Low birth weight is recognized to be caused by genetic, demographic, and environmental factors. 1–4 Because of the well-known relation of birth weight to infant mortality and adverse effects in later life, 5 environmental factors that contribute to reduced birth weight are of great public health concern. 6 Although air pollution has not been considered as an important environmental determinant of pregnancy outcomes, 7 there has been a growing concern about its relation to reproductive health hazards. 8,9 Only a few studies, however, have investigated a possible association between ambient air pollution and birth outcomes. Also, it is not known whether the associations between air pollution and pregnancy outcome found in some populations can be replicated in others. 10 Low birth weight for babies whose mothers lived in areas of high air pollution was first reported in Los Angeles, CA in the early 1970s. 6 In another Los Angeles study (1999), 8 exposure to high levels of ambient carbon monoxide (CO) (>5.5 ppm) during the last trimester was associated with an increased risk for low birth weight. A Beijing, China study reported exposure-response relation between maternal exposures to sulfur dioxide (SO2) and total suspended particles (TSP) during the third trimester of pregnancy and infant birth weight. 7 Recently, epidemiologic studies in the Czech Republic reported that the prevalence of low birth weight was positively associated with concentrations of SO2, and somewhat less strongly with TSP, in the first trimester. 11,12 However, in a Swedish study, which was ecological and did not use individual trimester-specific exposure measures, air pollution did not affect the odds ratios for low birth weight. 9
Therefore, the results of the epidemiologic studies for the relation between air pollution and low birth weight are not clear, especially with regard to consistency and biological mechanism. 11 Because fetuses are sensitive to damage by a variety of environmental toxicants 13 and the public health implications of exposure can be serious, the relation needs to be explored in different populations and sites. If low birth weight is associated with air pollution, lowering the concentrations of air pollution could reduce the associated health burden considerably.
We hypothesize that there might be a decreased in utero oxygen supply, resulting from a reduction of oxygen-carrying capacity or blood viscosity changes induced by air pollution. CO readily crosses the placenta to expose the fetus in utero, 14 leading to a rapid accumulation of carboxyhemoglobin and reducing the oxygen-carrying capacity of the blood. Another possibility is that the production of free radicals induced by air pollution might cause an inflammatory response, contributing to enhanced blood viscosity. 15,16 Suboptimal placenta perfusion from blood viscosity changes may cause adverse pregnancy outcomes, including low birth weight and preterm birth. 17
Our purpose for this study was to determine whether air pollution is associated with low birth weight, an important determinant of postneonatal infant mortality and morbidity. 18 To investigate this relation, we examined maternal exposure to air pollution in the first and trimester as a predictor of low birth weight.
Materials and Methods
We obtained birth certificates in the Seoul, South Korea area between January 1, 1996 and December 31, 1997 from the Korean National Birth Register. We determined the first and third trimester periods and ascertained birth weight and most of the covariates included in this study. Doctors or nurses recorded the birth certificates at delivery and registered them with regional public health centers. We extracted birth weight, infant gender, gestational age, birth order, marital status, parental age, and parental education for each birth record from the birth registry.
Of 286,474 births registered, we included 276,763 full-term singletons in the analysis, excluding preterm births (gestational age <37 weeks).
We selected the city of Seoul as the study area for several reasons. Seoul is the largest metropolitan city in the country and comprises 25 administrative areas. It has a distinct four-season climate, and the major air pollution source is automobile exhaust emission. Air pollution data were obtained from the Department of the Environment in Seoul. Exposure measurements during the study period were taken from 21 monitoring sites, which represented 84% of all administrative areas. Measurements of CO (nondispersive infrared photometry method), nitrogen dioxide (NO2, chemiluminiscence method), SO2 (ultraviolet photometry method), TSP (beta-ray absorption method), and ozone (O3, ultraviolet photometry method) were taken hourly. Twenty-four-hour averages of pollutant concentrations were constructed between measurement sites. In the case of O3, a daytime 8-hour average was used instead of a 24-hour average.
