Prenatal Exposure to Metal Mixture and Birth Weight; a Bayesian Kernel Machine Regression Analysis of Two Cohort Studies in Japan and Iran

doi:10.21203/rs.3.rs-3905217/v1

Download PDF

Research Article

Prenatal Exposure to Metal Mixture and Birth Weight; a Bayesian Kernel Machine Regression Analysis of Two Cohort Studies in Japan and Iran

https://doi.org/10.21203/rs.3.rs-3905217/v1

This work is licensed under a CC BY 4.0 License

You are reading this latest preprint version

Purpose

Potentially toxic metals can directly induce various adverse effects on reproductive organs or interrupt essential metals' physiological activities. Despite intensive efforts to reduce these metals in the environment, chronic and low-level exposure remains a public health problem. The present study aimed to investigate prenatal metal exposure, including arsenic (As), copper (Cu), lead (Pb), manganese (Mn), rubidium (Rb), selenium (Se), and zinc (Zn), effects on birth weight.

Methods

We collected 579 blood samples before the 16th week of gestation from apparently healthy women with singleton pregnancy in Iran and Japan. Blood metal concentrations were measured using inductively coupled plasma-mass spectrometry.

Results

Prenatal blood levels of As, Mn, Pb, and Zn were significantly higher, while Cu, Rb, and Se were significantly (p < 0.01) lower in Iranian participants than in Japanese. Adjusted linear regression analyses and Bayesian Kernel Machine Regression (BKMR) overall exposure-response functions showed inverse relationships between metals and birth weight.

Conclusion

The study findings, using data from geographically diverse countries, suggest prenatal blood metal exposure as a potential risk factor for lower birth weight. Therefore, women of reproductive age should minimize encountering to potentially toxic metals as much as possible.

blood metal

birth weight

pregnancy

mixture exposure

Bayesian Kernel Machine Regression (BKMR)

Since the Industrial Revolution, humans have been exposed to hundreds of dangerous chemicals through industrial activities and a polluted environment (air, water, and food chain) [1–3]. Potentially toxic metals, such as lead (Pb) and arsenic (As) are important subtypes of these pollutants, causing various adverse health effects at almost any concentration, individually or cumulatively with other elements [4]. In the recent decades, new legislations and intensive activities, like the banning of leaded gasoline since 1980 in Japan, have gradually reduced toxic metal exposure globally [2, 5]. Despite these efforts, chronic and low-level metal exposure remains a public health problem [6]. On the other hand, essential metals at the proper dose, such as zinc (Zn), manganese (Mn), selenium (Se), and copper (Cu) are necessary for normal biological activities but may also cause adverse health effects at high or low levels (deficiency) [7, 8]. Additionally, some toxic metals can disrupt the physiological activities of these essential metals [9, 10]. Therefore, studying metal mixtures can be a better research strategy to understand interaction and competition among them in the biological environment.

Pregnant women, a vulnerable group to various toxicants, are usually exposed to metals in the same manner of general population [11, 12]. Many epidemiologic studies, including our previous studies, have suggested various metals at different concentrations inducing adverse pregnancy outcomes, such as pregnancy induced-hypertension, preterm birth, low birth weight, preterm rupture of membranes, and impaired infancy neurodevelopment abilities [13–22]. Among these outcomes, birth weight is a well–established indicator of neonatal health status and predicting newborns' future health conditions [23, 24].

So far, the “single pollution” investigation has been the dominant research method in most epidemiological studies on metal toxicities during pregnancy [25]. This approach investigates effects of one metal on the human reproductive system without considering interaction, competition, and/or synergy among them. However, the lack of establishing an appropriate statistical method was a major challenge in studying mixture exposures [26]. Recently, the USA National Institute of Environmental Health Sciences launched a specific initiative called "Powering Research Through Innovative Methods for Mixtures in Epidemiology" to develop study method in this filed [27].

