Effects of prenatal exposure to ambient air pollutant PM10 on ultrasound-measured fetal growth

Effects of prenatal exposure to ambient air pollutant PM10 on ultrasound-measured fetal growth Abstract Background Limited epidemiological studies have investigated the relationship between prenatal exposure to ambient particulate matter and risk of abnormal fetal growth, and have reached inconclusive results. No study has been conducted in areas with very high air pollution levels. We investigated the hypothesis that exposure to high levels of particulate matter with aerodynamic diameter no larger than 10 µm (PM10) during pregnancy increases the risk of abnormal fetal growth. Methods A birth cohort study was performed in Lanzhou, China, 2010–12, including 8877 pregnant women with 18 583 ultrasound measurements of four fetal growth parameters during pregnancy, including biparietal diameter (BPD), femur length (FL), head circumference (HC) and abdominal circumference (AC). Mixed-effects modelling was used to examine the associations between PM10 exposure and risk of abnormal fetal growth. Results When average PM10 exposure from conception until the ultrasound examination exceeded 150 µg/m3, there were significant increases in standardized FL (β = 0.095, P = 0.0012) and HC (β = 0.090, P = 0.0078) measures. When average PM10 exposure was treated as continuous variable, we found a significant decrease in standardized BPD (β = −0.018, P = 0.0016) as per 10 µg/m3 increase in PM10. After examining the associations by various exposure windows, positive associations between higher levels of PM10 and fetal overgrowth were consistently seen for HC measures. Conclusions Our study suggested that prenatal exposure to high levels of ambient PM10 increased the risk of abnormal fetal growth. The findings from our study have important public health implications and also call for future studies to explore the underlying mechanisms and post-natal consequences of these findings. Fetal growth, PM10, air pollution, ultrasound measures, pregnancy Key Messages No study has been conducted in areas with very high air pollution levels. Exposure to high levels of ambient PM10 increases the risk of abnormal fetal growth. The risk may vary by different fetal growth measurements and fetal growth patterns. Positive associations between higher levels of PM10 and fetal over- growth were consistently seen for head circumference measures by various exposure windows. Introduction A growing body of evidence indicates that fetal growth restriction (FGR), also known as intrauterine growth restriction or fetal under-growth, is an important predictor of neonatal death.1–4 Over the past decade, numerous studies have indicated that ambient air pollution during pregnancy influences fetal growth.5,6 A majority of these previous epidemiological studies have associated maternal air pollution exposure with fetal growth by using various neonatal anthropometric measurements at birth, such as birthweight and length, size for gestational age, head circumference (HC) and abdominal circumference (AC).7–17 However, the late assessment of growth restriction potentially introduces bias,18 and patterns of fetal growth during pregnancy cannot be assessed in these studies.5 In addition, clinical studies have suggested that birthweight as a proxy for fetal growth may not be the best indicator, and the conditions for growth restriction and small size for gestational age are not synonymous.19,20 Furthermore, others suggest that birthweight poorly reflects fetal under-growth during the first two trimesters, because growth impairment and imbalance that may have occurred during early period may be compensated for in the remaining pregnancy period.21 Assessing fetal growth in utero using ultrasound measures could provide more accurate classification of restricted or excessive growth by reducing the time between exposure and outcome assessment.18,22 A small number of studies have investigated the association between exposure to various air pollutants [i.e. nitrogen monoxide, nitrogen dioxide (NO2), sulphur dioxide (SO2), ozone, carbon monoxide and particulate matter with aerodynamic diameter < 10 µm (PM10)] during pregnancy and fetal growth measured by ultrasound.5,22–29 Only three of these have focused on the effects of exposure to PM10 during pregnancy, and these studies reached inconsistent results.5,22,23 All of these studies investigated different exposure time windows. In addition, all of them were conducted in Europe or the USA, where the concentrations of ambient PM10 were lower than in Asia, with mean levels of PM10 ranging from 17.8 µg/m3 to 40.7 µg/m35,22,23 versus 140.8 µg/m3 in the current study (Table 1; and Table A1, available as Supplementary data at IJE online). Table 1 Distributions of selected characteristics of the study population (N = 8877) Characteristics  N  Percent (%)  Maternal age (years)   < 30  5636  63.5   ≥ 30  3239  36.5  Highest education level   < College  3309  37.3   ≥ College  5568  62.7  Family monthly income (RMB per capita)   < 3000  5111  57.6   ≥ 3000  3766  42.4  Employment during pregnancy   No  4101  46.2   Yes  4776  53.8  Smoking (active and passive) during pregnancy   No  7189  81.0   Yes  1688  19.0  Alcohol consumption during pregnancy   No  8857  99.8   Yes  20  0.2  Birthweight (g)   < 2500  438  4.9   2500–4000  7806  87.9   ≥ 4000  633  7.1  Pre-pregnancy BMIa   ≤ 18.5  1822  20.5   18.5–24.0  6118  68.9   ≥ 24.0  937  10.6  Weight gain during pregnancy (kg)   ≤ 15.0  3247  36.6   15–18.5  2089  23.5   > 18.5  3248  36.6  Parity   Primiparous  6669  75.1   Multiparous  2208  24.9  Caesarean section   No  5599  63.1   Yes  3278  36.9  Preeclampsia   No  8673  97.7   Yes  204  2.3  Maternal diabetes   No  8774  98.8   Yes  103  1.2  Season of conceptionb   Autumn  2431  27.4   Winter  2025  22.8   Spring  1854  20.9   Summer  2567  28.9  Cooking fuel   Gas or electricity  7635  86.0   Biomass or coal  1242  14.0  Newborn gender   Boys  4656  52.5   Girls  4221  47.5    Continuous variable  Mean (SD)  Range    Gestational age (weeks)  38.7 (1.7)  22–44  Averaged daily PM10 during pregnancy (µg/m3)  140.8 (24.8)  66.5–242.6  Averaged daily SO2 during pregnancy (µg/m3)  49.0 (13.1)  14.7–87.1  Averaged daily NO2 during pregnancy (µg/m3)  43.9 (6.0)  26.0–64.6  Averaged daily temperature during pregnancy (°C)  10.3 (3.1)  1.7–20.9  Birth AC (mm)  324.0 (25.2)  200–530  Birth HC (mm)  343.6 (19.0)  250–510  Characteristics  N  Percent (%)  Maternal age (years)   < 30  5636  63.5   ≥ 30  3239  36.5  Highest education level   < College  3309  37.3   ≥ College  5568  62.7  Family monthly income (RMB per capita)   < 3000  5111  57.6   ≥ 3000  3766  42.4  Employment during pregnancy   No  4101  46.2   Yes  4776  53.8  Smoking (active and passive) during pregnancy   No  7189  81.0   Yes  1688  19.0  Alcohol consumption during pregnancy   No  8857  99.8   Yes  20  0.2  Birthweight (g)   < 2500  438  4.9   2500–4000  7806  87.9   ≥ 4000  633  7.1  Pre-pregnancy BMIa   ≤ 18.5  1822  20.5   18.5–24.0  6118  68.9   ≥ 24.0  937  10.6  Weight gain during pregnancy (kg)   ≤ 15.0  3247  36.6   15–18.5  2089  23.5   > 18.5  3248  36.6  Parity   Primiparous  6669  75.1   Multiparous  2208  24.9  Caesarean section   No  5599  63.1   Yes  3278  36.9  Preeclampsia   No  8673  97.7   Yes  204  2.3  Maternal diabetes   No  8774  98.8   Yes  103  1.2  Season of conceptionb   Autumn  2431  27.4   Winter  2025  22.8   Spring  1854  20.9   Summer  2567  28.9  Cooking fuel   Gas or electricity  7635  86.0   Biomass or coal  1242  14.0  Newborn gender   Boys  4656  52.5   Girls  4221  47.5    Continuous variable  Mean (SD)  Range    Gestational age (weeks)  38.7 (1.7)  22–44  Averaged daily PM10 during pregnancy (µg/m3)  140.8 (24.8)  66.5–242.6  Averaged daily SO2 during pregnancy (µg/m3)  49.0 (13.1)  14.7–87.1  Averaged daily NO2 during pregnancy (µg/m3)  43.9 (6.0)  26.0–64.6  Averaged daily temperature during pregnancy (°C)  10.3 (3.1)  1.7–20.9  Birth AC (mm)  324.0 (25.2)  200–530  Birth HC (mm)  343.6 (19.0)  250–510  a BMI, body mass index. b Autumn: September, October and November; winter: December, January and February; spring: March, April and May; summer: June, July, and August. Table 1 Distributions of selected characteristics of the study population (N = 8877) Characteristics  N  Percent (%)  Maternal age (years)   < 30  5636  63.5   ≥ 30  3239  36.5  Highest education level   < College  3309  37.3   ≥ College  5568  62.7  Family monthly income (RMB per capita)   < 3000  5111  57.6   ≥ 3000  3766  42.4  Employment during pregnancy   No  4101  46.2   Yes  4776  53.8  Smoking (active and passive) during pregnancy   No  7189  81.0   Yes  1688  19.0  Alcohol consumption during pregnancy   No  8857  99.8   Yes  20  0.2  Birthweight (g)   < 2500  438  4.9   2500–4000  7806  87.9   ≥ 4000  633  7.1  Pre-pregnancy BMIa   ≤ 18.5  1822  20.5   18.5–24.0  6118  68.9   ≥ 24.0  937  10.6  Weight gain during pregnancy (kg)   ≤ 15.0  3247  36.6   15–18.5  2089  23.5   > 18.5  3248  36.6  Parity   Primiparous  6669  75.1   Multiparous  2208  24.9  Caesarean section   No  5599  63.1   Yes  3278  36.9  Preeclampsia   No  8673  97.7   Yes  204  2.3  Maternal diabetes   No  8774  98.8   Yes  103  1.2  Season of conceptionb   Autumn  2431  27.4   Winter  2025  22.8   Spring  1854  20.9   Summer  2567  28.9  Cooking fuel   Gas or electricity  7635  86.0   Biomass or coal  1242  14.0  Newborn gender   Boys  4656  52.5   Girls  4221  47.5    Continuous variable  Mean (SD)  Range    Gestational age (weeks)  38.7 (1.7)  22–44  Averaged daily PM10 during pregnancy (µg/m3)  140.8 (24.8)  66.5–242.6  Averaged daily SO2 during pregnancy (µg/m3)  49.0 (13.1)  14.7–87.1  Averaged daily NO2 during pregnancy (µg/m3)  43.9 (6.0)  26.0–64.6  Averaged daily temperature during pregnancy (°C)  10.3 (3.1)  1.7–20.9  Birth AC (mm)  324.0 (25.2)  200–530  Birth HC (mm)  343.6 (19.0)  250–510  Characteristics  N  Percent (%)  Maternal age (years)   < 30  5636  63.5   ≥ 30  3239  36.5  Highest education level   < College  3309  37.3   ≥ College  5568  62.7  Family monthly income (RMB per capita)   < 3000  5111  57.6   ≥ 3000  3766  42.4  Employment during pregnancy   No  4101  46.2   Yes  4776  53.8  Smoking (active and passive) during pregnancy   No  7189  81.0   Yes  1688  19.0  Alcohol consumption during pregnancy   No  8857  99.8   Yes  20  0.2  Birthweight (g)   < 2500  438  4.9   2500–4000  7806  87.9   ≥ 4000  633  7.1  Pre-pregnancy BMIa   ≤ 18.5  1822  20.5   18.5–24.0  6118  68.9   ≥ 24.0  937  10.6  Weight gain during pregnancy (kg)   ≤ 15.0  3247  36.6   15–18.5  2089  23.5   > 18.5  3248  36.6  Parity   Primiparous  6669  75.1   Multiparous  2208  24.9  Caesarean section   No  5599  63.1   Yes  3278  36.9  Preeclampsia   No  8673  97.7   Yes  204  2.3  Maternal diabetes   No  8774  98.8   Yes  103  1.2  Season of conceptionb   Autumn  2431  27.4   Winter  2025  22.8   Spring  1854  20.9   Summer  2567  28.9  Cooking fuel   Gas or electricity  7635  86.0   Biomass or coal  1242  14.0  Newborn gender   Boys  4656  52.5   Girls  4221  47.5    Continuous variable  Mean (SD)  Range    Gestational age (weeks)  38.7 (1.7)  22–44  Averaged daily PM10 during pregnancy (µg/m3)  140.8 (24.8)  66.5–242.6  Averaged daily SO2 during pregnancy (µg/m3)  49.0 (13.1)  14.7–87.1  Averaged daily NO2 during pregnancy (µg/m3)  43.9 (6.0)  26.0–64.6  Averaged daily temperature during pregnancy (°C)  10.3 (3.1)  1.7–20.9  Birth AC (mm)  324.0 (25.2)  200–530  Birth HC (mm)  343.6 (19.0)  250–510  a BMI, body mass index. b Autumn: September, October and November; winter: December, January and February; spring: March, April and May; summer: June, July, and August. Abnormal fetal growth includes both under-growth and over-growth. There is accumulating evidence that macrosomia, or being large-for-gestational age, as a result of excessive and unbalanced intrauterine growth places the individual at increased risk for metabolic short-term complications and can influence metabolic disorders in adulthood.30,31 It has been suggested that air pollution causes weight gain and obesity through oxidative stress and inflammatory response.32–34 However, to the best of our knowledge, no study has explored the effect of air pollution on fetal over-growth. In light of the limited studies investigating prenatal exposure to PM10 and risk of fetal under-growth, and the lack of studies investigating the association between air pollution and fetal over-growth, we analysed data from the Lanzhou Birth Cohort Study to investigate the hypothesis that exposure to high levels of PM10 during pregnancy increases the risk of abnormal fetal growth. Methods Study population The study population has been described previously.35–39 In brief, pregnant women who came to the Gansu Provincial Maternity and Child Care Hospital (GPMCCH) in Lanzhou, China, for delivery in 2010–12, and who were 18 years or older with a gestational age of ≥ 20 weeks, were eligible. Individuals with mental illness would have limited ability to answer questions. During the study period, a total of 13 pregnant women had a diagnosis of mental illness (including autism, bipolar disorder, depression, obsessive-compulsive disorder) and were excluded from the beginning of the study recruitment. A total of 10 542 (73.4%) women participated in the study. All study procedures were approved by the Human Investigation Committees at the GPMCCH and Yale University. After obtaining written consent, an in-person interview was conducted at the hospital by trained study interviewers using a standardized and structured questionnaire to collect information on demographics, reproductive and medical history, lifestyle factors, occupation and residential history. Information on birth outcomes and maternal complications were abstracted from the medical records. Gestational age was estimated using the date of the last menstrual period reported at recruitment and confirmed by the first ultrasound examination.39 Fetal growth characteristics The repeated ultrasound measurements of HC, AC, femur length (FL) and biparietal diameter (BPD) during pregnancy were abstracted from the medical records. All readings were recorded in millimetres. Since an ultrasound examination was not routinely performed during each of the three trimesters and not all examination records contained four fetal growth parameters, we restricted our analyses to pregnant women who had at least one ultrasound measurement. Of 8969 subjects who had singleton live births and lived in Lanzhou City,36 we excluded participants: who had no records of ultrasound measures; whose records were considered as unreasonable [values below or above 3 standard deviations (SD) from the normal ranges of fetal growth standards, N = 62, 23, 41 and 33 for HC, AC, FL and BPD, respectively];40 or who had babies with birthweights < 500 g (N = 38) or gestational age ≥ 309 days (N = 1); which yielded a final sample size of 8877. Similar distributions of maternal characteristics were observed between pregnant women with and without ultrasound measurement information (data not shown). Across different times during pregnancy, study participants had up to three ultrasound measurements available for FL and BPD, and up to four ultrasound measurements available for HC and AC, because HC and AC were also measured by ultrasound just before birth. The time of ultrasound examination for each fetal growth parameter