | 281 (10.51%) | 262 (11.26%) | 148 (15.71%)ab | ||
| No | 3338 (92.11%) | 2789 (91.53%) | 1098 (90.22%) | 2391 (89.49%) | 2064 (88.74%) | 794 (84.29%) | ||
| Intrahepatic cholestasis (%) | 0.689 | 0.352 | ||||||
| Yes | 262 (7.23%) | 235 (7.71%) | 95 (7.80%) | 272 (10.18%) | 253 (10.88%) | 114 (12.02%) | ||
| No | 3362 (92.77%) | 2812 (92.29%) | 1122 (92.20%) | 2400 (89.82%) | 2073 (89.12%) | 828 (87.98%) | ||
| Placenta Previa (%) | 0.998 | 0.985 | ||||||
| Complete | 5 (0.14%) | 4 (0.13%) | 2 (0.17%) | 23 (0.86%) | 27 (1.16%) | 9 (0.95%) | ||
| Partial | 17 (0.47%) | 16 (0.52%) | 9 (0.74%) | 47 (1.76%) | 51 (2.19%) | 21 (2.23%) | ||
| Marginal | 32 (0.88%) | 30 (0.98%) | 17 (1.40%) | 91 (3.41%) | 93 (4.00%) | 38 (4.03%) | ||
| Premature rupture of membranes (%) | 0.056 | 0.903 | ||||||
| Yes | 314 (8.67%) | 298 (9.78%) | 132 (10.85%) | 276 (10.33%) | 246 (10.57%) | 102 (10.83%) | ||
| No | 3310 (91.33%) | 2749 (90.22%) | 1085 (89.15%) | 2396 (89.67%) | 2080 (89.43%) | 840 (89.17%) | ||
| Polyhydramnios (%) | 0.360 | 0.527 | ||||||
| Yes | 121 (3.34%) | 119 (3.91%) | 49 (4.03%) | 133 (5.00%) | 128 (5.50%) | 55 (5.84%) | ||
| No | 3503 (96.66%) | 2928 (96.09%) | 1168 (95.97%) | 2539 (95.00%) | 2198 (94.50%) | 887 (94.16%) | ||
| Oligohydramnios (%) | 0.366 | 0.006 | ||||||
| Yes | 178 (4.89%) | 172 (5.64%) | 68 (5.58%) | 165 (6.17%) | 157 (6.75%) | 87 (9.23%) | ||
| No | 3446 (95.11%) | 2875 (94.36%) | 1149 (94.42%) | 2507 (93.83%) | 2169 (93.25%) | 855 (90.77%) | ||
| Postpartum hemorrhage (%) | 0.623 | 0.370 | ||||||
| Yes | 132 (3.64%) | 123 (4.04%) | 43 (3.53%) | 146 (5.46%) | 139 (5.97%) | 63 (6.69%) | ||
| No | 3492 (96.36%) | 2924 (95.96%) | 1174 (96.47%) | 2526 (94.54%) | 2187 (94.03%) | 879 (93.31%) | ||
1 Data are shown as percentages (%)
2 Chi-squared test
aP < 0.05, vs. PM2.5 < 50 µg/m3;
bP < 0.05, vs. PM2.5 50–75 µg/m3;
For the vaginal delivery group, significant differences in birth weights were evident (p = 0.004), with lower birth weights noted in the higher PM2.5 exposure cohorts. No significant differences were found in the rates of preterm delivery, Gestational age, macrosomia, 1-minute and 5-minute Apgar scores ≤ 7, or the incidence of birth defects across the PM2.5 exposure groups. In the cesarean delivery group, a highly significant difference was noted in birth weight (p < 0.001), with progressively lower birth weights in higher PM2.5 exposure groups. No statistically significant differences were observed in other Variables (Table 3).
Table 3.
