Advertisement
Research Article

Air Pollution and Newly Diagnostic Autism Spectrum Disorders: A Population-Based Cohort Study in Taiwan

  • Chau-Ren Jung,

    Affiliation: Department of Occupational Safety and Health, College of Public Health, China Medical University, Taichung, Taiwan

    X
  • Yu-Ting Lin,

    Affiliation: Department of Occupational Safety and Health, College of Public Health, China Medical University, Taichung, Taiwan

    X
  • Bing-Fang Hwang mail

    bfhwang@mail.cmu.edu.tw

    Affiliation: Department of Occupational Safety and Health, College of Public Health, China Medical University, Taichung, Taiwan

    X
  • Published: September 25, 2013
  • DOI: 10.1371/journal.pone.0075510

Abstract

There is limited evidence that long-term exposure to ambient air pollution increases the risk of childhood autism spectrum disorder (ASD). The objective of the study was to investigate the associations between long-term exposure to air pollution and newly diagnostic ASD in Taiwan. We conducted a population-based cohort of 49,073 children age less than 3 years in 2000 that were retrieved from Taiwan National Insurance Research Database and followed up from 2000 through 2010. Inverse distance weighting method was used to form exposure parameter for ozone (O3), carbon monoxide (CO), nitrogen dioxide (NO2), sulfur dioxide (SO2), and particles with aerodynamic diameter less than 10 µm (PM10). Time-dependent Cox proportional hazards (PH) model was performed to evaluate the relationship between yearly average exposure air pollutants of preceding years and newly diagnostic ASD. The risk of newly diagnostic ASD increased according to increasing O3, CO, NO2, and SO2 levels. The effect estimate indicating an approximately 59% risk increase per 10 ppb increase in O3 level (95% CI 1.42–1.79), 37% risk increase per 10 ppb in CO (95% CI 1.31–1.44), 340% risk increase per 10 ppb increase in NO2 level (95% CI 3.31–5.85), and 17% risk increase per 1 ppb in SO2 level (95% CI 1.09–1.27) was stable with different combinations of air pollutants in the multi-pollutant models. Our results provide evident that children exposure to O3, CO, NO2, and SO2 in the preceding 1 year to 4 years may increase the risk of ASD diagnosis.

Introduction

There is growing concern about the potential effect of air pollution on autism spectrum disorders (ASD). A group of neurodevelopmental disorders contains autistic disorder, Asperger's disorder, and pervasive developmental disorder not otherwise specified (PDD-NOS) is called ASD [1]. Abnormal development or impairment in social interaction, impaired verbal and non-verbal communication, and markedly restricted activity and interests are major characteristics of ASD [1], [2]. There is a large worldwide variation in the prevalence of ASD and there is also evidence that the prevalence has been increasing [3]. Both genetic and environmental factors play important roles in the aetiology of childhood ASD, and there is probably also genetic susceptibility to the effects of air pollution. Short-term changes in the occurrence of ASD are more likely influenced by changes in the environment than by change in the genetic pool. From preventive perspective, information on environmental, dietary, and behavioural factors is crucial. Identification of indicators of genetic susceptibility to environmental exposures could be useful from preventive point of view.

There is limited evidence that long-term exposure to ambient air pollution increases the risk of childhood ASD, but the role of different sources and components needs further elaboration [4][7]. Windham and colleagues in California studied hazardous air pollutants and ASD indicated an increased risk of autism with higher ambient air concentration of particularly metal (cadmium, mercury, and nickel), diesel particulate matter, and chlorinated solvents [4]. Kalkbrenner and colleagues examined children exposed to 35 hazardous air pollutants in North Carolina and West Virginia, and reported increased risk of ASD with increased concentration of methylene chloride, quinoline, and styrene [5]. Volk and colleagues found an association between early life living near a freeway within 309 m and risk of autism in California [6]. In addition, they reported traffic-related air pollution, nitrogen dioxide, particles with aerodynamic diameter less than 2.5 µm (PM2.5) and 10 µm (PM10) during pregnancy and during the first year of life were positively associated with autism [7]. Finding from previous studies of the effects of exposure to air pollution on the risk of ASD mainly has been cross-sectional or prevalent case-controls studies, where temporality constitutes problems.