was presented as week of gestation and was commonly performed after the first trimester (see Figure A1, available as Supplementary data at IJE online). We used the SGPLOT procedure to generate the smooth nonparamatric curves; the fetal growth patterns of BPD, FL and HC during pregnancy against the ultrasound examination time showed a nonlinear growth pattern (see Figure A2, available as Supplementary data at IJE online), which was consistent with the standard fetal growth curves worldwide.40 To avoid results that cannot be compared for each gestational week, the standardized fetal parameters (Z scores) were used as an outcome variable based on recently updated international standards for all four fetal growth measurements.40 The Z scores were calculated through the formula (actual value-mean)SD, where actual value was the observed ultrasound measures, and mean and SD were the standard values in the same weeks of gestation provided by Papageorghiou et al.40 Exposure assessment As described in previous studies,36–38 data on ambient air pollutants were obtained from the Gansu Provincial Environmental Monitoring Central Station, which collects 24-h average concentration for PM10, sulphur dioxide (SO2) and nitrogen dioxide (NO2) through an automated data reporting system from four monitoring stations in Lanzhou city (1088 km²). The 24-h average PM10 was measured for the period 1 April 2009 to 31 December 2012 for two stations, and for 1 January 2011 to 31 December 2012 for the two additional stations. The monitors were located in the southern part of Lanzhou in the metropolitan area with high population density (see Figure A3, available as Supplementary data at IJE online), the spread/density of the study population is similar to that of the total population (see Figure A4, available as Supplementary data at IJE online). Each subject’s residences throughout pregnancy, as measured by move-in and move-out dates, and work addresses were collected. We used the earth online sharing website provided by Google (www.earthol.com) to obtain longitude and latitude coordinates for each subject's home and work addresses. The mean and standard deviation of distance from maternal residence to each air monitor was 14.8(±10.4) km, 9.6(±10.7) km, 9.6(±12.1) km, and 9.4 (±12.4) km, respectively. The majority (93%) of the participants lived within 12 km from the nearest monitor. We calculated daily PM10 concentrations incorporating each subject’s home and work addresses, using all four monitors, or two monitors with the inverse distance weighting (IDW) approach or the nearest monitor. Subjects’ individual exposure levels were averaged: (i) from date of conception to the date of ultrasound examinations; (ii) 1 month before ultrasound examinations; (iii) 2 weeks before ultrasound examinations; and (iv) 1 week before ultrasound examinations. PM10 exposure levels for different exposure windows, estimated based on an average of the four monitors, two monitors or the nearest monitor, were similar. Therefore, we presented our results based on the four monitors. We calculated the correlations between average PM10 levels during the different exposure windows, and found that the PM10 levels averaged during exposure windows (ii), (iii) and (iv) were highly correlated (Pearson correlation coefficients ranging from 0.80 to 0.84), whereas the correlation between (i) and (ii) or (iii) or (iv) only ranged from 0.08 to 0.17. It suggested a temporal heterogeneity between long-term and short-term exposure windows. Statistical analysis All analyses were performed separately for each fetal parameter. The standard cutoffs were used to categorize under-growth (< 3rd centile, Z < −1.88), normal growth (3rd centile–97th centile, −1.88 ≤ Z ≤ 1.88) and over-growth (> 97th centile, Z > 1.88) (dotted horizontal lines in Figure 1). Similar results were obtained using other centiles for setting the normal range, which were the 5th and 95th centiles (Z = ±1.645), and the 10th and 90th centiles (Z = ±1.28) (results not shown). Thus, we presented more conservative results using the 3rd and 97th centiles as the boundaries of the normal range for the four fetal growth measurements. The subject-specific effect due to repeated measures across time during pregnancy was considered using mixed-effects models. Because ultrasound measures were recorded irregularly for each subject (one to four scans), the time intervals between measurements would be more or less unique to each subject. Under these circumstances, a continuous-time model is needed to describe the covariance among the errors; PROC MIXED with the spatial power covariance function (repeated/type = SP(POW) (gestational weeks at ultrasound)) was used for this purpose.41,42 The linear mixed model was fitted to calculate the mean change and the 95% confidence interval (CI) of the Z score of fetal growth measurements which was associated with maternal exposure to PM10. Then the generalized linear mixed model (mixed effects logistic regression) was fitted to calculate the odds ratio (OR) and 95% CI of the associations between maternal exposure to PM10 and risks of under-growth and over-growth. Normal growth was treated as the reference group for each of the analyses. Exposures of each subject to NO2 and SO2 were estimated in the same manner as PM10, and the average levels of SO2 (49.0 µg/m3) and NO2 (43.9 µg/m3) during pregnancy in our study area were lower than the China NAAQS standard (80 µg/m3 and 50 µg/m3),43 indicating these two pollutants were not the main components of air pollution in this city. Therefore, models were corrected for SO2, NO2 and daily temperature by using the averages as the continuous variables over the same exposure windows as for PM10. Exposure to PM10 was analysed as a continuous variable and, in a separate model, as a binary variable meeting or exceeding the US 24-h, health-based National Ambient Air Quality Standard (NAAQS, 150 µg/m3, equivalent to the China NAAQS Grade II level).43,44 Figure 1 View largeDownload slide The Scatter Plots of Associations Between Standardized Fetal Measurements (Z Scores) and PM10 Exposure Levels (μg/m3) During Pregnancy. Horizontal refernce lines at 3rd, 50th, and 97th centiles showed the cutoffs of undergrowth, normal growth, and overgrowth for fetal biparietal diameter (A), fetal femur length (B), fetal abdominal circumference (C), and fetal head circumference (D) measured by ultrasound according to gestational age. Dark black lines are the smooth nonparametric curves. Opens gray dots showed actual observations. Figure 1 View largeDownload slide The Scatter Plots of Associations Between Standardized Fetal Measurements (Z Scores) and PM10 Exposure Levels (μg/m3) During Pregnancy. Horizontal refernce lines at 3rd, 50th, and 97th centiles showed the cutoffs of undergrowth, normal growth, and overgrowth for fetal biparietal diameter (A), fetal femur length (B), fetal abdominal circumference (C), and fetal head circumference (D) measured by ultrasound according to gestational age. Dark black lines are the smooth nonparametric curves. Opens gray dots showed actual observations. All models were first adjusted for known determinants of fetal growth characteristics as shown in Table 1. For the final model selection, the Akaike information criterion was used with forward-selection procedure on covariates considered to be determinants of growth.45 We checked for the multi-collinearity of all variables in the model, especially for the effects of multiple pollutants including SO2 and NO2 exposures, as well as the daily temperature. The variation inflation factors varied from 1.02 to 2.57, demonstrating a lack of multi-collinearity. The same exposure windows were used for SO2, NO2 and temperature as for PM10 exposure in corresponding models. All analyses were performed using SAS software, version 9.4 (SAS Institute Inc., Cary, NC). Results There were 8770, 8537, 7733 and 8009 participants with ultrasound measurements of HC, AC, FL and BPD, respectively. The study included a total of 14 560, 18 583, 12 383 and 13 619 scans during the entire pregnancy for ultrasound measures of HC, AC, FL and BPD, respectively. According to the international standard fetal growth measurement cutoffs set at the 3rd and 97th centiles, for the standardized fetal growth measurements (Z scores): 515 HC (3.5%), 1316 AC (7.1%), 208 FL (1.7%) and 557 BPD (4.1%) scans were catergorized as under-growth; 11 012 HC (75.6%), 17 007 AC (91.5%), 9395 FL (75.9%) and 12 266 BPD (90.1%) scans were catergorized as normal growth; and 3033 HC (20.8%), 1576 AC (8.4%), 2779 FL (22.4%) and 796 BPD (5.8%) scans were catergorized as over-growth. Table 1 shows the distributions of selected characteristics of the study population. More than half of the women were younger than 30 years (63.5%), nulliparous (75.1%) and non-smokers during pregnancy (81.0%), and had higher education levels (62.7%) and normal pre-pregnancy body mass index (BMI) (68.9%). Table 2 presents the associations between PM10 exposures averaged by different exposure windows, and ultrasound measurement Z-scores. When average PM10 exposure from conception until the ultrasound examination exceeded 150 µg/m3, there were significant increases in standardized FL (β = 0.095, P = 0.0012) and HC (β = 0.090, P = 0.0078) measures, compared with PM10 < 150 µg/m3. When we treated the PM10 exposure as continuous variable, we found a significant decrease in BPD (β = −0.018, P = 0.0016) as per 10 µg/m3 increase in PM10. Table 2 Association between averaged PM10 exposure during the different exposure windows and the mean change in standardized fetal ultrasound measurements (Z scores) Exposure windows  BPD (n = 8009)   FL (n = 7733)   AC (n = 8537)   HC (n = 8770)   βa  95% CI  βa  95% CI  βa  95% CI  βa  95% CI  Mean change comparing exposures ≥ 150 µg/m3 PM10 with those < 150 μg/m3  From conception to ultrasound  0.016  −0.035, 0.067  0.095**  0.038, 0.153  0.001  −0.046, 0.049  0.090**  0.024, 0.156  One month before ultrasound  −0.052  −0.097, 0.007  −0.007  −0.051, 0.038  −0.044  −0.082, 0.003  0.092*  0.020, 0.151  Two weeks before ultrasound  −0.037  −0.083, 0.008  −0.043  −0.089, 0.002  −0.037  −0.076, 0.002  0.020  −0.040, 0.080  One week before ultrasound  −0.034  −0.079, 0.011  −0.005  −0.050, 0.040  -0.025  −0.064, 0.014  0.070*  0.010, 0.130  Mean change per 10 µg/m3 increase in PM10  From conception to ultrasound  −0.018**  −0.030, -0.007  0.008  −0.006, 0.022  0.005  −0.016, 0.006  0.011  −0.006, 0.029  One month before ultrasound  −0.004  −0.008, 0.000  −0.001  −0.005, 0.003  0.002  −0.007, 0.004  0.005*  0.000, 0.009  Two weeks before ultrasound  −0.002  −0.006, 0.002  0.002  −0.006, 0.003  0.002  −0.006, 0.001  0.008*  0.001, 0.011  One week before ultrasound  −0.002  −0.006, 0.002  0.000  −0.004, 0.004  −0.002  −0.005, 0.002  0.004*  0.000, 0.009  Exposure windows  BPD (n = 8009)   FL (n = 7733)   AC (n = 8537)   HC (n = 8770)   βa  95% CI  βa  95% CI  βa  95% CI  βa  95% CI  Mean change comparing exposures ≥ 150 µg/m3 PM10 with those < 150 μg/m3  From conception to ultrasound  0.016  −0.035, 0.067  0.095**  0.038, 0.153  0.001  −0.046, 0.049  0.090**  0.024, 0.156  One month before ultrasound  −0.052  −0.097, 0.007  −0.007  −0.051, 0.038  −0.044  −0.082, 0.003  0.092*  0.020, 0.151  Two weeks before ultrasound  −0.037  −0.083, 0.008  −0.043  −0.089, 0.002  −0.037  −0.076, 0.002  0.020  −0.040, 0.080  One week before ultrasound  −0.034  −0.079, 0.011  −0.005  −0.050, 0.040  -0.025  −0.064, 0.014  0.070*  0.010, 0.130  Mean change per 10 µg/m3 increase in PM10  From conception to ultrasound  −0.018**  −0.030, -0.007  0.008  −0.006, 0.022  0.005  −0.016, 0.006  0.011  −0.006, 0.029  One month before ultrasound  −0.004  −0.008, 0.000  −0.001  −0.005, 0.003  0.002  −0.007, 0.004  0.005*  0.000, 0.009  Two weeks before ultrasound  −0.002  −0.006, 0.002  0.002  −0.006, 0.003  0.002  −0.006, 0.001  0.008*  0.001, 0.011  One week before ultrasound  −0.002  −0.006, 0.002  0.000  −0.004, 0.004  −0.002  −0.005, 0.002  0.004*  0.000, 0.009  a Adjusted for maternal age (< 30, ≥ 30 years), education (< college, ≥ college), family monthly income per capita (< 3000, ≥ 3000 RMB), smoking during pregnancy (yes, no), pre-pregnancy BMI (≤ 18.5, 18.5–24.0, ≥ 24.0), weight gain du.ring pregnancy (≤ 15 kg, 15–18.5 kg, > 18.5 kg), parity (primiparous, multiparous), season of conception (spring/summer, autumn/winter), cooking fuel (gas/electricity, coal/biomass), estimated SO2, NO2 by different corresponding exposure windows, and averaged temperature during different corresponding exposure windows. * P < 0.05; ** P < 0.01. Table 2 Association between averaged PM10 exposure during the different exposure windows and the mean change in standardized fetal ultrasound measurements (Z scores) Exposure windows  BPD (n = 8009)   FL (n = 7733)   AC (n = 8537)   HC (n = 8770)   βa  95% CI  βa  95% CI  βa  95% CI  βa  95% CI  Mean change comparing exposures ≥ 150 µg/m3 PM10 with those < 150 μg/m3  From conception to ultrasound  0.016  −0.035, 0.067  0.095**  0.038, 0.153  0.001  −0.046, 0.049  0.090**  0.024, 0.156  One month before ultrasound  −0.052  −0.097, 0.007  −0.007  −0.051, 0.038  −0.044  −0.082, 0.003  0.092*  0.020, 0.151  Two weeks before ultrasound  −0.037  −0.083, 0.008  −0.043  −0.089, 0.002  −0.037  −0.076, 0.002  0.020  −0.040, 0.080  One week before ultrasound  −0.034  −0.079, 0.011  −0.005  −0.050, 0.040  -0.025  −0.064, 0.014  0.070*  0.010, 0.130  Mean change per 10 µg/m3 increase in PM10  From conception to ultrasound  −0.018**  −0.030, -0.007  0.008  −0.006, 0.022  0.005  −0.016, 0.006  0.011  −0.006, 0.029  One month before ultrasound  −0.004  −0.008, 0.000  −0.001  −0.005, 0.003  0.002  −0.007, 0.004  0.005*  0.000, 0.009  Two weeks before ultrasound  −0.002  −0.006, 0.002  0.002  −0.006, 0.003  0.002  −0.006, 0.001  0.008*  0.001, 0.011  One week before ultrasound  −0.002  −0.006, 0.002  0.000  −0.004, 0.004  −0.002  −0.005, 0.002  0.004*  0.000, 0.009  Exposure windows  BPD (n = 8009)   FL (n = 7733)   AC (n = 8537)   HC (n = 8770)   βa  95% CI  βa  95% CI  βa  95% CI  βa  95% CI  Mean change comparing exposures ≥ 150 µg/m3 PM10 with those < 150 μg/m3  From conception to ultrasound  0.016  −0.035, 0.067  0.095**  0.038, 0.153  0.001  −0.046, 0.049  0.090**  0.024, 0.156  One month before ultrasound  −0.052  −0.097, 0.007  −0.007  −0.051, 0.038  −0.044  −0.082, 0.003  0.092*  0.020, 0.151  Two weeks before ultrasound  −0.037  −0.083, 0.008  −0.043  −0.089, 0.002  −0.037  −0.076, 0.002  0.020  −0.040, 0.080  One week before ultrasound  −0.034  −0.079, 0.011  −0.005  −0.050, 0.040  -0.025  −0.064, 0.014  0.070*  0.010, 0.130  Mean change per 10 µg/m3 increase in PM10  From conception to ultrasound  −0.018**  −0.030, -0.007  0.008  −0.006, 0.022  0.005  −0.016, 0.006  0.011  −0.006, 0.029  One month before ultrasound  −0.004  −0.008, 0.000  −0.001  −0.005, 0.003  0.002  −0.007, 0.004  0.005*  0.000, 0.009  Two weeks before ultrasound  −0.002  −0.006, 0.002  0.002  −0.006, 0.003  0.002  −0.006, 0.001  0.008*  0.001, 0.011  One week before ultrasound  −0.002  −0.006, 0.002  0.000  −0.004, 0.004  −0.002  −0.005, 0.002  0.004*  0.000, 0.009  a Adjusted for maternal age (< 30, ≥ 30 years), education (< college, ≥ college), family monthly income per capita (< 3000, ≥ 3000 RMB), smoking during pregnancy (yes, no), pre-pregnancy BMI (≤ 18.5, 18.5–24.0, ≥ 24.0), weight gain du.ring pregnancy (≤ 15 kg, 15–18.5 