Prenatal exposure to residential PM2.5 and anogenital distance in infants at birth: A birth cohort study from Shanghai, China

Spatio-temporal fusion of meteorological factors for multi-site PM2.5 prediction: A deep learning and time-variant graph approach

In the field of environmental science, traditional methods for predicting PM2.5 concentrations primarily focus on singular temporal or spatial dimensions. This approach presents certain limitations when it comes to deeply mining the joint influence of multiple monitoring sites and their inherent connections with meteorological factors. To address this issue, we introduce an innovative deep-learning-based multi-graph model using Beijing as the study case. This model consists of two key modules: firstly, the 'Meteorological Factor Spatio-Temporal Feature Extraction Module'. This module deeply integrates spatio-temporal features of hourly meteorological data by employing Graph Convolutional Networks (GCN) and Long Short-Term Memory (LSTM) for spatial and temporal encoding respectively. Subsequently, through an attention mechanism, it retrieves a feature tensor associated with air pollutants. Secondly, these features are amalgamated with PM2.5 concentration values, allowing the 'PM2.5 Concentration Prediction Module' to predict with enhanced accuracy the joint influence across multiple monitoring sites. Our model exhibits significant advantages over traditional methods in processing the joint impact of multiple sites and their associated meteorological factors. By providing new perspectives and tools for the in-depth understanding of urban air pollutant distribution and optimization of air quality management, this model propels us towards a more comprehensive approach in tackling air pollution issues.

Associations of Exposure to Air Pollution during the Male Programming Window and Mini-Puberty with Anogenital Distance and Penile Width at Birth and at 1 Year of Age in the Multicenter U.S. TIDES Cohort

BACKGROUND

Ambient air pollution may be a developmental endocrine disruptor. In animal models, gestational and perinatal exposure to diesel exhaust and concentrated particulate matter alters anogenital distance (AGD), a marker of prenatal androgen activity, in both sexes. Little is known in humans.

OBJECTIVES

We examined exposure to fine particulate matter (PM 2.5 ) and nitrogen dioxide (NO 2 ) in relation to human AGD at birth and at 1 year of age, focusing on exposures during critical windows of reproductive development: the male programming window (MPW; gestational weeks 8-14) and mini-puberty (postnatal months 1-3).

METHODS

The Infant Development and Environment Study (TIDES) recruited first trimester pregnant women (n = 687 ) at four U.S. sites (Minneapolis, Minnesota; Rochester, New York; San Francisco, California; and Seattle, Washington) from 2010 to 2012. We measured anus to clitoris (AGD-AC) and anus to fourchette (AGD-AF) in female infants at birth; in males, we measured anus to penis (AGD-AP), anus to scrotum (AGD-AS), and penile width at birth and at 1 year of age. Using advanced spatiotemporal models, we estimated maternal exposure to PM 2.5 and NO 2 in the MPW and mini-puberty. Covariate-adjusted, sex-stratified linear regression models examined associations between PM 2.5 and NO 2 and AGD.

RESULTS

In males, a 1 - μ g / m 3 increase in PM 2.5 exposure during the MPW was associated with shorter AGD at birth, but a longer AGD at 1 year of age (e.g., birth AGD-AP: β = - 0.35 mm ; 95% CI: - 0.62 , - 0.07 ; AGD-AS: β = 0.37 mm ; 95% CI: 0.02, 0.73). Mini-pubertal PM 2.5 exposure was also associated with shorter male AGD-AP (β = - 0.50 mm ; 95% CI: - 0.89 , - 0.11 ) at 1 year of age. Although not associated with male AGD measures, 1 -ppb increases in NO 2 exposure during the MPW (β = - 0.07 mm ; 95% CI: - 0.02 , - 0.12 ) and mini-puberty (β = - 0.04 mm ; 95% CI: - 0.08 , 0.01) were both associated with smaller penile width at 1 year of age. Results were similar in multipollutant models, where we also observed that in females AGD-AC was inversely associated with PM 2.5 exposure, but positively associated with NO 2 exposure.

