Skip to main content An official website of the United States government Here's how you know Here's how you know Official websites use .gov A .gov website belongs to an official government organization in the United States. Secure .gov websites use HTTPS A lock ( Lock Locked padlock icon ) or https:// means you've safely connected to the .gov website. Share sensitive information only on official, secure websites. Search Log in Dashboard Publications Account settings Log out Search… Search NCBI Primary site navigation Search Logged in as: Dashboard Publications Account settings Log in Search PMC Full-Text Archive Search in PMC Journal List User Guide PERMALINK Copy As a library, NLM provides access to scientific literature. Inclusion in an NLM database does not imply endorsement of, or agreement with, the contents by NLM or the National Institutes of Health. Learn more: PMC Disclaimer | PMC Copyright Notice BMC Med . 2026 Mar 6;24:239. doi: 10.1186/s12916-026-04750-x Search in PMC Search in PubMed View in NLM Catalog Add to search Association between early life constituent-specific PM 2.5 exposure and child development and the moderating role of greenness: a nationwide study in China Ruili Li Ruili Li 1 Capital Center for Children’s Health, Capital Medical University, Capital Institute of Pediatrics, Beijing, China Find articles by Ruili Li 1 , Xiaoguo Zheng Xiaoguo Zheng 1 Capital Center for Children’s Health, Capital Medical University, Capital Institute of Pediatrics, Beijing, China Find articles by Xiaoguo Zheng 1 , Huimin Yang Huimin Yang 1 Capital Center for Children’s Health, Capital Medical University, Capital Institute of Pediatrics, Beijing, China Find articles by Huimin Yang 1 , Delu Yin Delu Yin 1 Capital Center for Children’s Health, Capital Medical University, Capital Institute of Pediatrics, Beijing, China Find articles by Delu Yin 1 , Tao Yin Tao Yin 1 Capital Center for Children’s Health, Capital Medical University, Capital Institute of Pediatrics, Beijing, China Find articles by Tao Yin 1 , Lihong Wang Lihong Wang 1 Capital Center for Children’s Health, Capital Medical University, Capital Institute of Pediatrics, Beijing, China Find articles by Lihong Wang 1 , Bowen Chen Bowen Chen 1 Capital Center for Children’s Health, Capital Medical University, Capital Institute of Pediatrics, Beijing, China Find articles by Bowen Chen 1, ✉ , Qingli Zhang Qingli Zhang 2 Ministry of Education - Shanghai Key Laboratory of Children’s Environmental Health, Xinhua Hospital Affiliated to Shanghai Jiao Tong University School of Medicine, Shanghai, China Find articles by Qingli Zhang 2, ✉ , Xiaoning Lei Xiaoning Lei 3 School of Public Health, Shanghai Jiao Tong University School of Medicine, Shanghai, China Find articles by Xiaoning Lei 3, ✉ Author information Article notes Copyright and License information 1 Capital Center for Children’s Health, Capital Medical University, Capital Institute of Pediatrics, Beijing, China 2 Ministry of Education - Shanghai Key Laboratory of Children’s Environmental Health, Xinhua Hospital Affiliated to Shanghai Jiao Tong University School of Medicine, Shanghai, China 3 School of Public Health, Shanghai Jiao Tong University School of Medicine, Shanghai, China ✉ Corresponding author. Received 2025 Sep 2; Accepted 2026 Feb 23; Collection date 2026. © The Author(s) 2026 Open Access This article is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License, which permits any non-commercial use, sharing, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if you modified the licensed material. You do not have permission under this licence to share adapted material derived from this article or parts of it. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by-nc-nd/4.0/ . PMC Copyright notice PMCID: PMC13077890 PMID: 41792754 Abstract Background Early life exposure to fine particulate matter (PM 2.5 ) is linked to child development, but limited studies have focused on the specific constituents of PM 2.5 , and the potential moderating role of residential greenness remains unknown. Objectives The aim of this study is to investigate the effects of constituent-specific PM 2.5 on child development and explore the moderating role of greenness. Methods This study included 8327 children aged 1–6 years from a nationwide cross-sectional survey in China. We assessed PM 2.5 and its five compositions across six exposure windows, covering the preconception, prenatal, and the year before the developmental assessment. Developmental quotients (DQ) across five domains were examined using the Children’s Developmental Scale of China to assess the child development. Greenness was estimated using the Normalized Difference Vegetation Index within a 1000-m buffer (NDVI 1000m ). Multivariable linear regression models were used to evaluate the associations of PM 2.5 , its constituents, and child DQ, as well as the effect modification by NDVI 1000m . Results PM 2.5 and its constituents were negatively associated with child DQ in motor, adaptability, and social skills. These effects varied by DQ domains and constituents, fine and gross motor domains showed more consistent negative associations, and stronger associations were observed for black carbon (BC), organic matter, and sulfates. Especially, PM 2.5 and all five compositions across six windows were consistently associated with lower fine motor DQ [e.g., β = − 4.66, 95% CI: − 5.45, − 3.87, for a 1-ln μg/m 3 increase in PM 2.5 in the year preceding the test]. Additionally, significant interactions were found between PM 2.5 , its constituents, and greenness. For example, BC exposure during the year before the test was related to greater decrements in fine motor DQ among children with low NDVI 1000m ( β = − 5.01, 95% CI: − 6.01, − 4.00), compared to those with high NDVI 1000m ( β = − 1.74, 95% CI: − 3.12, − 0.36). Conclusions PM 2.5 and its constituents were inversely linked to child development across motor, adaptability, social domains, with greenness potentially serving as a protective factor. Graphical Abstract Supplementary Information The online version contains supplementary material available at 10.1186/s12916-026-04750-x. Keywords: PM 2.5 , Chemical constituents, Child development, Greenness, Motor function Highlights PM 2.5 components were linked to poorer early life motor, adaptability, social skills. Early life children’s motor development showed more consistent negative associations. Stronger associations were observed for black carbon, organic matter, and sulfates. Greenness may mitigate the impacts of PM 2.5 and its constituents on child development. Supplementary Information The online version contains supplementary material available at 10.1186/s12916-026-04750-x. Background Early childhood is a critical developmental period that encompasses motor, language, cognitive, and social skills, which are key determinants of later health, educational attainment, and overall well-being. Children, especially during their early years, beginning from the in-utero period, are more vulnerable to surrounding environmental factors due to their immature defense systems, such as the nose-brain and blood–brain barriers [ 1 ]. Numerous environmental exposures, including air pollution, have been identified as major contributors to adverse child development and cognitive outcomes. In particular, accumulating evidence suggests links between fine particulate matter (PM 2.5 ) and structural changes in the brain [ 2 ], cognitive impairment [ 3 , 4 ], IQ loss [ 5 ], and developmental delay [ 6 ]. Nevertheless, some heterogeneity in the evidence to date remains [ 7 – 9 ], which may be partly due to geographic variation in regional chemical constituents [ 10 – 12 ]. PM 2.5 is a mixture consisting of widely variable constituents with diverse physical and chemical properties, such as black carbon (BC), organic matter (OM), and sulfates (SO 4 2− ) [ 13 ]. Several studies have explored the associations between exposure to specific PM 2.5 constituents and child development [ 14 – 20 ]. However, research remains limited, with most focusing on maternal exposure during pregnancy. The human brain undergoes rapid development before the age of 6 [ 21 , 22 ], and is highly sensitive to the exposure windows due to time-dependent nature of brain development [ 7 ]. Therefore, further studies exploring the hazardous constituents of PM 2.5 , while considering exposure in multiple critical windows of susceptibility, may provide valuable insights for targeted interventions to prevent developmental delay. Simultaneously, residential green space (e.g., grasslands, forests, parks, and gardens) represents another important surrounding environment factor. In recent years, research on its relationship with child development has grown significantly, including studies on white matter microstructure [ 23 ], cognitive performance [ 24 , 25 ], motor function, externalizing and internalizing symptoms [ 26 – 28 ], and autism spectrum disorders [ 29 ]. Notably, areas with higher greenspace levels generally exhibit lower concentrations of air pollutants, including total PM 2.5 and its constituents such as BC [ 24 , 30 , 31 ]. This is likely attributable to the ability of vegetation to capture