kg, > 18.5 kg), parity (primiparous, multiparous), season of conception (spring/summer, autumn/winter), cooking fuel (gas/electricity, coal/biomass), estimated SO2, NO2 by different corresponding exposure windows, and averaged temperature during different corresponding exposure windows. * P < 0.05; ** P < 0.01. Table 3 shows the associations between PM10 exposures averaged by different exposure windows, and risks of under-growth, and over-growth. When PM10 exposure was estimated from conception until the ultrasound, compared with PM10 < 150 µ/m3, PM10 ≥ 150 µg/m3 was associated with an increased risk of overgrowth [odds ratio (OR), 1.22; 95% CI, 1.04, 1.42 for FL, and 1.19; 1.07, 1.33 for HC]. PM10 was associated with a reduced risk of BPD overgrowth when PM10 was treated as < 150 µg/m3 versus ≥ 150 µg/m3 or a continuous variable during 1 month or 2 weeks before ultrasound examinations. Table 3 Association between averaged PM10 during the different exposure windows and risks of fetal under-growth and over-growth Exposure windows  BPD (n = 8009)   FL (n = 7733)   AC (n = 8537)   HC (n = 8770)   ORa  95% CI  ORa  95% CI  ORa  95% CI  ORa  95% CI  Number of pregnancies in each model  Normal growth  7005  5662  6989  6378   Under-growth  438  189  393  367   Over-growth  566  1882  1155  1665  Risk comparing exposures ≥ 150 µg/m3 PM10 with those < 150 μg/m3  From conception to ultrasound   Under-growth  1.13  0.89, 1.43  1.04  0.72, 1.50  0.86  0.73, 1.01  1.02  0.81, 1.30   Overg-rowth  1.00  0.81, 1.23  1.22*  1.04, 1.42  1.03  0.87, 1.21  1.19**  1.07, 1.33  One month before ultrasound   Under-growth  1.04  0.85, 1.26  0.87  0.62, 1.22  1.08  0.84, 1.39  0.98  0.78, 1.23   Over-growth  0.83*  0.69, 0.99  0.97  0.87, 1.08  0.95  0.83, 1.09  1.08*  1.00, 1.19  Two weeks before ultrasound   Under-growth  0.80  0.64, 1.01  1.08  0.78, 1.50  1.09  0.94, 1.26  0.97  0.76, 1.24   Over-growth  0.79**  0.66, 0.94  0.91  0.81, 1.02  0.91  0.79, 1.04  1.03  0.93, 1.14  One week before ultrasound   Under-growth  0.89  0.72, 1.09  0.86  0.61, 1.20  1.09  0.94, 1.26  0.99  0.77, 1.26   Over-growth  0.87  0.73, 1.04  0.96  0.86, 1.07  0.97  0.85, 1.11  1.11*  1.00, 1.23  Risk per 10 µg/m3 increase in PM10  From conception to ultrasound   Under-growth  1.03  0.98, 1.07  0.99  0.92, 1.07  0.98  0.93, 1.02  0.99  0.92, 1.06   Over-growth  0.98  0.94, 1.02  1.01  0.98, 1.04  1.01  0.98, 1.05  1.02  0.99, 1.05  One month before ultrasound   Under-growth  1.00  0.98, 1.02  0.99  0.96, 1.02  1.00  0.99, 1.02  1.00  0.98, 1.02   Over-growth  0.98*  0.97, 1.00  1.00  0.99, 1.01  0.99  0.98, 1.01  1.01*  1.00, 1.02  Two weeks before ultrasound   Under-growth  0.99  0.97, 1.01  1.00  0.97, 1.03  1.00  0.99, 1.02  1.00  0.97, 1.02   Over-growth  0.98**  0.96, 0.99  0.99  0.98, 1.01  1.00  0.99, 1.01  1.01*  1.00, 1.02  One week before ultrasound   Under-growth  0.99  0.97, 1.01  0.99  0.96, 1.02  1.00  0.99, 1.01  1.00  0.97, 1.02   Over-growth  0.99  0.97, 1.01  1.00  0.99, 1.01  1.00  0.98, 1.01  1.01*  1.00, 1.02  Exposure windows  BPD (n = 8009)   FL (n = 7733)   AC (n = 8537)   HC (n = 8770)   ORa  95% CI  ORa  95% CI  ORa  95% CI  ORa  95% CI  Number of pregnancies in each model  Normal growth  7005  5662  6989  6378   Under-growth  438  189  393  367   Over-growth  566  1882  1155  1665  Risk comparing exposures ≥ 150 µg/m3 PM10 with those < 150 μg/m3  From conception to ultrasound   Under-growth  1.13  0.89, 1.43  1.04  0.72, 1.50  0.86  0.73, 1.01  1.02  0.81, 1.30   Overg-rowth  1.00  0.81, 1.23  1.22*  1.04, 1.42  1.03  0.87, 1.21  1.19**  1.07, 1.33  One month before ultrasound   Under-growth  1.04  0.85, 1.26  0.87  0.62, 1.22  1.08  0.84, 1.39  0.98  0.78, 1.23   Over-growth  0.83*  0.69, 0.99  0.97  0.87, 1.08  0.95  0.83, 1.09  1.08*  1.00, 1.19  Two weeks before ultrasound   Under-growth  0.80  0.64, 1.01  1.08  0.78, 1.50  1.09  0.94, 1.26  0.97  0.76, 1.24   Over-growth  0.79**  0.66, 0.94  0.91  0.81, 1.02  0.91  0.79, 1.04  1.03  0.93, 1.14  One week before ultrasound   Under-growth  0.89  0.72, 1.09  0.86  0.61, 1.20  1.09  0.94, 1.26  0.99  0.77, 1.26   Over-growth  0.87  0.73, 1.04  0.96  0.86, 1.07  0.97  0.85, 1.11  1.11*  1.00, 1.23  Risk per 10 µg/m3 increase in PM10  From conception to ultrasound   Under-growth  1.03  0.98, 1.07  0.99  0.92, 1.07  0.98  0.93, 1.02  0.99  0.92, 1.06   Over-growth  0.98  0.94, 1.02  1.01  0.98, 1.04  1.01  0.98, 1.05  1.02  0.99, 1.05  One month before ultrasound   Under-growth  1.00  0.98, 1.02  0.99  0.96, 1.02  1.00  0.99, 1.02  1.00  0.98, 1.02   Over-growth  0.98*  0.97, 1.00  1.00  0.99, 1.01  0.99  0.98, 1.01  1.01*  1.00, 1.02  Two weeks before ultrasound   Under-growth  0.99  0.97, 1.01  1.00  0.97, 1.03  1.00  0.99, 1.02  1.00  0.97, 1.02   Over-growth  0.98**  0.96, 0.99  0.99  0.98, 1.01  1.00  0.99, 1.01  1.01*  1.00, 1.02  One week before ultrasound   Under-growth  0.99  0.97, 1.01  0.99  0.96, 1.02  1.00  0.99, 1.01  1.00  0.97, 1.02   Over-growth  0.99  0.97, 1.01  1.00  0.99, 1.01  1.00  0.98, 1.01  1.01*  1.00, 1.02  a Adjusted for maternal age (< 30, 30 years), education (< college, college), family monthly income per capita (< 3000, ≥ 3000 RMB), smoking during pregnancy (yes, no), pre-pregnancy BMI (≤ 18.5, 18.5–24.0, ≥ 24.0), weight gain during pregnancy (≤ 15 kg, 15–18.5 kg, > 18.5 kg), parity (primiparous, multiparous), season of conception (spring/summer, autumn/winter), cooking fuel (gas/electricity, coal/biomass), estimated SO2, NO2 by different corresponding exposure windows, and averaged temperature during different corresponding exposure windows. * P ≤ 0.05; ** P < 0.01. Table 3 Association between averaged PM10 during the different exposure windows and risks of fetal under-growth and over-growth Exposure windows  BPD (n = 8009)   FL (n = 7733)   AC (n = 8537)   HC (n = 8770)   ORa  95% CI  ORa  95% CI  ORa  95% CI  ORa  95% CI  Number of pregnancies in each model  Normal growth  7005  5662  6989  6378   Under-growth  438  189  393  367   Over-growth  566  1882  1155  1665  Risk comparing exposures ≥ 150 µg/m3 PM10 with those < 150 μg/m3  From conception to ultrasound   Under-growth  1.13  0.89, 1.43  1.04  0.72, 1.50  0.86  0.73, 1.01  1.02  0.81, 1.30   Overg-rowth  1.00  0.81, 1.23  1.22*  1.04, 1.42  1.03  0.87, 1.21  1.19**  1.07, 1.33  One month before ultrasound   Under-growth  1.04  0.85, 1.26  0.87  0.62, 1.22  1.08  0.84, 1.39  0.98  0.78, 1.23   Over-growth  0.83*  0.69, 0.99  0.97  0.87, 1.08  0.95  0.83, 1.09  1.08*  1.00, 1.19  Two weeks before ultrasound   Under-growth  0.80  0.64, 1.01  1.08  0.78, 1.50  1.09  0.94, 1.26  0.97  0.76, 1.24   Over-growth  0.79**  0.66, 0.94  0.91  0.81, 1.02  0.91  0.79, 1.04  1.03  0.93, 1.14  One week before ultrasound   Under-growth  0.89  0.72, 1.09  0.86  0.61, 1.20  1.09  0.94, 1.26  0.99  0.77, 1.26   Over-growth  0.87  0.73, 1.04  0.96  0.86, 1.07  0.97  0.85, 1.11  1.11*  1.00, 1.23  Risk per 10 µg/m3 increase in PM10  From conception to ultrasound   Under-growth  1.03  0.98, 1.07  0.99  0.92, 1.07  0.98  0.93, 1.02  0.99  0.92, 1.06   Over-growth  0.98  0.94, 1.02  1.01  0.98, 1.04  1.01  0.98, 1.05  1.02  0.99, 1.05  One month before ultrasound   Under-growth  1.00  0.98, 1.02  0.99  0.96, 1.02  1.00  0.99, 1.02  1.00  0.98, 1.02   Over-growth  0.98*  0.97, 1.00  1.00  0.99, 1.01  0.99  0.98, 1.01  1.01*  1.00, 1.02  Two weeks before ultrasound   Under-growth  0.99  0.97, 1.01  1.00  0.97, 1.03  1.00  0.99, 1.02  1.00  0.97, 1.02   Over-growth  0.98**  0.96, 0.99  0.99  0.98, 1.01  1.00  0.99, 1.01  1.01*  1.00, 1.02  One week before ultrasound   Under-growth  0.99  0.97, 1.01  0.99  0.96, 1.02  1.00  0.99, 1.01  1.00  0.97, 1.02   Over-growth  0.99  0.97, 1.01  1.00  0.99, 1.01  1.00  0.98, 1.01  1.01*  1.00, 1.02  Exposure windows  BPD (n = 8009)   FL (n = 7733)   AC (n = 8537)   HC (n = 8770)   ORa  95% CI  ORa  95% CI  ORa  95% CI  ORa  95% CI  Number of pregnancies in each model  Normal growth  7005  5662  6989  6378   Under-growth  438  189  393  367   Over-growth  566  1882  1155  1665  Risk comparing exposures ≥ 150 µg/m3 PM10 with those < 150 μg/m3  From conception to ultrasound   Under-growth  1.13  0.89, 1.43  1.04  0.72, 1.50  0.86  0.73, 1.01  1.02  0.81, 1.30   Overg-rowth  1.00  0.81, 1.23  1.22*  1.04, 1.42  1.03  0.87, 1.21  1.19**  1.07, 1.33  One month before ultrasound   Under-growth  1.04  0.85, 1.26  0.87  0.62, 1.22  1.08  0.84, 1.39  0.98  0.78, 1.23   Over-growth  0.83*  0.69, 0.99  0.97  0.87, 1.08  0.95  0.83, 1.09  1.08*  1.00, 1.19  Two weeks before ultrasound   Under-growth  0.80  0.64, 1.01  1.08  0.78, 1.50  1.09  0.94, 1.26  0.97  0.76, 1.24   Over-growth  0.79**  0.66, 0.94  0.91  0.81, 1.02  0.91  0.79, 1.04  1.03  0.93, 1.14  One week before ultrasound   Under-growth  0.89  0.72, 1.09  0.86  0.61, 1.20  1.09  0.94, 1.26  0.99  0.77, 1.26   Over-growth  0.87  0.73, 1.04  0.96  0.86, 1.07  0.97  0.85, 1.11  1.11*  1.00, 1.23  Risk per 10 µg/m3 increase in PM10  From conception to ultrasound   Under-growth  1.03  0.98, 1.07  0.99  0.92, 1.07  0.98  0.93, 1.02  0.99  0.92, 1.06   Over-growth  0.98  0.94, 1.02  1.01  0.98, 1.04  1.01  0.98, 1.05  1.02  0.99, 1.05  One month before ultrasound   Under-growth  1.00  0.98, 1.02  0.99  0.96, 1.02  1.00  0.99, 1.02  1.00  0.98, 1.02   Over-growth  0.98*  0.97, 1.00  1.00  0.99, 1.01  0.99  0.98, 1.01  1.01*  1.00, 1.02  Two weeks before ultrasound   Under-growth  0.99  0.97, 1.01  1.00  0.97, 1.03  1.00  0.99, 1.02  1.00  0.97, 1.02   Over-growth  0.98**  0.96, 0.99  0.99  0.98, 1.01  1.00  0.99, 1.01  1.01*  1.00, 1.02  One week before ultrasound   Under-growth  0.99  0.97, 1.01  0.99  0.96, 1.02  1.00  0.99, 1.01  1.00  0.97, 1.02   Over-growth  0.99  0.97, 1.01  1.00  0.99, 1.01  1.00  0.98, 1.01  1.01*  1.00, 1.02  a Adjusted for maternal age (< 30, 30 years), education (< college, college), family monthly income per capita (< 3000, ≥ 3000 RMB), smoking during pregnancy (yes, no), pre-pregnancy BMI (≤ 18.5, 18.5–24.0, ≥ 24.0), weight gain during pregnancy (≤ 15 kg, 15–18.5 kg, > 18.5 kg), parity (primiparous, multiparous), season of conception (spring/summer, autumn/winter), cooking fuel (gas/electricity, coal/biomass), estimated SO2, NO2 by different corresponding exposure windows, and averaged temperature during different corresponding exposure windows. * P ≤ 0.05; ** P < 0.01. Compared with the results of cumulative exposure from conception to the ultrasound examination, similar positive associations were only observed for HC measures for other exposure windows, 1 month before ultrasound examinations and 1 or 2 weeks before ultrasound examinations, when PM10 was treated as < 150 µg/m3 versus ≥ 150 µg/m3 or a continuous variable (Tables 2 and 3). Discussion Our results support the hypothesis that prenatal exposure to high levels of ambient PM10 is associated with increased risk of fetal under- and over-growth. Our study also suggests a stronger and consistent association of over-growth for HC over the different exposure windows compared with other fetal growth parameters. To the best of our knowledge, our study was the first one to examine the associations between prenatal exposure to PM10 and risk of fetal over-growth. We found that exposure to higher levels of PM10 increased the risk of fetal HC over-growth. In early studies conducted in Australia, Europe and the USA, Hansen et al.5 found reductions of FL, AC and HC associated with increased exposure to PM10 during gestational days 91–120, and a reduction of FL measurement during 0–30 days of gestation; van den Hooven et al.22 observed reduction in HC in the third trimester; and Ritz et al.23 reported that no association was observed between PM10 exposure and HC, FL and AC. Although van den Hooven et al. did not highlight the findings that there were significant increases in HC measures during the second trimester in the second and third quartiles of PM10 exposure, and that there were significant increases in fetal weight in higher quartiles (second, third and fourth) of PM10 exposure, these findings were similar to ours. Although the underlying mechanisms are currently unclear, a growing body of literature suggests that increased oxidative stress and inflammatory response to PMs can influence energy expenditure, resulting in increased dietary intake.32,33 There was an increasing awareness that changes in placental metabolism and nutrient transport capacity might contribute to altered fetal growth.46 Human studies showed that higher cord blood leptin levels caused by greater weight gain were associated with increased fetal size,47 and altered expression of maternal insulin-like growth factor due to excessive gestational weight gain contributed to increased fetal growth through increased placental nutrient transporter expression.48,49 It is possible that the positive relationship between ambient air pollutants and fetal over-growth was mediated by maternal BMI/weight gain. In our study, pregnant women carrying an overgrown fetus were more likely to be in the highest category of maternal weight gain and to be pre-pregnancy obese. Comparing the study results among the published studies and ours was challenging. The outcome measures varied across different studies, including differences in ultrasound measurement methods and potential of inter-sonographer variations. The exposure assessment also varied across the studies. van den Hooven et al.22 performed dispersion modelling based on home addresses only, though allowing for residential mobility; Hansen et al.5 estimated monthly exposure to nearest monitors using IDW over the first 4 months and restricted analysis through four exposure windows (0–30, 31–60, 61–90, 91–120 days of gestation). Ritz et al.23 considered a 2-weeks exposure window around each measurement time (at 19.1, 28.8 and 36.7 weeks of gestation). In addition, all these published studies were conducted in low-air-pollution areas as compared with our study. The compositions of PM10 in different studies might also be different and could cause a different impact on fetal growth. Our study included large sample size (N = 8877) with a relatively large number of ultrasound scans for each fetal parameter (N ∼18 583), compared with previous studies.5,22,23 It allowed exploration of association by subgroups based on the growth status. Our study assessed exposure based on both home and work addresses with consideration of residential mobility, as well as PM10 exposure through different exposure windows, which minimized potential exposure misclassification and ability to explore critical exposure time windows on fetal growth. Detailed information on potential confounding factors such as smoking, cooking fuels and other co-exposure air pollutants were collected and included in our analyses. All participants were recruited from the same hospital and the information on ultrasound measurements was obtained from medical records, so potential misclassifications of fetal measurements were minimized. Additionally, with a potential U-shape relationship between air pollution and fetal growth, we employed the latest international fetal growth standard to examine the association with both under-growth and over-growth, in addition to treating growth measurements as continuous variables as done by earlier studies.5,22,23 This potential heterogeneity of PM10 exposure effects on fetal growth measurements still calls for future study with larger sample size, to replicate and confirm our findings. Limitations should be