Comparison of neonatal outcomes of different concentrations of PM2.5
| Delivery mode | Vaginal delivery (n = 7888) | Cesarean delivery (n = 5940) | ||||||
|---|---|---|---|---|---|---|---|---|
| PM2.5 Group | < 50 µg/m3 (n = 3624)1 | 50–100 µg/m3 (n = 3047)1 | > 100 µg/m3 (n = 1217)1 | p-value2 | < 50 µg/m3 (n = 2672)1 | 50–100 µg/m3 (n = 2326)1 | > 100 µg/m3 (n = 942)1 | p-value2 |
| Gestational age(week) | 39.00 [37.00, 41.00] | 38.00 [36.00, 40.00] | 38.00 [36.00, 40.00] | 0.359 | 38.00 [37.00, 40.00] | 38.00 [36.00, 39.00] | 38.00 [36.00, 40.00] | 0.108 |
| Birth weight (g) | 3273.73 [2878.63, 3782.74] | 3232.28 [2713.81, 3712.09] | 3020.15 [2662.92, 3568.12] | 0.004 | 3219.71 [2617.29, 3716.53] | 3123.15 [2752.14, 3572.59] | 2936.62 [2573.05, 3404.37] | < 0.001 |
| Preterm delivery rate (%) | 0.394 | 0.372 | ||||||
| Yes | 308 (8.50%) | 271 (8.90%) | 119 (9.78%) | 231 (8.64%) | 202 (8.69%) | 95 (10.09%) | ||
| No | 3316 (91.50%) | 2776 (91.10%) | 1098 (90.22%) | 2441 (91.36%) | 2124 (91.31%) | 847 (89.91%) | ||
| Ncidence of macrosomia (%) | 0.983 | 0.802 | ||||||
| Yes | 212 (5.85%) | 176 (5.78%) | 72 (5.92%) | 163 (6.10%) | 151 (6.49%) | 62 (6.58%) | ||
| No | 3412 (94.15%) | 2871 (94.22%) | 1145 (94.08%) | 2509 (93.90%) | 2175 (93.51%) | 880 (93.42%) | ||
| 1 min Apgar score ≤ 7 (%) | 163 (4.50%) | 143 (4.70%) | 58 (4.76%) | 0.648 | 122 (4.56%) | 117 (5.03%) | 49 (5.20%) | 0.643 |
| 5 min Apgar score ≤ 7 (%) | 65 (1.80%) | 53 (1.74%) | 21 (1.72%) | 0.981 | 43 (1.61%) | 37 (1.59%) | 17 (1.80%) | 0.903 |
| Incidence of birth defects (%) | 0.473 | 0.701 | ||||||
| Yes | 26 (0.72%) | 23 (0.75%) | 13 (1.07%) | 19 (0.71%) | 16 (0.69%) | 9 (0.95%) | ||
| No | 3598 (99.28%) | 3024 (99.25%) | 1204 (98.93%) | 2653 (99.29%) | 2310 (99.31%) | 933 (99.05%) | ||
1 Data are shown as Median [IQR]; or n (%)
2 Kruskal-Wallis rank sum test; or Pearson’s Chi-squared test
aP < 0.05, vs. PM2.5 < 50 µg/m3;
bP < 0.05, vs. PM2.5 50–75 µg/m3;
Regression analysis was conducted to assess the association between PM2.5 exposure and perinatal complications, with the inclusion of delivery mode as a potential confounder. The results indicated a significant association between increased PM2.5 exposure and higher risk of gestational hypertension (OR = 1.294, 95% CI: 1.191–1.405, p < 0.001). There was no statistical significance between PM2.5 exposure and oligohydramnios (OR = 1.066, 95% CI: 0.998–1.138, p = 0.058). A significant negative association was found between PM2.5 exposure and birth weight (β=-0.219, 95% CI: -0.392 to -0.042, p < 0.001), Women undergoing cesarean delivery demonstrated significantly higher prevalence of gestational hypertension compared to those with vaginal delivery (OR = 0.683, 95% CI: 0.569–0.715, p < 0.001). Similarly, the incidence of oligohydramnios was notably elevated in the cesarean cohort versus the vaginal delivery group (OR = 0.610, 95% CI: 0.555–0.671, p < 0.001) (Table 4). The adjusted GAM suggested a non-linear association between PM2.5 exposure and both gestational hypertension and birth weight, indicating potential thresholds or cutoff points may exist among these variables (Fig. 1).
Table 4.