On the basis of the gestational age and birth date of each newborn, we estimated the first and third trimester exposure by averaging daily ambient air pollution concentrations during the corresponding days.
Initial analysis estimated means of birth weight by categories of covariates to evaluate the relations between birth weight and the variables controlling for other covariates. We used Pearson’s correlation coefficients to examine the relation among air pollutant concentrations. We evaluated the associations between ambient air pollution and low birth weight with a generalized additive logistic regression adjusting for gestational age, maternal age, parental educational level, infant’s birth order, and gender.
We used a loess smooth function of time trends to capture seasonal change. We computed the relative risk of low birth weight (<2,500 gm) for each interquartile range change of each pollutant in the first and third trimester of pregnancy. The exposure variables were evaluated singly and in combination as predictors of low birth weight. Birth weight was also analyzed as a continuous variable to estimate the reduction of birth weight by interquartile changes of each pollutant. We also used smoothing plots with a generalized additive model to analyze the exposure-response relation for each pollutant.
Table 1 presents the distribution of means of birth weight by infantile and maternal characteristics. The overall prevalence rate of low birth weight is 2.8%, excluding preterm birth, and mean birth weight is 3,310 gm (standard deviation = 410) for 276,763 singletons. Birth weight varied with maternal age, parental education, season, and infant’s birth order and gender after adjusting for other covariates. Young maternal age, less education of parents, early order of birth, summer season, and female gender are risk factors for low birth weight. Table 2 shows the average concentrations of air pollutants for the first and third trimester of pregnancy of study subjects for each quartile change. There was only a slight difference between air pollutant levels of first vs third trimesters of pregnancy. Table 3 shows the correlation matrix among air pollutant concentrations during the study period. The concentrations of CO, NO2, SO2, and TSP were positively correlated with each other (0.67 ≤r ≤ 0.83). The concentration of O3, however, was negatively correlated with other pollutants.
Concentrations of air pollutants during the study period showed seasonal patterns. To avoid seasonal confounding, we used a smooth function of time trend in the logistic regression model for low birth weight using the generalized additive model. 19,20 In the models that included concentrations of each pollutant during the first trimester of pregnancy, CO, NO2, SO2, and TSP were risk factors of low birth weight. The adjusted relative risk was 1.08 (95% CI = 1.04–1.12) for each interquartile increase in CO concentrations. The risk of low birth weight was increased to 1.07 (95% CI = 1.03–1.11) for NO2, 1.06 (95% CI = 1.02–1.10) for SO2, and 1.04 (95% CI = 1.00–1.08) for TSP for each interquartile change (Table 4). In contrast, the risk was lower at higher O3 levels. For the third trimester of pregnancy, O3 was a risk factor of low birth weight, whereas other pollutants seemed to lower the risk. O3 increased the relative risk to 1.09 (95% CI = 1.04–1.14) for each interquartile increase of the concentration.
When we entered concentrations of each pollutant during the first and third trimester of pregnancy together in the model, the relative risk of low birth weight for each air pollutant during the first trimester of pregnancy remained constant. In the third trimester, however, the risks changed toward the null for all pollutants.
Table 5 shows the reduction of birth weight for interquartile changes of each air pollutant. Each interquartile increase of CO concentration during the first trimester reduced 11.55 gm of birth weight. NO2, SO2, and TSP also decreased birth weight 8.41, 8.06, and 6.06 gm, respectively.
There was clearly a negative relation between birth weight and concentrations of CO, NO2, SO2, and TSP during the first trimester (Figure 1). The relations are relatively linear, without thresholds for concentrations of the pollutants.
We found that ambient CO, NO2, SO2, and TSP concentrations during the first trimester of pregnancy were associated with low birth weight after adjusting for time trends, gestational age, maternal age, parental educational level, parity, and infant’s birth order and gender. O3 concentration during the third trimester of pregnancy was also associated with low birth weight, but this association disappeared when first- and third-trimester exposures were examined together.
The relation of air pollution to pregnancy outcome, especially low birth weight, is a highly controversial issue. Only a few studies have investigated this relation. Moreover, the studies vary in terms of exposure and pollutants, and the results are inconsistent.
