Placental levels of metals and associated factors in urban and sub-urban areas of Seville (Spain)

Depósito de investigación de la Universidad de Sevilla https://idus.us.es/ “This is an Accepted Manuscript of an article published by Elsevier in: Journal of Trace Elements in Medicine and Biologyon on July 2019, available at: https://doi.org/10.1016/j.jtemb.2019.03.006” 1 Placental levels of metals and associated factors in urban and sub-urban areas of Seville 1 (Spain) 2 Cerrillos L1, Fernández R1, Machado MJ1, Morillas I1, Dahiri B2, Gonzalez-Weller D3, Gutiérrez 3 A3, Rubio C3, Hardisson A3, Moreno I2*, Fernández-Palacín A4 4 1 Department of Genetics, Reproduction and Fetal Medicine, Hospital Universitario Virgen del 5 Rocío, Avda. Manuel Siurot, 41013 Sevilla, 6 2 Obstetrics, Gynaecology, Paediatrics, Preventive Medicine and Public Health, Toxicology and 7 Legal Medicine. Facultad de Medicina, Universidad de La Laguna. Campus de Ofra s/n, 38071 8 La Laguna, Tenerife, Spain. 9 3 Nutrition, Food Chemistry and Toxicology, Facultad de Farmacia, Universidad de 10 Sevilla. C/ Profesor García González 2, 41012 Sevilla, Spain 11 4Area of Preventive Medicine and Public Health, Facultad de Medicina, Avda. Dr. Fedriani, s/n, 12 41009, Sevilla, Spain 13 14 15 16 *Corresponding author. C/ Profesor García González 2, 41012 Sevilla, Spain 17 Tel.: 0034954556762. Fax: 0034954233765. E-mail: [email protected] 18 19 20 Declarations of interest: none 21 Conflict of interest: there is no conflict of interest 22 23 2 24 Abstract 25 Environmental exposure to metals among women, revealed their adverse effects on 26 pregnancy. The fetus is exposed to these toxic elements only via the placenta which are able to 27 accumulate there or cross it, compromising the protective functions of this organ. Numerous 28 studies have shown associations between the prenatal exposition to some metals and an 29 impact on cognitive, motor and intellectual development of the child. Sixty two placental 30 samples were taken at delivery to determine the mineral content (Al, B, Ba, Ca, Cd, Cr, Cu, Fe, 31 K, Li, Mg, Mn, Mo, Na, Ni, Pb, Sr, V, Zn) by ICP-OES. Among these metals, essential ones (B, Ca, 32 Cu, Fe, Mg, Mn, Mo, Na, Zn) can have health beneficial effects at low levels however, in high 33 concentration are potentially toxic. On the other hand, elements such as Al, Cd, Pb, are 34 classified as toxic metals, no matter what its concentration is. The aim of this study is to find 35 the potential relationships between these metals levels, newborn's parameters, pregnancy 36 details and the epidemiologic information obtained using a questionnaire data from the 37 participant pregnant women from Seville (Spain). 38 The main maternal determinant of detectable placenta Cd levels was smoking during 39 pregnancy. Other maternal factors that may affect placenta metal levels were gestational age 40 (Al, B, Ba, and Pb) or dietary supplement (Fe). It has to be stressed that our results have to be 41 interpreted with caution, because of the small study group and the low exposure levels, along 42 with the lack of information on potential sources of exposure to these metals. The use of 43 placenta samples obtained at delivery can be considered strength of this study since the 44 concentration of some metals in placenta can indicate the extent of maternal exposure during 45 gestation. 46 47 Keywords: metals, placenta, associated factors, pregnant women, Spain 48 49 3 Introduction 50 Pregnant women are exposed to a wide variety of foreign chemicals through their diet, 51 maternal medication, lifestyle factors, such as smoking, drug abuse, and alcohol consumption, 52 or occupational and environmental sources [1]. Prenatal period is considered to be the most 53 sensitive phase in human development. During this time fetal cells are the subjects of division 54 and differentiation. This is the reason of high susceptibility of fetus to neurotoxic stressors 55 which can easily cause developmental alterations [2]. The fetus is exposed to these toxic 56 contaminants only via the placenta. When substances reach the placenta they can be 57 metabolized and accumulated, thus acting as an efficient or