considered when interpreting the study results. The accuracy of exposure estimation might be limited by the lack of monitors in rural areas in Lanzhou city. However, more than 90% of the women lived within 5.5 km of a monitor. Sensitivity analyses using the data from different exposure assessment approaches (nearest monitors, IDW using two or four monitors, using data for the women living within 5.5 km or 12.9 km of a monitor) showed consistent results. PM10 exposure was estimated using data from government monitors which, though the most commonly used method in air pollution epidemiology, may not represent the actual individual exposure level which would include differences in indoor/outdoor activity patterns, and may not consider intra-urban gradients in pollutants. However, measurements through personal air monitors in this setting (pregnancy, in a large population) may be impractical as well as expensive, and numerous studies have used the above monitoring approach successfully. To evaluate whether the four monitors provided reasonable spatial coverage in our study, we calculated the coefficient of divergence, which provides the diversity between concentrations at sampling site-pairs.50 We found that coefficients of divergence for all site-pairs were lower than 0.20, indicating low spatial heterogeneity of PM10 in our study area. However, future studies might consider additional monitor coverage, in order to address the issues of spatial heterogeneity and improve estimates of exposure. Although the exact sources of PM10 in Lanzhou city are still unclear, a factory of the China Petroleum and Chemical Corporation (Sinopec Limited) has been located in the west, an upwind location of this city. Future work may investigate PM sources, which have different chemical structures and possibly different health impacts. The measurement errors with respect to the outcomes would be another concern both in previous large epidemiological studies and in the current study, because it is possible that the ultrasound examinations at different gestational weeks were performed by different clinicians/technicians. However, GPMCCH is the largest maternity and child care hospital in Lanzhou, China. The hospital performs internal quality control in order to provide high-quality ultrasound examinations, and the historical records of reproducibility also demonstrated that the measurement errors would be likely minimized. In conclusion, our study supports the hypothesis that prenatal exposure to high levels of ambient PM10 increases the risk of abnormal fetal growth, and stronger and consistent associations were mainly seen for HC over-growth by different exposure windows compared with other fetal growth measurements. This study suggests that the associations between maternal exposure to PM10 and fetal growth may vary by different fetal growth measurements (BPD, FL, AC and HC) and fetal growth patterns (under-/over-growth). Future multi-site studies with different types of air pollutants are needed to fully understand the impacts of air pollution combinations on fetal growth. Our study also calls for future studies to explore the underlying mechanisms and post-natal consequences of these findings Supplementary Data Supplementary data are available at IJE online. Funding The work was supported by internal funding from the Gansu Provincial Maternity and Child Care Hospital, Gansu Provincial Science and Technology Department Grant (1204WCGA021), National Institutes of Health Grants (K02HD70324, R01ES019587) and the U.S. Environmental Protection Agency (EPA-G2014-STAR-J1). Author Contributions J.Q., Q.L. and Y.Z. designed the research; X.L., Z.T. and H.Z. conducted the birth cohort study; Y.Z. collected air pollution and meteorology data; N.Z. and H.H. managed the study and cleaned the database; N.Z., S.M., Y.H. and N.M. performed statistical analysis; N.Z. constructed the figures; N.Z., M.B. and Y.Z. wrote the first draft, and all authors contributed to the final draft and approved the manuscript. Conflict of interest: None declared. References 1 Lawn JE, Blencowe H, Pattinson R et al.   Stillbirths: Where? When? Why? How to make the data count? Lancet  2011; 377: 1448– 63. Google Scholar CrossRef Search ADS PubMed  2 Liu L, Johnson HL, Cousens S et al.   Global, regional, and national causes of child mortality: an updated systematic analysis for 2010 with time trends since 2000. 2012; 379: 2151– 61. 3 World Health Organization. A Global Review of the Key Interventions Related to Reproductive, Maternal, Newborn and Child Health (RMNCH) . Geneva: WHO, 2011. 4 Ouyang F, Zhang J, Betran AP, Yang Z, Souza JP, Merialdi M. Recurrence of adverse perinatal outcomes in developing countries. Bull World Health Organ  2013; 91: 357– 67. Google Scholar CrossRef Search ADS PubMed  5 Hansen CA, Barnett AG, Pritchard G. The effect of ambient air pollution during early pregnancy on fetal ultrasonic measurements during mid-pregnancy. Environ Health Perspect  2008; 116: 362– 69. Google Scholar CrossRef Search ADS PubMed  6 Glinianaia SV, Rankin J, Bell R, Pless-Mulloli T, Howel D. Particulate air pollution and fetal health: a systematic review of the epidemiologic evidence. Epidemiology  2004; 15: 36– 45. Google Scholar CrossRef Search ADS PubMed  7 Shah PS, Balkhair T. Knowledge Synthesis Group on Determinants of Preterm LBWb. Air pollution and birth outcomes: a systematic review. Environ Int  2011; 37: 498– 516. Google Scholar CrossRef Search ADS PubMed  8 Woodruff TJ, Parker JD, Darrow LA et al.   Methodological issues in studies of air pollution and reproductive health. Environ Res  2009; 109: 311– 20. Google Scholar CrossRef Search ADS PubMed  9 Choi H, Jedrychowski W, Spengler J et al.   International studies of prenatal exposure to polycyclic aromatic hydrocarbons and fetal growth. Environ Health Perspect  2006; 114: 1744– 50. Google Scholar PubMed  10 Hansen C, Neller A, Williams G, Simpson R. Low levels of ambient air pollution during pregnancy and fetal growth among term neonates in Brisbane, Australia. Environ Res  2007; 103: 383– 89. Google Scholar CrossRef Search ADS PubMed  11 Jedrychowski W, Bendkowska I, Flak E et al.   Estimated risk for altered fetal growth resulting from exposure to fine particles during pregnancy: an epidemiologic prospective cohort study in Poland. Environ Health Perspect  2004; 112: 1398– 402. Google Scholar CrossRef Search ADS PubMed  12 Salam MT, Millstein J, Li YF, Lurmann FW, Margolis HG, Gilliland FD. Birth outcomes and prenatal exposure to ozone, carbon monoxide, and particulate matter: results from the Children's Health Study. Environ Health Perspect  2005; 113: 1638– 44. Google Scholar CrossRef Search ADS PubMed  13 Hannam K, McNamee R, Baker P, Sibley C, Agius R. Air pollution exposure and adverse pregnancy outcomes in a large UK birth cohort: use of a novel spatio-temporal modelling technique. Scand J Work Environ Health  2014; 40: 518– 30. Google Scholar CrossRef Search ADS PubMed  14 Winckelmans E, Cox B, Martens E, Fierens F, Nemery B, Nawrot TS. Fetal growth and maternal exposure to particulate air pollution - More marked effects at lower exposure and modification by gestational duration. Environ Res  2015; 140: 611– 18. Google Scholar CrossRef Search ADS PubMed  15 Fleisch AF, Rifas-Shiman SL, Koutrakis P et al.   Prenatal exposure to traffic pollution: associations with reduced fetal growth and rapid infant weight gain. Epidemiology  2015; 26: 43– 50. Google Scholar CrossRef Search ADS PubMed  16 Liu S, Krewski D, Shi Y, Chen Y, Burnett RT. Association between maternal exposure to ambient air pollutants during pregnancy and fetal growth restriction. J Expo Sci Environ Epidemiol  2007; 17: 426– 32. Google Scholar CrossRef Search ADS PubMed  17 Dejmek J, Selevan SG, Benes I, Solansky I, Sram RJ. Fetal growth and maternal exposure to particulate matter during pregnancy. Environ Health Perspect  1999; 107: 475– 80. Google Scholar CrossRef Search ADS PubMed  18 Smarr MM, Vadillo-Ortega F, Castillo-Castrejon M, O'Neill MS. The use of ultrasound measurements in environmental epidemiological studies of air pollution and fetal growth. Curr Opin Pediatr  2013; 25: 240– 46. Google Scholar CrossRef Search ADS PubMed  19 Gardosi J, Figueras F, Clausson B, Francis A. The customised growth potential: an international research tool to study the epidemiology of fetal growth. Paediatr Perinat Epidemiol  2011; 25: 2– 10. Google Scholar CrossRef Search ADS PubMed  20 Zhang J, Mikolajczyk R, Grewal J, Neta G, Klebanoff M. Prenatal application of the individualized fetal growth reference. Am J Epidemiol  2011; 173: 539– 43. Google Scholar CrossRef Search ADS PubMed  21 Hemachandra AH, Klebanoff MA, Duggan AK, Hardy JB, Furth SL. The association between intrauterine growth restriction in the full-term infant and high blood pressure at age 7 years: results from the Collaborative Perinatal Project. Int J Epidemiol  2006; 35: 871– 77. Google Scholar CrossRef Search ADS PubMed  22 van den Hooven EH, Pierik FH, de Kluizenaar Y et al.   Air pollution exposure during pregnancy, ultrasound measures of fetal growth, and adverse birth outcomes: a prospective cohort study. Environ Health Perspect  2012; 120: 150– 56. Google Scholar CrossRef Search ADS PubMed  23 Ritz B, Qiu J, Lee PC et al.   Prenatal air pollution exposure and ultrasound measures of fetal growth in Los Angeles, California. Environ Res  2014; 130: 7– 13. Google Scholar CrossRef Search ADS PubMed  24 Slama R, Thiebaugeorges O, Goua V et al.   Maternal personal exposure to airborne benzene and intrauterine growth. Environ Health Perspect  2009; 117: 1313– 21. Google Scholar CrossRef Search ADS PubMed  25 Aguilera I, Garcia-Esteban R, Iniguez C et al.   Prenatal exposure to traffic-related air pollution and ultrasound measures of fetal growth in the INMA Sabadell cohort. Environ Health Perspect  2010; 118: 705– 11. Google Scholar CrossRef Search ADS PubMed  26 Iniguez C, Ballester F, Estarlich M et al.   Prenatal exposure to traffic-related air pollution and fetal growth in a cohort of pregnant women. Occup Environ Med  2012; 69: 736– 44. Google Scholar CrossRef Search ADS PubMed  27 Patelarou E, Kelly FJ. Indoor exposure and adverse birth outcomes related to fetal growth, miscarriage and prematurity - a systematic review. Int J Environ Res Public Health  2014; 11: 5904– 33. Google Scholar CrossRef Search ADS PubMed  28 Iniguez C, Esplugues A, Sunyer J et al.   Prenatal exposure to NO2 and ultrasound measures of fetal growth in the Spanish INMA Cohort. Environ Health Perspect  2016; 124: 235– 42. Google Scholar PubMed  29 Malmqvist E, Liew Z, Kallen K et al.   Fetal growth and air pollution - A study on ultrasound and birth measures. Environ Res  2017; 152: 73– 80. Google Scholar CrossRef Search ADS PubMed  30 Langer O. Fetal macrosomia: etiologic factors. Clin Obstet Gynecol  2000; 43: 283– 97. Google Scholar CrossRef Search ADS PubMed  31 Hod M, Pardo J. Environmental rather than genetic fetal overgrowth: defining the difference and hints for diagnosis and management. Ultrasound Obstet Gynecolt  2009; 33: 375– 78. Google Scholar CrossRef Search ADS   32 Xu Z, Xu X, Zhong M et al.   Ambient particulate air pollution induces oxidative stress and alterations of mitochondria and gene expression in brown and white adipose tissues. Particle Fibre Toxicol  2011; 8: 20. Google Scholar CrossRef Search ADS   33 Bolton JL, Smith SH, Huff NC et al.   Prenatal air pollution exposure induces neuroinflammation and predisposes offspring to weight gain in adulthood in a sex-specific manner. FASEB J  2012; 26: 4743– 54. Google Scholar CrossRef Search ADS PubMed  34 Thayer KA, Heindel JJ, Bucher JR, Gallo MA. Role of environmental chemicals in diabetes and obesity: a National Toxicology Program workshop review. Environ Health Perspect  2012; 120: 779– 89. Google Scholar CrossRef Search ADS PubMed  35 Qiu J, He X, Cui H et al.   Passive smoking and preterm birth in urban China. Am J Epidemiol  2014; 180: 94– 102. Google Scholar CrossRef Search ADS PubMed  36 Zhao N, Qiu J, Zhang Y et al.   Ambient air pollutant PM10 and risk of preterm birth in Lanzhou, China. Environ Int  2015; 76: 71– 77. Google Scholar CrossRef Search ADS PubMed  37 Jin L, Qiu J, Zhang Y et al.  . Ambient air pollution and congenital heart defects in Lanzhou, China. Environ Res Lett  2015; 10: 074005. Google Scholar CrossRef Search ADS   38 Huang X, Qiu J, Zhang Y et al.   Ambient air pollutant PM10 and risk of pregnancy-induced hypertension in urban China. Environ Res Lett  2015; 10: 084025. Google Scholar CrossRef Search ADS   39 Jiang M, Qiu J, Zhou M et al.   Exposure to cooking fuels and birthweight in Lanzhou, China: a birth cohort study. BMC Public Health  2015; 15: 712. Google Scholar CrossRef Search ADS PubMed  40 Papageorghiou AT, Ohuma EO, Altman DG et al.   International standards for fetal growth based on serial ultrasound measurements: the Fetal Growth Longitudinal Study of the INTERGROWTH-21st Project. Lancet  2014; 384: 869– 79. Google Scholar CrossRef Search ADS PubMed  41 Pinheiro JC, Bates DM. Mixed-Effects Models in S and S-Plus  New York, NY: Springer, 2000. Google Scholar CrossRef Search ADS   42 Moser EB. Repeated Measures Modeling With PROC MIXED . Louisiana State University, Baton Rouge, LA: SAS Institute, 2004. 43 Zhang Y, Li M, Bravo M et al.   Air quality in Lanzhou, a major industrial city in China: characteristics of air pollution and review of existing evidence from air pollution and health studies. Water, Air, & Soil Pollution  2014; 225: 1– 17. 44 Environmental Protection Agency. National Ambient Air Quality Standards (NAAQS) 2015. https://www.epa.gov/pm-pollution/table-historical-particulate-matter-pm-national-ambient-air-quality-standards-naaqs (15 January 2013, date last accessed). 45 Aho K, Derryberry D, Peterson T. Model selection for ecologists: the worldviews of AIC and BIC. Ecology  2014; 95: 631– 36. Google Scholar CrossRef Search ADS PubMed  46 Higgins L, Greenwood SL, Wareing M, Sibley CP, Mills TA. Obesity and the placenta: A consideration of nutrient exchange mechanisms in relation to aberrant fetal growth. Placenta  2011; 32: 1– 7. Google Scholar CrossRef Search ADS PubMed  47 Karakosta P, Georgiou V, Fthenou E et al.   Maternal weight status, cord blood leptin and fetal growth: a prospective mother-child cohort study (Rhea study). Paediatr Perinat Epidemiol  2013; 27: 461– 71. Google Scholar CrossRef Search ADS PubMed  48 Ferraro ZM, Qiu Q, Gruslin A, Adamo KB. Excessive gestational weight gain and obesity contribute to altered expression of maternal insulin-like growth factor binding protein-3. Int J Womens Health  2013; 5: 657– 65. Google Scholar CrossRef Search ADS PubMed  49 Tung E, Roberts CT, Heinemann GK et al.   Increased placental nutrient transporter expression at midgestation after maternal growth hormone treatment in pigs: a placental mechanism for increased fetal growth. Biol Reprod  2012; 87: 126. Google Scholar CrossRef Search ADS PubMed  50 Pinto JP, Lefohn AS, Shadwick DS. Spatial variability of PM2.5 in urban areas in the United States. Journal of the Air & Waste Management Association  2004; 54: 440– 49. Google Scholar CrossRef Search ADS   © The Author(s) 2018; all rights reserved. Published by Oxford University Press on behalf of the International Epidemiological Association http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png International Journal of Epidemiology Oxford University Press

