Prenatal exposure to VOCs and NOx and lung function in preschoolers

Abstract

Background:

Several studies have shown that exposure to air pollutants affects lung growth and development and can result in poor respiratory health in early life.

Methods:

We included a subsample of 772 Mexican preschoolers whose mothers participated in a Prenatal Omega-3 fatty acid Supplements, GRowth, And Development birth cohort study with the aim to evaluate the impact of prenatal exposure to volatile organic compounds and nitrogen oxides on lung function measured by oscillation tests. The preschoolers were followed until 5 years of age. Anthropometric measurements and forced oscillation tests were performed at 36, 48, and 60 months of age. Information on sociodemographic and health characteristics was obtained during follow up. Prenatal exposure to volatile organic compounds and nitrogen oxides was evaluated using a land use regression models and the association between them was tested using a lineal regression and longitudinal linear mixed effect models adjusting for potential confounders.

Results:

Overall, the mean (standard deviation) of the measurements of respiratory system resistance and respiratory system reactance at 6, 8, and 10 Hz during the follow up period was, 11.3 (2.4), 11.1 (2.4), 10.3 (2.2) and −5.2 (1.6), −4.8 (1.7) and −4.6 hPa.s.L⁻¹ (1.6), respectively. We found a significantly positively association between respiratory resistance (β_Rrs6 = 0.011 95%CI: 0.001, 0.023) (p<0.05) and prenatal exposure to nitrogen dioxide and a marginally negatively association between respiratory reactance (β_Xrs6 = −11.40 95%CI: −25.26, 1.17 and β_Xrs8 = −11.91 95%CI: −26.51, 1.43) (p=0.07) and prenatal exposure to xylene.

Conclusion:

Prenatal exposure to air pollutants was significantly associated with the alteration of lung function measured by oscillation tests in these preschool children.

INTRODUCTION

Lung function is an accurate measure of respiratory health and is a predictor of cardiorespiratory morbidity¹ and mortality, especially in children^1–3. Spirometry is the gold standard method to assess lung function in children and adults, however, obtaining acceptable and repeatable spirometry measurements requires significant efforts and high level of cooperation^4–6. Therefore, the diagnosis and management of respiratory diseases, such as asthma, remain suboptimal in young children who cannot perform an acceptable forced expiratory maneuvers. One potential lung function assessment method suitable for use in young children is the forced oscillation technique (FOT). The FOT was developed by DuBois et al⁷ in 1956, to measure the mechanical behavior of the respiratory system. Over the years, the FOT has been used in research and more recently in clinical practice^8–10. Standardized approaches for the collection of FOT outcomes have been established by the European Respiratory Society (ERS) and the American Thoracic Society in preschool children¹¹. Moreover, long-term exposure to ambient air pollution has been associated with adverse health effects in children, especially in lung function^12–15. These pollutants include; ozone (O₃), nitrogen dioxide (NO₂) and particle matters (<25 μm in diameter PM_2.5 and PM₁₀) mainly. However, studies that assess lung function by oscillation test in preschoolers to study prenatal exposure to air pollutants as ambient levels of volatile organic compounds (VOCs) and nitrogen oxides (NOx) are incipient.

VOCs and NOx are constantly present indoor and outdoor environment as a result of several industrial processes and commonly used cleaning, cosmetics, and paint products. One of the main sources in automobile exhaust, derived from the combustion of gasoline^16–18. Diane Gray and collaborators¹⁹, evaluated the impact of antenatal factors on early lung growth and lung function using the FOT in a cohort of 504 infants. The results showed that infants exposed to high household benzene levels had a lower ratio of time to peak tidal expiratory flow to total expiratory time, which may be related to airway obstruction or breathing patterns. To our knowledge, our study is the first to evaluate the impact of prenatal exposure of these contaminants on lung function in young children using the forced oscillation technique. Therefore, the aim of the current study was to evaluate the impact of prenatal exposure to VOCs and NOx on lung function in preschoolers measured by oscillation test.