partially protective barrier [3] or 58 cross it by passive diffusion or by active transport, pinocytosis or filtration. Some lipophilic 59 compounds of low molecular weight are able to cross the placenta without restriction and may 60 impact physiological transport mechanisms, compromising the protective functions of this 61 organ [4, 5]. Placental tissue is considered as non-invasive exposure biomarker for different 62 organic and inorganic pollutants. Thus, Cd concentrations were 10 times higher than that 63 detected in maternal blood and up to 100 times higher than that of cord blood indicate that Cd 64 accumulates in the placenta during gestation. This tissue usually discarded after birth, is easy 65 to obtain and may reflect the simultaneous exposure status to these contaminants of mothers 66 and their fetus as well as the first lifetime exposure time point of the fetus [3, 6]. 67 Heavy metals have been found to have an impact, in particular on cognitive, motor and 68 intellectual development after a prenatal exposure [7, 8]. Cadmium (Cd), and lead (Pb) are 69 developmental toxicants with no known biological function that are transported across the 70 placenta [7].Thus, has been reported that prenatal exposure to Pb and Cd is associated with 71 reduce birth weight, birth length and head circumference [9]. Also, infertility, spontaneous 72 abortions, and fetal and neonatal death have been reported after either male or female 73 occupational exposure to Pb [10]. Cd exposition is associated mainly with cigarette smoke and 74 its ovarian concentration increases with age, and has been associated with retardation of 75 4 trophoblastic outgrowth and development, placental necrosis and suppression of steroid 76 biosynthesis, and altered handling of nutrient metals by the placenta all contribute to 77 implantation delay and possible early pregnancy loss [11]. Maternal iron (Fe) status has been a 78 critical factor for pregnancy outcomes because maternal anemia as well as iron deficiency 79 increases the risk of adverse pregnancy outcomes such as preterm delivery and low birth 80 weight [10]. Treatment with Lithium (Li), as well as exposure to the metal via drinking water 81 and other environmental sources, is associated with adverse health outcomes [12]. 82 On the other hand, human organism has requirements of appropriate amounts of essential 83 metals such as K, Ca, Co, Cr, Cu, Mg, Mn, and Zn to maintain optimum health, though outside 84 of beneficial intake ranges, either elevated or reduced levels of these metals can adversely 85 affect health [12-13]. The physiological role of Mn in the developing organism in the animal 86 model is well documented, but little is known about the adverse effects of Mn deficiency or 87 overexposure in the human, including pregnancy outcome and birth weight [14]. Dietary zinc 88 deficiencies also cause developmental problems throughout pregnancy [10]. 89 It is therefore important to assess the nature and degree of exposure in pregnant women 90 because also is the exposure of the unborn child. Many various factors may have a significant 91 influence on the prenatal concentration of toxic metals, e.g. smoking during pregnancy, 92 maternal age, gestational age [15]. It seems to be useful to know about the relationship 93 between the above-specified factors and the concentration of metals in placenta. The aim of 94 the study was the simultaneous determination by ICP-OES of Al, B, Ba, Ca, Cd, Cr, Cu, Fe, K, Li, 95 Mg, Mn, Mo, Na, Ni, Pb, Sr, V, Zn in placental samples taken at delivery to find the potential 96 relationships between these metals concentration, newborn's parameters, pregnancy details 97 and the epidemiologic information obtained using a questionnaire data from the participant 98 pregnant women from Seville (Spain). 99 Material and methods 100 Study population 101 5 Sixty two healthy women living in Seville, south of Spain, participated in this research study in 102 year 2016. Only women who had been residents in the study area and who were older than 18 103 years old at the time of delivery were recruited for this study. Women who followed any 104 program of assisted reproduction or who suffered from chronic hypertension, any type of 105 diabetes, thyroid disorders or chronic renal or cardiac disease during pregnancy were 106 excluded. The study was approved by Coordinating Committee for the Ethics of Biomedical 107 Research of Andalusia. The consents were obtained from all the pregnant who agreed to 108 participate in this study and who were admitted. 