Effects of prenatal exposure to ambient air pollutant PM10 on ultrasound-measured fetal growth

Loading next page...
 
/lp/ou_press/effects-of-prenatal-exposure-to-ambient-air-pollutant-pm10-on-iS0NLMb30D
Publisher
Oxford University Press
Copyright
© The Author(s) 2018; all rights reserved. Published by Oxford University Press on behalf of the International Epidemiological Association
ISSN
0300-5771
eISSN
1464-3685
D.O.I.
10.1093/ije/dyy019
Publisher site
See Article on Publisher Site

Abstract

Abstract Background Limited epidemiological studies have investigated the relationship between prenatal exposure to ambient particulate matter and risk of abnormal fetal growth, and have reached inconclusive results. No study has been conducted in areas with very high air pollution levels. We investigated the hypothesis that exposure to high levels of particulate matter with aerodynamic diameter no larger than 10 µm (PM10) during pregnancy increases the risk of abnormal fetal growth. Methods A birth cohort study was performed in Lanzhou, China, 2010–12, including 8877 pregnant women with 18 583 ultrasound measurements of four fetal growth parameters during pregnancy, including biparietal diameter (BPD), femur length (FL), head circumference (HC) and abdominal circumference (AC). Mixed-effects modelling was used to examine the associations between PM10 exposure and risk of abnormal fetal growth. Results When average PM10 exposure from conception until the ultrasound examination exceeded 150 µg/m3, there were significant increases in standardized FL (β = 0.095, P = 0.0012) and HC (β = 0.090, P = 0.0078) measures. When average PM10 exposure was treated as continuous variable, we found a significant decrease in standardized BPD (β = −0.018, P = 0.0016) as per 10 µg/m3 increase in PM10. After examining the associations by various exposure windows, positive associations between higher levels of PM10 and fetal overgrowth were consistently seen for HC measures. Conclusions Our study suggested that prenatal exposure to high levels of ambient PM10 increased the risk of abnormal fetal growth. The findings from our study have important public health implications and also call for future studies to explore the underlying mechanisms and post-natal consequences of these findings. Fetal growth, PM10, air pollution, ultrasound measures, pregnancy Key Messages No study has been conducted in areas with very high air pollution levels. Exposure to high levels of ambient PM10 increases the risk of abnormal fetal growth. The risk may vary by different fetal growth measurements and fetal growth patterns. Positive associations between higher levels of PM10 and fetal over- growth were consistently seen for head circumference measures by various exposure windows. Introduction A growing body of evidence indicates that fetal growth restriction (FGR), also known as intrauterine growth restriction or fetal under-growth, is an important predictor of neonatal death.1–4 Over the past decade, numerous studies have indicated that ambient air pollution during pregnancy influences fetal growth.5,6 A majority of these previous epidemiological studies have associated maternal air pollution exposure with fetal growth by using various neonatal anthropometric measurements at birth, such as birthweight and length, size for gestational age, head circumference (HC) and abdominal circumference (AC).7–17 However, the late assessment of growth restriction potentially introduces bias,18 and patterns of fetal growth during pregnancy cannot be assessed in these studies.5 In addition, clinical studies have suggested that birthweight as a proxy for fetal growth may not be the best indicator, and the conditions for growth restriction and small size for gestational age are not synonymous.19,20 Furthermore, others suggest that birthweight poorly reflects fetal under-growth during the first two trimesters, because growth impairment and imbalance that may have occurred during early period may be compensated for in the remaining pregnancy period.21 Assessing fetal growth in utero using ultrasound measures could provide more accurate classification of restricted or excessive growth by reducing the time between exposure and outcome assessment.18,22 A small number of studies have investigated the association between exposure to various air pollutants [i.e. nitrogen monoxide, nitrogen dioxide (NO2), sulphur dioxide (SO2), ozone, carbon monoxide and particulate matter with aerodynamic diameter < 10 µm (PM10)] during pregnancy and fetal growth measured by ultrasound.5,22–29 Only three of these have focused on the effects of exposure to PM10 during pregnancy, and these studies reached inconsistent results.5,22,23 All of these studies investigated different exposure time windows. In addition, all of them were conducted in Europe or the USA, where the concentrations of ambient PM10 were lower than in Asia, with mean levels of PM10 ranging from 17.8 µg/m3 to 40.7 µg/m35,22,23 versus 140.8 µg/m3 in the current study (Table 1; and Table A1, available as Supplementary data at IJE online). Table 1 Distributions of selected characteristics of the study population (N = 8877) Characteristics  N  Percent (%)  Maternal age (years)   < 30  5636  63.5   ≥ 30  3239  36.5  Highest education level   < College  3309  37.3   ≥ College  5568  62.7  Family monthly income (RMB per capita)   < 3000  5111  57.6   ≥ 3000  3766  42.4  Employment during pregnancy   No  4101  46.2   Yes  4776  53.8  Smoking (active and passive) during pregnancy   No  7189  81.0   Yes  1688  19.0  Alcohol consumption during pregnancy   No  8857  99.8   Yes  20  0.2  Birthweight (g)   < 2500  438  4.9   2500–4000  7806  87.9   ≥ 4000  633  7.1  Pre-pregnancy BMIa   ≤ 18.5  1822  20.5   18.5–24.0  6118  68.9   ≥ 24.0  937  10.6  Weight gain during pregnancy (kg)   ≤ 15.0  3247  36.6   15–18.5  2089  23.5   > 18.5  3248  36.6  Parity   Primiparous  6669  75.1   Multiparous  2208  24.9  Caesarean section   No  5599  63.1   Yes  3278  36.9  Preeclampsia   No  8673  97.7   Yes  204  2.3  Maternal diabetes   No  8774  98.8   Yes  103  1.2  Season of conceptionb   Autumn  2431  27.4   Winter  2025  22.8   Spring  1854  20.9   Summer  2567  28.9  Cooking fuel   Gas or electricity  7635  86.0   Biomass or coal  1242  14.0  Newborn gender   Boys  4656  52.5   Girls  4221  47.5    Continuous variable  Mean (SD)  Range    Gestational age (weeks)  38.7 (1.7)  22–44  Averaged daily PM10 during pregnancy (µg/m3)  140.8 (24.8)  66.5–242.6  Averaged daily SO2 during pregnancy (µg/m3)  49.0 (13.1)  14.7–87.1  Averaged daily NO2 during pregnancy (µg/m3)  43.9 (6.0)  26.0–64.6  Averaged daily temperature during pregnancy (°C)  10.3 (3.1)  1.7–20.9  Birth AC (mm)  324.0 (25.2)  200–530  Birth HC (mm)  343.6 (19.0)  250–510  Characteristics  N  Percent (%)  Maternal age (years)   < 30  5636  63.5   ≥ 30  3239  36.5  Highest education level   < College  3309  37.3   ≥ College  5568  62.7  Family monthly income (RMB per capita)   < 3000  5111  57.6   ≥ 3000  3766  42.4  Employment during pregnancy   No  4101  46.2   Yes  4776  53.8  Smoking (active and passive) during pregnancy   No  7189  81.0   Yes  1688  19.0  Alcohol consumption during pregnancy   No  8857  99.8   Yes  20  0.2  Birthweight (g)   < 2500  438  4.9   2500–4000  7806  87.9   ≥ 4000  633  7.1  Pre-pregnancy BMIa   ≤ 18.5  1822  20.5   18.5–24.0  6118  68.9   ≥ 24.0  937  10.6  Weight gain during pregnancy (kg)   ≤ 15.0  3247  36.6   15–18.5  2089  23.5   > 18.5  3248  36.6  Parity   Primiparous  6669  75.1   Multiparous  2208  24.9  Caesarean section   No  5599  63.1   Yes  3278  36.9  Preeclampsia   No  8673  97.7   Yes  204  2.3  Maternal diabetes   No  8774  98.8   Yes  103  1.2  Season of conceptionb   Autumn  2431  27.4   Winter  2025  22.8   Spring  1854  20.9   Summer  2567  28.9  Cooking fuel   Gas or electricity  7635  86.0   Biomass or coal  1242  14.0  Newborn gender   Boys  4656  52.5   Girls  4221  47.5    Continuous variable  Mean (SD)  Range    Gestational age (weeks)  38.7 (1.7)  22–44  Averaged daily PM10 during pregnancy (µg/m3)  140.8 (24.8)  66.5–242.6  Averaged daily SO2 during pregnancy (µg/m3)  49.0 (13.1)  14.7–87.1  Averaged daily NO2 during pregnancy (µg/m3)  43.9 (6.0)  26.0–64.6  Averaged daily temperature during pregnancy (°C)  10.3 (3.1)  1.7–20.9  Birth AC (mm)  324.0 (25.2)  200–530  Birth HC (mm)  343.6 (19.0)  250–510  a BMI, body mass index. b Autumn: September, October and November; winter: December, January and February; spring: March, April and May; summer: June, July, and August. Table 1 Distributions of selected characteristics of the study population (N = 8877) Characteristics  N  Percent (%)  Maternal age (years)   < 30  5636  63.5   ≥ 30  3239  36.5  Highest education level   < College  3309  37.3   ≥ College  5568  62.7  Family monthly income (RMB per capita)   < 3000  5111  57.6   ≥ 3000  3766  42.4  Employment during pregnancy   No  4101  46.2   Yes  4776  53.8  Smoking (active and passive) during pregnancy   No  7189  81.0   Yes  1688  19.0  Alcohol consumption during pregnancy   No  8857  99.8   Yes  20  0.2  Birthweight (g)   < 2500  438  4.9   2500–4000  7806  87.9   ≥ 4000  633  7.1  Pre-pregnancy BMIa   ≤ 18.5  1822  20.5   18.5–24.0  6118  68.9   ≥ 24.0  937  10.6  Weight gain during pregnancy (kg)   ≤ 15.0  3247  36.6   15–18.5  2089  23.5   > 18.5  3248  36.6  Parity   Primiparous  6669  75.1   Multiparous  2208  24.9  Caesarean section   No  5599  63.1   Yes  3278  36.9  Preeclampsia   No  8673  97.7   Yes  204  2.3  Maternal diabetes   No  8774  98.8   Yes  103  1.2  Season of conceptionb   Autumn  2431  27.4   Winter  2025  22.8   Spring  1854  20.9   Summer  2567  28.9  Cooking fuel   Gas or electricity  7635  86.0   Biomass or coal  1242  14.0  Newborn gender   Boys  4656  52.5   Girls  4221  47.5    Continuous variable  Mean (SD)  Range    Gestational age (weeks)  38.7 (1.7)  22–44  Averaged daily PM10 during pregnancy (µg/m3)  140.8 (24.8)  66.5–242.6  Averaged daily SO2 during pregnancy (µg/m3)  49.0 (13.1)  14.7–87.1  Averaged daily NO2 during pregnancy (µg/m3)  43.9 (6.0)  26.0–64.6  Averaged daily temperature during pregnancy (°C)  10.3 (3.1)  1.7–20.9  Birth AC (mm)  324.0 (25.2)  200–530  Birth HC (mm)  343.6 (19.0)  250–510  Characteristics  N  Percent (%)  Maternal age (years)   < 30  5636  63.5   ≥ 30  3239  36.5  Highest education level   < College  3309  37.3   ≥ College  5568  62.7  Family monthly income (RMB per capita)   < 3000  5111  57.6   ≥ 3000  3766  42.4  Employment during pregnancy   No  4101  46.2   Yes  4776  53.8  Smoking (active and passive) during pregnancy   No  7189  81.0   Yes  1688  19.0  Alcohol consumption during pregnancy   No  8857  99.8   Yes  20  0.2  Birthweight (g)   < 2500  438  4.9   2500–4000  7806  87.9   ≥ 4000  633  7.1  Pre-pregnancy BMIa   ≤ 18.5  1822  20.5   18.5–24.0  6118  68.9   ≥ 24.0  937  10.6  Weight gain during pregnancy (kg)   ≤ 15.0  3247  36.6   15–18.5  2089  23.5   > 18.5  3248  36.6  Parity   Primiparous  6669  75.1   Multiparous  2208  24.9  Caesarean section   No  5599  63.1   Yes  3278  36.9  Preeclampsia   No  8673  97.7   Yes  204  2.3  Maternal diabetes   No  8774  98.8   Yes  103  1.2  Season of conceptionb   Autumn  2431  27.4   Winter  2025  22.8   Spring  1854  20.9   Summer  2567  28.9  Cooking fuel   Gas or electricity  7635  86.0   Biomass or coal  1242  14.0  Newborn gender   Boys  4656  52.5   Girls  4221  47.5    Continuous variable  Mean (SD)  Range    Gestational age (weeks)  38.7 (1.7)  22–44  Averaged daily PM10 during pregnancy (µg/m3)  140.8 (24.8)  66.5–242.6  Averaged daily SO2 during pregnancy (µg/m3)  49.0 (13.1)  14.7–87.1  Averaged daily NO2 during pregnancy (µg/m3)  43.9 (6.0)  26.0–64.6  Averaged daily temperature during pregnancy (°C)  10.3 (3.1)  1.7–20.9  Birth AC (mm)  324.0 (25.2)  200–530  Birth HC (mm)  343.6 (19.0)  250–510  a BMI, body mass index. b Autumn: September, October and November; winter: December, January and February; spring: March, April and May; summer: June, July, and August. Abnormal fetal growth includes both under-growth and over-growth. There is accumulating evidence that macrosomia, or being large-for-gestational age, as a result of excessive and unbalanced intrauterine growth places the individual at increased risk for metabolic short-term complications and can influence metabolic disorders in adulthood.30,31 It has been suggested that air pollution causes weight gain and obesity through oxidative stress and inflammatory response.32–34 However, to the best of our knowledge, no study has explored the effect of air pollution on fetal over-growth. In light of the limited studies investigating prenatal exposure to PM10 and risk of fetal under-growth, and the lack of studies investigating the association between air pollution and fetal over-growth, we analysed data from the Lanzhou Birth Cohort Study to investigate the hypothesis that exposure to high levels of PM10 during pregnancy increases the risk of abnormal fetal growth. Methods Study population The study population has been described previously.35–39 In brief, pregnant women who came to the Gansu Provincial Maternity and Child Care Hospital (GPMCCH) in Lanzhou, China, for delivery in 2010–12, and who were 18 years or older with a gestational age of ≥ 20 weeks, were eligible. Individuals with mental illness would have limited ability to answer questions. During the study period, a total of 13 pregnant women had a diagnosis of mental illness (including autism, bipolar disorder, depression, obsessive-compulsive disorder) and were excluded from the beginning of the study recruitment. A total of 10 542 (73.4%) women participated in the study. All study procedures were approved by the Human Investigation Committees at the GPMCCH and Yale University. After obtaining written consent, an in-person interview was conducted at the hospital by trained study interviewers using a standardized and structured questionnaire to collect information on demographics, reproductive and medical history, lifestyle factors, occupation and residential history. Information on birth outcomes and maternal complications were abstracted from the medical records. Gestational age was estimated using the date of the last menstrual period reported at recruitment and confirmed by the first ultrasound examination.39 Fetal growth characteristics The repeated ultrasound measurements of HC, AC, femur length (FL) and biparietal diameter (BPD) during pregnancy were abstracted from the medical records. All readings were recorded in millimetres. Since an ultrasound examination was not routinely performed during each of the three trimesters and not all examination records contained four fetal growth parameters, we restricted our analyses to pregnant women who had at least one ultrasound measurement. Of 8969 subjects who had singleton live births and lived in Lanzhou City,36 we excluded participants: who had no records of ultrasound measures; whose records were considered as unreasonable [values below or above 3 standard deviations (SD) from the normal ranges of fetal growth standards, N = 62, 23, 41 and 33 for HC, AC, FL and BPD, respectively];40 or who had babies with birthweights < 500 g (N = 38) or gestational age ≥ 309 days (N = 1); which yielded a final sample size of 8877. Similar distributions of maternal characteristics were observed between pregnant women with and without ultrasound measurement information (data not shown). Across different times during pregnancy, study participants had up to three ultrasound measurements available for FL and BPD, and up to four ultrasound measurements available for HC and AC, because HC and AC were also measured by ultrasound just before birth. The time of ultrasound examination for each fetal growth parameter was presented as week of gestation and was commonly performed after the first trimester (see Figure A1, available as Supplementary data at IJE online). We used the SGPLOT procedure to generate the smooth nonparamatric curves; the fetal growth patterns of BPD, FL and HC during pregnancy against the ultrasound examination time showed a nonlinear growth pattern (see Figure A2, available as Supplementary data at IJE online), which was consistent with the standard fetal growth curves worldwide.40 To avoid results that cannot be compared for each gestational week, the standardized fetal parameters (Z scores) were used as an outcome variable based on recently updated international standards for all four fetal growth measurements.40 The Z scores were calculated through the formula (actual value-mean)SD, where actual value was the observed ultrasound measures, and mean and SD were the standard values in the same weeks of gestation provided by Papageorghiou et al.40 Exposure assessment As described in previous studies,36–38 data on ambient air pollutants were obtained from the Gansu Provincial Environmental Monitoring Central Station, which collects 24-h average concentration for PM10, sulphur dioxide (SO2) and nitrogen dioxide (NO2) through an automated data reporting system from four monitoring stations in Lanzhou city (1088 km²). The 24-h average PM10 was measured for the period 1 April 2009 to 31 December 2012 for two stations, and for 1 January 2011 to 31 December 2012 for the two additional stations. The monitors were located in the southern part of Lanzhou in the metropolitan area with high population density (see Figure A3, available as Supplementary data at IJE online), the spread/density of the study population is similar to that of the total population (see Figure A4, available as Supplementary data at IJE online). Each subject’s residences throughout pregnancy, as measured by move-in and move-out dates, and work addresses were collected. We used the earth online sharing website provided by Google (www.earthol.com) to obtain longitude and latitude coordinates for each subject's home and work addresses. The mean and standard deviation of distance from maternal residence to each air monitor was 14.8(±10.4) km, 9.6(±10.7) km, 9.6(±12.1) km, and 9.4 (±12.4) km, respectively. The majority (93%) of the participants lived within 12 km from the nearest monitor. We calculated daily PM10 concentrations incorporating each subject’s home and work addresses, using all four monitors, or two monitors with the inverse distance weighting (IDW) approach or the nearest monitor. Subjects’ individual exposure levels were averaged: (i) from date of conception to the date of ultrasound examinations; (ii) 1 month before ultrasound examinations; (iii) 2 weeks before ultrasound examinations; and (iv) 1 week before ultrasound examinations. PM10 exposure levels for different exposure windows, estimated based on an average of the four monitors, two monitors or the nearest monitor, were similar. Therefore, we presented our results based on the four monitors. We calculated the correlations between average PM10 levels during the different exposure windows, and found that the PM10 levels averaged during exposure windows (ii), (iii) and (iv) were highly correlated (Pearson correlation coefficients ranging from 0.80 to 0.84), whereas the correlation between (i) and (ii) or (iii) or (iv) only ranged from 0.08 to 0.17. It suggested a temporal heterogeneity between long-term and short-term exposure windows. Statistical analysis All analyses were performed separately for each fetal parameter. The standard cutoffs were used to categorize under-growth (< 3rd centile, Z < −1.88), normal growth (3rd centile–97th centile, −1.88 ≤ Z ≤ 1.88) and over-growth (> 97th centile, Z > 1.88) (dotted horizontal lines in Figure 1). Similar results were obtained using other centiles for setting the normal range, which were the 5th and 95th centiles (Z = ±1.645), and the 10th and 90th centiles (Z = ±1.28) (results not shown). Thus, we presented more conservative results using the 3rd and 97th centiles as the boundaries of the normal range for the four fetal growth measurements. The subject-specific effect due to repeated measures across time during pregnancy was considered using mixed-effects models. Because ultrasound measures were recorded irregularly for each subject (one to four scans), the time intervals between measurements would be more or less unique to each subject. Under these circumstances, a continuous-time model is needed to describe the covariance among the errors; PROC MIXED with the spatial power covariance function (repeated/type = SP(POW) (gestational weeks at ultrasound)) was used for this purpose.41,42 The linear mixed model was fitted to calculate the mean change and the 95% confidence interval (CI) of the Z score of fetal growth measurements which was associated with maternal exposure to PM10. Then the generalized linear mixed model (mixed effects logistic regression) was fitted to calculate the odds ratio (OR) and 95% CI of the associations between maternal exposure to PM10 and risks of under-growth and over-growth. Normal growth was treated as the reference group for each of the analyses. Exposures of each subject to NO2 and SO2 were estimated in the same manner as PM10, and the average levels of SO2 (49.0 µg/m3) and NO2 (43.9 µg/m3) during pregnancy in our study area were lower than the China NAAQS standard (80 µg/m3 and 50 µg/m3),43 indicating these two pollutants were not the main components of air pollution in this city. Therefore, models were corrected for SO2, NO2 and daily temperature by using the averages as the continuous variables over the same exposure windows as for PM10. Exposure to PM10 was analysed as a continuous variable and, in a separate model, as a binary variable meeting or exceeding the US 24-h, health-based National Ambient Air Quality Standard (NAAQS, 150 µg/m3, equivalent to the China NAAQS Grade II level).43,44 Figure 1 View largeDownload slide The Scatter Plots of Associations Between Standardized Fetal Measurements (Z Scores) and PM10 Exposure Levels (μg/m3) During Pregnancy. Horizontal refernce lines at 3rd, 50th, and 97th centiles showed the cutoffs of undergrowth, normal growth, and overgrowth for fetal biparietal diameter (A), fetal femur length (B), fetal abdominal circumference (C), and fetal head circumference (D) measured by ultrasound according to gestational age. Dark black lines are the smooth nonparametric curves. Opens gray dots showed actual observations. Figure 1 View largeDownload slide The Scatter Plots of Associations Between Standardized Fetal Measurements (Z Scores) and PM10 Exposure Levels (μg/m3) During Pregnancy. Horizontal refernce lines at 3rd, 50th, and 97th centiles showed the cutoffs of undergrowth, normal growth, and overgrowth for fetal biparietal diameter (A), fetal femur length (B), fetal abdominal circumference (C), and fetal head circumference (D) measured by ultrasound according to gestational age. Dark black lines are the smooth nonparametric curves. Opens gray dots showed actual observations. All models were first adjusted for known determinants of fetal growth characteristics as shown in Table 1. For the final model selection, the Akaike information criterion was used with forward-selection procedure on covariates considered to be determinants of growth.45 We checked for the multi-collinearity of all variables in the model, especially for the effects of multiple pollutants including SO2 and NO2 exposures, as well as the daily temperature. The variation inflation factors varied from 1.02 to 2.57, demonstrating a lack of multi-collinearity. The same exposure windows were used for SO2, NO2 and temperature as for PM10 exposure in corresponding models. All analyses were performed using SAS software, version 9.4 (SAS Institute Inc., Cary, NC). Results There were 8770, 8537, 7733 and 8009 participants with ultrasound measurements of HC, AC, FL and BPD, respectively. The study included a total of 14 560, 18 583, 12 383 and 13 619 scans during the entire pregnancy for ultrasound measures of HC, AC, FL and BPD, respectively. According to the international standard fetal growth measurement cutoffs set at the 3rd and 97th centiles, for the standardized fetal growth measurements (Z scores): 515 HC (3.5%), 1316 AC (7.1%), 208 FL (1.7%) and 557 BPD (4.1%) scans were catergorized as under-growth; 11 012 HC (75.6%), 17 007 AC (91.5%), 9395 FL (75.9%) and 12 266 BPD (90.1%) scans were catergorized as normal growth; and 3033 HC (20.8%), 1576 AC (8.4%), 2779 FL (22.4%) and 796 BPD (5.8%) scans were catergorized as over-growth. Table 1 shows the distributions of selected characteristics of the study population. More than half of the women were younger than 30 years (63.5%), nulliparous (75.1%) and non-smokers during pregnancy (81.0%), and had higher education levels (62.7%) and normal pre-pregnancy body mass index (BMI) (68.9%). Table 2 presents the associations between PM10 exposures averaged by different exposure windows, and ultrasound measurement Z-scores. When average PM10 exposure from conception until the ultrasound examination exceeded 150 µg/m3, there were significant increases in standardized FL (β = 0.095, P = 0.0012) and HC (β = 0.090, P = 0.0078) measures, compared with PM10 < 150 µg/m3. When we treated the PM10 exposure as continuous variable, we found a significant decrease in BPD (β = −0.018, P = 0.0016) as per 10 µg/m3 increase in PM10. Table 2 Association between averaged PM10 exposure during the different exposure windows and the mean change in standardized fetal ultrasound measurements (Z scores) Exposure windows  BPD (n = 8009)   FL (n = 7733)   AC (n = 8537)   HC (n = 8770)   βa  95% CI  βa  95% CI  βa  95% CI  βa  95% CI  Mean change comparing exposures ≥ 150 µg/m3 PM10 with those < 150 μg/m3  From conception to ultrasound  0.016  −0.035, 0.067  0.095**  0.038, 0.153  0.001  −0.046, 0.049  0.090**  0.024, 0.156  One month before ultrasound  −0.052  −0.097, 0.007  −0.007  −0.051, 0.038  −0.044  −0.082, 0.003  0.092*  0.020, 0.151  Two weeks before ultrasound  −0.037  −0.083, 0.008  −0.043  −0.089, 0.002  −0.037  −0.076, 0.002  0.020  −0.040, 0.080  One week before ultrasound  −0.034  −0.079, 0.011  −0.005  −0.050, 0.040  -0.025  −0.064, 0.014  0.070*  0.010, 0.130  Mean change per 10 µg/m3 increase in PM10  From conception to ultrasound  −0.018**  −0.030, -0.007  0.008  −0.006, 0.022  0.005  −0.016, 0.006  0.011  −0.006, 0.029  One month before ultrasound  −0.004  −0.008, 0.000  −0.001  −0.005, 0.003  0.002  −0.007, 0.004  0.005*  0.000, 0.009  Two weeks before ultrasound  −0.002  −0.006, 0.002  0.002  −0.006, 0.003  0.002  −0.006, 0.001  0.008*  0.001, 0.011  One week before ultrasound  −0.002  −0.006, 0.002  0.000  −0.004, 0.004  −0.002  −0.005, 0.002  0.004*  0.000, 0.009  Exposure windows  BPD (n = 8009)   FL (n = 7733)   AC (n = 8537)   HC (n = 8770)   βa  95% CI  βa  95% CI  βa  95% CI  βa  95% CI  Mean change comparing exposures ≥ 150 µg/m3 PM10 with those < 150 μg/m3  From conception to ultrasound  0.016  −0.035, 0.067  0.095**  0.038, 0.153  0.001  −0.046, 0.049  0.090**  0.024, 0.156  One month before ultrasound  −0.052  −0.097, 0.007  −0.007  −0.051, 0.038  −0.044  −0.082, 0.003  0.092*  0.020, 0.151  Two weeks before ultrasound  −0.037  −0.083, 0.008  −0.043  −0.089, 0.002  −0.037  −0.076, 0.002  0.020  −0.040, 0.080  One week before ultrasound  −0.034  −0.079, 0.011  −0.005  −0.050, 0.040  -0.025  −0.064, 0.014  0.070*  0.010, 0.130  Mean change per 10 µg/m3 increase in PM10  From conception to ultrasound  −0.018**  −0.030, -0.007  0.008  −0.006, 0.022  0.005  −0.016, 0.006  0.011  −0.006, 0.029  One month before ultrasound  −0.004  −0.008, 0.000  −0.001  −0.005, 0.003  0.002  −0.007, 0.004  0.005*  0.000, 0.009  Two weeks before ultrasound  −0.002  −0.006, 0.002  0.002  −0.006, 0.003  0.002  −0.006, 0.001  0.008*  0.001, 0.011  One week before ultrasound  −0.002  −0.006, 0.002  0.000  −0.004, 0.004  −0.002  −0.005, 0.002  0.004*  0.000, 0.009  a Adjusted for maternal age (< 30, ≥ 30 years), education (< college, ≥ college), family monthly income per capita (< 3000, ≥ 3000 RMB), smoking during pregnancy (yes, no), pre-pregnancy BMI (≤ 18.5, 18.5–24.0, ≥ 24.0), weight gain du.ring pregnancy (≤ 15 kg, 15–18.5 kg, > 18.5 kg), parity (primiparous, multiparous), season of conception (spring/summer, autumn/winter), cooking fuel (gas/electricity, coal/biomass), estimated SO2, NO2 by different corresponding exposure windows, and averaged temperature during different corresponding exposure windows. * P < 0.05; ** P < 0.01. Table 2 Association between averaged PM10 exposure during the different exposure windows and the mean change in standardized fetal ultrasound measurements (Z scores) Exposure windows  BPD (n = 8009)   FL (n = 7733)   AC (n = 8537)   HC (n = 8770)   βa  95% CI  βa  95% CI  βa  95% CI  βa  95% CI  Mean change comparing exposures ≥ 150 µg/m3 PM10 with those < 150 μg/m3  From conception to ultrasound  0.016  −0.035, 0.067  0.095**  0.038, 0.153  0.001  −0.046, 0.049  0.090**  0.024, 0.156  One month before ultrasound  −0.052  −0.097, 0.007  −0.007  −0.051, 0.038  −0.044  −0.082, 0.003  0.092*  0.020, 0.151  Two weeks before ultrasound  −0.037  −0.083, 0.008  −0.043  −0.089, 0.002  −0.037  −0.076, 0.002  0.020  −0.040, 0.080  One week before ultrasound  −0.034  −0.079, 0.011  −0.005  −0.050, 0.040  -0.025  −0.064, 0.014  0.070*  0.010, 0.130  Mean change per 10 µg/m3 increase in PM10  From conception to ultrasound  −0.018**  −0.030, -0.007  0.008  −0.006, 0.022  0.005  −0.016, 0.006  0.011  −0.006, 0.029  One month before ultrasound  −0.004  −0.008, 0.000  −0.001  −0.005, 0.003  0.002  −0.007, 0.004  0.005*  0.000, 0.009  Two weeks before ultrasound  −0.002  −0.006, 0.002  0.002  −0.006, 0.003  0.002  −0.006, 0.001  0.008*  0.001, 0.011  One week before ultrasound  −0.002  −0.006, 0.002  0.000  −0.004, 0.004  −0.002  −0.005, 0.002  0.004*  0.000, 0.009  Exposure windows  BPD (n = 8009)   FL (n = 7733)   AC (n = 8537)   HC (n = 8770)   βa  95% CI  βa  95% CI  βa  95% CI  βa  95% CI  Mean change comparing exposures ≥ 150 µg/m3 PM10 with those < 150 μg/m3  From conception to ultrasound  0.016  −0.035, 0.067  0.095**  0.038, 0.153  0.001  −0.046, 0.049  0.090**  0.024, 0.156  One month before ultrasound  −0.052  −0.097, 0.007  −0.007  −0.051, 0.038  −0.044  −0.082, 0.003  0.092*  0.020, 0.151  Two weeks before ultrasound  −0.037  −0.083, 0.008  −0.043  −0.089, 0.002  −0.037  −0.076, 0.002  0.020  −0.040, 0.080  One week before ultrasound  −0.034  −0.079, 0.011  −0.005  −0.050, 0.040  -0.025  −0.064, 0.014  0.070*  0.010, 0.130  Mean change per 10 µg/m3 increase in PM10  From conception to ultrasound  −0.018**  −0.030, -0.007  0.008  −0.006, 0.022  0.005  −0.016, 0.006  0.011  −0.006, 0.029  One month before ultrasound  −0.004  −0.008, 0.000  −0.001  −0.005, 0.003  0.002  −0.007, 0.004  0.005*  0.000, 0.009  Two weeks before ultrasound  −0.002  −0.006, 0.002  0.002  −0.006, 0.003  0.002  −0.006, 0.001  0.008*  0.001, 0.011  One week before ultrasound  −0.002  −0.006, 0.002  0.000  −0.004, 0.004  −0.002  −0.005, 0.002  0.004*  0.000, 0.009  a Adjusted for maternal age (< 30, ≥ 30 years), education (< college, ≥ college), family monthly income per capita (< 3000, ≥ 3000 RMB), smoking during pregnancy (yes, no), pre-pregnancy BMI (≤ 18.5, 18.5–24.0, ≥ 24.0), weight gain du.ring pregnancy (≤ 15 kg, 15–18.5 kg, > 18.5 kg), parity (primiparous, multiparous), season of conception (spring/summer, autumn/winter), cooking fuel (gas/electricity, coal/biomass), estimated SO2, NO2 by different corresponding exposure windows, and averaged temperature during different corresponding exposure windows. * P < 0.05; ** P < 0.01. Table 3 shows the associations between PM10 exposures averaged by different exposure windows, and risks of under-growth, and over-growth. When PM10 exposure was estimated