The present study aimed to study prenatal blood metal mixture (As, Cu, Pb, Mn, Rb, Se, and Zn) effects on birth weight, using two birth cohort studies in Iran and Japan. Firstly, we conducted descriptive statistical analyses on participants’ characteristics and their blood metals levels, following by comparing these variables between two countries. Then, the linear regression model examined associations between individual blood metals concentrations and birth weight, adjusting for covariates. Finally, we employed Bayesian Kernel Machine Regression (BKMR), a novel statistical model implemented to analyze the effects of mixed components on the outcome based on exposure-response functions [26, 28].

Study Population

After obtaining ethical committee permissions (IR.TUMS.REC.1394.1180 and Ethics Committee of Juntendo University Hospital authorization number 632) and informed constant from all participants, birth cohort studies were initiated in Tehran, Iran (2016–2017) and Tokyo, Japan (2010–2012), with nearly identical study design. The inclusion criteria were singleton pregnancy, gestational age < 16 weeks for prenatal blood sampling, and the absence of chronic medical conditions, such as cancer, diabetes mellitus, hypertension, congestive heart failure, and so on. Prenatal blood samples, participants' anthropometric characteristics, and medical/pregnancy information were collected uniformly for both studies [13]. Blood metals measurements were conducted at the Faculty of Medicine Juntendo University, Tokyo, Japan, using the same devices and analytical protocol.

Metal Exposure Assessment

Blood samples were collected from the cubital vein using vacuum tubes (Venoject VP-H070K, Terumo, Tokyo, Japan) and stored at − 70°C till the measurement. A long-time blood samples storage, which may increase RBC lazing percentage, lead to measure metal concentrations in whole blood instead of level in serum. We used an inductively coupled plasma mass spectrometer (ICP-MS, Eran DRC-II, PerkinElmer, Waltham, MA, USA) for the blood metals levels analysis. Measurements were repeated three times and the average was used for subsequent analysis. For instrument calibration at least 10% of the analyses were the external standard, and 5% were blank samples (pure water).

Quality control utilized Certified Reference Materials (CRMs) provided by the Institute for Reference Materials and Measurements, European Union. The CRMs were used as accuracy control samples and their metal concentrations were measured by the same method as the blood samples taken from study subjects. All results obtained for the CRMs agreed with certified values at the 95% confidence interval level using the Student t-test.

The limit of detection (LOD) and the limit of quantification (LOQ) were 0.03 and 0.09 g/dL, respectively. Values below LOQ were replaced by half the LOQ for analysis. The present study aimed to study a bigger set of metals, such as antimony, cobalt, and cadmium, which could induce adverse pregnancy outcomes [29–32]. However, we had to withdraw them from statistical analysis due to a significant number of undetectable concentrations. Due to some outliers’ value of metal concentrations and improving model fit (reduce outlier effect), we also used natural log-transformation of blood metals in statistical analyses.

Outcome and Covariates Assessment

Participants’ anthropometric characteristics were obtained in early pregnancy. The newborn’s weight (gr) was measured using an infant’s standard scale. The mother’s weight (kg) was measured using a standard scale and length in standing position, using a rigid stadiometer. Body mass index (BMI) was calculated as weight (kg) divided by square of length (m²). Pregnancy weight gain (kg) was calculated by subtracting the woman’s weight just before delivery, which was measured at the time of admission for labor, from the body weight at the study’s recruitment. The full gestational age (week) was calculated from the first day of the last menstrual period till baby delivery.

Structured questionnaires collected medical history, reproductive history, and socio-demographic information during a face-to-face interview by trained research staff. Pregnancy outcomes, including newborn anthropometrics characteristics, were collected by research collaborators. After excluding 13 incomplete observations, statistical analysis was performed on 579 mother-newborn pairs.

Statistical Analyses

Descriptive statistics, Pearson’s correlation coefficient, Student t-test, and Wilcoxon test were employed. Linear regression examined the associations between individual metal and birth weight, adjusting for covariates, which might associate with birth weight and/or blood metal concentrations, including maternal age [33], pregnancy weight gain, maternal BMI [34], gestational age, newborn sex [35], and maternal hematocrit [36].