ambient air pollutants through leaf stomata [ 32 ]. Nevertheless, previous studies often examined the effects of air pollution and greenspace on child development separately, focusing on their individual effects or treating greenspace as a confounder or mediator in the association between air pollution and child development. To date, only a limited number of studies have explored the interactive effects of total PM 2.5 , its constituents, and residential greenness on other health outcomes, including metabolic syndrome [ 33 ], childhood overweight or obesity [ 34 ], prediabetes and diabetes [ 35 ], and blood flow velocity [ 36 ]. Research investigating whether greenness can modify the impact of specific PM 2.5 constituents on child development remains lacking. To address these research gaps, we conducted a nationwide survey of children aged 1–6 years in China. Our objective was to provide a comprehensive analysis of the effects of PM 2.5 and its constituents during the preconception, prenatal, and the year before the test on childhood development. Additionally, we hypothesized that residential greenness might mitigate these harmful effects. Methods Study population This study was a cross-sectional study using data from a nationally representative survey on child development in China, part of the National Nutrition and Health Systematic Survey for Children, conducted between 2014 and 2021. A multi-stage stratified cluster random sampling strategy was employed to select a nationally representative sample of children aged 1–6 years. Detailed information on the survey design is available in a previous publication [ 37 ]. Eligibility criteria for inclusion in the survey were as follows: (1) residence in the survey area for more than 6 months; (2) no congenital, neurological, or genetic diseases; and (3) no acute or chronic illness. In total, 8843 eligible children were enrolled. In addition, 124 children from twin or multiple pregnancies and 154 children with extremely low (< 70, n = 63) or extremely high (> 130, n = 91) developmental quotient (DQ) scores were excluded. Twin and multiple pregnancies were excluded due to their higher likelihood of preterm birth, low birth weight, other perinatal complications, and distinct developmental outcomes, which differentiate them from singletons. To ensure sample homogeneity and the representativeness of the general child population, analyses were restricted to singleton births. Participants with missing values for key covariates used in the main regression analyses, including region ( n = 13), ethnicity ( n = 17), season at conception ( n = 29), and maternal age at conception ( n = 203), were also excluded. Finally, 8327 eligible children were enrolled, all of whom completed a questionnaire and underwent a developmental examination. The geographical distribution of participants spanned 16 provinces or municipalities (Additional file 1: Fig. S1). The study protocol was approved by the Ethics Committee of the National Institute for Nutrition and Health, Chinese Center for Disease Control and Prevention (No. 2019–009). Written informed consent was obtained from all participants or their legal guardians. Exposure assessment BC, OM, and secondary inorganic ions, including SO 4 2− , nitrates (NO 3 − ), and ammonium (NH 4 + ), are the main constituents of PM 2.5 . In this study, daily concentrations of PM 2.5 and its main constituents were derived from a full-coverage, near-real-time dataset of PM 2.5 chemical composition with a spatial resolution of 0.1° × 0.1° available since 2000. This dataset is part of the Tracking Air Pollution in China product ( http://tapdata.org.cn ) [ 38 , 39 ]. The concentrations were estimated by integrating the Weather Research and Forecasting–Community Multiscale Air Quality modeling system, ground observations, a machine learning algorithm, and multisource-fusion PM 2.5 data. The estimated concentrations of PM 2.5 and its constituents show good agreement with available observations, with correlation coefficients ranging from 0.67 to 0.80 for the period 2013 to 2020, and most normalized mean biases are within ± 20%. Concentrations of constituent-specific PM 2.5 were extracted for each participant based on the longitude and latitude of their residential address. Time-weighted average exposure levels for six exposure windows were calculated: one during the preconception period (12 weeks before pregnancy), four during the prenatal period (the first trimester, T1; the second trimester, T2; the third trimester, T3; and the entire pregnancy), and one representing recent exposure during the year prior to the development assessment (1 year before test). Outcomes In this study, child development was assessed by trained child health care physicians using the Children’s Developmental Scale of China (CDSC) together with its standardized toolbox. The CDSC (WS/T 580–2017) is a national health-industry standard developed by the Capital Institute of Pediatrics in China. It has shown good internal consistency (Cronbach’s α = 0.85–0.954) and split-half reliability (0.890–0.968) among Chinese children aged 0–6 years [ 40 , 41 ] and is widely used in pediatric and maternal-child health-care institutions nationwide. The scale is age-specific, consisting of discrete test sets for each age interval—monthly for 0–12 months, every 3 months for 1–3 years, and every 6 months for 4–6 years. Each age-specific set assesses five developmental domains: gross motor, fine motor, adaptability, language, and social behavior. Each domain includes 1–2 items, and each age-specific set contains 8 to 10 items, with the full scale comprising 261 items in total. Scoring is assigned by domain and varies by age group: 1 point per domain for 0–12 months, 3 points for 1–3 years, and 6 points for 4–6 years. Testing is dynamically administered—it begins with the child’s age-appropriate set (the “main test month-age”) and proceeds both backward and forward until the child passes or fails two consecutive age sets in each domain. The highest age level consistently passed defines that domain’s mental age (MA). The DQ was calculated as the ratio of the child’s MA to their chronological age, multiplied by 100 (DQ = MA/chronological age in month × 100). This score represents a relative developmental index rather than an absolute ability measure, reflecting each child’s developmental level compared with what is expected for their chronological age. It allows for standardized comparisons across ages and developmental domains, even when sensitivity may differ across domains. According to the CDSC, children with a DQ > 130 in any domain were considered to have high intellectual function, whereas those with a DQ < 70 were considered to have intellectual disability. To ensure that the analyses reflected associations within the range of typical development, these participants were excluded [ 40 ]. Potential confounders Sociodemographic information of study population was collected through interview-administrated questionnaires, including region (North vs. South China), ethnicity (Han vs. non-Han), maternal age at conception (continuous) [ 42 ], household monthly income per capita (≤ 1500, 1501–3000, 3001–5000, > 5000 Yuan/month), maternal education (Junior school or below; High school; College or above), season of conception (Spring, Summer, Autumn, Winter), breastfeeding duration (≥ 6 months vs. < 6 months), parity (Primiparous vs. Multiparous), delivery mode (Cesarean vs. Vaginal), gestational age (continuous in weeks), child sex (Boys vs. Girls), age at development examination (continuous), and birth weight (continuous). Meteorological factors, including daily mean ambient temperature (Temp) and relative humidity (RH), were also obtained from AgERA5 re-analysis dataset based on hourly ECMWF ERA5 surface data. The number and proportion of missing values for each variable are reported in Table 1 . Table 1. Characteristics of study population and child developmental quotient ( n = 8327) Variables n (%)/Mean ± SD Mother Maternal age at conception (years) 28.57 (4.45) Gestational age (weeks) 39.15 (1.35) Ethnicity Han 7668 (92%) Non-Han 659 (7.9%) Region North 4187 (50%) South 4140 (50%) Maternal education Junior school or below 1502 (18%) High school 3928 (47%) College or above 2897 (35%) Household incomes per capita (Yuan/month) ≤ 1500 281 (3.4%) 1501–3000 1022 (12%) 3001–5000 3856 (46%) > 5000 3168 (38%) Season at conception Spring (March–May) 2177 (26%) Summer (June–August) 2040 (24%) Autumn (September–November) 2285 (27%) Winter (December–February) 1825 (22%) Delivery mode Vaginal 4862 (58%) Cesarean 3465 (42%) Parity Primiparous 5705 (69%) Multiparous 2621 (31%) Gestational diabetes mellitus 120 (1.4%) Gestational hypertension 57 (0.7%) Child Age (years) 3.29 (1.65) Sex Boys 4144 (50%) Girls 4183 (50%) Birth weight (g) 3323 (440) Breastfeeding duration (months) < 6 1987 (24%) ≥ 6 6339 (76%) Year of the survey 2014 3495 (42%) 2019 1831 (22%) 2020 1549 (19%) 2021 1452 (17%) Developmental quotient Gross motor 104.86 (14.45) Fine motor 94.18 (14.07) Adaptability 99.01 (14.71) Language 99.21 (15.87) Social behavior 100.16 (15.93) Open in a new tab Continuous variables are presented as mean (± SD); categorical variables are presented as number (%). Missing values for breastfeeding duration ( n = 1), parity ( n = 1), birth weight ( n = 32) were observed Effect modifiers In this study, the satellite-derived Normalized Difference Vegetation Index (NDVI) was used as an indicator of greenness availability to assess residential surrounding greenness. Satellite images with a spatial resolution of 250 m over a 16-day period were obtained from a Resolution Imaging Spectroradiometer (MODIS) Vegetation Indices (MOD13Q1) [ 43 ]. NDVI is a dimensionless metric that quantifies the difference of absorption between near-infrared and red light, with values ranging from − 1 to 1. Higher values indicate denser vegetated zone [ 44 ]. The average NDVI was calculated within buffer radius of 500 m (NDVI 500m ), 1000 m (NDVI 1000m ), and 2000 m (NDVI 2000m ) surrounding the residential addresses of each child. Statistical analyses The characteristics of the study population and the distribution of DQ scores were described using mean ± standard deviation (SD) or n (%). The exposure distribution was presented by selected percentiles. Due to the right-skewed distribution, constituent-specific PM 2.5 concentrations were naturally log-transformed. In the main analyses, multivariable linear regression models were fitted to evaluate the association between constituent-specific PM 2.5 exposure and DQ scores in five domains for six exposure windows separately. We estimated the change in DQ scores and 95% confidence intervals (CIs) associated with one-unit increase in naturally log-transformed PM 2.5 and constituents concentrations. Similarly, multivariable linear regression models were also fitted to examine the association between NDVI 1000m (continuous) and DQ scores, with estimates reported as the change in DQ scores and 95% CIs for each 0.2 unit increase in NDVI 1000m . NDVI 1000m was selected as the primary indicator of residential greenness, reflecting both the neighborhood environment during pregnancy and the accessible outdoor spaces for young children after birth. This buffer size balances maternal and child environmental exposure and is consistent with previous studies [ 45 – 47 ]. As in previous studies [ 16 , 26 ], potential confounders, including region, ethnicity, maternal age at conception, household income, maternal education level, season of conception, Temp, RH, and child sex, were adjusted for in all models. The Benjamini–Hochberg false discovery rate (FDR) procedure was used for the multiple hypothesis testing correction, and an FDR value < 0.05 was considered statistically significant. We further investigated the interactive effects by incorporating an interaction term between NDVI 1000m and constituent-specific PM 2.5 , followed by stratified analyses based on NDVI 1000m exposure (low and high subgroups classified by the median). Similarly, we explored potential effect modifications by region and child sex using interaction terms and stratified analyses. Additionally, separate analyses were conducted for younger children (1–2 years, infant period) and older children (3–6 years, preschool period). To confirm the robustness of the main findings of associations between constituent-specific PM 2.5 exposure and child development, several sensitivity analyses were performed: (1) further adjusting for the calendar year of the survey, parity, delivery mode, breastfeeding duration, child age at the development examination, child’s birth weight to minimize its confounding effects; (2) excluding children born prematurely (gestational age < 37 weeks) or those with low birth weight (birth weight < 2500 g); (3) excluding mothers with pregnancy complications (gestational diabetes mellitus or gestational hypertension). Additionally, we evaluated the association between child DQ and greenness using two additional radial buffers (i.e., NDVI 500m , NDVI 2000m ). Exposure assessment and statistical analysis were conducted using ArcGIS Pro (version 3.3.2) and R software (version 4.3.1, R Development Core Team). A two-sided P value < 0.05 was considered statistically significant, including for interaction terms ( P int , P value for interaction terms). Results Characteristics of participants and child DQ Basic characteristics of the study population are summarized in Table 1 , with continuous variables presented as mean (SD). Among the 8327 children, half were from the northern region of China, and the majority were of Han ethnicity. The mean maternal age at conception was 28.57 (4.45) years, and the mean gestational age was 39.15 (1.35) weeks. Over one-third of the mothers had attained a college education or above (35%) and 38% had household monthly incomes per capita > 5000 Yuan. The majority of deliveries were vaginal (58%), and 69% of the mothers were primiparous. The sample included 4144 boys and 4183 girls, with a mean birth weight of 3323 ± 440 g and a mean age of 3.29 ± 1.65 years. As shown in Table 1 , the mean DQ scores for gross motor, fine motor, adaptability, language, and social behavior were 104.86 (14.45), 94.18 (14.07), 99.01 (14.71), 99.21 (15.87), and 100.16 (15.93), respectively. Exposure characteristics Additional file 1: Table S1 presents the concentrations of PM 2.5 and its compositions across different exposure windows. Due to the right-skewed distribution of PM 2.5 and its constituents, their concentrations were natural log-transformed prior to regression analyses. For interpretability, untransformed median concentrations are presented in Additional file 1: Table S1. The median PM 2.5 concentrations during preconception, pregnancy, and the year before the test were 56.78, 58.74, and 47.79 μg/m 3 , respectively, with corresponding values of 57.25, 56.86, and 54.69 μg/m 3 for T1, T2, and T3. Among the five compositions, OM had the highest concentrations (11.16 ~ 14.56 μg/m 3 ), followed by SO 4 2− (8.55 ~ 10.85 μg/m 3 ) and NO 3 − (8.66 ~ 10.52 μg/m 3 ), then NH 4 + (5.70 ~ 7.33 μg/m 3 ), and BC (2.09 ~ 2.99 μg/m 3 ) as the lowest. This trend was consistent across all exposure windows. The median NDVI 1000m was 0.24 in the year before the test and 0.23 in the other five windows (Additional file 1: Table S2). Associations between constituents-specific PM 2.5 and child DQ Figure 1 and Additional file 1: Table S3 show the associations between child DQ across five domains and exposure to PM 2.5 and its five compositions during six windows, which encompass the preconception, pregnancy, and recent periods. All effect estimates represent changes in DQ scores per 1-ln unit increase in exposure. Across all six exposure windows, PM₂.₅ was negatively associated with gross and fine motor DQ, but positively associated with language DQ. For adaptability, negative associations appeared from T2 onward, whereas for social behavior, they were mainly observed in late pregnancy (T3) and the year before the test, suggesting that these later periods may represent more sensitive exposure windows. Fig. 1. Open in a new tab Associations between PM 2.5 , its constituents, and developmental quotient among children aged 1–6 years in China. All models adjusted for region, ethnicity, maternal age at conception, household income, maternal education level, season of conception, Temp, RH, and child’s sex Consistent with this pattern, BC and OM exhibited broadly similar associations to PM 2.5 . Both constituents across all six windows were linked to lower DQ in the gross motor, fine motor, and adaptability domains, except for BC exposure during preconception, which showed no association with adaptability. In contrast, language DQ was positively associated with both constituents—for BC from preconception through all pregnancy windows, and for OM during mid to late pregnancy (T2–T3). For social behavior, OM exposure during late and entire pregnancy, and both BC and OM in the year before the test, were associated with lower DQ. Overall, OM exhibited broader and slightly stronger adverse associations than BC (Fig. 1 and Additional file 1: Table S3). As for the secondary inorganic ions (SO 4 2− , NO 3 − , and NH 4 + ), exposures across all six windows were associated with lower fine motor DQ for all ions and with lower gross motor DQ for SO 4 2− . NO 3 − and NH 4 + exposures were associated with lower gross motor DQ from T2 and T3 onward, respectively. Negative associations were also observed for adaptability and social behavior, especially in the year before the test for all three ions. Additionally, SO 4 2− and NO 3 − showed effects at T3 and throughout pregnancy for adaptability and social behavior, respectively. In contrast, language DQ was positively associated with all three ions across exposure windows (Fig. 1 and Additional file 1: Table S3). The moderating role of greenness Figure 2 and Additional file 1: Table S4 show the associations between NDVI 1000m during six exposure windows and child DQ across five domains. A 0.2 unit increment in NDVI 1000m across all six exposure windows was associated with higher DQ in the gross and fine motor domains. However, for the language domain, NDVI 1000m exposure during five of the six exposure windows, excluding the year before the test, was inversely associated with language DQ. No