from conception until the ultrasound, compared with PM10 < 150 µ/m3, PM10 ≥ 150 µg/m3 was associated with an increased risk of overgrowth [odds ratio (OR), 1.22; 95% CI, 1.04, 1.42 for FL, and 1.19; 1.07, 1.33 for HC]. PM10 was associated with a reduced risk of BPD overgrowth when PM10 was treated as < 150 µg/m3 versus ≥ 150 µg/m3 or a continuous variable during 1 month or 2 weeks before ultrasound examinations. Table 3 Association between averaged PM10 during the different exposure windows and risks of fetal under-growth and over-growth Exposure windows  BPD (n = 8009)   FL (n = 7733)   AC (n = 8537)   HC (n = 8770)   ORa  95% CI  ORa  95% CI  ORa  95% CI  ORa  95% CI  Number of pregnancies in each model  Normal growth  7005  5662  6989  6378   Under-growth  438  189  393  367   Over-growth  566  1882  1155  1665  Risk comparing exposures ≥ 150 µg/m3 PM10 with those < 150 μg/m3  From conception to ultrasound   Under-growth  1.13  0.89, 1.43  1.04  0.72, 1.50  0.86  0.73, 1.01  1.02  0.81, 1.30   Overg-rowth  1.00  0.81, 1.23  1.22*  1.04, 1.42  1.03  0.87, 1.21  1.19**  1.07, 1.33  One month before ultrasound   Under-growth  1.04  0.85, 1.26  0.87  0.62, 1.22  1.08  0.84, 1.39  0.98  0.78, 1.23   Over-growth  0.83*  0.69, 0.99  0.97  0.87, 1.08  0.95  0.83, 1.09  1.08*  1.00, 1.19  Two weeks before ultrasound   Under-growth  0.80  0.64, 1.01  1.08  0.78, 1.50  1.09  0.94, 1.26  0.97  0.76, 1.24   Over-growth  0.79**  0.66, 0.94  0.91  0.81, 1.02  0.91  0.79, 1.04  1.03  0.93, 1.14  One week before ultrasound   Under-growth  0.89  0.72, 1.09  0.86  0.61, 1.20  1.09  0.94, 1.26  0.99  0.77, 1.26   Over-growth  0.87  0.73, 1.04  0.96  0.86, 1.07  0.97  0.85, 1.11  1.11*  1.00, 1.23  Risk per 10 µg/m3 increase in PM10  From conception to ultrasound   Under-growth  1.03  0.98, 1.07  0.99  0.92, 1.07  0.98  0.93, 1.02  0.99  0.92, 1.06   Over-growth  0.98  0.94, 1.02  1.01  0.98, 1.04  1.01  0.98, 1.05  1.02  0.99, 1.05  One month before ultrasound   Under-growth  1.00  0.98, 1.02  0.99  0.96, 1.02  1.00  0.99, 1.02  1.00  0.98, 1.02   Over-growth  0.98*  0.97, 1.00  1.00  0.99, 1.01  0.99  0.98, 1.01  1.01*  1.00, 1.02  Two weeks before ultrasound   Under-growth  0.99  0.97, 1.01  1.00  0.97, 1.03  1.00  0.99, 1.02  1.00  0.97, 1.02   Over-growth  0.98**  0.96, 0.99  0.99  0.98, 1.01  1.00  0.99, 1.01  1.01*  1.00, 1.02  One week before ultrasound   Under-growth  0.99  0.97, 1.01  0.99  0.96, 1.02  1.00  0.99, 1.01  1.00  0.97, 1.02   Over-growth  0.99  0.97, 1.01  1.00  0.99, 1.01  1.00  0.98, 1.01  1.01*  1.00, 1.02  Exposure windows  BPD (n = 8009)   FL (n = 7733)   AC (n = 8537)   HC (n = 8770)   ORa  95% CI  ORa  95% CI  ORa  95% CI  ORa  95% CI  Number of pregnancies in each model  Normal growth  7005  5662  6989  6378   Under-growth  438  189  393  367   Over-growth  566  1882  1155  1665  Risk comparing exposures ≥ 150 µg/m3 PM10 with those < 150 μg/m3  From conception to ultrasound   Under-growth  1.13  0.89, 1.43  1.04  0.72, 1.50  0.86  0.73, 1.01  1.02  0.81, 1.30   Overg-rowth  1.00  0.81, 1.23  1.22*  1.04, 1.42  1.03  0.87, 1.21  1.19**  1.07, 1.33  One month before ultrasound   Under-growth  1.04  0.85, 1.26  0.87  0.62, 1.22  1.08  0.84, 1.39  0.98  0.78, 1.23   Over-growth  0.83*  0.69, 0.99  0.97  0.87, 1.08  0.95  0.83, 1.09  1.08*  1.00, 1.19  Two weeks before ultrasound   Under-growth  0.80  0.64, 1.01  1.08  0.78, 1.50  1.09  0.94, 1.26  0.97  0.76, 1.24   Over-growth  0.79**  0.66, 0.94  0.91  0.81, 1.02  0.91  0.79, 1.04  1.03  0.93, 1.14  One week before ultrasound   Under-growth  0.89  0.72, 1.09  0.86  0.61, 1.20  1.09  0.94, 1.26  0.99  0.77, 1.26   Over-growth  0.87  0.73, 1.04  0.96  0.86, 1.07  0.97  0.85, 1.11  1.11*  1.00, 1.23  Risk per 10 µg/m3 increase in PM10  From conception to ultrasound   Under-growth  1.03  0.98, 1.07  0.99  0.92, 1.07  0.98  0.93, 1.02  0.99  0.92, 1.06   Over-growth  0.98  0.94, 1.02  1.01  0.98, 1.04  1.01  0.98, 1.05  1.02  0.99, 1.05  One month before ultrasound   Under-growth  1.00  0.98, 1.02  0.99  0.96, 1.02  1.00  0.99, 1.02  1.00  0.98, 1.02   Over-growth  0.98*  0.97, 1.00  1.00  0.99, 1.01  0.99  0.98, 1.01  1.01*  1.00, 1.02  Two weeks before ultrasound   Under-growth  0.99  0.97, 1.01  1.00  0.97, 1.03  1.00  0.99, 1.02  1.00  0.97, 1.02   Over-growth  0.98**  0.96, 0.99  0.99  0.98, 1.01  1.00  0.99, 1.01  1.01*  1.00, 1.02  One week before ultrasound   Under-growth  0.99  0.97, 1.01  0.99  0.96, 1.02  1.00  0.99, 1.01  1.00  0.97, 1.02   Over-growth  0.99  0.97, 1.01  1.00  0.99, 1.01  1.00  0.98, 1.01  1.01*  1.00, 1.02  a Adjusted for maternal age (< 30, 30 years), education (< college, college), family monthly income per capita (< 3000, ≥ 3000 RMB), smoking during pregnancy (yes, no), pre-pregnancy BMI (≤ 18.5, 18.5–24.0, ≥ 24.0), weight gain during pregnancy (≤ 15 kg, 15–18.5 kg, > 18.5 kg), parity (primiparous, multiparous), season of conception (spring/summer, autumn/winter), cooking fuel (gas/electricity, coal/biomass), estimated SO2, NO2 by different corresponding exposure windows, and averaged temperature during different corresponding exposure windows. * P ≤ 0.05; ** P < 0.01. Table 3 Association between averaged PM10 during the different exposure windows and risks of fetal under-growth and over-growth Exposure windows  BPD (n = 8009)   FL (n = 7733)   AC (n = 8537)   HC (n = 8770)   ORa  95% CI  ORa  95% CI  ORa  95% CI  ORa  95% CI  Number of pregnancies in each model  Normal growth  7005  5662  6989  6378   Under-growth  438  189  393  367   Over-growth  566  1882  1155  1665  Risk comparing exposures ≥ 150 µg/m3 PM10 with those < 150 μg/m3  From conception to ultrasound   Under-growth  1.13  0.89, 1.43  1.04  0.72, 1.50  0.86  0.73, 1.01  1.02  0.81, 1.30   Overg-rowth  1.00  0.81, 1.23  1.22*  1.04, 1.42  1.03  0.87, 1.21  1.19**  1.07, 1.33  One month before ultrasound   Under-growth  1.04  0.85, 1.26  0.87  0.62, 1.22  1.08  0.84, 1.39  0.98  0.78, 1.23   Over-growth  0.83*  0.69, 0.99  0.97  0.87, 1.08  0.95  0.83, 1.09  1.08*  1.00, 1.19  Two weeks before ultrasound   Under-growth  0.80  0.64, 1.01  1.08  0.78, 1.50  1.09  0.94, 1.26  0.97  0.76, 1.24   Over-growth  0.79**  0.66, 0.94  0.91  0.81, 1.02  0.91  0.79, 1.04  1.03  0.93, 1.14  One week before ultrasound   Under-growth  0.89  0.72, 1.09  0.86  0.61, 1.20  1.09  0.94, 1.26  0.99  0.77, 1.26   Over-growth  0.87  0.73, 1.04  0.96  0.86, 1.07  0.97  0.85, 1.11  1.11*  1.00, 1.23  Risk per 10 µg/m3 increase in PM10  From conception to ultrasound   Under-growth  1.03  0.98, 1.07  0.99  0.92, 1.07  0.98  0.93, 1.02  0.99  0.92, 1.06   Over-growth  0.98  0.94, 1.02  1.01  0.98, 1.04  1.01  0.98, 1.05  1.02  0.99, 1.05  One month before ultrasound   Under-growth  1.00  0.98, 1.02  0.99  0.96, 1.02  1.00  0.99, 1.02  1.00  0.98, 1.02   Over-growth  0.98*  0.97, 1.00  1.00  0.99, 1.01  0.99  0.98, 1.01  1.01*  1.00, 1.02  Two weeks before ultrasound   Under-growth  0.99  0.97, 1.01  1.00  0.97, 1.03  1.00  0.99, 1.02  1.00  0.97, 1.02   Over-growth  0.98**  0.96, 0.99  0.99  0.98, 1.01  1.00  0.99, 1.01  1.01*  1.00, 1.02  One week before ultrasound   Under-growth  0.99  0.97, 1.01  0.99  0.96, 1.02  1.00  0.99, 1.01  1.00  0.97, 1.02   Over-growth  0.99  0.97, 1.01  1.00  0.99, 1.01  1.00  0.98, 1.01  1.01*  1.00, 1.02  Exposure windows  BPD (n = 8009)   FL (n = 7733)   AC (n = 8537)   HC (n = 8770)   ORa  95% CI  ORa  95% CI  ORa  95% CI  ORa  95% CI  Number of pregnancies in each model  Normal growth  7005  5662  6989  6378   Under-growth  438  189  393  367   Over-growth  566  1882  1155  1665  Risk comparing exposures ≥ 150 µg/m3 PM10 with those < 150 μg/m3  From conception to ultrasound   Under-growth  1.13  0.89, 1.43  1.04  0.72, 1.50  0.86  0.73, 1.01  1.02  0.81, 1.30   Overg-rowth  1.00  0.81, 1.23  1.22*  1.04, 1.42  1.03  0.87, 1.21  1.19**  1.07, 1.33  One month before ultrasound   Under-growth  1.04  0.85, 1.26  0.87  0.62, 1.22  1.08  0.84, 1.39  0.98  0.78, 1.23   Over-growth  0.83*  0.69, 0.99  0.97  0.87, 1.08  0.95  0.83, 1.09  1.08*  1.00, 1.19  Two weeks before ultrasound   Under-growth  0.80  0.64, 1.01  1.08  0.78, 1.50  1.09  0.94, 1.26  0.97  0.76, 1.24   Over-growth  0.79**  0.66, 0.94  0.91  0.81, 1.02  0.91  0.79, 1.04  1.03  0.93, 1.14  One week before ultrasound   Under-growth  0.89  0.72, 1.09  0.86  0.61, 1.20  1.09  0.94, 1.26  0.99  0.77, 1.26   Over-growth  0.87  0.73, 1.04  0.96  0.86, 1.07  0.97  0.85, 1.11  1.11*  1.00, 1.23  Risk per 10 µg/m3 increase in PM10  From conception to ultrasound   Under-growth  1.03  0.98, 1.07  0.99  0.92, 1.07  0.98  0.93, 1.02  0.99  0.92, 1.06   Over-growth  0.98  0.94, 1.02  1.01  0.98, 1.04  1.01  0.98, 1.05  1.02  0.99, 1.05  One month before ultrasound   Under-growth  1.00  0.98, 1.02  0.99  0.96, 1.02  1.00  0.99, 1.02  1.00  0.98, 1.02   Over-growth  0.98*  0.97, 1.00  1.00  0.99, 1.01  0.99  0.98, 1.01  1.01*  1.00, 1.02  Two weeks before ultrasound   Under-growth  0.99  0.97, 1.01  1.00  0.97, 1.03  1.00  0.99, 1.02  1.00  0.97, 1.02   Over-growth  0.98**  0.96, 0.99  0.99  0.98, 1.01  1.00  0.99, 1.01  1.01*  1.00, 1.02  One week before ultrasound   Under-growth  0.99  0.97, 1.01  0.99  0.96, 1.02  1.00  0.99, 1.01  1.00  0.97, 1.02   Over-growth  0.99  0.97, 1.01  1.00  0.99, 1.01  1.00  0.98, 1.01  1.01*  1.00, 1.02  a Adjusted for maternal age (< 30, 30 years), education (< college, college), family monthly income per capita (< 3000, ≥ 3000 RMB), smoking during pregnancy (yes, no), pre-pregnancy BMI (≤ 18.5, 18.5–24.0, ≥ 24.0), weight gain during pregnancy (≤ 15 kg, 15–18.5 kg, > 18.5 kg), parity (primiparous, multiparous), season of conception (spring/summer, autumn/winter), cooking fuel (gas/electricity, coal/biomass), estimated SO2, NO2 by different corresponding exposure windows, and averaged temperature during different corresponding exposure windows. * P ≤ 0.05; ** P < 0.01. Compared with the results of cumulative exposure from conception to the ultrasound examination, similar positive associations were only observed for HC measures for other exposure windows, 1 month before ultrasound examinations and 1 or 2 weeks before ultrasound examinations, when PM10 was treated as < 150 µg/m3 versus ≥ 150 µg/m3 or a continuous variable (Tables 2 and 3). Discussion Our results support the hypothesis that prenatal exposure to high levels of ambient PM10 is associated with increased risk of fetal under- and over-growth. Our study also suggests a stronger and consistent association of over-growth for HC over the different exposure windows compared with other fetal growth parameters. To the best of our knowledge, our study was the first one to examine the associations between prenatal exposure to PM10 and risk of fetal over-growth. We found that exposure to higher levels of PM10 increased the risk of fetal HC over-growth. In early studies conducted in Australia, Europe and the USA, Hansen et al.5 found reductions of FL, AC and HC associated with increased exposure to PM10 during gestational days 91–120, and a reduction of FL measurement during 0–30 days of gestation; van den Hooven et al.22 observed reduction in HC in the third trimester; and Ritz et al.23 reported that no association was observed between PM10 exposure and HC, FL and AC. Although van den Hooven et al. did not highlight the findings that there were significant increases in HC measures during the second trimester in the second and third quartiles of PM10 exposure, and that there were significant increases in fetal weight in higher quartiles (second, third and fourth) of PM10 exposure, these findings were similar to ours. Although the underlying mechanisms are currently unclear, a growing body of literature suggests that increased oxidative stress and inflammatory response to PMs can influence energy expenditure, resulting in increased dietary intake.32,33 There was an increasing awareness that changes in placental metabolism and nutrient transport capacity might contribute to altered fetal growth.46 Human studies showed that higher cord blood leptin levels caused by greater weight gain were associated with increased fetal size,47 and altered expression of maternal insulin-like growth factor due to excessive gestational weight gain contributed to increased fetal growth through increased placental nutrient transporter expression.48,49 It is possible that the positive relationship between ambient air pollutants and fetal over-growth was mediated by maternal BMI/weight gain. In our study, pregnant women carrying an overgrown fetus were more likely to be in the highest category of maternal weight gain and to be pre-pregnancy obese. Comparing the study results among the published studies and ours was challenging. The outcome measures varied across different studies, including differences in ultrasound measurement methods and potential of inter-sonographer variations. The exposure assessment also varied across the studies. van den Hooven et al.22 performed dispersion modelling based on home addresses only, though allowing for residential mobility; Hansen et al.5 estimated monthly exposure to nearest monitors using IDW over the first 4 months and restricted analysis through four exposure windows (0–30, 31–60, 61–90, 91–120 days of gestation). Ritz et al.23 considered a 2-weeks exposure window around each measurement time (at 19.1, 28.8 and 36.7 weeks of gestation). In addition, all these published studies were conducted in low-air-pollution areas as compared with our study. The compositions of PM10 in different studies might also be different and could cause a different impact on fetal growth. Our study included large sample size (N = 8877) with a relatively large number of ultrasound scans for each fetal parameter (N ∼18 583), compared with previous studies.5,22,23 It allowed exploration of association by subgroups based on the growth status. Our study assessed exposure based on both home and work addresses with consideration of residential mobility, as well as PM10 exposure through different exposure windows, which minimized potential exposure misclassification and ability to explore critical exposure time windows on fetal growth. Detailed information on potential confounding factors such as smoking, cooking fuels and other co-exposure air pollutants were collected and included in our analyses. All participants were recruited from the same hospital and the information on ultrasound measurements was obtained from medical records, so potential misclassifications of fetal measurements were minimized. Additionally, with a potential U-shape relationship between air pollution and fetal growth, we employed the latest international fetal growth standard to examine the association with both under-growth and over-growth, in addition to treating growth measurements as continuous variables as done by earlier studies.5,22,23 This potential heterogeneity of PM10 exposure effects on fetal growth measurements still calls for future study with larger sample size, to replicate and confirm our findings. Limitations should be considered when interpreting the study results. The accuracy of exposure estimation might be limited by the lack of monitors in rural areas in Lanzhou city. However, more than 90% of the women lived within 5.5 km of a monitor. Sensitivity analyses using the data from different exposure assessment approaches (nearest monitors, IDW using two or four monitors, using data for the women living within 5.5 km or 12.9 km of a monitor) showed consistent results. PM10 exposure was estimated using data from government monitors which, though the most commonly used method in air pollution epidemiology, may not represent the actual individual exposure level which would include differences in indoor/outdoor activity patterns, and may not consider intra-urban gradients in pollutants. However, measurements through personal air monitors in this setting (pregnancy, in a large population) may