BKMR analysis assessed effects of metal mixtures (using exposure-response function) on birth weight, considering covariates (maternal age, pregnancy weight gain, maternal BMI, newborn sex, hematocrit, gestational age, and country). This statistical method can fit the high-dimensional non-linear and non-additive associations of mixture predictors on outcomes. The analysis was done by “bkmr function” via implements the Markov Chain Monte Carlo (MCMC) sampler to fit the BKMR. A detailed computational and theoretical background of the BKMR is available elsewhere [26, 28]. We set the argument for 20,000 iterations of the MCMC procedure. In addition, we set link function (family = "gaussian") and the argument variable selection (varsel = TRUE) to calculate the posterior inclusion probability (PIPs; range from 0 to 1) values for each metal at component-wise variable selection. Individual metals with a PIPs ≥ 0.5 considered important in the model [37, 38].

To visualize results of the high-dimensional exposure-response functions and interactive associations between metals and birth weight, statistical findings presented in graphical categories using “ggplot2 package” in R software. 1) The overall effects of all metals together (from quantiles 25th to 75th ) on birth weight were compared to zero change when metals fixed at their median (50th quantile). 2) Using the bivariate predictor-response function we computed potential interactions between a single metal (predictor fixed at its median) and the second metal placed at its 25th, 50th, or 75th quantiles, on birth weight. 3) Univariate exposure-response function was run to reveal relationship between each metal and birth weight, where all the other metals were fixed at the 50th quantile (median value). 4) Furthermore, the BKMR analyses calculated the contribution of a single metal on birth weight when the metal was fixed at its 25th compared to the 75th quantiles, where all other metals were fixed at their 25th, 50th, or 75th quantiles.

All statistical analyses were conducted using SAS (version 9.4; SAS Institute Inc., Cary, NC) and R statistical software (version 4.2.2, R Foundation for Statistical Computing Platform). The accepted statistical significance level (p-value) was ≤ 0.05.

Participants’ Characteristics Comparison

Comparison of maternal and newborn demographic information was conducted between Iranian and Japanese participants (Table 1). On average, participants were 31 years old; Japanese pregnant women were 9.1 years older than Iranian counterparts (p < 0.01). Maternal weight, BMI, and pregnancy weight gain were significantly higher in Iranian women than Japanese (Δ = 11.9 kg, 4.2 kg/m², and 1.1 kg, respectively; p < 0.05). Similarly, Iranian newborns had a significantly higher birth weight than Japanese babies (Δ = 216 gr, p < 0.01).

Table 1

Maternal and newborns characteristics and prenatal blood metals levels, stratified by country of origin
Variables *	Total (n = 579)	Japan (n = 386)	Iran (n = 193)	p value^$
Maternal
age (year)	31.4 ± 6.3	34.5 ± 4.8	25.4 ± 4.3	< 0.01
height (cm)	159.7 ± 5.4	159.3 ± 5.1	160.7 ± 5.9	< 0.01
Body mass index (kg/m²)	22.2 ± 3.7	20.8 ± 2.7	25.0 ± 3.9	< 0.01
pregnancy weight gain (kg)	10.8 ± 4.2	10.5 ± 3.3	11.6 ± 5.5	< 0.05
hematocrit	37.3 ± 3.2	36.3 ± 2.7	39.4 ± 3.1	< 0.05
early pregnancy weight (kg)	56.8 ± 10.2	52.8 ± 7.1	64.7 ± 10.9	< 0.01
Newborn
weight (gr)	3133 ± 415	3060.8 ± 381.1	3277.5 ± 443.8	< 0.01
height (cm)	49.0 ± 2.3	48.2 ± 2.1	50.5 ± 2.1	< 0.01
gestational age at delivery (week)	39.3 ± 1.2	39.3 ± 1.3	39.2 ± 1.0	0.297
Prenatal blood metal levels (µg/L)				p value^†
arsenic (As)	11.4 ± 4.9	9.9 ± 4.6	14.5 ± 4.1	< 0.01
copper (Cu)	1270 ± 249	1303 ± 248	1204 ± 239	< 0.01
lead (Pb)	18.7 ± 17.5	9.8 ± 5.5	36.4 ± 19.6	< 0.01
manganese (Mn)	12.4 ± 3.8	11.1 ± 3.0	15.1 ± 3.8	< 0.01
rubidium (Rb)	1602 ± 437	1857 ± 281	1094 ± 166	< 0.01
selenium (Se)	198 ± 90	231 ± 91	130 ± 28	< 0.01
zinc (Zn)	4399 ± 1037	4200 ± 1015	4799 ± 966	< 0.01
* mean ± standard deviation, $ Student t−test, † Wilcoxon test