We conducted a prospective 11-year population-based cohort study to access the association between postnatal exposure to air pollutants and newly diagnostic ASD in Taiwan. The design enabled us at the same time to verify an appropriate temporality between the hypothesized exposure and outcome and to eliminate the possibility that the presence of outcome would influence the assessment of exposure.

Methods

Study Design

We conducted a population-based cohort study by retrieving all individuals from the longitudinal health insurance database 2000 (LHID2000). The LHID2000 includes registration files and claim data of ambulatory care expenditures or inpatient expenditures for 1,000,000 individuals systematically and randomly selected from the year 2000 registry of beneficiaries of National Insurance Research Database (NHIRD). NHIRD comprised of detail health care information from more than twenty three million enrollees, representing more than 99% entire population in Taiwan. National Health Research Institutes (NHRI) confirmed there was no significant difference in the gender distribution, age distribution, newborn number every year, and average insurance payroll amount between LHID2000 and the NHIRD [8]. The accuracy of diagnosis of major diseases, such as acute coronary syndrome and ischemic stroke had been validated [9], [10]. Because the data was analyzed anonymously, the Institute Review Board specifically waived the need for consent by each subject. This study has been approved by the institute review board of China Medical University Hospital (CMU-REC-101-012), and it complied with the principles outlined in the Helsinki Declaration.

Study Population

We included one millions individuals at the baseline in 2000. We focus on children that age less than 3 years at the baseline. The baseline study population constituted of 49,833 individuals. We conducted a prospective cohort study directed at all the individuals of the cohort from January 1, 2000 through December 31, 2010. The follow-up time began from each individual's first date of enrollment until the first diagnosis of ASD. In the present analyses, we excluded children, who had experienced ASD at the baseline (N = 4), and those who had missing information on air pollution (N = 756). Thus the final study population constituted a total of 49,073 individuals. There were 535 individuals (1.09%) lost to follow-up during the study period. The response rate is up to 98.91%.

Outcome of interest

The database provides personal diagnosis codes according to International Classification of Diseases, Ninth Revision, Clinical Modification (ICD-9-CM). The bureau of National Health Insurance verify the validity and quality of diagnosis by random sampling a constant ratio of claims from every hospital each year and through strict review by an independent group of doctors [11]. To ensure the diagnostic validity of ASD in present study, only children with at least two consensus diagnosis of autistic disorder (code 299.0), Asperger syndrome (ICD-9-CM code 299.8), or PDD-NOS (ICD-9-CM 299.9)[12] between January 1, 2000 and December 31, 2010 that were selected as our outcome of interest. The newly diagnostic ASD was defined as the diagnosis code first appeared in the claims.

Exposure assessment

Carbon monoxide (CO), nitrogen dioxide (NO2), sulfur dioxide (SO2) and PM10 monthly average data and Ozone (O3) monthly average of daily maximum value were obtained from 70 Taiwan Environmental Protection Agency (EPA) monitoring station on Taiwan's main island. All of these air pollutants measured hourly—CO by nondispersive infrared absorption, NO2 by chemiluminescence, O3 by ultraviolet absorption, SO2 by ultraviolet fluorescence, and PM10 by beta-gauge—and continuously.

The locations of the monitoring stations and air pollution sources were identified and managed by geographic information system (GIS) (ArcGIS version 10; ESRI, Redlands, CA, USA). The monitoring data were integrated into yearly point data and interpolated to pollutant surfaces using inverse distance weighting method (IDW). For the IDW approach, we used suitable spatial resolution (100 m) [13] and inverse squared distance (1/squared distance) weighted average of the three nearest monitors within 25 kilometer to compute yearly mean concentration. Then, the yearly air pollution data were assigned to individuals by post-code levels. The postal code typically corresponded to one block face in urban areas (mean±SD: 17±8.56 square kilometer) but was larger in rural areas (mean±SD: 154±104.39 square kilometer) with low population density. We averaged exposure to CO, NO2, O3, PM10, and SO2 over several years before newly diagnostic ASD: preceding 1 year, 2 years, 3 years, and 4 years.

Covariates

The baseline age in 2000 and gender (male or female) of children were available from NHIRD and municipal-level socioeconomic status (SES) of parents was determined based on distribution of annual average incomes data from Directorate-general of Budget, Account, Statistics which was classified to four level: >75th percentile, 75th to 50th percentile, 50th to 25th percentile, and <25th percentile.