significant associations were observed between NDVI 1000m exposure and DQ in the adaptability or social behavior domains for any exposure windows. Fig. 2. Open in a new tab Associations between NDVI 1000m greenness exposure and child developmental quotient across six exposure windows. All models adjusted for region, ethnicity, maternal age at conception, household income, maternal education level, season of conception, Temp, RH, and child’s sex As shown in Fig. 3 and Additional file 1: Table S5, significant interaction effects between PM 2.5 and NDVI 1000m on child DQ were observed during pregnancy and the year before the test in the gross motor ( P int : 0.004 ~ 0.005), fine motor ( P int : < 0.001 ~ 0.017), and adaptability domains ( P int : < 0.001 ~ 0.047). Similar interaction effects were also found between all PM 2.5 constituents and NDVI 1000m , especially BC, NO 3 − , and SO 4 2− . Additionally, significant interaction effects were observed during the preconception period (for BC, NO 3 − , and SO 4 2− ), as well as for the language (BC, OM, NH 4 + , and SO 4 2− ) and social behavior domains (BC, NH 4 + , and SO 4 2− ). Fig. 3. Open in a new tab Modification of NDVI on the association between PM 2.5 , its constituents, and child developmental quotient across six exposure windows. All models adjusted for region, ethnicity, maternal age at conception, household income, maternal education level, season of conception, Temp, RH, and child’s sex Greenness stratification analyses further revealed that children with low NDVI 1000m exposure were more sensitive to PM 2.5 and its constituents. For instance, a 1-ln unit increase in PM 2.5 and BC during the year before the test was related to greater decrements in fine motor DQ among children with low NDVI 1000m exposure (− 6.42, 95% CI: − 7.64, − 5.21 for PM 2.5 ; − 5.01, 95% CI: − 6.01, − 4.00 for BC), compared to those with high NDVI 1000m exposure (− 2.72, 95% CI: − 4.15, − 1.28 for PM 2.5 ; − 1.74, 95% CI: − 3.12, − 0.36 for BC). For language DQ, PM 2.5 and its major constituents were predominantly positively associated across most exposure windows, particularly in the high NDVI 1000m group. In the low NDVI 1000m group, positive associations were mainly observed during pregnancy, whereas inverse associations appeared in the year before the test. Additionally, secondary inorganic ions during the 1 st (NH 4 + and SO 4 2− ) and 2nd (NO 3 − , and NH 4 + ) trimesters showed positive associations with child adaptability DQ among children with high NDVI 1000m exposure. Stratified and sensitive analyses Stratification analyses indicated that the associations between PM 2.5 and its constituents and child DQ varied by the region, sex, and age (Additional file 1: Tables S6–S8). For example, the effects of PM 2.5 exposure during the year before the test on fine motor DQ were stronger in children living in northern China (− 7.37, 95% CI: − 9.02, − 5.72 for the north vs. − 4.17, 95% CI: − 5.18, − 3.15 for the south), boys (− 5.46, 95% CI: − 6.58, − 4.34 for boys vs. − 3.85, 95% CI: − 4.96, − 2.74 for girls), and older children (− 7.29, 95% CI: − 8.27, − 6.31 for 3 − 6 years vs. − 2.01, 95% CI: − 3.22, − 0.79 for 1–2 years). These trends were consistent across PM 2.5 constituents, developmental domains, and exposure windows. The main findings remained largely unchanged after FDR correction (Additional file 1: Table S3), after additional confounder adjustment (Additional file 1: Table S9), and after excluding preterm or low-birth-weight children (Additional file 1: Table S10), or mothers with pregnancy complications (Additional file 1: Table S11). A similar association between NDVI 1000m and child DQ was observed when using other radial buffers for NDVI estimation (i.e., NDVI 500m and NDVI 2000m ) (Additional file 1: Table S12). Discussion In this study, we demonstrate negative associations of PM 2.5 and its five constituents with child DQ across multiple domains, with the exception of language, among children aged 1–6 years. These effects varied by constituents, exposure windows, and DQ domains. Collectively, compared to other domains, the fine and gross motor domains showed more consistent negative associations. Among the five constituents, stronger associations were observed for BC, OM, and SO 4 2− . In addition, this is the first study to provide evidence on the modifying effects of greenness on the associations between constituent-specific PM 2.5 and child development. More apparent associations were observed in children exposed to lower level of greenness, suggesting that greenness may play a protective role in child development. We assessed a wide range of child developmental domains, including gross motor, fine motor, adaptability, language, and social behavior. Our findings revealed a negative association between PM 2.5 exposure and four of these domains (except language), with highly consistent results in fine and gross motor domains across all six exposure windows. These results are consistent with a growing body of epidemiological studies worldwide [ 16 , 48 – 54 ]. The consistent and robust associations observed in the fine and gross motor domains may indicate that these domains are more strongly associated with air pollution exposure during early development. Early motor development depends heavily on the integrity and rapid maturation of the central and peripheral nervous systems, including processes such as myelination, synaptogenesis, and neuromuscular coordination. These processes occur intensively during infancy and early childhood and are therefore highly susceptible to air-pollution–induced oxidative stress, systemic inflammation, and hypoxia, which can disrupt these neurophysiological processes and impair the integrity of neural pathways involved in motor control [ 55 – 57 ]. Unexpectedly, we observed positive associations between PM 2.5 exposure across all six windows and language function in the overall population and the younger subgroup (1–2 years), whereas a trend toward a negative association emerged in the older subgroup (3–6 years). One plausible explanation is that language development may be largely influenced by genetic factors and the early-life environmental conditions, including sociodemographic factors, cognitive stimulation, and sensory input [ 58 ]. The positive associations observed may also partly reflect residual confounding related to urbanicity or socioeconomic status, since families living in more urbanized or advantaged areas often experience higher air pollution levels but also have greater access to educational resources, healthcare, and language-rich environments, potentially enhancing children’s language skills. Moreover, potential selection bias or exposure misclassification may also contribute—for example, children living in more polluted areas might spend more time indoors engaging in educational or reading activities, potentially offsetting the adverse effects of pollution exposure. In addition, measurement error in the language domain of the CDSC cannot be entirely ruled out, as this assessment partly relies on parental report and examiner interpretation. Our results partially align with previous studies. For instance, in older children (3–6 years), a birth cohort study from the INMA project in Spain found that PM 2.5 exposure during pregnancy had a negative effect on verbal development in boys at the 4–6 year follow-up [ 59 ]. In contrast, for younger children (< 2 years), the ECLIPSES study in Spain reported a nonsignificant positive association between maternal exposure to PM 2.5 absorbance during pregnancy and language function in infants at 40 days of age [ 8 ]. However, a birth cohort study in Mexico city revealed that each 1 µg/m 3 increase in PM 2.5 exposure during pregnancy was associated with a decrease in language function by − 0.38 points (95% CI: − 0.77, − 0.01) in children up to 24 months of age [ 60 ]. The discrepancy in studies examining language outcomes among children under 2 years may be attributed to the greater variability in developmental trajectories at this early stage. We further investigated the effects of PM 2.5 constituents, including BC, OM, and secondary inorganic ions (NO 3 − , NH 4 + , SO 4 2− ), which vary in particle size, pollution concentration, emission source, and toxicity. Similar to total PM 2.5 , each constituent exhibited negative effects on child gross motor, fine motor, adaptability, and social behavior. Although the patterns varied by constituent, outcome domain, and exposure windows, constituents related to fossil fuel combustion (i.e., BC, OM, and SO 4 2− ) consistently demonstrated negative associations with child DQ in gross and fine motor domains across all six windows. These results were partially consistent with most of the limited existing literature on the associations between gestational PM 2.5 constituents and child development [ 16 , 17 , 19 , 61 ]. For example, a recent birth cohort study in China revealed that PM 2.5 and its constituents (i.e., BC, OM, NO 3 − , NH 4 + , SO 4 2− ) during the early postnatal period were associated with developmental delays in gross motor, fine motor, and problem-solving domains among children under 2 years of