International Journal of Hygiene and Environmental Health xxx (2013) xxx– xxx
Contents lists available at SciVerse ScienceDirect
International Journal of Hygiene andEnvironmental Health
journa l h om epage: www.elsev ier .com/ locate / i jheh
irth outcome measures and maternal exposure to heavy metalslead, cadmium and mercury) in Saudi Arabian population
man Al-Saleha,∗, Neptune Shinwaria, Abdullah Mashhoura, Abdullah Rabahb
Environmental Health Section, Biological & Medical Research Department, King Faisal Specialist Hospital & Research Centre, PO Box 3354, Riyadh, Saudi ArabiaDepartment of Pediatrics, King Khalid Hospital-Al-Kharj, Saudi Arabia
a r t i c l e i n f o
rticle history:eceived 3 September 2012eceived in revised form 30 March 2013ccepted 22 April 2013
This cross-sectional study was conducted to assess the association between exposure to heavy metals(lead, cadmium and mercury) during pregnancy and birth outcomes in 1578 women aged 16–50 yearswho delivered in Al-Kharj hospital, Saudi Arabia, in 2005 and 2006. The levels of lead, cadmium and mer-cury were measured in umbilical cord blood, maternal blood and the placenta. Outcome variables wereanthropometric measures taken at birth, along with the risk of being small-for-gestational age (SGA). Weselected the 10th percentile as the cutoff for dichotomizing measures of birth outcome. Cadmium, despiteits partial passage through the placenta had the most prominent effect on several measures of birth out-come. After adjustment for potential confounders, logistic regression models revealed that crown-heellength (p = 0.034), the Apgar 5-minute score (p = 0.004), birth weight (p = 0.015) and SGA (p = 0.049) wereinfluenced by cadmium in the umbilical cord blood. Significant decreases in crown-heel length (p = 0.007)and placental thickness (p = 0.022) were seen with higher levels of cadmium in maternal blood. As placen-tal cadmium increased, cord length increased (p = 0.012) and placental thickness decreased (p = 0.032).Only lead levels in maternal blood influenced placental thickness (p = 0.011). Mercury in both umbilicalcord and maternal blood was marginally associated with placental thickness and placental weight, respec-tively. Conversely, placental mercury levels significantly influenced head circumference (p = 0.017), theApgar 5-minute score (p = 0.01) and cord length (p = 0.026). The predictions of these models were furtherassessed with the area under the curve (AUC) of the receiver operating curves (ROCs), which were mod-
est (larger than 0.5 and smaller than 0.7). The independence of gestational age or preterm births on theobserved effect of metals on some measures of birth outcome, suggested detrimental effects of exposureon fetal development. The magnitude of the estimated effects might not necessarily be of clinical sig-nificance for infants but may have a considerable public-health relevance given the high prevalence ofexposure to heavy metals. Further research should be conducted to confirm these findings and to evaluate
ny.
their long-term risks, if a
ntroduction
Even though in utero exposure to heavy metals has been wellnvestigated over the last few decades (Yoshida, 2002; Bellinger,005; Thompson and Bannigan, 2008), our knowledge of the threatso the fetus at low levels of exposure remains either limited ornconsistent (Rahman and Hakeem, 2003; Jelliffe-Pawlowski et al.,006; Tian et al., 2009; Gundacker et al., 2010; Shirai et al., 2010).
Please cite this article in press as: Al-Saleh, I., et al., Birth outcome measuresin Saudi Arabian population. Int. J. Hyg. Environ. Health (2013), http://dx.d
mbilical cord blood, as a noninvasive sample, has been frequentlyested for assessing prenatal exposure to a variety of metals;ther biological tissues such as maternal blood, human milk, urine,
∗ Corresponding author at: Environmental Health Section, Biological & Medicalesearch Department, King Faisal Specialist Hospital & Research Centre, PO Box354, Riyadh 11211, Saudi Arabia. Tel.: +966 11442 4772; fax: +966 11442 4971.