Some important comorbidities related to ASD were also included [14], [15]. We included anxiety (ICD-9-CM code 300.0), bipolar disorder (296.4–296.8), depressive disorder (296.2, 296.3 and 311), intellectual disabilities (317–319), obsessive-compulsive disorder (300.3), and phobic disorder (300.2). Besides, children were classified into preterm (765.0 or 765.1) and full-term (children without a code of 765.x) [16].

Statistical methods

Extended Cox proportional hazards (PH) model incorporated time-dependent and time-independent variables as predictors that were used to investigate the association between air pollutants and ASD. For the analysis of time to newly diagnostic ASD, each individual's survey time was censored at the year when the person was quite insurance, or at the end of the follow-up. Any variable for a given individual change over time is called a time dependent variables [17]. We applied an extended Cox PH model: where a baseline hazard function h0(t) is multiplied by an exponential function which contain Xi, time-independent variables, and Xj, time-dependent variables. The X(t) is the entire collection of predictor [17]. Air pollutants such as CO, NO2, O3, PM10 and SO2 were treated as time-dependent variables in the Cox PH model, representing average exposure over several years before the newly diagnostic ASD: preceding 1 year, 2 years, 3 years, and 4 years. The children whose age of newly diagnostic ASD was younger than deductive years prior to newly diagnostic ASD were removed from analyses (e.g., we excluded the children whose age of newly diagnostic ASD was younger than four years old when assessing four years prior to first diagnosis of ASD).

We first applied chi-square tests and t-tests including the covariates listed above to identify covariates associated with newly diagnostic ASD and air pollution, which were subsequently included in Cox PH model investigating the association between air pollution exposures and ASD. Further, we also selected potential confounders a priori and included them into the final model.

We fitted one pollutant models and two pollutants models to estimate the effect average exposure of CO, NO2, O3, PM10 and SO2 on newly diagnostic ASD while controlling second pollutant in the model. The effect of each pollutant on the risk of newly diagnostic ASD was estimated as hazard ratio (HR) per 100-ppb change for CO, 10-ppb change for NO2 and O3, 10-µg/m3 change for PM10, and 1-ppb for SO2, along with their 95% confidence interval (CI). We also compared the risk of newly diagnostic ASD in four exposure categories based on the distribution of each pollutant representing high (>75th percentile), medium (75th to 50th percentile), low exposure (<50th to 25th percentile) and <25th percentile as the reference category. All analyses were performed with SAS version 9.2 for Windows (PROC PHREG).

Results

Within the cohort of 49,073 children from January 1, 2000 to December 31, 2010, a total of 342 children developed ASD. In our study, most children developed ASD were male (83.2%), younger children (baseline age 1.01±0.95 years; age of newly diagnostic ASD 6.26±2.91 years) and with higher SES (≥1,095,192). The common comorbidities were intellectual disabilities (35.67% of children with newly diagnostic ASD), and anxiety (13.16%) (Table 1).

thumbnail

Table 1. Demographic data of the study cohort from January 1, 2000 to December 31, 2010.

doi:10.1371/journal.pone.0075510.t001

Effect of air pollution before newly diagnostic ASD

We investigated the relationship between newly diagnostic ASD and average exposure over several years prior to newly diagnostic ASD. The adjusted HR of newly diagnostic ASD increased with raised concentration of O3 per 10-ppb change, the tendency was similar from preceding 1 year to 4 years, varying between 1.27 and 1.60 (Table 2). The risk of newly diagnostic ASD also positively associated with a 100-ppb change in CO (adjusted HR varying between 1.37 and 1.43) and 10-ppb change in NO2 (adjusted HR varying between 3.14 and 4.75). There is no association between the risk of newly diagnostic ASD and PM10 from preceding 1 year to 4 years (adjusted HR varying between 0.92 and 1.08). The risk of newly diagnostic ASD was positively associated with 1-ppb change of SO2 in the preceding 1 year and 2 years (adjusted HR varying between 1.13 and 1.18), but the effect estimates were decreased substantially in the preceding 3 years and 4 years (adjusted HR varying between 0.88 and 0.94) (Table 2). Because of consistent pattern of all pollutants, we therefore presented average exposure of CO, NO2, O3, PM10, and SO2 of preceding 1 year before newly diagnostic ASD in the subsequent analysis.

thumbnail

Table 2. Crude and adjusted hazard ratios (95% confidence interval) for newly diagnostic autism spectrum disorders (ASD) from single pollutant model over several years before newly diagnostic ASD.