Regression between pm2.5 and perinatal complications and neonatal outcomes
| Item | Gestational hypertension | Oligohydramnios | Birth weight | |||
|---|---|---|---|---|---|---|
| OR (95% CI) | p-value | OR (95% CI) | p-value | β (95% CI) | p-value | |
| Delivery mode | ||||||
| vaginal delivery | Reference | Reference | Reference | |||
| cesarean delivery | 0.683 (0.569 ~ 0.715) | < 0.001 | 0.610 (0.555 ~ 0.671) | < 0.001 | 0.479 (0.221 ~ 0.726) | 0.319 |
| PM2.5 | 1.294 (1.191 ~ 1.405) | < 0.001 | 1.066(0.998 ~ 1.138) | 0.058 | -0.219 (-0.392~ -0.042) | < 0.001 |
CI = confidence interval
Fig. 1.
Association between PM2.5 and gestational hypertension (A), birth weight (B). The adjusted GAM revealed a non-linear association between PM2.5 exposure and gestational hypertension and birth weight. Solid red line represents the smooth curve fit between variables. Blue line represents the 95% of confidence interval from the fit
Discussion
Through a stratified analysis by delivery mode and PM2.5 exposure levels, our findings revealed a significant association between higher PM2.5 exposure and increased incidence of gestational hypertension, oligohydramnios (in cesarean deliveries), and reduced birth weight. These results are partially consistent with prior studies that have reported adverse pregnancy outcomes associated with air pollution exposure [10–12]. However, most previous research has focused on the impact of air pollution on pregnancy outcomes such as pregnancy loss and preterm birth, with limited attention given to the specific effects of PM2.5 on perinatal complications such as gestational hypertension and perinatal outcomes, our study provides clear evidence of a significant association between PM2.5 exposure and these complications. Our study was conducted in China, which is one of the regions with the highest levels of air pollution globally [13]. This finding further supplements the evidence-based medical literature, underscoring the significant impact of air pollution on Perinatal Complications. It can provide evidentiary support for the formulation of public health policies and intervention measures.
The observed association between higher PM2.5 exposure and an increased risk of gestational hypertension is particularly concerning given the potentially serious health implications for both mother and fetus. A previous study on gestational hypertension in developing countries revealed that the incidence rates of gestational hypertension in India, Pakistan, Mozambique, and Nigeria were 10.3%, 9.3%, 10.9%, and 10.2%, respectively [14], which generally align with the incidence rate of gestational hypertension in our study. However, our research found that patients with lower exposure to PM2.5 had a lower incidence of gestational hypertension. The incidence of gestational hypertension was significantly lower in patients who underwent vaginal delivery compared to those undergoing cesarean section. This may be attributed to the fact that patients with gestational hypertension are more likely to undergo cesarean section. It is well - known that gestational hypertension can progress to preeclampsia, a condition that can cause organ damage, preterm birth, and even maternal and fetal mortality [15–17]. The mechanisms by which PM2.5 exposure may heighten the risk of gestational hypertension have not been fully elucidated but are posited to involve systemic inflammation, endothelial dysfunction, and oxidative stress [18]. Studies have shown that patients with gestational hypertension and preeclampsia have higher levels of inflammatory markers compared to normal patients. In another study, researchers found that exposure to PM2.5 during pregnancy was associated with elevated levels of inflammatory markers in maternal blood, and these elevated inflammatory markers may further impact fetal health and development [19]. Yin Jie et al. reported that PM2.5 could activate the inflammatory axis of COX-2/PGES/PGE2 in vascular endothelial cells, promoting cell apoptosis and an inflammatory response [20]. Due to their small size, PM2.5 particles can penetrate deep into the lungs and may even enter the bloodstream. Once in circulation, these particles can trigger systemic inflammation and oxidative stress [21–23]. In animal experiments, PM2.5 was found to increase the expression of inflammatory factors by activating the nuclear factor κB (NF-κB) signaling pathway, thereby exacerbating the inflammatory response in the cardiovascular system. This inflammatory response may be one of the important mechanisms leading to gestational hypertension [24, 25]. Additionally, emerging evidence suggests a potential correlation between systemic inflammatory mediators and gestational depression, proposing that psychological alterations in pregnant women might constitute a secondary pathway influencing blood pressure regulation [26]. This potential association warrants further investigation through targeted clinical studies.