age, although no significant associations were observed during the prenatal period [ 15 ]. Similarly, an urban pregnancy cohort in the Northeastern United States reported that exposure to ambient air pollutant mixtures was linked to poorer memory function and increased attention problems in school-age children at 6.5 years, with BC, OM, and SO 4 2− being the major contributors to these associations [ 14 ]. Additionally, a prospective cohort study of 4494 mother–child pairs demonstrated that exposure to PM 2.5 and its four main components (i.e., OM, NO 3 − , NH 4 + , SO 4 2− ) during all three trimesters increased the risk of developmental and behavioral problems in children, with OM and SO 4 2− being the primary contributors to these effects [ 20 ]. On the other hand, similar to total PM 2.5 , all of its constituents showed positive effects on child language DQ, with the three secondary inorganic ions (NO 3 − , NH 4 + , SO 4 2− ) demonstrating particularly consistent associations across all six windows in this study. Further research is warranted to replicate and clarify the constituent-specific effects of PM 2.5 on child development, particularly in the language domain. Most studies have reported that exposure to greenness is associated with improved neurocognitive function in children [ 25 , 27 , 28 , 62 , 63 ], although some studies have found inverse or nonsignificant associations [ 24 ]. For example, a multi-centric birth cohort study conducted across six European countries revealed that greater green space exposure was associated with higher inattentiveness scores [ 64 ]. A recent study of three Canadian birth cohorts found no evidence of an association between NDVI 500m during pregnancy and cognitive or motor function, or with externalizing and internalizing symptoms in children aged 2 to 5 years [ 26 ]. The reasons for the heterogeneity of these findings remain unclear. In the present study, our findings demonstrated that NDVI 1000m exposure across all six exposure windows was associated with improved fine and gross motor function, although its association with language ability was less consistent, showing significantly lower scores across all exposure windows except the year before the test. Furthermore, to the best of our knowledge, our study is the first to reveal substantial interactions between constituent-specific PM 2.5 and greenness in relation to child development. Specifically, children living in less green areas showed stronger negative associations between PM 2.5 (and its five constituents) and developmental outcomes across multiple domains, with the language domain showing predominantly significant and positive associations in areas with higher greenness, whereas these associations were weaker or even negative in less green areas. Although the precise mechanisms underlying the modifying role of greenness are not yet fully understood, several potential pathways have been proposed. Green vegetation may mitigate some developmental risks linked to air pollution by reducing airborne deleterious substances through dry deposition processes. Beyond pollutant reduction, greenness may also buffer developmental harm by mitigating maternal stress during pregnancy through cortisol regulation, involving modulation of the hypothalamic–pituitary–adrenal axis, and by providing sensory stimulation within natural environments, which offer rich visual, auditory, and olfactory inputs that foster emotional restoration, attention recovery, and cognitive engagement [ 65 , 66 ]. Furthermore, exposure to green spaces may strengthen resilience to environmental stressors by encouraging outdoor physical activity, exercise, and social interactions [ 67 , 68 ]. Taken together, these pathways suggest that greenness may directly or indirectly buffer developmental harm and warrant further investigation into its modifying role in the association between air pollution and child development. In line with the majority of previous epidemiological and experimental studies [ 58 , 59 , 69 – 72 ], our sex-specific analyses revealed that PM 2.5 and its constituents exhibited stronger adverse effects among boys. This may be linked to oxidative stress and inflammation, the primary toxic mechanisms of PM 2.5 exposure [ 57 ]. Animal studies suggest that air pollution could lead to sexually dimorphic fetal programming of neuroinflammatory responses [ 73 ], and sex hormones may further contribute to these sex-specific developmental differences [ 74 ]. To our knowledge, few studies have compared the effects of PM 2.5 and its constituents on child development across different age groups. Older children (3–6 years) appeared to be at higher vulnerability to developmental risk compared to younger children (1–2 years) in our study, which may be explained by the fact that child development processes become more sophisticated and refined in older age groups, requiring advanced neural coordination across multiple brain regions [ 75 ]. This may also reflect the cumulative adverse effects of prolonged early-life exposure. Moreover, we observed more pronounced effects in children living in northern region of China, which could be attributed to spatial variations in PM 2.5 concentrations and chemical constituent characteristics between the northern and southern regions of China [ 39 ]. These findings may help to identify subpopulations that are more strongly associated with adverse outcomes. Our findings suggest potential implications for public health and environmental policy. The observed modifying effect of surrounding greenness suggests that enhancing urban vegetation may be a practical strategy to mitigate the adverse effects of air pollution on early child development, particularly in highly polluted areas. In parallel, children living in areas with lower greenness or higher PM 2.5 concentrations—especially boys and those residing in northern China—may constitute more vulnerable populations who warrant closer attention in public health planning. These findings underscore the importance of integrating air quality improvement efforts with equitable access to green space, as well as the need for further longitudinal and mechanistic research to better elucidate these relationships and inform future interventions. The present study has several notable strengths. First, we conducted a large-scale, nationally representative survey over the period from 2014 to 2021, encompassing an ethnically and geographically diverse population from both northern and southern regions of China. This diversity, coupled with the extended study period, facilitated substantial heterogeneity in exposure levels to PM 2.5 , its constituents, and greenness, thereby enhancing the statistical power and generalizability of our findings across China. Second, child development was assessed using age-appropriate, validated assessment tools specifically tailored to Chinese children, covering a range of developmental domains. To the best of our knowledge, this study is the first to examine the moderating role of residential greenspace in the association between constituent-specific PM 2.5 exposure and child development, offering novel insights into the field. However, we also acknowledge several limitations in this study. First, a major limitation of this study is the age heterogeneity of the participants, aged 1–6 years, a critical period marked by rapid and dynamic neurodevelopment, which may lead to differential susceptibility to environmental exposures across distinct developmental stages. The use of the CDSC’s age-specific test sets and scoring system, along with our age-stratified analyses by developmental stage, helps mitigate the potential impact of this age variability. However, future research with more homogeneous age groups could provide more refined insights into the developmental impacts of environmental exposures. Second, exposure assessment was based on outdoor concentration at child residential addresses, and we only enrolled participants who had lived at the reported location for at least six months before the survey. However, the lack of indoor exposure data and potential residential relocation prior to this 6-month period were not accounted for, which may have resulted in incomplete assessment of actual exposure levels. Third, the use of NDVI as a proxy for greenspace may not fully reflect the accessibility, types, and quality of the surrounding green spaces, and could not distinguish true vegetation from artificial or non-natural green surfaces, potentially limiting the accuracy of our greenspace assessment. This limitation also prevents analyzing how greenspace characteristics influence child development or modify the effects of air pollution. Fourth, we used window-specific and single-pollutant models to examine the associations between PM 2.5 exposures and child developmental outcomes. However, mutual adjustment across exposure windows or among PM 2.5 components was not performed. As such, potential confounding between exposure windows, or by other constituents and total PM 2.5 mass, cannot be entirely ruled out. The observed associations should therefore be interpreted with caution, and future studies are warranted to disentangle the independent