doi:10.1371/journal.pone.0075510.t002

Air pollution

The seasonal concentrations of O3 are higher in spring, summer, and fall, and lower in winter. The concentrations of CO, NO2, PM10, and SO2 are lowest in summer, and higher in both side (Figure 1). We examined the results of spearman correlation for average concentration of preceding 1 year before newly diagnostic ASD (Table 3). The concentration of CO was highly and positively correlated with NO2 (r = 0.95), which represent the common source of motor vehicles, and moderately correlated with SO2 (r = 0.32). O3 was highly correlated with PM10 (r = 0.65).

thumbnail

Figure 1. The seasonal variation of five main air pollutants during the study period.

doi:10.1371/journal.pone.0075510.g001
thumbnail

Table 3. Spearman correlation of average concentration of air pollutants of preceding year before end of follow-up.

doi:10.1371/journal.pone.0075510.t003

Air pollution and newly diagnostic ASD

In the single-pollutant model, the risk of newly diagnostic ASD was increased in relation to the O3 levels per 10-ppb changes (adjusted HR = 1.59; 95% CI 1.42–1.78). The adjusted HR was 1.72 (95% CI 1.20–2.47) for low, 2.55 (95% CI 1.72–3.78) for medium and 5.88 (95% CI 4.05–8.52) for high O3 exposure and shows an exposure-response pattern (Table 4). In the two-pollutant models, the effects estimates for O3 exposure were stable for the three different combinations of pollutants, varying between 1.53 and 1.57 (Table 5).

thumbnail

Table 4. Crude and adjusted hazard ratios (95% confidence interval) for newly diagnostic autism spectrum disorders (ASD) from single pollutant model.

doi:10.1371/journal.pone.0075510.t004
thumbnail

Table 5. Adjusted hazard ratios (95% confidence interval) of two pollutant mode for newly diagnostic autism spectrum disorders (ASD).

doi:10.1371/journal.pone.0075510.t005

The risk of newly diagnostic ASD was also associated with a 100-ppb change in CO (adjusted HR = 1.37, 95% CI1.31–1.44) (Table 4), and inclusion of both of the traffic-related pollutants and O3 did not change the effect estimate substantially (Table 5). The effect estimates for newly diagnostic ASD for CO were significantly elevated with an exposure-response pattern (adjusted HR low CO = 2.59, 95% CI 1.53–4.37; adjusted HR medium CO = 6.08, 95% CI 3.65–10.13; adjusted HR high CO = 13.67, 95% CI 8.14–22.97) (Table 4) and the estimates changed little when different second pollutants were added (Table 5). As for CO, the adjusted hazard ratio for 10-ppb change in NO2 for ASD in the single pollutant model were 4.43 (95% CI 3.33–5.90) (Table 4) and the estimates changed little when different second pollutants were added (Table 5). The risk of newly diagnostic ASD was elevated with an exposure-response pattern in the single pollutant model and inclusion both of the O3 and stationary fossil fuel combustion-related air pollutants (SO2 and PM10) did not change the effect estimate substantially (Table 5). The adjusted HR for 1-ppb change in SO2 for ASD in the single-pollutant model was 1.18 (95% CI 1.09–1.28), but did not show an exposure-response pattern (Table 4). Surprisingly, an inverse association between PM10 exposure and newly diagnostic ASD was found.

Discussion

The risk of newly diagnostic ASD increased according to increasing O3, CO, NO2, and SO2 levels. The effect estimate indicating an approximately 59% risk increase per 10-ppb increase in O3 level, 37% risk increase per 10-ppb in CO, 343% risk increase per 10-ppb increase in NO2 level, and 18% risk increase per 1-ppb in SO2 level was stable with different combinations of air pollutants in the two-pollutant models. The results provide evidence that children exposure to O3, CO, NO2, and SO2 in the preceding 1 year to 4 years may increase the risk of newly diagnostic ASD. A negative or weak association between the risk of newly diagnostic ASD and PM10 was found.