METHODS

Study design and study population

The analysis of the present study is based on the Prenatal Omega-3 fatty acid Supplements, GRowth, And Development birth cohort study; a large double-blind, randomized controlled trial of prenatal docosa hexaenoic aid (DHA) supplementation. The study design of the clinical trial has been described in detail elsewhere²⁰. Briefly, eligible women were 18-35 years of age and in week 18-22 of gestation. All participants had normal pregnancies, without any pregnancy-related diseases or complications. Of the 1040 women who started treatment, 978 completed the study and 973 live infants were delivered. For the present report, we included a subsample of 772 mother-child pairs, which 379 were from the placebo group and 393 from the supplementation group, all with complete anthropometric and lung function data at 5 years of age. The women gave informed consent for themselves and their children to participate in the study. The Ethics and Investigation Committees of the National Institute of Public Health and Emory University Ethics Committee approved the investigation protocol, as well as the letters of informed consent.

Data Collection

Respiratory function by forced oscillation test

Pulmonary function was measured using the Forced Oscillation Technique (FOT)^8–10 at 36, 48, and 60 months of age. The FOT is a noninvasive method that allows for the assessment of the resistive and reactive properties of the respiratory system and provides a tidal breathing-based measurement of airway mechanics⁷. The respiratory system resistance (Rrs) and respiratory system reactance (Xrs) were measured using commercially available equipment (COSMED12; Italy) according to the ERS recommendations¹¹. The commercial FOT equipment is based on the equipment prototype described by Landser et al^8,10. Pulmonary test methods and validation processes have been described in detail in our previous report²¹. All the tests were conducted by highly trained technicians and the collected data were checked for quality. The quality of the measurements was examined by observing the coherence at each individual frequency; a coherence of more than equal to 0.95 was considered acceptable. Measurements in which three or more individual frequencies had a coherence of less than equal to 0.95 were excluded from the analysis. In addition to the quality control processes integrated into the computer programme of the device, which uses the coherence function between the signal emitted and the data recorded at each frequency. During each annual visit the child performed about seven tests, however, the three best tests carried out during each visit were selected. We included the averages of Rrs and Xrs at 6, 8, and 10 Hz. in the final analysis. It is also important to highlight that all the children included in the study were healthy and did not have any respiratory symptoms for 15 days before to lung function testing.

Prenatal exposure assessment to COVś and NOx

For each mother participant, we estimated the individual nitrogen oxide (NO₂, NO, and NO_X) and volatile organic compounds (benzene, toluene and xylene) exposure during pregnancy at home, using a standardized area-specific land use regression models. During the second and part of the third trimester of pregnancy, ambient levels of air pollutants were measured continuously, using Ogawa and 3M passive samplers for two consecutive weeks in 60 different sites throughout the state of Morelos. The samplers were positioned outdoors, near the participants’ homes (e.g. roofs, light-poles) making sure they were not being blocked by any object that could obstruct the flow of air (e.g. trees, buildings). The samplers were cleaned before use and transported in sealed amber-color containers before and after the measurement. After 2 weeks of continuous monitoring in different periods, the samples were collected and again placed in resealable bags in amber-colored containers and transported at 5ᵒ C at laboratory of the Mexico National Institute of Public Health where, inside a glove box, the pads were extracted and stored in refrigeration until their analysis. As part of quality control, 10% were blanks and duplicates. Air pollutants concentrations were determined in the laboratory of Harvard School of Public Health, using spectrophotometry²². The main predictors included in the final land use regression models were; vehicular traffic, land use, topography, population density and distance to avenues. These models explained 70% of the variability of measured air pollution levels. Details of the land use regression models have been described in detail in our previous report²³.

Information on other variables

At baseline, the participating women responded to a general health questionnaire that included information on weight and height prior to delivery, sociodemographic characteristics, health, and gynecological antecedents. In addition, an environmental and dietary questionnaire was administered during pregnancy and a respiratory health questionnaire recorded for all children before the lung function tests. Growth (height and weight) was measured each year during the follow-up.