effects of specific exposure windows and individual PM 2.5 constituents. Fifth, although we adjusted for a range of confounders, information on certain individual lifestyle, behavioral, and psychosocial factors (e.g., maternal smoking, alcohol consumption, physical exercise, and maternal stress) as well as environmental co-exposures (e.g., noise) was unavailable, which may have contributed to residual confounding. Lastly, although our findings are based on a large and diverse Chinese sample, generalizability to other regions (e.g., high-income countries) should be interpreted with caution. Variations in PM 2.5 sources and composition (e.g., coal combustion vs. traffic emissions), vegetation types, and urban infrastructure may lead to differences in the observed associations between air pollution, greenness, and child development. Conclusions In summary, our findings demonstrate that increased exposure to PM 2.5 and all five of its constituents is associated with lower scores in multiple domains of child development, except language, with particularly strong associations observed for BC, OM, and SO 4 2− . Notably, fine motor function was consistently affected by PM 2.5 and its five constituents across all six exposure windows, spanning the preconception, prenatal, and the year before the test. Furthermore, residential greenspace was associated with weaker adverse associations, suggesting a potential protective role. These results indicate that reducing exposure to PM 2.5 , particularly its toxic constituents, along with enhancing greenspace, may be associated with improved children’s neurocognitive development. Supplementary Information 12916_2026_4750_MOESM1_ESM.docx (2.2MB, docx) Additional file 1: Table S1. Distribution of PM 2.5 and its compositions across six exposure windows. Table S2. Distribution of greenness, temperature, and relative humidity across six exposure windows. Table S3. Associations between PM 2.5 , its constituents, and child developmental quotient. Table S4. Associations between NDVI 1000m greenness exposure and child developmental quotient across six exposure windows. Table S5. Modification of NDVI on the association between PM 2.5 , its constituents and child developmental quotient across six exposure windows. Table S6. Regional modification effects on the association between PM 2.5 , its compositions, and child developmental quotient. Table S7. Sex modification effects on the association between PM 2.5 , its compositions, and child developmental quotient. Table S8. Age-stratified associations between PM 2.5 , its compositions, and child developmental quotient. Table S9. Associations between PM 2.5 , its compositions, and child developmental quotient in regression models adjusted for additional confounders. Table S10. Associations between PM 2.5 , its compositions, and developmental quotient among children excluding those with premature birth or low birth weight. Table S11. Associations between PM 2.5 , its compositions, and developmental quotient among children whose mothers had no pregnancy complications. Table S12. Associations between greenness within different buffer zones and child developmental quotient across six exposure windows. Figure S1. Residential address locations of study population. Acknowledgements We would like to express our gratitude to all the study participants and the investigators involved in this study. Abbreviations BC Black carbon CDSC The Children’s Developmental Scale of China CI Confidence interval DQ Developmental quotients FDR False discovery rate MA Mental age NH 4 + Ammonium NO 3 − Nitrates NDVI Normalized Difference Vegetation Index OM Organic matter P int P Value for interaction terms PM 2.5 Fine particulate matter RH Relative humidity SD Standard deviation SO 4 2 − Sulfates TAP Tracking Air Pollution Temp Temperature T1 The first trimester T2 The second trimester T3 The third trimester Authors’ contributions B.C., Q.Z., and X.L. were responsible for conceptualization; R.L., D.Y., T.Y., B.C., and Q.Z. performed data curation; X.Z., H.Y., Q.Z., and X.L. developed the methodology; R.L., L.W., and X.L. carried out the investigation; R.L. and X.L. performed the visualization; R.L., B.C., Q.Z., and X.L. secured funding acquisition; R.L. managed project administration; B.C., Q.Z., and X.L. provided supervision; R.L. and X.L. wrote the original draft of the manuscript and contributed to review and editing. All authors read and approved the final manuscript. Funding This work was supported by grants from the National Natural Science Foundation of China (22206127), the National Science and Technology Planning Project of China (2012BAI03B01), the Ministry of Science and Technology Basic Resources Survey Project of China (2017FY101106, 2017FY101100), the Clinical Innovative research of Xinhua Hospital (23XHCR16C), the Project of the Integration of Industry, Education and Research of the Shanghai Municipal Education Commission (BJ1-3000–24-0061), and the National Key Research and Development Program of China (2022YFC2702900). Data availability The data that support the findings of this study are available from the corresponding author upon reasonable request. Declarations Ethics approval and consent to participate The study protocol was approved by the Ethics Committee of the National Institute for Nutrition and Health, Chinese Center for Disease Control and Prevention (No. 2019–009). Written informed consent was obtained from all participants or their legal guardians. Consent for publication Not applicable. Competing interests The authors declare no competing interests. Footnotes Publisher’s Note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. Contributor Information Bowen Chen, Email: [email protected]. Qingli Zhang, Email: [email protected]. Xiaoning Lei, Email: [email protected]. References 1. Block ML, Elder A, Auten RL, Bilbo SD, Chen H, Chen JC, et al. The outdoor air pollution and brain health workshop. Neurotoxicology. 2012;33(5):972–84. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 2. Peterson BS, Bansal R, Sawardekar S, Nati C, Elgabalawy ER, Hoepner LA, et al. Prenatal exposure to air pollution is associated with altered brain structure, function, and metabolism in childhood. J Child Psychol Psychiatry. 2022;63(11):1316–31. [ DOI ] [ PubMed ] [ Google Scholar ] 3. Bansal E, Hsu HH, de Water E, Martínez-Medina S, Schnaas L, Just AC, et al. Prenatal PM2.5 exposure in the second and third trimesters predicts neurocognitive performance at age 9–10 years: a cohort study of Mexico City children. Environ Res. 2021;202:111651. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 4. Chauhan R, Dande S, Hood DB, Chirwa SS, Langston MA, Grady SK, et al. Particulate matter2.5 (PM 2.5 )-associated cognitive impairment and morbidity in humans and animal models: a systematic review. J Toxicol Environ Health B. 2025;28(4):233–63. [ DOI ] [ PubMed ] 5. Alter NC, Whitman EM, Bellinger DC, Landrigan PJ. Quantifying the association between PM 2.5 air pollution and IQ loss in children: a systematic review and meta-analysis. Environ Health. 2024;23(1):101. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 6. Liu Y, Zhang L, Wang J, Sui X, Li J, Gui Y, et al. Prenatal PM2.5 exposure associated with neonatal gut bacterial colonization and early children’s cognitive development. Environ Health. 2024;2(11):802–15. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 7. Volk HE, Perera F, Braun JM, Kingsley SL, Gray K, Buckley J, et al. Prenatal air pollution exposure and neurodevelopment: a review and blueprint for a harmonized approach within ECHO. Environ Res. 2021;196:110320. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 8. Iglesias-Vázquez L, Binter AC, Canals J, Hernández-Martínez C, Voltas N, Ambros A, et al. Maternal exposure to air pollution during pregnancy and child’s cognitive, language, and motor function: ECLIPSES study. Environ Res. 2022;212:113501. [ DOI ] [ PubMed ] [ Google Scholar ] 9. Whitworth KW, Rector-Houze AM, Chen WJ, Ibarluzea J, Swartz M, Symanski E, et al. Relation of prenatal and postnatal PM2.5 exposure with cognitive and motor function among preschool-aged children. Int J Hyg Environ Health. 2024;256:114317. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 10. Masselot P, Sera F, Schneider R, Kan H, Lavigne E, Stafoggia M, et al. Differential mortality risks associated with PM 2.5 components: a multi-country, multi-city study. Epidemiology. 2022;33:167–75. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 11. Song J, Saathoff H, Gao L, Gebhardt R, Jiang F, Vallon M, et al. Variations of PM2.5 sources in the context of meteorology and seasonality at an urban street canyon in Southwest Germany. Atmos Environ. 2022;282:119147. [ Google Scholar ] 12. Van Donkelaar A, Martin RV, Li C, Burnett RT. Regional estimates of chemical composition of fine particulate matter using a combined geoscience-statistical method with information from satellites, models, and monitors. Environ Sci Technol. 2019;53(5):2595–611. [ DOI ] [ PubMed ] [ Google Scholar ] 13. Zhong P, Liu J, Qiu X, Jiang X, Zhu T. Non-targeted screening of urban PM 2.5 components based on cellular effects: influence from molecular features and sources. J Environ Exposure Assess. 