A Czech study found elevated levels of SO2 and TSP, especially in the first trimester, to be associated with an increased risk of low birth weight. 10 Another Czech study of birth outcomes found that the risk of intrauterine growth retardation was increased in full-term births when mothers were exposed to high levels of PM10 in the first month of pregnancy after controlling for maternal characteristics. 21 In a Chinese study, there was a strong exposure-response relation between third-trimester maternal exposures to SO2 and TSP and low birth weight. 7 Ritz and Yu 8 also showed exposure-response relations between third-trimester exposure to CO and birth weight. More recently, air pollution exposure based on annual average atmospheric concentrations of TSP and SO2 without considering specific trimester was reported to be associated with very low birth weight (less than 1,500 gm). 22
Notwithstanding the above positive findings between air pollution and low birth weight, there are also studies that do not support the conclusion that there is a relation between the risk of low birth weight and maternal exposure to air pollution. 9,23
Although these previous studies have not provided conclusive results regarding the relation between birth weight and air pollution, our findings add weight to the conclusion that air pollutants negatively impact fetal development.
There is also a controversy over the effect of specific pollutants. A study based in Los Angeles, where the levels of ambient CO pollutants in the early 1970s routinely exceeded 300 ppm, reported a lower mean birth weight for babies of mothers who lived in areas of higher air pollution. 6 There was no clear relation, however, between ambient levels of CO for the third trimester of pregnancy and low birth weight in Denver, where the levels of CO were considerably lower than those in Los Angeles. 23 Our results show that ambient concentrations of CO for the first trimester of pregnancy confer risk of low birth weight even in the lower range of exposure.
CO is well known as a reproductive toxicant that can interfere with oxygen delivery to the fetus. CO shifts the oxyhemoglobin dissociation equilibrium and displaces oxygen from hemoglobin for a given partial pressure of oxygen. 14 It has also been shown that CO crossed the placental barrier 24 and that the fetus is particularly vulnerable to CO poisoning because of 10–15% higher accumulation in fetal blood than maternal levels. 25 Its elimination is slower in fetal blood than in maternal circulation. 26 Another possible toxic mechanism of CO is that it can also affect leukocytes, platelets, and the endothelium, inducing a cascade of effects resulting in oxidative injury that contribute to the toxicity of other air pollutants. 27
The evidence for CO toxicity to fetus is supported by epidemiologic studies, which examined the relation of cigarette consumption and CO levels during pregnancy. They found associations between smoking and birth outcomes that include fetal growth retardation, neonatal deaths, and premature delivery. 28,29
In our study, NO2, SO2, and TSP concentrations of the first trimester are also associated with low birth weight. These pollutants are highly correlated with each other. In addition, secondary particles are formed in the atmosphere by chemical reactions involving NO2 and SO2. 30 Therefore, it is reasonable to consider these pollutants together rather than separately.
It has been shown that inflammation in the lung caused by air pollutants increases the coagulability of the blood. 15,31 Production of free radicals induced by pollutants might cause an inflammatory response, contributing to enhanced blood coagulation. Human volunteers exposed to diesel particles at 300 μg/m3 for an hour had increases in peripheral neutrophils and platelets as well as upregulation of endothelial adhesion molecules. 32 Decreased oxygen supply from blood viscosity changes by increasing coagulability may cause chronic hypoxic injury to fetus. This theory is supported by evidence of the role of elevated blood viscosity for impaired efficiency of maternal blood flow. 33,17
Although there appears to be a relation between CO, NO2, SO2, and TSP concentration and birth weight, little is known about low birth weight in relation to O3 exposure. A few animal studies have shown that O3 reduced the body weight of offspring. 34–36 We cannot explain, however, why the concentration of O3 in the third trimester of pregnancy is related to low birth weight whereas other air pollutants in the first trimester are related to the risk. Because O3 is a secondary photooxidant pollutant, it might be a proxy for other toxic chemicals from vehicle emissions in the air. Those chemicals could be related to the effect on growth rate during the third trimester rather than on organogenesis during the first trimester. Nevertheless, the relation between concentrations of O3 and low birth weight does not seem to be robust, because this relation disappeared when we used O3 concentrations of the first and third trimester together in the model.