Statistical analysis

Baseline characteristics of the study population were compared between the DHA-supplemented and placebo groups by the t-test for continuous variables or the χ² test for categorical variables. These included gender (%), height (cm), weight (kg), smoking status during pregnancy (%), maternal atopy and pulmonary function tests. We performed a cross-sectional and longitudinal analysis for 36, 48, and 60 months using a lineal regression models and lineal mixed effect models with a random intercept, age as randomized slope, and a covariance matrix without structure. This enabled us to appreciate the variability within and between subjects. The mixed effect model is as follows:

$$\text{Yi}={\text{X}}_{\mathrm{i}}\beta +{\text{Z}}_{\mathrm{i}}{\text{b}}_{\mathrm{i}}+{\epsilon }_{\mathrm{i}},$$

Where X_i = Z_i X_i is the appropriate (n_ix p) matrix of known covariates with fixed effects β and subject-specific effects b_i and ε_i in an n_i-dimentional vector of residual components. One advantage of the models used is that they do not discard subjects with incomplete data. All final models were adjusted for age, age squared, height, sex, supplementation group, smoking and maternal history of atopy. P< 0.5 was considered statistically significant. Other variables such as weight, socioeconomic status, diet during pregnancy and exposure to other indoor environmental factors (tobacco, household pets, fuel for cooking, etc.) were examined, but were not significant confounders and were not included in the final models. The goodness-of-fit of the models were evaluated by residuals and through the Akaike Information Criterion. All data analysis was performed using STATA software version 13.1²⁴.

RESULTS

Data from 772 binomial were analyzed, which 379 were from the placebo group and 393 from the supplementation group. A total of 1,464 measurements of lung function were taken during the 3 years of follow-up. All children who fulfilled the information requirements for the defined variables until 5-year of age were included. Preschool children who were not included in the analysis did not differ in terms of the main variables of the participants included in the analysis (data not shown). The sociodemographic characteristics and anthropometric measures of the participants by treatment group are shown in Table 1, we did not observe any significant difference by treatment group. Table 2 and Figure 1 show the results of the pulmonary function during each years or follow up and for the total period of the study; the mean (standard deviation) of the measurements of respiratory system resistance and respiratory system reactance at 6, 8, and 10 Hz during the follow up period was, 11.3 (2.4), 11.1 (2.4), 10.3 (2.2) and −5.2 (1.6), −4.8 (1.7) and −4.6 hPa.s.L⁻¹ (1.6), respectively, we did not observe any significant difference in lung function by treatment group. Table 3 shows the studied pollutants’ concentrations found during the period of the study, estimated by the land-use regression models and compared with the values reported for New York City of the United States. The mean of NOx concentration was 19.4 ppb (Interquartile range, IQR = 7.14 ppb), value below to the Environmental Protection Agency standard for health protection. The mean of benzene and xylene was 0.003 and 0.007 mg/m³, (IQR = 0.001 and IQR=0.006 mg/m³) respectively.

Table 1.

Sociodemographic characteristics of the 772 Preschool children who completed the study at 60 months according to treatment group.

Variables	Total n = 772	Placebo n = 379	DHA n = 393		p-value*
Boys’ n (%)	407(56.3)	179(47.2)	181(46.1)		0.887
Weight kg (mean, SD)	16.1(2.5)	16.1(2.5)	16.2(2.7)		0.664
Height cm (mean, SD)	103.2(5.8)	101.2(5.5)	101.2(5.6)		0.562
Smoking, Yes n (%)	145(18.7)	75(19.7)	70 (17.8)		0.887
Maternal atopy, Yes n (%)	233(30.2)	116(30.6)		117(29.7)	0.813

Note: Data are expressed as mean and standard deviation (SD) for height and weight, and n (%) for sex, passive smoking and maternal history of atopy.

Abbreviation: DHA, docosa hexaenoic acid.

^*P value of the Student t test for the difference of means by treatment group. χ² test for categorical variables

Table 2:

Lung function (resistance and reactance) measures stratified by age of follow up and treatment group.