meconium and amniotic fluid can also be used (Gundacker et al.,2010; Shirai et al., 2010; Caserta et al., 2011). The placentahas recently been used as a tool for investigating and predict-ing some aspects of fetal developmental toxicity. It acts as aselective fetal–maternal barrier allowing nutrients and oxygen topass to the fetus and is supposed to prevent potentially harm-ful compounds from crossing (Iyengar and Rapp, 2001). Heavymetals, however, do cross the placenta. Lead and mercury caneasily cross placenta and accumulate in fetal tissues, while cad-mium can partially cross (Iyengar and Rapp, 2001). Studies havesuggested that placental metallothionein might play a protec-tive role against cadmium toxicity by its binding to the metal(Kippler et al., 2010). The detection of cadmium in umbilical cord
and maternal exposure to heavy metals (lead, cadmium and mercury)oi.org/10.1016/j.ijheh.2013.04.009
blood as reported in some studies (Tian et al., 2009; Lin et al.,2011), however, indicates that the role of placental metalloth-ionein as a barrier to cadmium is inconclusive (Nakamura et al.,2012).
Associations have been reported between cadmium in umbili-al cord blood and detrimental effects on the head circumferencef newborns (Lin et al., 2011), thyroid hormone status at birthIijima et al., 2007) or child growth later in life (Tian et al., 2009).ifferent studies, though report conflicting results. Gundacker et al.
2010) found that placental and meconium lead were predictors ofirth height, while birth weight was affected by lead in the placentand maternal blood. Zhang et al. (2004) revealed that cadmium inmbilical cord blood, but not in maternal blood or the placenta,as negatively associated with neonatal birth height. On the otherand, Salpietro et al. (2002) reported that a decrease in birth weightas associated with cadmium in samples of maternal and umbilical
ord blood. Gundacker et al. (2010) found that mercury in hair, butot in umbilical cord blood, maternal blood or the placenta, wasssociated with birth height.
Our previous study on the same population provided evidencehat both the mothers and their newborns were substantiallyxposed to heavy metals, even though 88.7% of the participantsere housewives at the time of the study, and all were non-smokers
Al-Saleh et al., 2011). Lead, cadmium and mercury were detectedn the majority of the three compartments studied (placenta andmbilical cord and maternal blood), confirming their transplacentalransfer. In the present study, we have expanded our data analysesy examining the influence of lead, cadmium and mercury mea-ured in the placenta and in umbilical cord and maternal blood oneasures of birth outcome.
aterials and methods
The data evaluated in this study originated from samples anduestionnaires collected for the project “Exposure to environ-ental pollutants and its effect on pregnancy outcome” (Al-Saleh
t al., 2011). The source population included 1578 women whoelivered between June 2005 and 2006 in a main public hos-ital in the Al-Kharj area located about 80 km southeast of theapital city of Riyadh. None of the women smoked, but 26% hadither a husband or at least one household member who smokedcigarettes, sheesha and/or muaasal). Each woman was asked tooin the study and completed a consent form. The women answered
detailed questionnaire and were interviewed after delivery byrained health-care personnel. The response rate was 99%. Umbil-cal cord blood and the placenta were obtained at the time ofelivery, while maternal blood was collected within next hours ofelivery. The details of sample collection and analytical proceduresave been previously described elsewhere (Al-Saleh et al., 2011).he study protocol was approved by the Research Ethics Committeef King Faisal Specialist Hospital and Research Centre.
Birth anthropometric measurements. These data were obtained byhe obstetrician at birth included birth weight (g), birth height (cm),ead circumference (cm), crown-heel length (cm), Apgar 1-mincore, Apgar 5-minute score, placental weight, placental thickness,nd cord length. Ponderal index was calculated as birth weightkilograms) divided by birth height3 (m) cubed. The cephalizationndex was calculated as the ratio of head circumference (cm) toirth weight (g × 102). Gestational age was calculated from the lastenstrual period to the termination of pregnancy. The obstetrician
xamined each placenta and umbilical cord in the delivery roomnd recorded the placental weight, placental thickness and cordength.