Although elucidating biological pathways is important, an association between air pollution and low birth weight needs to be well established first. If these findings are confirmed in different populations, they would provide an important contribution to the debate on reducing the exposures to air pollution. 9
The main limitation of this study is the lack of information to adjust for some individual risk factors for low birth weight such as smoking and alcohol consumption. It is not reasonable, however, to believe that the individual risk factors are correlated with the ambient air pollution levels. Another limitation is the use of environmental monitoring data, which do not necessarily represent individual exposures. Although outdoor measurements of pollutants can represent indoor environments, they can result in misclassification of exposure because pregnant women may behave differently from the general population. They may tend to stay at home or make efforts to avoid air pollution, which might result in underestimation of the association of air pollutants on birth weight. Even though the proportion of population migration may not be enough to affect the results, this factor could also cause underestimation of the associations because the direction of migration was likely random.
Because air pollution exposure is universal in the general population, even a small shift in the mean birth weight distribution curve toward the left among those exposed means an increased number of low-birth weight infants, thus contributing to a significant etiologic fraction of low birth weight.
We thank Douglas W. Dockery for constructive reviews of manuscript.
1. Kramer MS. Determinants of low birth weight: methodological assessment and meta-analysis. Bull World Health Organ 1987; 65: 663–737.
2. Kramer MS. Intrauterine growth and gestational duration determinants. Pediatrics 1987; 80: 502–511.
3. Kramer MS, Olivier M, McLean FH, Dougherty GE, Willis DM, Usher RH. Determinants of fetal growth and body proportionality. Pediatrics 1990; 86: 18–26.
4. Silbergeld E, Tonat K. Investing in prevention: opportunities to prevent disease and reduce health care costs by identifying environmental and occupational causes of noncancer disease. Toxicol Ind Health 1994; 10: 675–827.
5. Joseph KS, Kramer MS. Review of the evidence on fetal and early childhood antecedents of adult chronic disease. Epidemiol Rev 1996; 18: 158–174.
6. Williams L, Spence AM, Tideman SC. Implication of the observed effect of air pollution
on birth weight. Soc Biol 1977; 24: 1–9.
7. Wang X, Ding H, Ryan L, Xu X. Association between air pollution
and low birth weight: a community-based study. Environ Health Perspect 1997; 105: 514–520.
8. Ritz B, Yu F. The effect of ambient carbon monoxide on low birth weight among children born in southern California between 1989 and 1993. Environ Health Perspect 1999; 107: 17–25.
9. Landgren O. Environmental pollution
and delivery outcome in southern Sweden: a study with central registries. Acta Paediatr 1996; 85: 1361–1364.
10. Bobak M. Outdoor air pollution
, low birth weight, and prematurity. Environ Health Perspect 2000; 108: 173–176.
11. Bobak M, Leon DA. Pregnancy outcomes and outdoor air pollution
: an ecological study in districts of the Czech Republic 1986–1988. Occup Environ Med 1999; 56: 539–543.
12. Bobak M, Leon DA. The effect of air pollution
on infant mortality appears specific for respiratory causes in the postneonatal period. Epidemiology 1999; 10: 666–670.
13. Perera FP, Jedrychowski W, Rauh V, Whyatt RM. Molecular epidemiologic research on the effects of environmental pollutants on the fetus. Environ Health Perspect 1999; 107 (suppl 3): 451–460.
14. Longo LD. The biological effects of carbon monoxide on the pregnant woman, fetus, and newborn infant. Am J Obstet Gynecol 1977; 129: 69–103.
15. Peters A, Doering A, Wichmann HE, Koening W. Increased plasma viscosity during an air pollution
episode: a link to mortality. Lancet 1997; 349: 1582–1587.