Variables	36 mo (n=153)			48 mo (n= 529)			60 mo (n=772)			Follow up
	Placebo n = 70	DHA n = 83	P-value*	Placebo n = 260	DHA n = 269	P-value*	Placebo n = 379	DHA n = 393.	P-value*	period n = 772
Rr6, mean (SD)	11.8(2.2)	11.9(2.1)	0.605	11.8(2.4)	11.6(2.4)	0.605	11.7(2.9)	11.6(2.8)	0.276	11.3 (2.4)
Rr8, mean (SD)	11.6(2.0)	11.6(1.9)	0.590	11.4(2.3)	11.3(2.4)	0.299	11.3(3.0)	11.1(2.8)	0.234	11.1 (2.4)
Rr10, mean (SD)	10.6(1.9)	10.7(1.7)	0.651	10.7(2.1)	10.6(2.3)	0.250	10.5(2.7)	10.4(2.6)	0.284	10.3 (2.2)
Xr6, mean (SD)	−5.8(1.5)	−5.7(1.6)	0.700	−5.3(1.5)	−5.3(1.6)	0.568	−5.6(1.9)	−5.3(1.9)	0.950	−5.2 (1.6)
Xr8, mean (SD)	−5.5(1.7)	−5.3(1.6)	0.739	−5.0(1.5)	−5.0(1.7)	0.437	−5.4(2.2)	−5.1(2.1)	0.945	−4.8 (1.7)
Xr10, mean (SD)	−5.1(1.5)	−5.0(1.5)	0.717	−4.7(1.5)	−4.7(1.6)	0.493	−5.0(2.2)	−4.8(2.0)	0.937	−4.6 (1.6)

Note: The measurement unit is hPa.s.L⁻¹, n = Number of participants by age.

Abbreviation: DHA = Docosa Hexaenoic Acid

^*P value of the Student’s t test for the difference of means by treatment group

Figure 1.

Lung function (resistance and reactance) measures stratified by age of follow up and treatment group

Table 3.

Concentrations of the pollutants (VOCs and NOx) during the prenatal study period compared with the values reported for New York City^a from United States.

Pollutants	Mean ± SD	Percentile 25%−75%	Mean ± SD	Percentile 25%−75%
Benzene (mg/m3)	0.003 (0.001)	0.002 – 0.003	0.004 (0.006)	0.0001 – 0.034
Toluene (mg/m3)	0.008 (0.006)	0.001 – 0.012	0.002 (0.001)	0.0008 – 0.002
Xylene (mg/m3)	0.007(0.009)	0.001 – 0.007	0.003 (0.002)	0.0009 – 0.012
NO2 (ppb)	19.01(20.31)	10.63 – 20.80	100 (22.6)	3.6 – 127
NOx (ppb)	19.37(6.48)	16.24 – 23.38	44.1 (17.3)	21.7 – 115

Abbreviations: NOx, nitrogen oxides; VOC, volatile organic compounds

^aToxical profile for toxic substances, U.S. Department of Health and Human Services, Public Health Service, ATSDR, 2007.

Table 4 and 5 show the results of the crude and multivariate analysis between the different measures of resistance and pulmonary reactance and the prenatal exposure to air pollutants, during each visit of follow up (36, 48 and 60 months) and for the total period of the study. We found that, overall, the prenatal exposure to these types of pollutants were associated with the pulmonary function in preschool children. However, we only found significant associations between prenatal exposure to NO₂ and prenatal exposure to xylene. Both for the crude models and for when they were adjusted by age, age squared, height, sex, supplementation group, smoking and maternal history of atopy. A significant increase in respiratory resistance (β_Rrs6 = 0.011, 95%CI: 0.001, 0.023) was observed for each ppb of increase in NO₂ during pregnancy and a marginally significant decrease in respiratory reactance for each mg/m³ of increase in xylene during pregnancy (β_Xrs6 = −11.40, 95%CI: 25.26, 1.17 and β_Xrs8 = −11.91, 95%CI: −26.51, 1.43), respectively.

Table 4.

Effect of prenatal exposure to VOC and NOx on selected measures of pulmonary function (resistance and reactance) (hPa.s.L⁻¹) by age of follow up and during total period of study in preschool children from Morelos, Mexico.