109 The samples of placenta were taken just after the delivery in the Department of Genetics, 110 Reproduction and Fetal Medicine in Hospital Virgen del Rocío (Sevilla). Five samples of 111 approximately 5 g each were taken from its four corners and one from the middle. All portions 112 of the samples were about 1 cm in width and length and 0.5 cm thickness. No gross 113 morphological abnormalities were observed. Samples were washed twice in cold PBS, cut into 114 smaller pieces and homogenized. Then the homogenized samples were immediately coded, frozen 115 and stored at -80ºC until analysis. Samples were supported with the newborn's parameters 116 (bodyweight, body length, Apgar score) and sex of the child. Also a detailed questionnaire with 117 demographic, socioeconomic, lifestyle, and environmental data was filled in a face to face 118 interview. 119 Chemical and Reagents 120 Panreac (Barcelona, España) standard solutions of about 1000 mgL-1 were used as stock 121 solution of each element for calibration. Other reagents used where of analytical grade. Milli-Q 122 treated water was used throughout. 123 Analytical Procedure 124 Placental samples with presence of an excess of blood were discarded to avoid the 125 contamination of the placental tissue though the maternal and fetal blood. A portion of 126 10.2±0.5 g of each homogenized sample was placed in porcelain crucibles that were desiccated 127 6 for 24 hours in an oven at 70ºC. The samples were then subjected to incineration in a muffle 128 furnace with a temperature-time program of 450±25ºC-18 to 24 hours, until the production of 129 white ashes which were dissolved in nitric acid 5% to a volume of 50 mL [16]. The final 130 dissolution factor of the sample was 1/5 (m/V) (10 g of sample diluted in 50 ml of acid). 131 132 The metals were determined by Inductively Coupled Plasma-Optical Emission Spectrometry 133 (ICP-OES) technique; model ICAP 6300 Duo Thermo Scientific. The instrumental conditions 134 were as follows: approximate RF power, 1150W; gas flow (nebulizer gas flow, auxiliary gas 135 flow), 0.5 L/min; injection of the sample to the pump flow, 50 rpm; stabilization time, 0 s. 136 Quality Controls 137 Detection and quantification limits under reproducibility conditions, which were estimated as 138 three and ten times the standard deviation (SD) resulting from the analysis of 15 blanks [17] 139 were the lowest for Cd and Pb LOD: 0.003 µg/g and LOQ: 0.008 µg/g and the highest for Na: 140 LOD: 9.142 µg/g and LOQ: 30.46 µg/g, Ca: LOD: 4.833 µg/g and LOQ: 16.29 µg/g and K: LOD: 141 4.708 µg/g and LOQ: 15.70 µg/g. 142 Quality controls were performed to verify the accuracy of the analytical procedure. These 143 controls were based on the study of the recovery percentage obtained with blank samples and 144 reference material measured under reproducible conditions. . Reference materials (SRM 145 1566b Oyster Tissue, SRM 1573a Tomato Leaves and SRM 1515 Apple Leaves from the 146 National Institute for Standards and Technology (NIST)) were processed by digestion as the 147 same way as placenta samples. The recoveries obtained were all upper than 92%. Along the 148 analytical procedures, each batch of 20 samples was analyzed together with at least a blank 149 and a reference sample. Calibration was performed using the calibration curve method. 150 Statistical Analysis 151 Statistical analysis was performed using the statistical package IBM Statistics SPSS 24.0 152 (Statistical Package for the Social Sciences). To describe the characteristics of our participants, 153 7 descriptive statistical parameters were initially computed. Arithmetic mean ± standard 154 deviation, geometric mean, minimum and maximum values of concentration above the limit of 155 detection were calculated for potential continuous variables. Categorical variables were 156 described using frequencies and percentages. 