Among the seven metals, concentrations of As, Mn, Pb, and Zn were significantly higher (Δ = 4.6, 4.0, 26.6, and 599 µg/L, respectively, p < 0.01), while Cu, Rb, and Se levels were significantly lower (Δ = 101, 763, and 99 µg/L, respectively, p < 0.01) in Iranian participants compared to Japanese pregnant women (Table 1).

Pearson Correlation Coefficients

Pearson correlation coefficient analysis tested the relationship between prenatal blood metals levels and birth weight (r: As = 0.047, Cu=-0.079, Pb = 0.170, Mn = 0.086, Rb=-0.212, Se=-0.129, and Zn = 0.038). Blood Pb had positive correlations with As, Mn, and Zn (0.391, 0.358, and 0.286, respectively; p < 0.05), and negative correlations with Rb and Se (-0.544 and − 0.307, respectively; p < 0.01) (Fig. 1).

Linear Regression Analysis

Unadjusted and adjusted linear regression models were fitted separately for each metal (Table 2). Unadjusted models showed significantly negative correlations for Se and Rb and significant positive correlation for Mn and Pb with birth weight (coefficient; -0.6, -0.2, 9.3, and 4.0 respectively, p < 0.05). Adjusted models found a negative relationship between all seven metals and birth weight (p > 0.05). Maternal age, BMI, pregnancy weight gain, male gender, and gestational age were significantly correlated with birth weight in all models (not shown data).

Table 2

Association between prenatal blood metal levels and birth weight using the linear regression model (n = 597)
Prenatal blood metal levels (µg/L)	Unadjusted coefficient	95% CI	p value	Adjusted^† coefficient	95% CI	p value
arsenic (As)	3.94	-2.92 ~ 10.81	NS	-6.34	-13.07 ~ 0.40	NS
copper (Cu)	-0.13	-0.27 ~ 4.83	NS	-1.30	-0.13 ~ 0.10	NS
lead (Pb)	4.04	2.12 ~ 5.95	< 0.001	-0.25	-2.60 ~ 2.07	NS
manganese (Mn)	9.35	0.48 ~ 18.22	< 0.05	-6.67	-15.36 ~ 2.02	NS
rubidium (Rb)	-0.20	-0.28 ~ -0.13	< 0.001	-5.95	-0.18 ~ 6.12	NS
selenium (Se)	-0.59	-0.97 ~ -0.22	< 0.01	-4.92	-0.42 ~ 0.33	NS
zinc (Zn)	1.52	-0.02 ~ 4.80	NS	-1.51	-4.47 ~ 1.45	NS
† Adjusted for maternal age, body mass index, hematocrit, pregnancy weight gain, newborn sex, country, and gestational age
CI: confidence interval
NS: non−significant

Estimating the Exposure-Response Function using the BKMR Models

Exposure-response function and interactions between metal mixture and birth weight were investigated using the BKMR analysis. All seven metals were important in the MCMC sampling model (PIPs > 0.5). The BKMR overall function estimated a slight decrease in newborns' birth weight as prenatal blood metals levels (combined) elevated (from 12.9 ± 13.5gr to -6.3 ± 16.9gr, Δ = 19.2gr, p > 0.05), while metal mixture levels increased from their 25th to 75th quantiles (Fig. 2).

The univariate and bivariate predictor-response functions showed U-shape relationship between Pb and birth weight. In addition, prenatal blood Pb revealed dose-dependent response in the BKMR bivariate function (decreasing birth weight by increasing Pb percentile), when it added to the model as second metal (Figs. 3 and 4).

The single predictor function analysis presented in the Fig. 5. The largest effect was found between prenatal blood Pb and birth weight (mean ± SD; -14.1 ± 25.3, -16.1 ± 25.0, and − 17.3 ± 25.6, p > 0.05) when Pb level was at 25th, 50th, and 75th quantiles, respectively.