There were four previous studies accessed the associations between air pollutants and ASD, but they provided inconsistent results. Exposure to Metal, diesel particulate matter, chlorinated solvents, quinoline, and styrene of ambient air during prenatal, perinatal, or early childhood may increase the risk of ASD [4], [5]. By examining the associations between distance to nearest freeway or major road and autism, Volk and colleagues found living near a freeway (less than 309 m) may increase the risk of autism [6], and furthermore, they reported that higher local estimates of traffic-related air pollution (traffic-related pollutant mixture include elemental carbon, CO, and NO2) and regional measures of PM2.5, PM10, and NO2 during pregnancy and first year of life would increase the risk of autism [7]. Consistently, the present study found results the risk of ASD increased with increasing concentration of traffic-related pollutants, such as CO and NO2. In addition, our results also found O3 and SO2 were significantly positive correlated with ASD. Oxidative stress may contribute to the pathogenesis of ASD through lipid peroxidation, reduced antioxidant activity, elevated nitric oxide level [18][23] and inflammation [24], [25]. Ambient air is a complex of variety individual pollutants that are free radical such as NO2or have the ability to drive free radical reaction such as O3 and particulates [26]. Carbon monoxide is 218 times of tendency to bind with Hemoglobin (Hb) than with oxygen [27]. When CO bind to Hb to form carboxyhemoglobin (COHb), then the oxygen carrying capacity of the blood decreases may cause tissue hypoxia. Hypoxia may deprive of usual supply of oxygen and damage cells of the heart, muscle, brain and nervous system [28], [29]. Children exposed to CO and then reoxygenation may generate oxidative stress and damage brain attribute to the increased production of partially reduced oxygen species (PROS) [30] or lipid peroxides arise from oxidative degradation of brain lipids [31]. Exposure to NO2 may induce the release of proinflammatory mediators from the lung [32]. Chronic lung inflammation may result in systemic inflammation that impacts blood vessel [33]. Cytokine derived from Systemic inflammation may also across the blood brain barrier [34] and active microglia [35]. Actively microglia may release secondary inflammation mediators and intensify neuroinflammation, then lead to brain damage [35]. Acute or chronic O3 exposure may generate oxidative stress that lead to brain lipid peroxidation [36], [37], neuronal morphological and ultrastructural changes [38], and memory deterioration [39]. Further, expose to low doses of O3 can develop progressive dopaminergic neuronal death in the substantia nigra of animal model [40]. Superoxide dismutase (SOD), glutathione peroxidase (GSH-Px), and glutathione (GSH) are used to cope with free radicals in cell, SO2inhalation decrease activities of these enzymes and lead to lipid peroxidation may damage brain [41]. We did not find any association between PM10 and the risk of newly diagnostic ASD. It’s possible that PM10 was inversely correlated with CO level.

The study has several strengths. First, previous studies examined associations between air pollution and ASD by using cross-sectional or case-control study to estimate the effects of exposure to air pollution. The NHIRD covered 99% entire population in Taiwan which provide a sufficient, prospective, and population base database that allow us to test the relationship between air pollutants and ASD. In addition, we included 49,073 children hence the analysis approach may provide more statistical power than case-control design and provide benefits in studying rare outcome. Second, the longer follow-up period provide an opportunity to explore the risk of newly diagnostic ASD associated with postnatal exposure to long-term cumulative exposure. Third, this is the first study of the kind to be conducted in an Asian population. Race may be an important confounder that previous study had adjusted for this confounder in the relation between air pollution and ASD [4], [5]. This confounder may be neglected in Taiwan, because almost all residents in Taiwan are Han ethnicity. Our study also has several limitations. First, we were not be able to adjust for some confounders including occupational exposure, genetic factors, parental education, parental age, and birth weight [4], [5], [42][44], because there was no such information available in NHIRD. Second, we could not rule out possible misclassification of ASD. There errors are assumed to be non-differential with respect to lower exposure and higher exposure groups. Our nationwide population-based cohort study based on NHIRD has the advantage of having larger numbers of children which would reduce the uncertainly due to random error typical for smaller studies that collected detailed information on covariates. Third, previous studies have speculated that prenatal exposure to NO2 and O3 would disturb early neuromotor development, result in coordination deficits, reduce activity and reactivity, and affect neurobehavioral performances [45], [46]. Prenatal exposure to NO, NO2, PM2.5, PM10, and O3 were associated with an increased risk of autism [7], [47]. Duration of pregnancy can be seen as a sensitive period of ASD. Although, there was no detailed information on gestational age of children from NHIRD, we could not rule out the misclassification of prenatal exposure. Therefore, our primary objective is to assess the association between postnatal exposure to air pollution and the risk of newly diagnostic ASD.