DISCUSSION

PM 2.5 and NO 2 exposures during critical pre- and postnatal windows may disrupt reproductive development. More work is needed to confirm these novel results and clarify mechanisms. https://doi.org/10.1289/EHP12627.

Dynamic graph neural network with adaptive edge attributes for air quality prediction: A case study in China

Air quality prediction is a typical Spatiotemporal modeling problem, which always uses different components to handle spatial and temporal dependencies in complex systems separately. Previous models based on time series analysis and recurrent neural network (RNN) methods have only modeled time series while ignoring spatial information. Previous graph convolution neural networks (GCNs) based methods usually require providing spatial correlation graph structure of observation sites in advance. The correlations among these sites and their strengths are usually calculated using prior information. However, due to the limitations of human cognition, limited prior information cannot reflect the real station-related structure or bring more effective information for accurate prediction. To this end, we propose a novel Dynamic Graph Neural Network with Adaptive Edge Attributes (DGN-AEA) on the message passing network, which generates the adaptive bidirected dynamic graph by learning the edge attributes as model parameters. Unlike prior information to establish edges, our method can obtain adaptive edge information through end-to-end training without any prior information. Thus reducing the complexity of the problem. Besides, the hidden structural information between the stations can be obtained as model by-products, which can help make some subsequent decision-making analyses. Experimental results show that our model received state-of-the-art performance than other baselines.

Prenatal exposure to fine particulate matter and newborn anogenital distance: a prospective cohort study

BACKGROUND

Considerable attention has been paid to reproductive toxicity of fine particulate matter (PM_2.5). However, the relationship between prenatal PM_2.5 exposure and anogenital distance (AGD) has not been well studied. We aim to investigate the potential effects of prenatal exposure to PM_2.5 on newborn AGD.

METHODS

Prenatal PM_2.5 exposure of 2332 participates in Shanghai (2013-2016) was estimated using high-performance machine learning models. Anoscrotal distance (AGDas) in male infants and anofourchette distance (AGDaf) in female infants were measured by well-trained examiners within 3 days after birth. We applied multiple linear regression models and multiple informant models to estimate the association between prenatal PM_2.5 exposure and AGD.

RESULTS

Multiple linear regression models showed that a 10 μg/m³ increase in PM_2.5 exposure during full pregnancy, the second and third trimesters was inversely associated with AGDas (adjusted beta = - 1.76, 95% CI: - 2.21, - 1.31; - 0.73, 95% CI: - 1.06, - 0.40; and - 0.52; 95% CI: - 0.87, - 0.18, respectively) in males. A 10 μg/m³ increase in PM_2.5 exposure during the full pregnancy, the first, second, and third trimesters was inversely associated with AGDaf (adjusted beta = - 4.55; 95% CI: - 5.18, - 3.92; - 0.78; 95% CI: - 1.10, - 0.46; - 1.11; 95% CI: - 1.46, - 0.77; - 1.45; 95% CI: - 1.78, - 1.12, respectively) in females after adjusting for potential confounders. Multiple informant models showed consistent but slightly attenuated associations.

CONCLUSION

Our study observed a significant association between gestational PM_2.5 exposure during pregnancy and shortened AGD in newborns, and provided new evidence on potential reproductive toxicity of prenatal PM_2.5 exposure.

Associations between maternal urinary isoflavone concentrations and anogenital distance of offspring throughout infancy: a prospective cohort study

STUDY QUESTION

Are maternal urinary isoflavone (ISO) concentrations during pregnancy associated with anogenital distance (AGD) in infants at birth, and at 6 and 12 months of age?

SUMMARY ANSWER

Higher maternal urinary ISO concentrations during pregnancy were associated with longer AGD in infants of both sexes, and equol (EQU) and daidzein (DAD) were identified as the important ISO mixture components in the observed associations.