be impractical as well as expensive, and numerous studies have used the above monitoring approach successfully. To evaluate whether the four monitors provided reasonable spatial coverage in our study, we calculated the coefficient of divergence, which provides the diversity between concentrations at sampling site-pairs.50 We found that coefficients of divergence for all site-pairs were lower than 0.20, indicating low spatial heterogeneity of PM10 in our study area. However, future studies might consider additional monitor coverage, in order to address the issues of spatial heterogeneity and improve estimates of exposure. Although the exact sources of PM10 in Lanzhou city are still unclear, a factory of the China Petroleum and Chemical Corporation (Sinopec Limited) has been located in the west, an upwind location of this city. Future work may investigate PM sources, which have different chemical structures and possibly different health impacts. The measurement errors with respect to the outcomes would be another concern both in previous large epidemiological studies and in the current study, because it is possible that the ultrasound examinations at different gestational weeks were performed by different clinicians/technicians. However, GPMCCH is the largest maternity and child care hospital in Lanzhou, China. The hospital performs internal quality control in order to provide high-quality ultrasound examinations, and the historical records of reproducibility also demonstrated that the measurement errors would be likely minimized. In conclusion, our study supports the hypothesis that prenatal exposure to high levels of ambient PM10 increases the risk of abnormal fetal growth, and stronger and consistent associations were mainly seen for HC over-growth by different exposure windows compared with other fetal growth measurements. This study suggests that the associations between maternal exposure to PM10 and fetal growth may vary by different fetal growth measurements (BPD, FL, AC and HC) and fetal growth patterns (under-/over-growth). Future multi-site studies with different types of air pollutants are needed to fully understand the impacts of air pollution combinations on fetal growth. Our study also calls for future studies to explore the underlying mechanisms and post-natal consequences of these findings Supplementary Data Supplementary data are available at IJE online. Funding The work was supported by internal funding from the Gansu Provincial Maternity and Child Care Hospital, Gansu Provincial Science and Technology Department Grant (1204WCGA021), National Institutes of Health Grants (K02HD70324, R01ES019587) and the U.S. Environmental Protection Agency (EPA-G2014-STAR-J1). Author Contributions J.Q., Q.L. and Y.Z. designed the research; X.L., Z.T. and H.Z. conducted the birth cohort study; Y.Z. collected air pollution and meteorology data; N.Z. and H.H. managed the study and cleaned the database; N.Z., S.M., Y.H. and N.M. performed statistical analysis; N.Z. constructed the figures; N.Z., M.B. and Y.Z. wrote the first draft, and all authors contributed to the final draft and approved the manuscript. Conflict of interest: None declared. References 1 Lawn JE, Blencowe H, Pattinson R et al.   Stillbirths: Where? When? Why? How to make the data count? Lancet  2011; 377: 1448– 63. Google Scholar CrossRef Search ADS PubMed  2 Liu L, Johnson HL, Cousens S et al.   Global, regional, and national causes of child mortality: an updated systematic analysis for 2010 with time trends since 2000. 2012; 379: 2151– 61. 3 World Health Organization. A Global Review of the Key Interventions Related to Reproductive, Maternal, Newborn and Child Health (RMNCH) . Geneva: WHO, 2011. 4 Ouyang F, Zhang J, Betran AP, Yang Z, Souza JP, Merialdi M. Recurrence of adverse perinatal outcomes in developing countries. Bull World Health Organ  2013; 91: 357– 67. Google Scholar CrossRef Search ADS PubMed  5 Hansen CA, Barnett AG, Pritchard G. The effect of ambient air pollution during early pregnancy on fetal ultrasonic measurements during mid-pregnancy. Environ Health Perspect  2008; 116: 362– 69. Google Scholar CrossRef Search ADS PubMed  6 Glinianaia SV, Rankin J, Bell R, Pless-Mulloli T, Howel D. Particulate air pollution and fetal health: a systematic review of the epidemiologic evidence. Epidemiology  2004; 15: 36– 45. Google Scholar CrossRef Search ADS PubMed  7 Shah PS, Balkhair T. Knowledge Synthesis Group on Determinants of Preterm LBWb. Air pollution and birth outcomes: a systematic review. Environ Int  2011; 37: 498– 516. Google Scholar CrossRef Search ADS PubMed  8 Woodruff TJ, Parker JD, Darrow LA et al.   Methodological issues in studies of air pollution and reproductive health. Environ Res  2009; 109: 311– 20. Google Scholar CrossRef Search ADS PubMed  9 Choi H, Jedrychowski W, Spengler J et al.   International studies of prenatal exposure to polycyclic aromatic hydrocarbons and fetal growth. Environ Health Perspect  2006; 114: 1744– 50. Google Scholar PubMed  10 Hansen C, Neller A, Williams G, Simpson R. Low levels of ambient air pollution during pregnancy and fetal growth among term neonates in Brisbane, Australia. Environ Res  2007; 103: 383– 89. Google Scholar CrossRef Search ADS PubMed  11 Jedrychowski W, Bendkowska I, Flak E et al.   Estimated risk for altered fetal growth resulting from exposure to fine particles during pregnancy: an epidemiologic prospective cohort study in Poland. Environ Health Perspect  2004; 112: 1398– 402. Google Scholar CrossRef Search ADS PubMed  12 Salam MT, Millstein J, Li YF, Lurmann FW, Margolis HG, Gilliland FD. Birth outcomes and prenatal exposure to ozone, carbon monoxide, and particulate matter: results from the Children's Health Study. Environ Health Perspect  2005; 113: 1638– 44. Google Scholar CrossRef Search ADS PubMed  13 Hannam K, McNamee R, Baker P, Sibley C, Agius R. Air pollution exposure and adverse pregnancy outcomes in a large UK birth cohort: use of a novel spatio-temporal modelling technique. Scand J Work Environ Health  2014; 40: 518– 30. Google Scholar CrossRef Search ADS PubMed  14 Winckelmans E, Cox B, Martens E, Fierens F, Nemery B, Nawrot TS. Fetal growth and maternal exposure to particulate air pollution - More marked effects at lower exposure and modification by gestational duration. Environ Res  2015; 140: 611– 18. Google Scholar CrossRef Search ADS PubMed  15 Fleisch AF, Rifas-Shiman SL, Koutrakis P et al.   Prenatal exposure to traffic pollution: associations with reduced fetal growth and rapid infant weight gain. Epidemiology  2015; 26: 43– 50. Google Scholar CrossRef Search ADS PubMed  16 Liu S, Krewski D, Shi Y, Chen Y, Burnett RT. Association between maternal exposure to ambient air pollutants during pregnancy and fetal growth restriction. J Expo Sci Environ Epidemiol  2007; 17: 426– 32. Google Scholar CrossRef Search ADS PubMed  17 Dejmek J, Selevan SG, Benes I, Solansky I, Sram RJ. Fetal growth and maternal exposure to particulate matter during pregnancy. Environ Health Perspect  1999; 107: 475– 80. Google Scholar CrossRef Search ADS PubMed  18 Smarr MM, Vadillo-Ortega F, Castillo-Castrejon M, O'Neill MS. The use of ultrasound measurements in environmental epidemiological studies of air pollution and fetal growth. Curr Opin Pediatr  2013; 25: 240– 46. Google Scholar CrossRef Search ADS PubMed  19 Gardosi J, Figueras F, Clausson B, Francis A. The customised growth potential: an international research tool to study the epidemiology of fetal growth. Paediatr Perinat Epidemiol  2011; 25: 2– 10. Google Scholar CrossRef Search ADS PubMed  20 Zhang J, Mikolajczyk R, Grewal J, Neta G, Klebanoff M. Prenatal application of the individualized fetal growth reference. Am J Epidemiol  2011; 173: 539– 43. Google Scholar CrossRef Search ADS PubMed  21 Hemachandra AH, Klebanoff MA, Duggan AK, Hardy JB, Furth SL. The association between intrauterine growth restriction in the full-term infant and high blood pressure at age 7 years: results from the Collaborative Perinatal Project. Int J Epidemiol  2006; 35: 871– 77. Google Scholar CrossRef Search ADS PubMed  22 van den Hooven EH, Pierik FH, de Kluizenaar Y et al.   Air pollution exposure during pregnancy, ultrasound measures of fetal growth, and adverse birth outcomes: a prospective cohort study. Environ Health Perspect  2012; 120: 150– 56. Google Scholar CrossRef Search ADS PubMed  23 Ritz B, Qiu J, Lee PC et al.   Prenatal air pollution exposure and ultrasound measures of fetal growth in Los Angeles, California. Environ Res  2014; 130: 7– 13. Google Scholar CrossRef Search ADS PubMed  24 Slama R, Thiebaugeorges O, Goua V et al.   Maternal personal exposure to airborne benzene and intrauterine growth. Environ Health Perspect  2009; 117: 1313– 21. Google Scholar CrossRef Search ADS PubMed  25 Aguilera I, Garcia-Esteban R, Iniguez C et al.   Prenatal exposure to traffic-related air pollution and ultrasound measures of fetal growth in the INMA Sabadell cohort. Environ Health Perspect  2010; 118: 705– 11. Google Scholar CrossRef Search ADS PubMed  26 Iniguez C, Ballester F, Estarlich M et al.   Prenatal exposure to traffic-related air pollution and fetal growth in a cohort of pregnant women. Occup Environ Med  2012; 69: 736– 44. Google Scholar CrossRef Search ADS PubMed  27 Patelarou E, Kelly FJ. Indoor exposure and adverse birth outcomes related to fetal growth, miscarriage and prematurity - a systematic review. Int J Environ Res Public Health  2014; 11: 5904– 33. Google Scholar CrossRef Search ADS PubMed  28 Iniguez C, Esplugues A, Sunyer J et al.   Prenatal exposure to NO2 and ultrasound measures of fetal growth in the Spanish INMA Cohort. Environ Health Perspect  2016; 124: 235– 42. Google Scholar PubMed  29 Malmqvist E, Liew Z, Kallen K et al.   Fetal growth and air pollution - A study on ultrasound and birth measures. Environ Res  2017; 152: 73– 80. Google Scholar CrossRef Search ADS PubMed  30 Langer O. Fetal macrosomia: etiologic factors. Clin Obstet Gynecol  2000; 43: 283– 97. Google Scholar CrossRef Search ADS PubMed  31 Hod M, Pardo J. Environmental rather than genetic fetal overgrowth: defining the difference and hints for diagnosis and management. Ultrasound Obstet Gynecolt  2009; 33: 375– 78. Google Scholar CrossRef Search ADS   32 Xu Z, Xu X, Zhong M et al.   Ambient particulate air pollution induces oxidative stress and alterations of mitochondria and gene expression in brown and white adipose tissues. Particle Fibre Toxicol  2011; 8: 20. Google Scholar CrossRef Search ADS   33 Bolton JL, Smith SH, Huff NC et al.   Prenatal air pollution exposure induces neuroinflammation and predisposes offspring to weight gain in adulthood in a sex-specific manner. FASEB J  2012; 26: 4743– 54. Google Scholar CrossRef Search ADS PubMed  34 Thayer KA, Heindel JJ, Bucher JR, Gallo MA. Role of environmental chemicals in diabetes and obesity: a National Toxicology Program workshop review. Environ Health Perspect  2012; 120: 779– 89. Google Scholar CrossRef Search ADS PubMed  35 Qiu J, He X, Cui H et al.   Passive smoking and preterm birth in urban China. Am J Epidemiol  2014; 180: 94– 102. Google Scholar CrossRef Search ADS PubMed  36 Zhao N, Qiu J, Zhang Y et al.   Ambient air pollutant PM10 and risk of preterm birth in Lanzhou, China. Environ Int  2015; 76: 71– 77. Google Scholar CrossRef Search ADS PubMed  37 Jin L, Qiu J, Zhang Y et al.  . Ambient air pollution and congenital heart defects in Lanzhou, China. Environ Res Lett  2015; 10: 074005. Google Scholar CrossRef Search ADS   38 Huang X, Qiu J, Zhang Y et al.   Ambient air pollutant PM10 and risk of pregnancy-induced hypertension in urban China. Environ Res Lett  2015; 10: 084025. Google Scholar CrossRef Search ADS   39 Jiang M, Qiu J, Zhou M et al.   Exposure to cooking fuels and birthweight in Lanzhou, China: a birth cohort study. BMC Public Health  2015; 15: 712. Google Scholar CrossRef Search ADS PubMed  40 Papageorghiou AT, Ohuma EO, Altman DG et al.   International standards for fetal growth based on serial ultrasound measurements: the Fetal Growth Longitudinal Study of the INTERGROWTH-21st Project. Lancet  2014; 384: 869– 79. Google Scholar CrossRef Search ADS PubMed  41 Pinheiro JC, Bates DM. Mixed-Effects Models in S and S-Plus  New York, NY: Springer, 2000. Google Scholar CrossRef Search ADS   42 Moser EB. Repeated Measures Modeling With PROC MIXED . Louisiana State University, Baton Rouge, LA: SAS Institute, 2004. 43 Zhang Y, Li M, Bravo M et al.   Air quality in Lanzhou, a major industrial city in China: characteristics of air pollution and review of existing evidence from air pollution and health studies. Water, Air, & Soil Pollution  2014; 225: 1– 17. 44 Environmental Protection Agency. National Ambient Air Quality Standards (NAAQS) 2015. https://www.epa.gov/pm-pollution/table-historical-particulate-matter-pm-national-ambient-air-quality-standards-naaqs (15 January 2013, date last accessed). 45 Aho K, Derryberry D, Peterson T. Model selection for ecologists: the worldviews of AIC and BIC. Ecology  2014; 95: 631– 36. Google Scholar CrossRef Search ADS PubMed  46 Higgins L, Greenwood SL, Wareing M, Sibley CP, Mills TA. Obesity and the placenta: A consideration of nutrient exchange mechanisms in relation to aberrant fetal growth. Placenta  2011; 32: 1– 7. Google Scholar CrossRef Search ADS PubMed  47 Karakosta P, Georgiou V, Fthenou E et al.   Maternal weight status, cord blood leptin and fetal growth: a prospective mother-child cohort study (Rhea study). Paediatr Perinat Epidemiol  2013; 27: 461– 71. Google Scholar CrossRef Search ADS PubMed  48 Ferraro ZM, Qiu Q, Gruslin A, Adamo KB. Excessive gestational weight gain and obesity contribute to altered expression of maternal insulin-like growth factor binding protein-3. Int J Womens Health  2013; 5: 657– 65. Google Scholar CrossRef Search ADS PubMed  49 Tung E, Roberts CT, Heinemann GK et al.   Increased placental nutrient transporter expression at midgestation after maternal growth hormone treatment in pigs: a placental mechanism for increased fetal growth. Biol Reprod  2012; 87: 126. Google Scholar CrossRef Search ADS PubMed  50 Pinto JP, Lefohn AS, Shadwick DS. Spatial variability of PM2.5 in urban areas in the United States. Journal of the Air & Waste Management Association  2004; 54: 440– 49. Google Scholar CrossRef Search ADS   © The Author(s) 2018; all rights reserved. Published by Oxford University Press on behalf of the International Epidemiological Association

Journal

International Journal of EpidemiologyOxford University Press

Published: Feb 26, 2018

There are no references for this article.

You’re reading a free preview. Subscribe to read the entire article.


DeepDyve is your
personal research library

It’s your single place to instantly
discover and read the research
that matters to you.

Enjoy affordable access to
over 18 million articles from more than
15,000 peer-reviewed journals.

All for just $49/month

Explore the DeepDyve Library

Search

Query the DeepDyve database, plus search all of PubMed and Google Scholar seamlessly

Organize

Save any article or search result from DeepDyve, PubMed, and Google Scholar... all in one place.

Access

Get unlimited, online access to over 18 million full-text articles from more than 15,000 scientific journals.

Your journals are on DeepDyve

Read from thousands of the leading scholarly journals from SpringerNature, Elsevier, Wiley-Blackwell, Oxford University Press and more.

All the latest content is available, no embargo periods.

See the journals in your area

DeepDyve

Freelancer

DeepDyve

Pro

Price

FREE

$49/month
$360/year

Save searches from
Google Scholar,
PubMed

Create lists to
organize your research

Export lists, citations

Read DeepDyve articles

Abstract access only

Unlimited access to over
18 million full-text articles

Print

20 pages / month

PDF Discount

20% off