In conclusion, the present study found a statistically significant association between O3, CO, and NO2 exposure and newly diagnostic ASD with an exposure-response pattern. Our finding suggests that improve ambient air quality, especially in traffic-related air pollutants and O3, might decrease the risk of newly diagnostic ASD.

Acknowledgments

We thank the Taiwan Environmental Protection Agency (EPA) for providing air pollutant monitoring data and the Bureau of National Health Insurance, Department of Health for accessing to the data of National Insurance Research Database.

Author Contributions

Conceived and designed the experiments: CRJ BFH. Performed the experiments: CRJ. Analyzed the data: CRJ YTL. Contributed reagents/materials/analysis tools: BFH. Wrote the paper: CRJ BFH.

References

  1. 1. Newschaffer CJ, Croen LA, Daniels J, Giarelli E, Grether JK, et al. (2007) The epidemiology of autism spectrum disorders. Annu Rev Public Health 28: 235–258. doi: 10.1146/annurev.publhealth.28.021406.144007
  2. 2. American Psychiatirc Association (2000) Diagnostic and Statistical Manual of Mental Disorders, 4th edition, Text Revision. Washington, DC: American Psychiatirc Association Press.
  3. 3. Rutter M (2005) Incidence of autism spectrum disorders: changes over time and their meaning. Acta Pædiatr 94: 2–15. doi: 10.1080/08035250410023124
  4. 4. Windham GC, Zhang L, Gunier R, Croen LA, Grether JK (2006) Autism spectrum disorders in relation to distribution of hazardous air pollutants in the San Francisco Bay Area. Environ Health Perspect 114: 1438–1444. doi: 10.1289/ehp.9120
  5. 5. Kalkbrenner AE, Daniels JL, Chen JC, Poole C, Emmch M, et al. (2010) Perinatal exposure to hazardous air pollutants and autism spectrum disorders at age 8. Epidemiology 21: 631–641. doi: 10.1097/ede.0b013e3181e65d76
  6. 6. Volk HE, Hertz-Picciotto I, Delwiche L, Lurmann F, McConnell R (2012) Residential proximity to freeways and in the CHARGE study. Environ Health Perspect 119: 873–877. doi: 10.1289/ehp.1002835
  7. 7. Volk HE, Lurmann F, Penfold B, Hertz-Picciotto I, McConnell R (2013) Traffic-related air pollution, particulate matter, and autism. Arch Gen Psychiatry 70: 71–77. doi: 10.1001/jamapsychiatry.2013.266
  8. 8. Bureau of National Health Insurance (2012) National Health Insurance Research Database, Taiwan. Available: http://www.nhri.org.tw/nhird/en/index.ht​m. Accessed 28 December 2012.
  9. 9. Wu CY, Chan FKL, Wu MS, Kuo KN, Wang CB, et al. (2010) Histamine2-receptor antagonists are an alternative to proton pump inhibitor in patients receiving clopidogrel. Gastroenterolgy 139: 1165–1171. doi: 10.1053/j.gastro.2010.06.067
  10. 10. Cheng CL, Kao YHY, Lin SJ, Lee CH, Lai ML (2011) Validation of the National Health Insurance Research Database with ischemic stroke cases in Taiwan. Pharmacoepidemiol Drug Saf 20: 236–242. doi: 10.1002/pds.2087
  11. 11. Lin HC, Chen YH, Lee HC, Lin HC (2010) Increased risk of acute myocardial infarction after acute episode of schizophrenia: 6 year follow-up study. Aust N Z J Psychiatry 44: 273–279. doi: 10.3109/00048670903487209
  12. 12. Rosen NJ, Yoshida CK, Croen LA (2007) Infection in the first 2 years of life and autism spectrum disorders. Pediatrics 119: e61–e69. doi: 10.1542/peds.2006-1788
  13. 13. Stroh E, Harrie L, Gustafsson S (2007) A study of spatial resolution in pollution exposure modelling. Int J Health Geogr 6: 19–19. doi: 10.1186/1476-072x-6-19