WHAT IS KNOWN ALREADY

Evidence of the association of prenatal exposure to ISO with offspring's AGD is mainly derived from animal studies, which used different study designs and had inconsistent results. Only one human study has been reported and it found null associations between maternal ISO exposure during pregnancy and AGD among boys at birth, with a small sample size and a wide range of exposure windows. No human study on girls was found.

STUDY DESIGN, SIZE, DURATION

Prospective cohort study (Shanghai-Minhang Birth Cohort Study), with pregnant women recruited at 12-16 weeks of gestation in Shanghai, China between April and December 2012. One thousand two hundred and twenty-five live singletons were left in the cohort at delivery of which 480 mother-infant pairs had data on both maternal urinary ISO concentrations and at least one AGD measurement and were included in the present study. Anopenile distance (AGDAP) and anoscrotal distance (AGDAS) of boys and anoclitoral distance (AGDAC) and anofourchette distance (AGDAF) of girls were measured at birth and at 6 and 12 months of age.

PARTICIPANTS/MATERIALS, SETTING, METHODS

Multiple linear regression models were used to examine the associations between maternal ISO concentrations and AGD. Bayesian kernel machine regression (BKMR) was implemented to examine both the overall effects of ISO mixture and the single effect of each ISO and identify important components of ISO mixture.

MAIN RESULTS AND THE ROLE OF CHANCE

A general profile of higher concentrations of maternal ISO associated with longer AGD in infants of both sexes was observed, when maternal education, parity, BMI before pregnancy (BMI, categorical variable), passive smoking during early pregnancy, age at delivery, gestational weeks and infant body size were adjusted for. Among boys, EQU was associated with increased AGDAS at birth and at 6 and 12 months, and DAD was associated with increased AGDAP at birth. Among girls, the associations of EQU and DAD with increased AGDAC and AGDAF at birth were found. When gestational weight gain and feeding patterns of infants in the first 6 months were additionally adjusted for, and maternal BMI was adjusted for as a continuous variable, more pronounced associations were observed, especially for associations of genistein (GEN), DAD and glycitein (GLY) with increased AGDAP and AGDAS at 6 months in boys. However, these associations were not always observed in the highest tertile group, and no consistent dose-response relationships were found. Similar results were observed in BKMR models, showing positive correlations of concentration of ISO mixture with increased AGDAS at both 6 and 12 months among boys, and increased AGDAC and AGDAF at birth among girls. Statistically significant increments of 4.96 mm (95% credible interval (CrI): 1.40, 8.52) and 1.07 mm (95% CrI: 0.02, 2.13) in AGDAS at 6 months among boys and AGDAC at birth among girls, respectively, were observed at the 75th percentile of ISO mixture, compared with 25th percentile. EQU and DAD were identified as the important components among ISO-AGD associations.

LIMITATIONS, REASONS FOR CAUTION

First, due to the short half-lives of ISO, the accuracy of a single spot urine sample reflecting ISO exposure during pregnancy may be limited, and thus may cause non-differential misclassification. Second, despite the adjustments for several important covariates in the study, unmeasured and residual confounding factors may remain a concern. Third, false discovery due to multiple testing may remain. Finally, the reduced sample sizes attributed to the loss of follow-up and missing data of confounders may limit our ability to detect an association, if any existed.

WIDER IMPLICATIONS OF THE FINDINGS

Prenatal ISO exposure may affect the reproductive development of offspring. As ISO can be widely detected in pregnant women, especially in Eastern countries, more studies are warranted to provide evidence of the effects of prenatal ISO exposure on long-term reproductive outcomes.

STUDY FUNDING/COMPETING INTEREST(S)

This work was supported by grants from the National Key Research and Development Program of China (2021YFC2701003), the National Natural Science Foundation of China (22076123), the Science and Technology Commission of Shanghai Municipality (21ZR1454700 and 20ZR1448000), the Shanghai Municipal Health Commission (20194Y0160) and Innovation-oriented Science and Technology Grant from NHC Key Laboratory of Reproduction Regulation (CX2022-04). The authors have no conflicts of interest to declare.