This is the first time examine prenatal blood metal mixture using data from two countries with variation in race, socioeconomic status, and environmental exposure. The present research revealed differences in metal levels and anthropometric characteristics between the two populations. Pearson’s correlation coefficients and unadjusted linear regression analysis found some correlations between prenatal blood metals concentrations and birth weight. Adjusted linear regression analysis found a weak reverse relationship between metals and birth weight. Similarly, the BKMR overall exposure-response function showed decreasing birth weight in response to increase metal levels.

Lead, a well-known toxic metal, extensively studied for toxicities on reproductive system [18, 19, 39]. In the present study, the average blood Pb was three times lower in Japanese women than Iranian (Table 2). Lower Pb in Japanese women could be influenced by earlier phased-out leaded gasoline in Japan than Iran and better protection regulations for pregnant women. The BKMR bivariate analysis showed a U-shape relationship between Pb and birth weight. In addition, the single-variable predictor model found the largest adverse effect on birth weight by blood Pb. Although the underlying mechanisms of adverse health effects are not fully understood, Pb might involve disruption of reproductive hormones [40], induction reactive oxygen species production [41], and interaction with essential elements functions [4].

On average, Rb level was 70% (763µg/L) higher in Japanese women than Iranian. Pearson’s and linear regression analyses showed a significant negative correlation between Rb and birth weight. Although the reproductive system toxicities of Rb have not been sufficiently studied, a few reports have indicated pregnancy concentration increasing [42], fetal growth decreasing [43], and increasing the risk of pregnancy induced-hypertension [44] in Rb exposure groups.

Selenium, critical for human reproduction, thyroid hormone metabolism, DNA synthesis, and protection from oxidative damage [45], showed 76% higher level in Japanese women than Iranian, which may relate to higher tuna fish consumption, one of the popular food in Japan [45]. Our statistical analyses showed a weak negative correlation between Se and birth weight. Similarly, a few previous studies have reported inverse association between birth weight and Se level in blood, serum, and the placenta [46, 47].

Concentrations of blood Mn and As were higher in Iranian pregnant women than Japanese participants. There was not a strong correlation between these two elements and birth weight. On the other hand, some researchers, including our previous study, have reported adverse pregnancy outcomes, including low birth size, induced by prenatal blood Mn [7, 15, 17, 48, 49] and As [50, 51].

The present study also investigated Zn and Cu, necessary for fetal growth and development via many physiological pathways, such as enzyme activities, and protein synthesis [52–54]. Blood concentrations of these two metals may vary based on factors like gender, age, diets, and supplements intakes [55]. The comparison revealed higher Zn and lower Cu levels (roughly 10%) in Iranian pregnant women than Japanese.

While many previous studies have examined the relationship between prenatal metal exposure and birth weight, this study is unique in employing data from two different study populations. Variations in subject characteristics enhance the study’s power to generalize findings to the target population. In addition, the current work considered metal that has been not studied (Rb) enough for effects on reproductive system. However, our findings may be influenced by variations in characteristics between the two populations. Additionally, the study used first-trimester blood metals concentrations for exposure assessment instead of serial measurements. Many studies have reported a higher probability of adverse pregnancy outcomes with early-stage gestation metal exposure [16, 30]. However, blood metals’ half-life is from days to weeks, and concentrations can change during subsequent trimesters. Finally, none of study participants were occupationally exposed to metal, thus, they might be encountered from their diet and polluted environment, but, the study did not measure metal concentrations in their foods and/or environment.

The present study estimated a slightly negative relationship between blood metal levels and birth weight in adjusted linear regression model and a continuously decreasing birth weight in response to increasing prenatal blood metals levels in the overall function of the BKMR. These results might indicate prenatal exposure to potentially toxic metals at current levels unsafe for human reproductive system. Thus, pregnant women should avoid encountering to them as much as possible.