  14. 14. Leyfer OT, Folstein SE, Bacalman S, Davis NO, Dinh E, et al. (2006) Comorbid psychiatric disorders in children with autism: interview development and rates of disorders. J Autism Dev Disord 36: 849–861. doi: 10.1007/s10803-006-0123-0
  15. 15. Matson JL, Nebel-Schwalm MS (2007) Comorbid psychopathology with autism spectrum disorder in children: An overview. Res Dev Disabil 28: 341–352. doi: 10.1016/j.ridd.2005.12.004
  16. 16. Hwang YS, Weng SF, Cho CY, Tsai WH (2013) Higher prevalence of autism in Taiwanese children born prematurely: a nationwide population-based study. Res Dev Disabil 34: 2462–2468. doi: 10.1016/j.ridd.2013.05.019
  17. 17. Kleinbaum DG, Klein M (2005) Survival Analysis: A Self-Learning Text, 2nd ed. New York, NY: Springer Press.
  18. 18. James SJ, Rose S, Melnyk S, Jernigan S, Blossom S, et al. (2009) Cellular and mitochondrial glutathione redox imbalance in lymphoblastoid cells derived from children with autism. FASEB J 23: 2374–2383. doi: 10.1096/fj.08-128926
  19. 19. Chauhan A, Chauhan V, Brown WT, Cohen I (2004) Oxidative stress in autism: Increased lipid peroxidation and reduced serum levels of ceruloplasmin and transferrin - the antioxidant proteins. Life Sci 75: 2539–2549. doi: 10.1016/j.lfs.2004.04.038
  20. 20. Ming X, Stein TP, Brimacombe M, Johnson WG, Lambert GH, et al. (2005) Increased excretion of a lipid peroxidation biomarker in autism. Prostaglandins Leukot Essent Fatty Acids 73: 379–384. doi: 10.1016/j.plefa.2005.06.002
  21. 21. Zoroglu SS, Armutcu F, Ozen S, Gurel A, Sivasli E, et al. (2004) Increased oxidative stress and altered activities of erythrocyte free radical scavenging enzymes in autism. Eur Arch Psychiatry Clin Neurosci 254: 143–147. doi: 10.1007/s00406-004-0456-7
  22. 22. Yorbik O, Sayal A, Akay C, Akbiyik DI, Sohmen T (2002) Investigation of antioxidant enzymes in children with autistic disorder. Prostaglandins Leukot Essent Fatty Acids 67: 341–343. doi: 10.1054/plef.2002.0439
  23. 23. Söğüt S, Zoroğlu SS, Özyurt H, Yilmaz HR, Ozuğurlu F, et al. (2003) Changes in nitric oxide levels and antioxidant enzyme activities may have a role in the pathophysiological mechanisms involved in autism. Clin Chim Acta 331: 111–117. doi: 10.1016/s0009-8981(03)00119-0
  24. 24. Enstrom A, Krakowiak P, Onore C, Pessah IN, Hertz-Picciotto I, et al. (2009) Increased IgG4 levels in children with autism disorder. Brain Behav Immun 23: 389–395. doi: 10.1016/j.bbi.2008.12.005
  25. 25. Li X, Chauhan A, Sheikh AM, Patil S, Chauhan V, et al. (2009) Elevated immune response in the brain of autistic patients. J Neuroimmunol 207: 111–116. doi: 10.1016/j.jneuroim.2008.12.002
  26. 26. Kelly FJ (2003) Oxidative stress: its role in air pollution and adverse health effects. Occup Environ Med 60: 612–616. doi: 10.1136/oem.60.8.612
  27. 27. Rodkey FL, O′neal JD, Collison HA (1969) Oxygen and carbon monoxide equilibria of human adult hemoglobin at atmospheric and elevated pressure. Blood 33: 57–65.
  28. 28. Karalliedde L, Keshishian C (2012) Carbon monoxide poisoning. In: Baker D, Karalliedde L, Murray V, Maynard R, Parkinson NHT, editors. Essentials of Toxicology for Health Protection: A Handbook for field professionals, 2nd edition. UK: Oxford University Press. pp. 157–167.
  29. 29. Prockop LD, Chichkova RI (2007) Carbon monoxide intoxication: an updated review. J Neurol Sci 262: 122–130. doi: 10.1016/j.jns.2007.06.037
  30. 30. Zhang J, Piantadosi CA (1992) Mitochondrial Oxidative Stress after Carbon Monoxide Hypoxia in the Rat Brain. J Clin Invest 90: 1193–1199. doi: 10.1172/jci115980