Additionally, a previous study has indicated that the incidence of oligohydramnios varies across different research studies, and it is generally believed to occur in approximately 5% of pregnancies, which is slightly lower than the incidence observed in patients with higher PM2.5 exposure in our study [27]. Our findings revealed a positive correlation between high PM2.5 exposure and the incidence of oligohydramnios in patients undergoing cesarean section. However, in patients with vaginal delivery and in the multivariate logistic regression analysis, no statistically significant association was found between PM2.5 exposure and oligohydramnios. This discrepancy may be attributable to the higher prevalence of underlying comorbidities in patients undergoing cesarean procedures, which could predispose them to amniotic fluid abnormalities. However, this requires further clarification in future studies.
Our study found that exposure to PM2.5 leads to a reduction in neonatal birth weight, which is consistent with previous research [28, 29]. The mechanisms underlying the reduction in birth weight due to PM2.5 exposure are complex and multifaceted. The decrease in birth weight may stem from external environmental disturbances, intrauterine growth restriction, or placental insufficiency [30–32]. Previous studies have demonstrated a close association between exposure to air pollution (including PM2.5, NOx, etc.) and reduced birth weight [33, 34]. Hao et al. reported a 2% increased risk of term low birth weight per 5 µg/m³ increment in PM2.5 exposure during pregnancy, establishing a significant exposure-response relationship [34]. A comprehensive meta-analysis by Sun et al. further corroborated these findings, suggesting that late gestational exposure to PM2.5 constituents may exert particularly detrimental effects on fetal development due to their enhanced toxicokinetic potential during critical growth periods [35]. These studies, including our current results, support the hypothesis that PM2.5 constituents likely mediate fetal growth restriction through maternal hematological pathways. Additionally, research has shown that mothers with low socioeconomic status and low educational levels often face higher reproductive risks, which may result in lower birth weights and higher infant mortality rates [36]. Moreover, low socioeconomic status often implies poorer access to acceptable medical care. However, in this study, all patients were recruited from two fixed hospitals, and there were no statistically significant differences in the educational levels of patients across groups in the baseline data, thus excluding differences in neonatal weight due to variations in medical care and educational levels.
Our study revealed a nonlinear association between PM2.5 exposure and both gestational hypertension and birth weight. These findings suggest potential thresholds beyond which the risk of adverse outcomes disproportionately escalates, carrying significant implications for public health strategies aimed at mitigating air pollution’s impact on pregnancy outcomes. However, given the exclusive recruitment of participants from two geographical regions and inherent limitations of retrospective study design, precise determination of threshold values remains scientifically unestablished. Additional multi-regional prospective investigations are required to accurately delineate these critical exposure levels. Identification of specific PM2.5 concentration thresholds associated with elevated risks would enable targeted protective measures to safeguard pregnant individuals and their fetuses from airborne pollutants.
The strengths of this study include a large sample size, the collection of retrospective data from two hospitals, and a comprehensive assessment of perinatal complications and neonatal outcomes, it also has several limitations. While our analytical models accounted for key covariates including maternal age, educational status, and body mass index, potential residual confounding persists due to unquantified variables. These may include behavioral modifications during pregnancy, maternal dietary patterns, and environmental mitigation strategies such as reduced outdoor activities in high-pollution zones and residential air purification system utilization. Furthermore, dynamic physiological adaptations during gestation, unexplained biological variables, and potential residential relocation during pregnancy could introduce residual confounding. These unaccounted variables may introduce potential bias in our statistical analyses.
In conclusion, our study provides valuable evidence that elevated PM2.5 exposure levels are significantly associated with increased risks of gestational hypertension and reduced birth weight. Therefore, obstetricians should advise high-risk women in polluted regions to limit outdoor activities during peak pollution periods, promote air purifier use at home, and integrate air quality index monitoring into prenatal counseling. Policy-wise, governments must establish pregnancy-specific air quality thresholds, subsidize air purification for vulnerable populations, and enforce PM2.5 standards in maternal healthcare facilities. Future studies could further conduct intervention experiments targeting PM2.5 exposure during pregnancy and investigate the underlying mechanisms, to establish a more comprehensive understanding of the pathophysiological relationships between PM2.5 exposure and perinatal complications and neonatal outcomes.