2025. 10.20517/jeea.2024.60. 14. Chiu YHM, Wilson A, Hsu HHL, Jamal H, Mathews N, Kloog I, et al. Prenatal ambient air pollutant mixture exposure and neurodevelopment in urban children in the Northeastern United States. Environ Res. 2023;233:116394. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 15. Deng Y, Li X, Li X, Liu X, Lin L, Cai L, et al. Association of early life exposure to PM2.5 and its components with offspring neurodevelopment: a prospective birth cohort study. Environ Res. 2025;266:120552. [ DOI ] [ PubMed ] [ Google Scholar ] 16. Lei X, Zhang Y, Wang Z, Lu Z, Pan C, Zhang S, et al. Effects of prenatal exposure to PM 2.5 and its composition on cognitive and motor functions in children at 12 months of age: The Shanghai Birth Cohort Study. Environ Int. 2022;170:107597. [ DOI ] [ PubMed ] [ Google Scholar ] 17. Lu ZH, Liu C, Chen YJ, Chen YJ, Lei XN, Cai LJ, et al. Gestational exposure to PM 2.5 and specific constituents, meconium metabolites, and neonatal neurobehavioral development: a cohort study. Environ Sci Technol. 2024. 10.1021/es2045678. [ DOI ] [ PubMed ] 18. Sun X, Liu C, Ji H, Li W, Miao M, Yuan W, et al. Prenatal exposure to ambient PM2.5 and its chemical constituents and child intelligence quotient at 6 years of age. Ecotoxicol Environ Saf. 2023;255:114813. [ DOI ] [ PubMed ] [ Google Scholar ] 19. Xu X, Tao S, Huang L, Du J, Liu C, Jiang Y, et al. Maternal PM2.5 exposure during gestation and offspring neurodevelopment: findings from a prospective birth cohort study. Sci Total Environ. 2022;842:156778. [ DOI ] [ PubMed ] [ Google Scholar ] 20. Wang X, Li C, Zhou L, Liu L, Qiu X, Huang D, et al. Associations of prenatal exposure to PM 2.5 and its components with offsprings' neurodevelopmental and behavioral problems: a prospective cohort study from China. Ecotoxicol Environ Saf. 2024;282:116739. [ DOI ] [ PubMed ] 21. Haist F, Anzures G. Functional development of the brain’s face‐processing system. Wiley Interdiscip Rev Cogn Sci. 2017;8(1–2):e1423. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 22. Kim P, Leckman JF, Mayes LC, Feldman R, Wang X, Swain JE. The plasticity of human maternal brain: longitudinal changes in brain anatomy during the early postpartum period. Behav Neurosci. 2010;124(5):695–705. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 23. Binter AC, Granés L, Bannier E, de Castro M, Petricola S, Fossati S, et al. Urban environment during pregnancy and childhood and white matter microstructure in preadolescence in two European birth cohorts. Environ Pollut. 2024;346:123612. [ DOI ] [ PubMed ] [ Google Scholar ] 24. Garkov S, Dearden L, Armstrong B, Milojevic A. The relationship between PM2.5, greenness, and road noise exposures and children’s cognitive performance in England: the Millennium Cohort Study. Environments. 2024;11(10):213. [ Google Scholar ] 25. Opbroek J, Barboza EP, Nieuwenhuijsen M, Dadvand P, Mueller N. Urban green spaces and behavioral and cognitive development in children: a health impact assessment of the Barcelona “Eixos Verds” plan (Green Axis Plan). Environ Res. 2024;244:117909. [ DOI ] [ PubMed ] [ Google Scholar ] 26. Binter AC, Doiron D, Shareck M, Pitt T, McDonald SW, Subbarao P, et al. Urban environment during pregnancy, cognitive abilities, motor function, and externalizing and internalizing symptoms at 2–5 years old in 3 Canadian birth cohorts. Environ Int. 2025;195:109222. [ DOI ] [ PubMed ] [ Google Scholar ] 27. Hazlehurst MF, Hajat A, Tandon PS, Szpiro AA, Kaufman JD, Tylavsky FA, et al. Associations of residential green space with internalizing and externalizing behavior in early childhood. Environ Health. 2024;23(1):17. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 28. Yuchi W, Brauer M, Czekajlo A, Davies HW, Davis Z, Guhn M, et al. Neighborhood environmental exposures and incidence of attention deficit/hyperactivity disorder: a population-based cohort study. Environ Int. 2022;161:107120. [ DOI ] [ PubMed ] [ Google Scholar ] 29. Lavigne É, Abdulaziz KE, Murphy MS, Stanescu C, Dingwall-Harvey AL, Stieb DM, et al. Associations of neighborhood greenspace, and active living environments with autism spectrum disorders: a matched case-control study in Ontario, Canada. Environ Res. 2024;252:118828. [ DOI ] [ PubMed ] [ Google Scholar ] 30. Basagaña X, Esnaola M, Rivas I, Amato F, Alvarez-Pedrerol M, Forns J, et al. Neurodevelopmental deceleration by urban fine particles from different emission sources: a longitudinal observational study. Environ Health Perspect. 2016;124(10):1630–6. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 31. Matthaios VN, Holland I, Kang CM, Hart JE, Hauptman M, Wolfson JM, et al. The effects of urban green space and road proximity to indoor traffic-related PM2.5, NO2, and BC exposure in inner-city schools. J Expo Sci Environ Epidemiol. 2024. 10.1038/s41370-024-00669-8. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 32. Buraerah MF, Patandjengi B, Suryani S, Hamzah A, Demmalino EB. The effect of vegetation in reducing air pollution in an urban environment: a review. IOP Conf. Ser.: Earth Environ. Sci. 2023;1253: 012105. 33. Xiao Y, Liu C, Lei R, Wang Z, Wang X, Tian H, et al. Associations of PM 2.5 composition and green space with metabolic syndrome in a Chinese essential hypertensive population. Chemosphere. 2023;343:140243. [ DOI ] [ PubMed ] [ Google Scholar ] 34. Wang Y, Li W, Chen S, Zhang J, Liu X, Jiang J, et al. PM2.5 constituents associated with childhood obesity and larger BMI growth trajectory: a 14-year longitudinal study. Environ Int. 2024;183:108417. [ DOI ] [ PubMed ] [ Google Scholar ] 35. Cui Z, Pan R, Liu J, Yi W, Huang Y, Li M, et al. Green space and its types can attenuate the associations of PM2.5 and its components with prediabetes and diabetes: a multicenter cross-sectional study from eastern China. Environ Res. 2024;245:117997. [ DOI ] [ PubMed ] [ Google Scholar ] 36. Xu Z, Han Z, Wang J, Jin R, Li Z, Wu Z, et al. Association between long‐term exposure to fine particulate matter constituents and progression of cerebral blood flow velocity in Beijing: modifying effect of greenness. Geohealth. 2023;7(7):e2023GH000796. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 37. Yang ZY, Zhang Q, Yi ZHAI, Tao XU, Wang YY, Bo WC, et al. National nutrition and health systematic survey for children 0–17 years of age in China. Biomed Environ Sci. 2021;34(11):891–9. [ DOI ] [ PubMed ] [ Google Scholar ] 38. Geng G, Zhang Q, Tong D, Li M, Zheng Y, Wang S, et al. Chemical composition of ambient PM2.5 over China and relationship to precursor emissions during 2005–2012. Atmos Chem Phys. 2017;17(14):9187–203. [ Google Scholar ] 39. Liu S, Geng G, Xiao Q, Zheng Y, Liu X, Cheng J, et al. Tracking daily concentrations of PM2.5 chemical composition in China since 2000. Environ Sci Technol. 2022;56(22):16517–27. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 40. Li R, Zheng X, Yang H, Yin D, Yin T, Wang L, et al. Prenatal and postnatal exposure to different sizes of particulate matter and neurodevelopment in children across 16 provinces in China. Ecotoxicol Environ Saf. 2025;307:119373. [ DOI ] [ PubMed ] [ Google Scholar ] 41. Zheng X, Li R, Wang L, Yang H, Li L, Cui J, et al. Association between breastfeeding duration and neurodevelopment in Chinese children aged 2 to 3 years. Infant Behav Dev. 2024;77:101991. [ DOI ] [ PubMed ] [ Google Scholar ] 42. Jiang W, Huang Y, Jin H, Gan Y, Zhang Q, He X, et al. A prospective exposome-based gene-environment interaction study on the effects of prenatal environmental exposure on fetal growth in the Shanghai Birth Cohort. Environ Health Perspect. 2025. 10.1289/EHP15902. [ DOI ] [ PubMed ] [ Google Scholar ] 43. Didan K. MOD13Q1 MODIS/Terra Vegetation Indices 16-Day L3 Global 250m SIN Grid V006. 2015. 10.5067/MODIS/MOD13Q1.006. 44. Weier J, Herring D. Measuring vegetation (NDVI & EVI). NASA Earth Observatory. 2000. https://earthobservatory.nasa.gov/features/MeasuringVegetation . 45. Di N, Li S, Xiang H, Xie Y, Mao Z, Hou J, et al. Associations of residential greenness with depression and anxiety in rural Chinese adults. The Innovation. 2020;1(3):100054. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 46. Lamichhane DK, Ha E, Hong YC, Lee DW, Park MS, Song S, et al. Ambient particulate matter and surrounding greenness in relation to sleep quality among pregnant women: a nationwide cohort study. Heliyon. 2024;10(5):e26742. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 47. Wang R, Yang B, Yao Y, Bloom MS, Feng Z, Yuan Y, et al. Residential greenness, air pollution and psychological well-being among urban residents in Guangzhou, China. Sci Total Environ. 2020;711:134843. [ DOI ] [ PubMed ] [ Google Scholar ] 48. Cosemans C, Madhloum N, Sleurs H, Alfano R, Verheyen L, Wang C, et al. Prenatal particulate matter exposure is linked with neurobehavioural development in early life. Environ Res. 2024;252:118879. [ DOI ] [ PubMed ] [ Google Scholar ] 49. Guilbert A, Bernard JY, Peyre H, Costet N, Hough I, Seyve E, et al. Prenatal and childhood exposure to ambient air pollution and cognitive function in school-age children: examining sensitive windows and sex-specific associations. Environ Res. 2023;235:116557. [ DOI ] [ PubMed ] [ Google Scholar ] 50. Guxens M, Garcia-Esteban R, Giorgis-Allemand L, Forns J, Badaloni C, Ballester F, et al. Air pollution during pregnancy and childhood cognitive and psychomotor development: six European birth cohorts. Epidemiology. 