16. Bouthillier L, Vincent R, Goegan P, Adamson IY, Bjarnason S, Stewart M, Guenette J, Potvin M, Kumarathasan P. Acute effects of inhaled urban particles and ozone: lung morphology, macrophage activity, and plasma endothelin-1. Am J Pathol 1998; 153: 1873–1884.
17. Knottnerus JA, Delgado LR, Knipschild PG, Essed GG, Smiths F. Haematologic parameters and pregnancy outcome: A prospective cohort study in the third trimester. J Clin Epidemiol 1990; 43: 461–466.
18. McCormick MC. The contribution of low birth weight to infant mortality and childhood morbidity. N Engl J Med 1985; 312: 82–90.
19. Hastie T, Tibshirani R. Generalized additive models. London: Chapman and Hall, 1990.
20. Schwartz J. The distributed lag between air pollution
and daily deaths. Epidemiology 2000; 11: 320–326.
21. 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: 475–480.
22. Rogers JF, Thompson SJ, Addy CL, McKeown RE, Cowen DJ, Decoufle P. Association of very low birth weight with exposures to environmental sulfur dioxide and total suspended particulates. Am J Epidemiol 2000; 151: 602–613.
23. Beth WA, Anna EB, David AS. Maternal exposure to neighborhood carbon monoxide and risk of low infant birth weight. Public Health Rep 1985 : 410–414.
24. Bosley ARJ, Sibert JR, Newcombe RG. Effects of maternal smoking on fetal growth and nutrition. Arch Dis Child 1981; 56: 727–729.
25. Matthew JE. Ellenhorn’s Medical Toxicology. 2nd ed. Baltimore: Williams and Wilkins, 1997; 1465–1476.
26. Koren G, Sharav T, Pastuszak A, Garrettson LK, Hill K, Samson I, Rorem M, King A, Dolgin JE. A multicenter, prospective study of fetal outcome following accidental carbon monoxide poisoning in pregnancy. Reprod Toxicol 1991; 5: 397–403.
27. Hardy KR, Thom SR. Pathophysiology and treatment of carbon monoxide poisoning. Clin Toxicol 1994; 32: 613–629.
28. Weisberg E. Smoking and reproductive health. Clin Reprod Fertil 1985; 3: 175–186.
29. Secker-Walker RH, Vacek PM, Flynn BS, Mead PB. Smoking in pregnancy, exhaled carbon monoxide, and birth weight. Obstet Gynecol 1997; 89: 649–652.
30. Levy J, Spengler JD, Hlinka D, Sullivan D. Estimated Public Health Impacts of Criteria Pollutant Air Emissions from the Salem Harbor and Brayton Point Power Plants. Report commissioned by the Clean Air Task Force. Boston: Harvard School of Public Health, 2000.
31. Seaton A, MacNee W, Donaldson K, Godden D. Particulate air pollution
and acute health effects. Lancet 1995; 345: 176–178.
32. Salvi S, Blomberg A, Rudell B, Kelly F, Sandstrom T, Holgate ST, Frew A. Acute inflammatory responses in the airways and peripheral blood after short-term exposure to diesel exhaust in healthy human volunteers. Am J Respir Crit Care Med 1999; 159: 702–709.
33. Zondervan HA, Oosting J, Smorenberg-Schoorl ME, Treffers PE. Maternal whole blood viscosity in pregnancy. Gynecol Obstet Invest 1988; 25: 83–88.
34. Bignami G, Musi B, Dell’Omo G, Laviola G, Alleva E. Limited effects of ozone exposure during pregnancy on physical and neurobehavioral development of CD-1 mice. Toxicol Appl Pharmacol 1994; 129: 264–271.
35. Dell’Omo G, Fiore M, Petruzzi S, Alleva E, Bignami G. Neurobehavioral development of CD-1 mice after combined gestational and postnatal exposure to ozone. Arch Toxicol 1995; 69: 608–616.
36. Kavlock R, Daston G, Grabowski CT. Studies on the developmental toxicity of ozone. I. Prenatal effects. Toxicol Appl Pharmacol 1979; 48: 29–41.