157 Spearman's correlation analysis was conducted to assess relationships between metal 158 concentrations in placenta tissue. Normality of distributions for each continuous variable was 159 assessed with Shapiro-Wilks test; non-parametric Kruskal-Wallis test were performed in 160 bivariate analyses in not normally distributed variables. ANOVAs were used to compare 161 continuous variables between two groups. 162 Multiple linear regression models were conducted to examine potential factors contributing to 163 exposure levels, with logarithm-transformed concentrations of the metals as dependent 164 variables. Previously, it was made a selection of the variables that will be introduced in the 165 model by calculating Pearson's linear correlation coefficients. Due to the high number of 166 potentially candidate variables to be introduced in the multiple linear regression models, we 167 applied a previous pre-selection based on bivariate analysis, following the methodology used 168 in all the multivariate techniques of construction of regression models. All covariates were 169 tested in multivariate analysis and only were included in the final model those which were 170 significantly associated with metal levels in prior bivariate analyses (P≤0.1). The collinearity of 171 the independent variables also was considered by calculating the tolerance and the variance 172 inflation factor. 173 Results and discussion 174 Study participants 175 The age of the mothers at delivery was mainly ranging from 31 to 40 years. Most of the 176 participants were from Spain and just the 42% of all of them resided in the metropolitan area 177 (Seville), with the remaining participants spread across the sub urban area, predominantly 178 around the west of the province of Seville. Only the 45.2% of the participants had tertiary 179 8 education studies. Only 13% of the women were working during the pregnancy. The majority 180 of participating mothers did not smoke, though the 22.6% of the participants were smoking 181 along the pregnancy. Also it was considered if pregnant took or not supplemental Fe. 182 Factors potentially connected to metal levels and/or sociodemographic and lifestyle 183 characteristics were also considered. Thus, 92% of participants had not previous premature 184 delivery, neither previous misbirth. The gestational age was 39±1.4 weeks, the mean neonatal 185 weight around 3229±646 g and the apgar score at 10 minutes was in all the cases 10. 186 Profile of metals in placenta 187 The concentration and descriptive statistics for each metal in placenta are reported in Table 1. 188 No differences were detected in metal levels in placentas from male or female newborn in our 189 study. There are only a few recent studies that have investigated essential elements in human 190 placenta. Our results showed that almost 100% of placenta samples contained quantifiable 191 amounts of alkali and alkaline earth metals, with the exception of Ba and Ca (98 and 95%, 192 respectively). Metalloids and post-transition metals reached 100% of LOQ in placenta tissue. 193 The highest concentration were observed for K (geometric mean: 1185.75 µg/g), Na 194 (geometric mean: 1070.53 µg/g), Ca (geometric mean: 352.59 µg/g) and Mg (geometric mean: 195 77.94 µg/g). The concentrations of Ca and Al were similar to those seen in singleton pregnant 196 women in Rochester (NY) [5] but levels of Ca was lower when comparing with both fetal and 197 maternal placental from teenagers and adults women from Brazil [18], this differences may be 198 due to dietary patterns. However, levels of Ba, Mg and Sr were 44-fold, 3-fold and 4-fold 199 higher respectively, in comparison with de Angelis et al. [5] study. It should be noted the great 200 difference in Ba levels from the two locations (Rochester vs Seville). Taking into account that 201 this metal is not present in food, the possible explanation may be differences in geographical 202 pollution. Among transition metals Ni and Cr were the elements with the high percentages of 203 samples with concentration below quantification limits (79 and 77%, respectively) followed by 204 Mn and V with a 40% under the LOQ. Concentrations were particularly low for Molybdenum 205 15 [9] Gundatacker, C., Hengstschlager, M. The role of the placenta in fetal exposure to heavy 338 metals. Wien Med Wochenschr 162 (2012) 201–206. 