As: Arsenic

BKMR: Bayesian Kernel Machine Regression

CI: Confidence Interval

Cu: Copper

ICP-MS: Inductively Coupled Plasma Mass Spectrometry

IQR: Interquartile Range

LOD: Limits of Quantification

Mn: Manganese

Pb: Lead

PIP: Posterior Inclusion Probability

Rb: Rubidium

SD: Standard Deviation

Se: Selenium

Zn: Zinc

The authors declare that there are no conflicts of interest in this study. The present study was supported by the Japan Society for the Promotion of Science (JSPS; #L22517).

Compliance with Ethical Standards

The original studies were approved by local ethical committees in Tehran University of Medical Sciences (#IR.TUMS.REC.1394.1180) and in the Juntendo University Hospital (#632). Informed constant were obtained from all participants.

Acknowledgements

This study was conducted during Mohsen Vigeh stay at International University of Health and Welfare, Akasaka, Tokyo, under the Japan Society for the Promotion of Science (JSPS) Invitation Fellowships for Research in Japan (L22517).

Weiss, D., et al., Atmospheric Pb deposition since the industrial revolution recorded by five Swiss peat profiles: Enrichment factors, fluxes, isotopic composition, and sources. Environmental Science and Technology, 1999. 33(9): p. 1340-1352.
Candelone, J.P., et al., Post‐Industrial Revolution changes in large‐scale atmospheric pollution of the northern hemisphere by heavy metals as documented in central Greenland snow and ice. Journal of Geophysical Research: Atmospheres, 1995. 100(D8): p. 16605-16616.
Mahar, A., et al., Immobilization of Lead and Cadmium in Contaminated Soil Using Amendments: A Review. Pedosphere, 2015. 25(4): p. 555-568.
Peana, M., et al., Metal Toxicity and Speciation: A Review. Curr Med Chem, 2021. 28(35): p. 7190-7208.
Razzak, S.A., et al., A comprehensive review on conventional and biological-driven heavy metals removal from industrial wastewater. Environmental Advances, 2022. 7: p. 100168.
Centeno, J.A., et al., Environmental Pathology, in Essentials of Medical Geology: Revised Edition. 2013, Springer Netherlands. p. 569-596.
Vigeh, M., et al., Blood manganese concentrations and intrauterine growth restriction. Reprod Toxicol, 2008. 25(2): p. 219-23.
Irwinda, R., N. Wibowo, and A.S. Putri, The Concentration of Micronutrients and Heavy Metals in Maternal Serum, Placenta, and Cord Blood: A Cross-Sectional Study in Preterm Birth. Journal of Pregnancy, 2019. 2019.
Zhang, M., et al., In utero exposure to heavy metals and trace elements and childhood blood pressure in a u.S. urban, low-income, minority birth cohort. Environmental Health Perspectives, 2021. 129(6).
Kot, K., et al., Interactions between 14 elements in the human placenta, fetal membrane and umbilical cord. International Journal of Environmental Research and Public Health, 2019. 16(9).
Bommarito, P.A., et al., Urinary trace metals, maternal circulating angiogenic biomarkers, and preeclampsia: A single-contaminant and mixture-based approach. Environmental Health: A Global Access Science Source, 2019. 18(1).
Lee, M.S., et al., Umbilical cord blood metal mixtures and birth size in Bangladeshi children. Environmental Health Perspectives, 2021. 129(5).
Nishioka, E., et al., Evidence that birth weight is decreased by maternal lead levels below 5μg/dl in male newborns. Reprod Toxicol, 2014. 47: p. 21-6.
Vigeh, M., et al., Prenatal mercury exposure and birth weight. Reprod Toxicol, 2018. 76: p. 78-83.
Vigeh, M., et al., Increased prenatal blood manganese may induce gestational blood pressure. Hypertens Pregnancy, 2016. 35(4): p. 583-592.
Vigeh, M., et al., Low level prenatal blood lead adversely affects early childhood mental development. J Child Neurol, 2014. 29(10): p. 1305-11.
Vigeh, M., et al., Increase in blood manganese induces gestational hypertension during pregnancy. Hypertens Pregnancy, 2013. 32(3): p. 214-24.