aboratory methods
Please cite this article in press as: Al-Saleh, I., et al., Birth outcome measuresin Saudi Arabian population. Int. J. Hyg. Environ. Health (2013), http://dx.d
Samples of umbilical cords and maternal venous blood wereollected in Vacutainer tubes containing 10.5 mg of K3-EDTAtripotassium-ethylene diaminetetraacetic acid) as anticoagulant
PRESSnd Environmental Health xxx (2013) xxx– xxx
and stored at 4 ◦C until analysis. These tubes were tested formetal contamination before use. All blood and placental sampleswere analyzed for lead and cadmium using a Varian AA-280 Zee-man Atomic Absorption Spectrophotometer coupled to a GTA-120graphite furnace (Varian Techtron Pty. Ltd., Australia). Mercurywas analyzed by a Varian AA-880 Zeeman atomic absorption spec-trophotometer coupled to a Vapor Generation Accessory VGA-77(Varian Techtron Pty. Ltd., Australia). Details of the analytical pro-cedure have been reported previously (Al-Saleh et al., 2011). Theaccuracy of the methods were verified by the analysis of two lev-els of certified reference materials for lead and cadmium (CONTOXHeavy Metal Blood Control A, Kaulson Laboratories, NJ, USA). Theexperimental values agreed well with the certified recommendedvalues for both controls. The experimental values for levels I andIII lead were 19.67 ± 2.72 �g/l N = 88 and 45.76 ± 6.98 �g/l N = 87respectively, while the recommended values were 14.0–22.0 and39.0–51.0 �g/l respectively. The measured values for cadmium forlevel I (10.34 ± 1.89 �g/l, N = 79) and level II (15.79 ± 3.31 �g/l,N = 77) were within the recommended ranges of 4–12 �g/l and11–19 �g/l, respectively. Unfortunately, we were unable to use thereference materials for mercury analysis due to limited sample vol-umes. The bovine muscle powder (SRM 8414) reference materialfrom the National Institute of Standards and Technology (NIST)was used for placental metal analysis. The results of our determina-tions were (0.47 ± 0.132 �g/g, N = 44) for lead, (0.014 ± 0.0058 �g/g,N = 38) for cadmium and (0.0054 ± 0.002 �g/g, N = 23) for mercury.The results are in good agreement with the recommended cer-tified values for lead (0.38 ± 0.24 �g/g), (0.013 ± 0.011) �g/g forcadmium and (0.005 ± 0.003 �g/g) for mercury. The mean recov-eries for spiked blood samples ranged from 102% to 104% for lead,99% to 102% for cadmium and 102% to 107% for mercury with rela-tive standard deviation (%RSD) ranging from 5.6% to 9.1%. The meanrecoveries for placental samples were 99% to 103% for lead, 99% to102% for cadmium and 100% to 109% for mercury, with %RSDs inthe range of 6–11%. The method’s detection limits (MDLs) in bloodwere 0.397 �g/dl, 0.42 �g/l, and 0.25 �g/l for lead, cadmium andmercury, respectively. The MDLs in placental tissues were 0.25 �g/gdry wt., 0.025 �g/g dry wt. and 0.033 �g/g dry wt. for lead, cadmiumand mercury, respectively.
Due to the inaccuracy of self-reported smoking, urinary cotininewas measured as an index of smoking using the commercial Cotin-ine Direct ELISA Kit (Bio-Quant, Inc., USA). Values were correctedfor creatinine which was measured by colorimetric assays (OxfordBiomedical Research, MI, USA).
Statistical analyses. Data were analyzed using SPSS for Windows(version 17; SPSS Inc., Chicago, IL, USA). A p-value of <0.05 was setas the level of statistical significance.
Data are given as arithmetic means, standard deviations (SDs)or proportions (%) unless otherwise stated. When required, skeweddata were natural log-transformed to approximate normalitybefore statistical analyses.
We selected the 10th percentiles as cutoffs for dichotomizingbirth anthropometric measures (head circumference, heel-crownlength, Apgar 1-min score, Apgar 5-minute score, birth weight,birth height, placenta weight, placenta thickness, cord length, pon-deral index and cephalization index) to easily interpret the risk tofetal growth. A further binary outcome for small-for-gestational age(SGA) was created according to the method of Khanjani and Sim(2006) by comparing the birth weight below the 10th percentile ofeach newborn for that gestational age and gender.