Pollutants	Resistance						Reactance
	Rrs6		Rrs8		Rrs10		Xrs6		Xrs8		Xrs10
	β^	p-value	β^	p-value	β^	p-value	β^	p-value	β^	p-value	β^	p-value
36 moa
Xylene	−10.2	0.565	−9.47	0.534	−10.9	0.444	1.92	0.885	−0.570	0.966	0.257	0.983
NO2	−0.017	0.164	−0.019	0.079	−0.016	0.109	0.010	0.279	0.013	0.161	0.013	0.144
48 moa
Xylene	19.9	0.092*	24.1	0.037*	19.4	0.079*	−9.62	0.218	−12.1	0.129	−11.2	0.149
NO2	0.014	0.023*	0.013	0.031*	0.014	0.019*	−0.002	0.505	−0.003	0.432	−0.003	0.350
60 moa
Xylene	10.9	0.366	5.93	0.622	−1.09	0.922	−24.7	0.002*	−30.1	0.001*	−24.7	0.004*
NO2	0.008	0.034*	0.007	0.089*	0.006	0.078*	−0.004	0.088*	−0.004	0.131	−0.004	0.148
Follow upb
Benzene	27.69	0.686	6.11	0.926	−9.42	0.877	−43.72	0.365	−45.93	0.370	−28.2	0.560
Toluene	5.28	0.679	8.37	0.498	8.57	4.462	−3.03	0.736	−5.45	0.570	−4.98	0.582
Xylene	11.76	0.199	11.50	0.187	6.90	0.390	−12.10	0.062*	−13.51	0.046*	−11.28	0.079*
NO2	0.011	0.037*	0.006	0.155	0.006	0.070*	−0.003	0.323	−0.002	0.585	−0.002	0.550

^aCrude lineal regression models (bivariate).

^bCrude mixed effects models (bivariate).

.05 ≤ P ≤ .09.

^{^}β = Coefficients

Table 5.

Effect of prenatal exposure to VOC and NOx on selected measures of pulmonary function (resistance and reactance) (hPa.s.L⁻¹) by age of follow up and during total period of study in preschool children from Morelos, Mexico.

Pollutants	Resistance						Reactance
	Rrs6		Rrs8		Rrs10		Xrs6		Xrs8		Xrs10
	βa	p-value	βa	p-value	βa	p-value	βa	p-value	βa	p-value	βa	p-value
36 monthsb
Xylene	−15.5	0.422	−16.1	0.337	−20.1	0.207	1.30	0.930	1.40	0.927	1.22	0.933
NO2	−0.018	0.491	−0.015	0.179	−0.013	0.191	0.003	0.780	0.007	0.448	0.007	0.414
48 monthsb
Xylene	20.8	0.070*	25.4	0.025*	20.3	0.064*	−9.33	0.199	−12.6	0.089*	−11.7	0.177
NO2	0.016	0.011*	0.014	0.022*	0.014	0.024*	−0.005	0.177	−0.006	0.131	−0.006	0.127
60 monthsb
Xylene	1.12	0.928	−3.56	0.779	−9.41	0.427	−21.0	0.012*	−24.5	0.012*	−19.6	0.032*
NO2	0.008	0.300	0.004	0.568	0.003	0.654	−0.010	0.045*	−0.010	0.089*	−0.008	0.139
Follow upc
Benzene	26.99	0.704	−40.89	0.551	−50.60	0.427	−13.23	0.789	−13.70	0.798	1.61	0.975
Toluene	−1.26	0.927	1.26	0.925	1.92	0.878	−3.05	0.751	−3.42	0.744	−2.85	0.776
Xylene	9.48	0.304	9.08	0.303	4.37	0.593	−11.40	0.074*	−11.91	0.083*	−10.14	0.124
NO2	0.011	0.050*	0.011	0.209	0.011	0.193	−0.011	0.101	−0.014	0.228	−0.013	0.260

^aAll model adjusted for age, age 2, height, sex, supplementation group, passive smoking and maternal history of atopy.

^bLineal regression models.

^cLineal mixed effects models.

^*.05 ≤ P ≤.08.