Vigeh, M., et al., Blood lead at currently acceptable levels may cause preterm labour. Occup Environ Med, 2011. 68(3): p. 231-4.
Vigeh, M., et al., Early pregnancy blood lead levels and the risk of premature rupture of the membranes. Reprod Toxicol, 2010. 30(3): p. 477-80.
Rahman, M.L., et al., Early pregnancy exposure to metal mixture and birth outcomes - A prospective study in Project Viva. Environ Int, 2021. 156: p. 106714.
Dou, Y., et al., Maternal exposure to metal mixtures during early pregnancy and fetal growth in the Jiangsu Birth Cohort, China. Environ Res, 2022. 215(Pt 2): p. 114305.
Michael, T., et al., Prenatal exposure to heavy metal mixtures and anthropometric birth outcomes: a cross-sectional study. Environ Health, 2022. 21(1): p. 139.
Kramer, M.S. and C.G. Victora, Low Birth Weight and Perinatal Mortality, in Nutrition and Health in Developing Countries, R.D. Semba and M.W. Bloem, Editors. 2001, Humana Press: Totowa, NJ. p. 57-69.
Green, L.R. and M.A. Hanson, 56 - Developmental Effects on the Fetal Circulation, in Fetal and Neonatal Physiology (Fifth Edition), R.A. Polin, et al., Editors. 2017, Elsevier. p. 584-591.e3.
Padula, A.M., Z. Rivera-Núñez, and E.S. Barrett, Combined Impacts of Prenatal Environmental Exposures and Psychosocial Stress on Offspring Health: Air Pollution and Metals. Current Environmental Health Reports, 2020. 7(2): p. 89-100.
Bobb, J.F., et al., Statistical software for analyzing the health effects of multiple concurrent exposures via Bayesian kernel machine regression. Environ Health, 2018. 17(1): p. 67.
Joubert, B.R., et al., Powering Research through Innovative Methods for Mixtures in Epidemiology (PRIME) Program: Novel and Expanded Statistical Methods. Int J Environ Res Public Health, 2022. 19(3).
Bobb, J.F., et al., Bayesian kernel machine regression for estimating the health effects of multi-pollutant mixtures. Biostatistics, 2015. 16(3): p. 493-508.
Lee, Y.M., et al., Association of paternal cadmium and other heavy metal exposure to birth outcomes using propensity score matching. Chemosphere, 2023. 311(Pt 1): p. 136792.
Liu, C., et al., Association of both prenatal and early childhood multiple metals exposure with neurodevelopment in infant: A prospective cohort study. Environ Res, 2022. 205: p. 112450.
Howe, C.G., et al., Prenatal metal(loid) mixtures and birth weight for gestational age: A pooled analysis of three cohorts participating in the ECHO program. Environ Int, 2022. 161: p. 107102.
Howe, C.G., et al., Prenatal metal mixtures and fetal size in mid-pregnancy in the MADRES study. Environ Res, 2021. 196: p. 110388.
Bulka, C.M., P.A. Bommarito, and R.C. Fry, Predictors of toxic metal exposures among US women of reproductive age. Journal of Exposure Science and Environmental Epidemiology, 2019. 29(5): p. 597-612.
Tela, F.G., A.M. Bezabih, and A.K. Adhanu, Effect of pregnancy weight gain on infant birth weight among mothers attending antenatal care from private clinics in Mekelle City, Northern Ethiopia: A facility based follow-up study. PLoS One, 2019. 14(3): p. e0212424.
Freire, C., et al., Placental metal concentrations and birth outcomes: The Environment and Childhood (INMA) project. International Journal of Hygiene and Environmental Health, 2019. 222(3): p. 468-478.
Kaur, G., et al., Human red blood cell uptake and sequestration of arsenite and selenite: Evidence of seleno-bis(S-glutathionyl) arsinium ion formation in human cells. Biochem Pharmacol, 2020. 180: p. 114141.
Cathey, A.L., et al., Individual and joint effects of phthalate metabolites on biomarkers of oxidative stress among pregnant women in Puerto Rico. Environment International, 2021. 154: p. 106565.
Barbieri, M.M. and J.O. Berger, Optimal predictive model selection. The Annals of Statistics, 2004. 32(3): p. 870-897, 28.