TRIAL REGISTRATION NUMBER

N/A.

A dual-path dynamic directed graph convolutional network for air quality prediction

Accurate air quality prediction can help cope with air pollution and improve the life quality. With the development of the deployments of low-cost air quality sensors, increasing data related to air quality has provided chances to find out more accurate prediction methods. Air quality is affected by many external factors such as the position, wind, meteorological information, and so on. Meanwhile, these factors are spatio-temporal dynamic and there are many dynamic contextual relationships between them. Many methods for air quality prediction do not consider these complex spatio-temporal correlations and dynamic contextual relationships. In this paper, we propose a dual-path dynamic directed graph convolutional network (DP-DDGCN) for air quality prediction. We first create a dual-path transposed dynamic directed graph according to static distance relationships of stations and the dynamic relationships generated by wind speed and directions. Then based on the dual-path dynamic directed graph, we can capture the dynamic spatial dependencies more comprehensively. After that we apply gated recurrent units (GRUs) and add the future meteorological features, to extract the complex temporal dependencies of historical air quality data. Using dual-path dynamic directed graph blocks and the GRUs, we finally construct a dynamic spatio-temporal gated recurrent block to capture the dynamic spatio-temporal contextual correlations. Based on real-world datasets, which record a large amount of PM_2.5 concentration data, we compare the proposed model with the benchmark models. The experimental results show that our proposed model has the best performance in predicting the PM_2.5 concentrations.

Association Between Neonatal Thyroid Function and Anogenital Distance from Birth to 48 Months of Age

BACKGROUND

Evidence from animal studies has indicated that neonatal thyroid function is vital for the reproductive development. Anogenital distance (AGD), a sensitive biomarker of the fetal hormonal milieu, can be used to predict adult reproductive disorders. However, few human studies have examined the association between neonatal thyroid function and AGD. We aimed to explore their associations in a birth cohort study.

METHODS

Concentrations of thyroid stimulating hormone (TSH) and thyroid hormones (THs), including total triiodothyronine (TT₃), total thyroxine (TT₄), free triiodothyronine (FT₃), and free thyroxine (FT₄) were measured in cord plasma in the Shanghai-Minhang Birth Cohort. The offspring AGD (AGD_AP [anus-penis] and AGD_AS [anus-scrotum] for boys and AGD_AC [anus-clitoris] and AGD_AF [anus-fourchette] for girls), body weight and anogenital index (AGI = AGD/weight [mm/kg]) were obtained at each follow-up visit. In total, 344 children (194 boys and 150 girls) with cord plasma concentrations of THs and TSH and at least one AGD measurement at birth and at 6, 12, and 48 months of age were included. Multiple linear regression and generalized estimating equation (GEE) models were used to examine the associations of cord plasma concentrations of THs and TSH with AGI.

RESULTS

Multiple linear regression models showed inverse associations of TT₄, FT₃, and FT₄ with female AGI, although statistical significance was only reached at birth, 6 and 48 months of age. These associations were also found in GEE models: higher TT₄ and FT₄ concentrations were associated with lower AGI_AC (TT₄: β = -0.27, 95% CI: -0.50, -0.03 for middle vs. lowest tertile; FT₄: β = -0.38, 95% CI: -0.61, -0.16 for middle and β = -0.30, 95% CI: -0.55, -0.04 for highest vs. lowest tertile). Besides, girls with the highest tertile of FT₃ concentrations had lower AGI_AF than those with the lowest tertile (the highest vs. lowest tertile: β = -0.22, 95% CI: -0.36, -0.08). Positive associations between TSH and AGI at birth and at 12 months of age were observed in boys.

CONCLUSIONS

This study provides further evidence on the effects of neonatal thyroid function on reproductive development at an early life stage.