Acknowledgements
The authors thank the patients who participated in this study and all the physicians and nurses at the Reproductive Medicine Center, the First People’s Hospital of Shangqiu, China, Departments of Neonatology, Children’s Hospital Affiliated to Zhengzhou University, and the Home of researchers for their support in collecting the data.
Author contributions
F.L. and X.L.L. conceived of and designed the experiments. S.G, X.Y.G and P.L. selected and supervised suitable patients. F.L, Y.M.G, W.J.Z and Y.J.K. provided overall supervision. F.L., Y.C. drafted the manuscript. All authors reviewed this manuscript.
Funding
This work was supported by National Natural Science Foundation of China (U1904165), the Key Science and Technology Foundation of Henan Province (LHGJ20240822), the Medical Education Research Project of Henan Province (WJLX2024138), The Science and Technology Development Project of Shangqiu (2024013 and 2024107), Hospital fund of the First People’s Hospital of Shangqiu City (YJ202302001).
Data availability
The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation, all data is available from the corresponding author on reasonable request. The important and representative information are available in the quotations and tables available in the manuscript.
Declarations
Ethics approval and consent to participate
The study protocol adhered to the ethical principles of the Declaration of Helsinki and was approved by the Institutional Review Board of The First People’s Hospital of Shangqiu. As a retrospective study, the informed consent was waived.
Consent for publication
Not applicable.
Competing interests
The authors declare no competing interests.
Footnotes
Publisher’s note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
References
- 1.Skakkebæk NE, Lindahl-Jacobsen R, Levine H, Andersson AM, Jørgensen N, Main KM, Lidegaard Ø, Priskorn L, Holmboe SA, Bräuner EV, et al. Environmental factors in declining human fertility. Nat Reviews Endocrinol. 2022;18(3):139–57. [DOI] [PubMed] [Google Scholar]
- 2.You YA, Park S, Kwon E, Kim YA, Hur YM, Lee GI, Kim SM, Song JM, Kim MS, Kim YJ et al. Maternal PM2.5 exposure is associated with preterm birth and gestational diabetes mellitus, and mitochondrial OXPHOS dysfunction in cord blood. Environmental science and pollution research international 2024. [DOI] [PMC free article] [PubMed]
- 3.Thurston GD. Moving beyond PM2.5 mass to more effectively protect health. Am J Respir Crit Care Med 2023. [DOI] [PMC free article] [PubMed]
- 4.Kohlhepp LM, Hollerich G, Vo L, Hofmann-Kiefer K, Rehm M, Louwen F, Zacharowski K, Weber CF. [Physiological changes during pregnancy]. Anaesthesist. 2018;67(5):383–96. [DOI] [PubMed] [Google Scholar]
- 5.Kapraun DF, Wambaugh JF, Setzer RW, Judson RS. Empirical models for anatomical and physiological changes in a human mother and fetus during pregnancy and gestation. PLoS ONE. 2019;14(5):e0215906. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Zhao L, Li T, Wang H, Fan YM, Xiao Y, Wang X, Wang S, Sun P, Wang P, Jiangcuo Z et al. Association of co-exposure to metal(loid)s during pregnancy with birth outcomes in the Tibetan plateau. Chemosphere 2023:140144. [DOI] [PubMed]
- 7.Pan SC, Huang CC, Chen BY, Chin WS, Guo YL. Risk of type 2 diabetes after diagnosed gestational diabetes is enhanced by exposure to PM2.5. Int J Epidemiol 2023. [DOI] [PubMed]