2014;25(5):636–47. [ DOI ] [ PubMed ] [ Google Scholar ] 51. Ha S, Yeung E, Bell E, Insaf T, Ghassabian A, Bell G, et al. Prenatal and early life exposures to ambient air pollution and development. Environ Res. 2019;174:170–5. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 52. Holm SM, Balmes JR, Gunier RB, Kogut K, Harley KG, Eskenazi B. Cognitive development and prenatal air pollution exposure in the CHAMACOS cohort. Environ Health Perspect. 2023;131(3):037007. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 53. Lertxundi A, Baccini M, Lertxundi N, Fano E, Aranbarri A, Martínez MD, et al. Exposure to fine particle matter, nitrogen dioxide and benzene during pregnancy and cognitive and psychomotor developments in children at 15 months of age. Environ Int. 2015;80:33–40. [ DOI ] [ PubMed ] [ Google Scholar ] 54. Li J, Liao J, Hu C, Bao S, Mahai G, Cao Z, et al. Preconceptional and the first trimester exposure to PM2.5 and offspring neurodevelopment at 24 months of age: examining mediation by maternal thyroid hormones in a birth cohort study. Environ Pollut. 2021;284:117133. [ DOI ] [ PubMed ] [ Google Scholar ] 55. Meredith RM. Sensitive and critical periods during neurotypical and aberrant neurodevelopment: a framework for neurodevelopmental disorders. Neurosci Biobehav Rev. 2015;50:180–8. [ DOI ] [ PubMed ] [ Google Scholar ] 56. Maciak M, Koszela K, Beniuk A, Woldańska-Okońska M. Development of postural-motor, coordination, and reflex functions in children in the first year of life. Pol Merkur Lekarski. 2024. 10.36740/Merkur202405105. [ DOI ] [ PubMed ] [ Google Scholar ] 57. Ye M, Yang J, Li J, Wang Y, Chen W, Zhu L, et al. Progress in mechanisms, pathways and cohort studies about the effects of PM2.5 exposure on the central nervous system. Rev Environ Contam Toxicol. 2023;261(1):7. [ Google Scholar ] 58. Ford ALB, Elmquist M, Merbler AM, Kriese A, Will KK, McConnell SR. Toward an ecobehavioral model of early language development. Early Child Res Q. 2020;50:246–58. [ Google Scholar ] 59. Lertxundi A, Andiarena A, Martínez MD, Ayerdi M, Murcia M, Estarlich M, et al. Prenatal exposure to PM2.5 and NO2 and sex-dependent infant cognitive and motor development. Environ Res. 2019;174:114–21. [ DOI ] [ PubMed ] [ Google Scholar ] 60. Hurtado-Díaz M, Riojas-Rodríguez H, Rothenberg SJ, Schnaas-Arrieta L, Kloog I, Just A, et al. Prenatal PM2.5 exposure and neurodevelopment at 2 years of age in a birth cohort from Mexico city. Int J Hyg Environ Health. 2021;233:113695. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 61. Pujol J, Martínez-Vilavella G, Macià D, Fenoll R, Alvarez-Pedrerol M, Rivas I, et al. Traffic pollution exposure is associated with altered brain connectivity in school children. Neuroimage. 2016;129:175–84. [ DOI ] [ PubMed ] [ Google Scholar ] 62. Lee KS, Kim BN, Cho J, Jang YY, Choi YJ, Lee WS, et al. Associations between surrounding residential greenness and intelligence quotient in 6-year-old children. Sci Total Environ. 2021;759:143561. [ DOI ] [ PubMed ] [ Google Scholar ] 63. Liao J, Zhang B, Xia W, Cao Z, Zhang Y, Liang S, et al. Residential exposure to green space and early childhood neurodevelopment. Environ Int. 2019;128:70–6. [ DOI ] [ PubMed ] [ Google Scholar ] 64. Julvez J, López-Vicente M, Warembourg C, Maitre L, Philippat C, Gützkow KB, et al. Early life multiple exposures and child cognitive function: a multi-centric birth cohort study in six European countries. Environ Pollut. 2021;284:117404. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 65. Nieuwenhuijsen MJ, Khreis H, Triguero-Mas M, Gascon M, Dadvand P. Fifty shades of green: pathway to healthy urban living. Epidemiology. 2017;28(1):63–71. [ DOI ] [ PubMed ] [ Google Scholar ] 66. Jones R, Tarter R, Ross AM. Greenspace interventions, stress and cortisol: a scoping review. Int J Environ Res Public Health. 2021;18(6):2802. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 67. Markevych I, Schoierer J, Hartig T, Chudnovsky A, Hystad P, Dzhambov AM, et al. Exploring pathways linking greenspace to health: theoretical and methodological guidance. Environ Res. 2017;158:301–17. [ DOI ] [ PubMed ] [ Google Scholar ] 68. Kumar P, Druckman A, Gallagher J, Gatersleben B, Allison S, Eisenman TS, et al. The nexus between air pollution, green infrastructure and human health. Environ Int. 2019;133:105181. [ DOI ] [ PubMed ] [ Google Scholar ] 69. Rahman MM, Shu YH, Chow T, Lurmann FW, Yu X, Martinez MP, et al. Prenatal exposure to air pollution and autism spectrum disorder: sensitive windows of exposure and sex differences. Environ Health Perspect. 2022;130(1):017008. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 70. Rivas I, Basagaña X, Cirach M, López-Vicente M, Suades-González E, Garcia-Esteban R, et al. Association between early life exposure to air pollution and working memory and attention. Environ Health Perspect. 2019;127(5):057002. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 71. Umezawa M, Onoda A, Korshunova I, Jensen AC, Koponen IK, Jensen KA, et al. Maternal inhalation of carbon black nanoparticles induces neurodevelopmental changes in mouse offspring. Part Fibre Toxicol. 2018;15:1–18. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 72. Yi C, Wang Q, Qu Y, Niu J, Oliver BG, Chen H. In-utero exposure to air pollution and early-life neural development and cognition. Ecotoxicol Environ Saf. 2022;238:113589. [ DOI ] [ PubMed ] [ Google Scholar ] 73. Bolton JL, Auten RL, Bilbo SD. Prenatal air pollution exposure induces sexually dimorphic fetal programming of metabolic and neuroinflammatory outcomes in adult offspring. Brain Behav Immun. 2014;37:30–44. [ DOI ] [ PubMed ] [ Google Scholar ] 74. Allen JL, Conrad K, Oberdörster G, Johnston CJ, Sleezer B, Cory-Slechta DA. Developmental exposure to concentrated ambient particles and preference for immediate reward in mice. Environ Health Perspect. 2013;121(1):32–8. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 75. Monk C, Lugo-Candelas C, Trumpff C. Prenatal developmental origins of future psychopathology: mechanisms and pathways. Annu Rev Clin Psychol. 2019;15(1):317–44. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] Associated Data This section collects any data citations, data availability statements, or supplementary materials included in this article. Supplementary Materials 12916_2026_4750_MOESM1_ESM.docx (2.2MB, docx) Additional file 1: Table S1. Distribution of PM 2.5 and its compositions across six exposure windows. Table S2. Distribution of greenness, temperature, and relative humidity across six exposure windows. Table S3. Associations between PM 2.5 , its constituents, and child developmental quotient. Table S4. Associations between NDVI 1000m greenness exposure and child developmental quotient across six exposure windows. Table S5. Modification of NDVI on the association between PM 2.5 , its constituents and child developmental quotient across six exposure windows. Table S6. Regional modification effects on the association between PM 2.5 , its compositions, and child developmental quotient. Table S7. Sex modification effects on the association between PM 2.5 , its compositions, and child developmental quotient. Table S8. Age-stratified associations between PM 2.5 , its compositions, and child developmental quotient. Table S9. Associations between PM 2.5 , its compositions, and child developmental quotient in regression models adjusted for additional confounders. Table S10. Associations between PM 2.5 , its compositions, and developmental quotient among children excluding those with premature birth or low birth weight. Table S11. Associations between PM 2.5 , its compositions, and developmental quotient among children whose mothers had no pregnancy complications. Table S12. Associations between greenness within different buffer zones and child developmental quotient across six exposure windows. Figure S1. Residential address locations of study population. Data Availability Statement The data that support the findings of this study are available from the corresponding author upon reasonable request. Articles from BMC Medicine are provided here courtesy of BMC ACTIONS View on publisher site PDF (2.3 MB) Cite Collections Permalink PERMALINK Copy RESOURCES Similar articles Cited by other articles Links to NCBI Databases Cite Copy Download .nbib .nbib Format: AMA APA MLA NLM Add to Collections Create a new collection Add to an existing collection Name your collection * Choose a collection Unable to load your collection due to an error Please try again Add Cancel Follow NCBI NCBI on X (formerly known as Twitter) NCBI on Facebook NCBI on LinkedIn NCBI on GitHub NCBI RSS feed Connect with NLM NLM on X (formerly known as Twitter) NLM on Facebook NLM on YouTube National Library of Medicine 8600 Rockville Pike Bethesda, MD 20894 Web Policies FOIA HHS Vulnerability Disclosure Help Accessibility Careers NLM NIH HHS USA.gov Back to Top