339 [10] Sengupta, P., Banerjee, R., Nath, S., Das, S. Banerjee, S. Metals and female reproductive 340 toxicity. Hum. Exp. Toxicol. 34 (2015) 679–697. 341 [11] Thompson, J., Bannigan, J. Cadmium: Toxic effects on the reproductive system and the 342 embryo. Reprod. Toxicol. 25 (2008) 304-315. 343 [12] Callan, A.C., Hinwood, A.L., Ramalingam, M., Boyce, M., Heyworth, J., McCafferty, P., 344 Odlandet., J.Ø. Maternal exposure to metals—Concentrations and predictors of exposure. 345 Environ. Res. 126 (2013) 111-117. 346 [13] Alvarez, M., Moreno, I.M., Pichardo, S., Cameán, A.M., González, A. G. Metallic profiles of 347 Sherry wines using inductively coupled plasma atomic emission spectrometry methods. Sci. 348 Alim. 27 (2007) 80-89. 349 [14] Vigeh, M., Yokoyama, K., Ramezanzadeh, F., Dahaghin, M., Fakhriazad, E., Seyedaghamiri, 350 Z., Araki, S. Blood manganese concentrations and intrauterine growth restriction. Reprod. 351 Toxicol. 25 (2008) 219-223. 352 [15] Wells, E., Jarret, J., Lin, Y., Caldwell, K.L., Hibbeln, J.R., Apelberg, B.J., Herbstman, J., 353 Halden, R.U., Witter, F.R., Goldman, L.R. Body burdens of mercury, lead, selenium and copper 354 among Baltimore newborns. Environ. Res. 111 (2011) 411-417. 355 [16] Hernández-Sánchez, C., Luis, G., Moreno, I., Cameán, A., González, A.G., González-Weller, 356 D., Castilla A., Gutiérrez, A., Rubio C., Hardisson A.,. Differentiation of mangoes (Magnifera 357 indica L.) conventional and organically cultivated according to their mineral content by using 358 support vector machines. Talanta 97 (2012) 325-330. 359 [17] IUPAC (International Union of Pure and Applied Chemistry). Nomenclature in evaluation of 360 analytical methods including detection and quantification capabilities. Pure Appl. Chem. 67 361 (1995) 1699–1723. 362 16 [18] de Moraes, M. L., de Faria Barbosa, R., Santo, R. E., da Silva Santos, F., de Almeida, L. B., 363 de Jesus, E. F. O., de Carvalho Sardinha, F. L., do Carmo, M. d. T. . Distribution of Calcium, Iron, 364 Copper, and Zinc in Two Portions of Placenta of Teenager and Adult Women. Biol Trace Elem 365 Res 143 (2011) 1271–1281. 366 [19] Kubala-Kukus, A., Banas, D., Braziewicz, J., Majewska, U., Pajek, M. Comparative study of 367 trace element contents in human full-term placenta and fetal membranes by total reflection X- 368 ray fluorescence. Spectroc. Acta Part B 58 (2003) 725–734. 369 [20] Amaya, E., Gil, F., Freire, C., Olmedo, P., Fernández-Rodríguez, M., Fernández, M.F., Olea, 370 N. Placental concentrations of heavy metals in a mother–child cohort. Environ. Res. 120 (2013) 371 63–70. 372 [21] Freire, C., Amaya, E., Gil, F., Fernández, M. F., Murcia, M., Llop, S., Andiarena, A., 373 Aurrekoetxea, J., Bustamante, M., Guxens, M., Ezama, E., Fernández-Tardón, G., Olea, N.,. 374 Prenatal co-exposure to neurotoxic metals and neurodevelopment in preschool children: The 375 Environment and Childhood (INMA) Project. Sci. Total Environ. 621 (2018) 340-35. 376 [22] Al-Saleh, I., Shinwari, N., Mashhour, A., Mohamed, G.E.D., Rabah, A. Heavy metals (lead, 377 cadmium and mercury) in maternal, cord blood and placenta of healthy women. Int. J. Hyg. 378 Environ. Health 214 (2011) 79–101. 379 [23] Sakamoto, M., Yasutake, A., Domingo, J.L., Chan, H.M., Kubota, M., Murata, K. 380 Relationships between trace elements concentrations in chorionic tissue of placenta and 381 umbilical cord tissue: Potential use as indicator for prenatal exposure. Environ. Int. 60 (2013) 382 106-111. 383 [24] Kantola, M., Purkunen, R., Kroger, P., Tooming, A., Juravskaja, J., Pasanen, M., Saarikoski, 384 S., Vartiainen, T. Accumulation of Cadmium, Zinc, and Copper in Maternal Blood and 385 Developmental Placental Tissue: Differences between Finland, Estonia, and St. Petersburg. 386 Environ. Res. Sect. A 83 (2000) 54-66. 387 17 [25] Ronco, A.M., Arguello, G., Muñoz, L., Gras, N., Llanos, M. Metals content in placentas from 388 moderate cigarette consumers: correlation with newborn birth weight. BioMetals 18 (2005) 389 233–241. 