- 8.Coogan PF, White LF, Yu J, Burnett RT, Seto E, Brook RD, Palmer JR, Rosenberg L, Jerrett M. PM2.5 and diabetes and hypertension incidence in the black women’s health study. Epidemiol (Cambridge Mass). 2016;27(2):202–10. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Koo EJ, Bae JG, Kim EJ, Cho YH. Correlation between exposure to fine particulate matter (PM2.5) during pregnancy and congenital anomalies: its surgical perspectives. J Korean Med Sci. 2021;36(38):e236. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Dai XC, Liang DT, Sun F. [PM2.5 exposure-caused damage to male reproductive function: progress in research]. Zhonghua Nan Ke Xue. 2021;27(4):361–5. [PubMed] [Google Scholar]
- 11.Zhou S, Xi Y, Chen Y, Zhang Z, Wu C, Yan W, Luo A, Wu T, Zhang J, Wu M, et al. Ovarian dysfunction induced by chronic Whole-Body PM2.5 exposure. Small. 2020;16(33):e2000845. [DOI] [PubMed] [Google Scholar]
- 12.Trusz A, Ghazal H, Piekarska K. Seasonal variability of chemical composition and mutagenic effect of organic PM2.5 pollutants collected in the urban area of Wrocław (Poland). Sci Total Environ. 2020;733:138911. [DOI] [PubMed] [Google Scholar]
- 13.Han L, Zhou W, Pickett ST, Li W, Qian Y. Multicontaminant air pollution in Chinese cities. Bull World Health Organ. 2018;96(4):233–e242. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Magee LA, Sharma S, Nathan HL, Adetoro OO, Bellad MB, Goudar S, Macuacua SE, Mallapur A, Qureshi R, Sevene E, et al. The incidence of pregnancy hypertension in India, Pakistan, Mozambique, and Nigeria: A prospective population-level analysis. PLoS Med. 2019;16(4):e1002783. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Grover S, Brandt JS, Reddy UM, Ananth CV. Chronic hypertension, perinatal mortality and the impact of preterm delivery: a population-based study. BJOG. 2022;129(4):572–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Sciatti E, Orabona R. Cardiovascular prevention after hypertensive disorders of pregnancy: do not forget fetal growth restriction! J Am Coll Cardiol. 2021;78(1):91. [DOI] [PubMed] [Google Scholar]
- 17.Tita AT, Szychowski JM, Boggess K, Dugoff L, Sibai B, Lawrence K, Hughes BL, Bell J, Aagaard K, Edwards RK, et al. Treatment for mild chronic hypertension during pregnancy. N Engl J Med. 2022;386(19):1781–92. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Çintesun E, Incesu Çintesun FN, Ezveci H, Akyürek F, Çelik Ç. Systemic inflammatory response markers in preeclampsia. J Lab Physicians. 2018;10(3):316–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Friedman C, Dabelea D, Thomas DSK, Peel JL, Adgate JL, Magzamen S, Martenies SE, Allshouse WB, Starling AP. Exposure to ambient air pollution during pregnancy and inflammatory biomarkers in maternal and umbilical cord blood: the healthy start study. Environ Res. 2021;197:111165. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Yin J, Xia W, Li Y, Guo C, Zhang Y, Huang S, Jia Z, Zhang A. COX-2 mediates PM2.5-induced apoptosis and inflammation in vascular endothelial cells. Am J Translational Res. 2017;9(9):3967–76. [PMC free article] [PubMed] [Google Scholar]
- 21.Chen CC, Wang YR, Liu JS, Chang HY, Guo YL, Chen PC. Burden of cardiovascular disease attributable to long-term exposure to ambient PM2.5 concentration and the cost-benefit analysis for the optimal control level. Sci Total Environ 2023:164767. [DOI] [PubMed]
- 22.Fujitani Y, Furuyama A, Hayashi M, Hagino H, Kajino M. Assessing oxidative stress induction ability and oxidative potential of PM(2.5) in cities in Eastern and Western Japan. Chemosphere. 2023;324:138308. [DOI] [PubMed] [Google Scholar]
- 23.Aguilera J, Konvinse K, Lee A, Maecker H, Prunicki M, Mahalingaiah S, Sampath V, Utz PJ, Yang E, Nadeau KC. Air pollution and pregnancy. Semin Perinatol 2023:151838. [DOI] [PMC free article] [PubMed]