DISCUSSION

The results of the present study indicate that the prenatal exposure to VOCs and NO₂ at low levels of exposure and estimated by a land-use regression models significantly affects the lung function measured by oscillation test in children at preschool age. To our knowledge, this is the first prospective study realized in small to middle income countries. This approach provides stronger results given that most of previous studies related with prenatal exposure to air pollutants and lung function, used results from spirometry test, assign exposure using data from fixed monitoring stations and mainly included a susceptible population, such as asthmatic children^25–29.

In this regard, Morales ³⁰ reported that for an interquartile range of increase in exposures during the second trimester of pregnancy, the forced expiratory volume in 1 second decrease −18.4 mL, 95% CI −34.8 to −2.1 for benzene exposure and −28.0 mL, 95% CI −52.9 to −3.2 for NO₂ exposure, results that are consistent with ours, however, that study evaluated the pulmonary function by spirometry test, the concentrations of the pollutants studied were more higher than those reported by us and the sample size was more smaller. In another study, the authors examined the relationship between maternal NO₂ exposure and lung function during the early lives of asthmatic children. They reported that increased indoor NO₂ exposure during the fetal development period could have the potential to result in poor lung function in early life³¹.

Only one previous study evaluated the impact of prenatal exposure to benzene on pulmonary function on 6-week-old infants using forced oscillation tests. They found a significant association between prenatal exposure to this pollutant and lower time to peak tidal expiratory flow over total expiratory time ratio¹⁸.

The development of the respiratory system involves the formation of a highly ordered system of ramifications of the respiratory passages^32,33 and during its development, there are different factors that can affect it, such as environmental exposures^32,34. A toxic environment during pregnancy are particularly important for the respiratory system, especially in the case of VOCs and oxides of nitrogen since these compounds have a high toxic potential.

For example, a fetus’s exposure to environmental tobacco smoke during pregnancy has been shown to decrease lung function at birth. This increase the risk of developing obstructive respiratory diseases at a younger ages^35–38.

In this sense, results from experimental studies indicate that some aromatic VOCs have been shown to cause oxidative stress in human lung epithelial cells and consequently, redox-sensitive intracellular signaling pathways are activated leading to the release of inflammation mediators such as the monocyte chemo-attractant protein, interleukin 8 and prostaglandins³⁹. Also, aromatics VOCs and other air pollutants have been show to induce Th2-driven immune responses, allergic sensitization and allergic airway inflammation⁴⁰. Similarly, the association between exposure to these pollutants and lung function can be explained by the direct toxicity of the contaminant, derived from these pollutants during pregnancy can alter the normal pattern of development of respiratory system, metabolic, immune and neurological functions that are constantly changing during pregnancy, as well as during postnatal growth^41,42.

This study, however, has some limitations that should be considered when interpreting the results. First, the number of sites monitored and the duration of the measurements limit the possible variations that could occur in the concentrations of the pollutants over the total study period, assuming that, the air concentrations had the same behavior throughout the entire time of pregnancy. Second, we did not obtain detailed information about the time spent of each participants and other possible indoor sources of these pollutants, instead, we presumed that exposure was primarily associated with the amount of time spent in outdoors. And the third limitation is the possibility that the results were a consequence of poor control of confounders. However, this is unlikely due to the fact that the design of the study excluded women with high-risk pregnancies and/or pre-existing illnesses and that the models were adjusted specifically for potential confounders.

Likewise, our study has important strengths; the longitudinal study design with a high participation rate and a very large sample size. Additional assets were (a) the fact that the measurement of pulmonary function was performed with a minimally invasive method (forced oscillation) and (b) the wealth of valuable information collected on other variables of importance (from the gestational stage to 5 years of age) gave greater support to the findings.

In summary, our results support the evidence that exposure to VOCs and NOx during pregnancy at levels below those established by the environmental protection agency standard for health protection, affect the lung function in healthy preschool children. Therefore, it is highly advisable to influence the revision and promulgation of norms that regulate their issuance and allow a better control of exposure especially at critical stages of development, such as pregnancy. Highlighting the need for further research evaluating the impact of antenatal and early exposure to indoor air pollutants on lung health.