  31. 31. Thom SR (1990) Carbon monoxide-mediated brain lipid peroxidation in the rat. J Appl Physiol 68: 997–1003.
  32. 32. Bayram H, Sapsford RJ, Abdelaziz MM, Khair OA (2001) Effect of ozone and nitrogen dioxide on the release of proinflammatory mediators from bronchial epithelial cells of nonatopic nonasthmatic individuals and atopic asthmatic patients in vitro. J Allergy Clin Immunol 107: 287–294. doi: 10.1067/mai.2001.111141
  33. 33. Tamagawa E, van Eeden SF (2006) Impaired lung function and risk for stroke. Chest 130: 1631–1633. doi: 10.1378/chest.130.6.1631
  34. 34. Banks WA, Farr SA, Morley JE (2002/2003) Entry of blood-borne cytokines in to the central nervous system: effects on cognitive processes. . Neuroimmunomodulation 10: 319–327. doi: 10.1159/000071472
  35. 35. Block ML, Zecca L, Hong JS (2007) Microglia-mediated neurotoxicity: uncovering the molecular mechanisms. Nat Rev Neurosci 8: 57–69. doi: 10.1038/nrn2038
  36. 36. Rivas-Arancibia S, Dorado-Martínez C, Colin-Barenque L, Kendrick KM, de la Riva C, et al. (2003) Effect of acute ozone exposure on locomotor behavior and striatal function. Pharmacol Biochem Behav 74: 891–900.
  37. 37. Pereyra-Muñoz N, Rugerio-Vargas C, Angoa-Pérez M, Borgonio-Pérez G, Rivas-Arancibia S (2006) Oxidative damage in substania nigra and striatum of rats chronically exposed to ozone. J Chem Neuroanat 31: 114–123. doi: 10.1016/j.jchemneu.2005.09.006
  38. 38. ColõÂn-Barenque L, Avila-Costa MR, Fortoul T, Rugerio-Vargas C, Machado-Salas JP, et al. (1999) Morphologic alteration of the olfactory bulb after acute ozone exposure in rats. Neurosci Lett 274: 1–4. doi: 10.1016/s0304-3940(99)00639-4
  39. 39. Avila-Costa MR, ColõÂn-Barenque L, Fortoul TI, Machado-Salas P, Espinosa-Villanueva J, et al. (1999) Memory deterioration in an oxidative stress model and its correlation with cytological changes on rat hippocampus CA1. Neurosci Lett 270: 107–109. doi: 10.1016/s0304-3940(99)00458-9
  40. 40. Santiago-López D, Bautista-Martínez JA, Reyes-Hernandez CI, Aguilar-Martínez M, Rivas-Arancibia S (2010) Oxidative stress, progressive damage in the substantia nigra and plasma dopamine oxidation, in rats chronically exposed to ozone. Toxicol Lett 197: 193–200. doi: 10.1016/j.toxlet.2010.05.020
  41. 41. Meng Z (2003) Oxidative damage of sulfur dioxide on various organs of mice: sulfur dioxide is a systemic oxidative damage agent. Inhal Toxicol 15: 181–195. doi: 10.1080/08958370304476
  42. 42. Maimburg RD, Væth M (2006) Perinatal risk factors and infantile autism. Acta Psychiatr Scand 114: 257–264. doi: 10.1111/j.1600-0447.2006.00805.x
  43. 43. Durkin MS, Maenner MJ, Newschaffer CJ, Lee LC, Cunniff CM, et al. (2008) Advanced parental age and the risk of autism spectrum disorder. Am J Epidemiol 168: 1268–1276. doi: 10.1093/aje/kwn250
  44. 44. Gillberg C (1998) Chromosomal disorder and autism. J Autism Dev Disord 28: 415–425. doi: 10.1023/a:1026004505764
  45. 45. Tabacova S, Nikiforov B, Balabaeva L (1985) Postnatal effects of maternal exposure to nitrogen dioxide. Neurobehav Toxicol Teratol 7: 785–789.
  46. 46. Sorace A, de Acetis L, Santucci D (2001) Prolonged exposure to low doses of ozone: short- and long-term changes in behavioral performance in mice. Environ Res 85: 122–134. doi: 10.1006/enrs.2000.4097
  47. 47. Becerra TA, Wilhelm M, Olsen J, Cockburn M, Ritz B (2013) Ambient air pollution and autism in Los Angeles county, California. Environ Health Perspect 121: 380–386. doi: 10.1289/ehp.1205827