Vigeh, M., et al., Lead and other trace metals in preeclampsia: a case-control study in Tehran, Iran. Environ Res, 2006. 100(2): p. 268-75.
Ren, J., et al., Low-level lead exposure is associated with aberrant sperm quality and reproductive hormone levels in Chinese male individuals: Results from the MARHCS study low-level lead exposure is associated with aberrant sperm quality. Chemosphere, 2020. 244: p. 125402.
Shilpa, O., et al., Lead (Pb) induced Oxidative Stress as a Mechanism to Cause Neurotoxicity in Drosophila melanogaster. Toxicology, 2021. 462: p. 152959.
Gu, T., et al., An Evaluation of Exposure to 18 Toxic and/or Essential Trace Elements Exposure in Maternal and Cord Plasma during Pregnancy at Advanced Maternal Age. International Journal of Environmental Research and Public Health, 2022. 19(21).
Callan, A.C., et al., Sex specific influence on the relationship between maternal exposures to persistent chemicals and birth outcomes. International Journal of Hygiene and Environmental Health, 2016. 219(8): p. 734-741.
McKeating, D.R., et al., Circulating trace elements for the prediction of preeclampsia and small for gestational age babies. Metabolomics, 2021. 17(10).
Sunde, R., Selenium, in Modern Nutrition in Health and Disease, A. Ross, et al., Editors. 2012, Lippincott Williams & Wilkins: Philadelphia, PA. p. 225-37.
Álvarez-Silvares, E., et al., Placental levels of essential and non-essential trace element in relation to neonatal weight in Northwestern Spain: application of generalized additive models. Environ Sci Pollut Res Int, 2023.
Mehta, R., et al., Associations between maternal and infant selenium status and child growth in a birth cohort from Dhaka, Bangladesh. Br J Nutr, 2023: p. 1-39.
Mullin, A.M., et al., Maternal blood arsenic levels and associations with birth weight-for-gestational age. Environmental research, 2019. 177: p. 108603.
Guan, H., et al., Prenatal exposure to arsenic and its effects on fetal development in the general population of Dalian. Biological trace element research, 2012. 149: p. 10-15.
Wang, Y., et al., Effects of prenatal exposure to arsenic on neonatal birth size in Wujiang, China. Chemosphere, 2022. 299: p. 134441.
Xu, F.X., et al., Association between gestational arsenic exposure and intrauterine growth restriction: the role of folate content. Environ Sci Pollut Res Int, 2022. 29(59): p. 89652-89661.
Collins, J., Copper, in Modern Nutrition in Health and Disease, A. Ross, et al., Editors. 2014, Lippincott Williams & Wilkins: Baltimore, MD. p. 206-16.
Prohaska, J., Copper, in Present Knowledge in Nutrition, J. Erdman, I. Macdonald, and Zeisel SH, Editors. 2012, Wiley-Blackwell: Washington, DC. p. 540-53.
MacDonald, R.S., The role of zinc in growth and cell proliferation. J Nutr, 2000. 130(5S Suppl): p. 1500s-8s.
Hennigar, S.R., et al., Serum Zinc Concentrations in the US Population Are Related to Sex, Age, and Time of Blood Draw but Not Dietary or Supplemental Zinc. J Nutr, 2018. 148(8): p. 1341-1351.

Download PDF

Reviewers agreed at journal
18 Mar, 2024
Reviewers invited by journal
13 Mar, 2024
Editor assigned by journal
05 Feb, 2024
First submitted to journal
30 Jan, 2024

You are reading this latest preprint version

Prenatal Exposure to Metal Mixture and Birth Weight; a Bayesian Kernel Machine Regression Analysis of Two Cohort Studies in Japan and Iran

Status:

Version 1

Abstract

Purpose

Methods

Results

Conclusion

Figures

Introduction

Methods

Study Population

Metal Exposure Assessment

Outcome and Covariates Assessment

Statistical Analyses

Results

Participants’ Characteristics Comparison

Pearson Correlation Coefficients

Linear Regression Analysis

Estimating the Exposure-Response Function using the BKMR Models

Discussion

Conclusions

Abbreviations

Declarations

References

Status:

Version 1