- 24.Chen H, Chen X, Hong X, Liu C, Huang H, Wang Q, Chen S, Chen H, Yang K, Sun Q. Maternal exposure to ambient PM(2.5) exaggerates fetal cardiovascular maldevelopment induced by homocysteine in rats. Environ Toxicol. 2017;32(3):877–89. [DOI] [PubMed] [Google Scholar]
- 25.Agrawal A, Wenger NK. Hypertension during pregnancy. Curr Hypertens Rep. 2020;22(9):64. [DOI] [PubMed] [Google Scholar]
- 26.La Verde M, Luciano M, Fordellone M, Sampogna G, Lettieri D, Palma M, Torella D, Marrapodi MM, Di Vincenzo M, Torella M. Postpartum depression and inflammatory biomarkers of Neutrophil-Lymphocyte ratio, Platelet-Lymphocyte ratio, and Monocyte-Lymphocyte ratio: A prospective observational study. Gynecol Obstet Invest. 2024;89(2):140–9. [DOI] [PubMed] [Google Scholar]
- 27.Madendag Y, Madendag IC, Sahin E, Aydin E, Sahin ME, Acmaz G. How well do the popular ultrasonic techniques estimate amniotic fluid volume and diagnose oligohydramnios, in fact?? Ultrasound Q. 2019;35(1):35–8. [DOI] [PubMed] [Google Scholar]
- 28.Campbell AG, Miranda PY. Breastfeeding trends among very low birth weight, low birth weight, and normal birth weight infants. J Pediatr. 2018;200:71–8. [DOI] [PubMed] [Google Scholar]
- 29.Kim M, Okunowo O, Ades AM, Fuller S, Rintoul NE, Naim MY. Single-Center comparison of outcomes following cardiac surgery in low birth weight and standard birth weight neonates. J Pediatr. 2021;238:161–e167161. [DOI] [PubMed] [Google Scholar]
- 30.Edmond K. Introduction to Evidence for Global Health Care Interventions for Preterm or Low Birth Weight Infants. Pediatrics 2022, 150(Suppl 1). [DOI] [PubMed]
- 31.Rumbajan JM, Yamaguchi Y, Nakabayashi K, Higashimoto K, Yatsuki H, Nishioka K, Matsuoka K, Aoki S, Toda S, Takeda S, et al. The HUS1B promoter is hypomethylated in the placentas of low-birth-weight infants. Gene. 2016;583(2):141–6. [DOI] [PubMed] [Google Scholar]
- 32.Nkwabong E, Kamgnia Nounemi N, Sando Z, Mbu RE, Mbede J. Risk factors and placental histopathological findings of term born low birth weight neonates. Placenta. 2015;36(2):138–41. [DOI] [PubMed] [Google Scholar]
- 33.O’Sharkey K, Xu Y, Cabison J, Rosales M, Yang T, Chavez T, Johnson M, Lerner D, Lurvey N, Corral CMT et al. Effects of In-Utero Personal Exposure to PM2.5 Sources and Components on Birthweight. Research square 2023. [DOI] [PMC free article] [PubMed]
- 34.Hao Y, Strosnider H, Balluz L, Qualters JR. Geographic variation in the association between ambient fine particulate matter (PM2.5) and term low birth weight in the united States. Environ Health Perspect. 2016;124(2):250–5. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35.Sun X, Luo X, Zhao C, Zhang B, Tao J, Yang Z, Ma W, Liu T. The associations between birth weight and exposure to fine particulate matter (PM2.5) and its chemical constituents during pregnancy: A meta-analysis. Environ Pollution (Barking Essex: 1987). 2016;211:38–47. [DOI] [PubMed] [Google Scholar]
- 36.Larrañaga I, Santa-Marina L, Molinuevo A, Álvarez-Pedrerol M, Fernández-Somoano A, Jimenez-Zabala A, Rebagliato M, Rodríguez-Bernal CL, Tardón A, Vrijheid M, et al. Poor mothers, unhealthy children: the transmission of health inequalities in the INMA study, Spain. Eur J Pub Health. 2019;29(3):568–74. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Data Availability Statement
The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation, all data is available from the corresponding author on reasonable request. The important and representative information are available in the quotations and tables available in the manuscript.