390 [26] Sabra, S., Malmqvist, E., Saborit, A., Gratacós, E., Gomez-Roig, M.D. Heavy metals 391 exposure levels and their correlation with different clinical forms of fetal growth restriction. 392 PLoS ONE 12, (2017) e0185645. 393 [27] Jin, L., Zhang, L., Li, Z., Liu, J-m., Ye, R., Ren, A. Placental concentrations of mercury, lead, 394 cadmium, and arsenic and the risk of neural tube defects in a Chinese population. Reprod. 395 Toxicol. 35 (2013) 25– 31. 396 397 18 TABLE LEGENDS 398 Table 1. Descriptive statistics of the metal concentrations in placenta*. Concentrations are in 399 µg/g unless otherwise noted. 400 Group & element N %>LOD %>LOQ Mean ± SD Geometric Mean Minimum-maximum Alkali Metals Na 62 100 100 1078.94±128.67 1070.53 580.26-1314.76 K 62 100 100 1205.37±206.12 1185.75 482.32-1643.07 Li 62 100 100 0.47±0.34 0.37 0.03-2.34 Alkaline Earth Metals Ba 62 100 98 0.39±0.22 0.34 0.12-1.07 Ca 62 100 95 567.74±697.89 352.60 110.92-2756.31 Mg 62 100 100 77.94±18.42 76.06 54.25-139.70 Sr 62 100 100 0.31±0.25 0.24 0.09-1.06 Metalloids B 62 100 100 0.39±0.11 0.37 0.19-0.63 Post transition metals Al 62 100 100 4.82±5.73 2.68 0.46-26.37 Pb (ng/g) 62 100 100 41.99±14.80 39.78 23.87-84.60 Transition Elements Cd (ng/g) 62 100 94 17.24±12.20 19.92 9.42-72.39 Cr (ng/g) 62 82 23 83.24±28.10 79.59 60.01-144.84 Cu 62 100 100 0.97±0.24 0.94 0.43-1.54 Fe 62 100 100 51.07±13.74 49.14 25.10-81.24 Mn 62 100 60 1.66±9.56 0.11 0.07-58.22 Mo (ng/g) 62 100 94 9.56±2.34 9.30 7.01-15.00 Ni (ng/g) 62 95 21 39.27±23.33 35.05 23.30-109.49 V (ng/g) 62 100 60 72.42±21.71 69.10 40.01-116.26 Zn 62 100 100 8.43±0.97 8.38 6.38-11.27 401 19 Table 2. Results of Multiple Linear regression models of factors influencing placenta metals 402 concentrations. 403 Metal Factors Standardized coefficient 95% confidence interval for beta R2 Adj. R2 Beta Sig Lower bound Upper bound Al Constant Gestational age (weeks) 0.253 0.083 0.047 -13.960 0.003 0.885 0.382 0.064 0.049 B Constant Gestational age (weeks) Age ≤30 >30 Medication (yes/no Fe) 0.286 -0.125 0.245 0.001 0.023 0.313 0.044 -5.496 0.009 -0.218 0.004 -1.454 0.110 0.071 0.281 0.180 0.138 Ba Constant Gestational age (weeks) Vegetable consumption ≤1 >1 0.366 -0.298 0.000 0.003 0.014 -9.436 0.048 -0.546 -2.808 0.217 -0.066 0.210 0.183 Ca Constant Gestational age (weeks) Medication (yes/no Fe) 0.248 0.223 0.719 0.055 0.083 -7.899 -0.004 -0.053 5.487 0.334 0.837 0.107 0.075 Cd Constant Gestational age (weeks) Age ≤30 >30 0.228 -0.156 0.000 0.091 0.245 -11.362 -0.014 -0.410 -3.528 0.184 0.107 0.095 0.062 Cr Constant Neonatal weight (g) -0.525 0.000 0.054 -2.563 0.000 -1.504 0.000 0.276 0.215 Cu Constant Age ≤30 >30 -0.243 0.814 0.057 -0.091 -0.266 0.115 0.004 0.059 0.043 Fe Constant Gestational age (weeks) Fe supplementation (yes/no) 0.226 0.263 0.058 0.069 0.035 -0.062 -0.004 0.011 3.824 0.095 0.288 0.123 0.093 K Constant Maternal education (yes/no university) 0.237 0.000 0.064 6.878 -0.007 7.095 0.234 0.056 0.040 Li Constant Medication (yes/no Fe) 0.214 0.000 0.095 -1.824 -0.048 -0.855 0.577 0.046 0.030 Mg Constant Maternal education (yes/no university) 0.266 0.000 0.036 4.093 0.009 4.336 0.281 0.071 0.055 Mo Constant Residence (Seville/other) Maternal education (yes/no university) 0.219 0.234 0.000 0.095 0.075 -5.043 -0.020 -0.015 -4.702 0.241 0.303 0.101 0.068 Na Constant Age living in this area (>/< 5 years) 0.205 0.000 0.109 6.742 -0.014 7.004 0.133 0.042 0.026 Ni Constant Smoking during pregnancy (yes/no) Neonatal weight (g) -0.105 -0.675 0.000 0.670 0.017 -3.146 -0.680 -0.534 -1.344 0.454 -0.065 0.541 0.457 Pb Constant Gestational age (weeks) Age ≤30 >30 0.275 -0.148 0.000 0.035 0.251 -7.931 0.005 -0.262 -3.321 0.121 0.070 0.118 0.087 Sr Constant Gestational age (weeks) Age ≤30 >30 0.213 -0.216 0.033 0.100 0.095 -10.083 -0.020 -0.645 -0.431 0.224 0.053 0.113 0.082 V Constant Vegetable consumption -0.317 0.000 0.113 -2.680 -0.450 -2.267 0.051 0.222 0.151 20 ≤1 >1 Water consumption ≤1.5 L >1.5 L -0.290 0.146 -0.436 0.069 Zn Constant Gestational age (weeks) Fe supplementation (yes/no) 0.211 0.251 0.001 0.092 0.046 0.608 -0.003 0.001 2.157 0.037 0.112 0.109 0.079 404 405