Information on potential confounding variables related todemographic, socioeconomic, environmental and maternal and
and maternal exposure to heavy metals (lead, cadmium and mercury)oi.org/10.1016/j.ijheh.2013.04.009
newborn health conditions were obtained from a detailed ques-tionnaire. We also calculated total maternal weight gain bysumming the average weight gains for the second and thirdtrimesters. Among the socioeconomic indicators, only the highest
evel of mothers’ education and the total family income were used.o assess potential confounders, we evaluated their associationsith birth outcome measures or with the metals, using Pear-
on correlation analyses for continuous variables and standards,2-tests, Student-t tests or one way analyses of variance (ANOVA)
or categorical variables followed by a Scheffe post hoc analysishen necessary. Univariate logistic regression analysis was firsterformed to determine the odds of low birth anthropometriceasures (<10th percentile) per log unit change in metal levels
ompared to those above the 10th percentile. A multivariate logisticegression analysis was subsequently constructed after includingnly the variables in the bivariate logistic regression model if the
value was less than or equal to 0.05. Odds ratios (ORs) and 95%onfidence intervals (CIs) are reported. We also repeated all anal-ses after adjusting for gestational age and or excluding pretermirths (delivered before 37 completed weeks of gestation). Finally,he receiver-operating characteristics (ROCs) analysis was appliedo estimate the discriminative value of only those metals that weretatistically significant in predicting the measures of birth outcome.stimates of predictive accuracy were obtained from the area underhe curve (AUC) and its 95% confidence intervals derived from theogistic regression analyses. The practical lower limit for the AUCf a diagnostic test is 0.5. Park et al. (2004) considered an AUCalue larger than 0.5 as having at least some ability to discriminateetween two groups with and without a particular disease.
In this study, undetectable measurements were reported aseros, and we also used quantifiable values below the MDL toptimize statistical power and to avoid biased estimates as recom-ended by Kim et al. (1995). Furthermore, Helsel (2005) suggested
hat substituting results below the MDL with one-half the detectionimit is recommended only if the proportion of values falls below0%. In our study, 10.7%, 24.5%, 29.8% and 52.2% of participantsad levels of maternal blood mercury, placental lead, placentaladmium and placental mercury, respectively, below their MDLs.
esults
eneral characteristics of the study population
Selected demographic, socioeconomic and maternal and neona-al clinical characteristics of the studied population are presentedn Table 1. The mean age of the mothers at the interviews was8.5 ± 6.0 years, mean gestational age was 37.96 ± 1.768 weeks,nd mean birth weight was 3.14 ± 0.54 kg. The mean pregnant bodyass index (BMI) was 27.0, 28.4 and 29.1 kg/m2 during the first,
econd and third trimesters, respectively. The average increase inaternal weight during the second trimester was 2.38 ± 3.9 kg and
eaked during the third trimester at 5.97 ± 5.6 kg. The proportionsf female and male babies in the population were relatively sim-lar (49.1% to 50.9%, respectively). The percentages of newbornsorn prematurely (<37 weeks) or with SGA (<10th percentile) were.9% and 10%, respectively. The average birth height and head cir-umference were 50.15 ± 2.97 cm (24–59 cm) and 34.11 ± 2.05 cm18–55 cm), respectively. Of the newborns, 85.6% were deliveredy normal vaginal birth, and 14.4% were delivered by cesarean sec-ion (C.S.). Labor had been induced in 5.2% of the mothers. Fetalistress during the first 24 hours after birth was observed in 7% ofhe newborns, 6.2% needed resuscitation, 6.7% needed medication,
ainly antibiotics, 3.5% had an abnormal color, 3.7% had abnormalry, 2.6% had abnormal movement, 1.7% had congenital malforma-ions, 10.2% were admitted to either the neonatal intensive care
Please cite this article in press as: Al-Saleh, I., et al., Birth outcome measuresin Saudi Arabian population. Int. J. Hyg. Environ. Health (2013), http://dx.d
nit or the special care baby unit, 0.38% had an injury and 13.2%ad various other clinical problems.
Of the 1578 mothers, 23.4% were either illiterate or attended lit-racy classes, 33.4% had completed either primary or intermediate
PRESSnd Environmental Health xxx (2013) xxx– xxx 3
school, 23.4% had completed high school and 19.9% had completedeither a college diploma or a university degree. The majority ofthe women (88.7%) either never worked or had worked previously.Only 11.3% were working during the period of the study. Of thewomen who worked, most were teachers (8.1%). Income levelsof the participant’s families were: 30.5% ≤5000 Saudi riyals (SR),20.5% in the range of SR5001-7500, 15.4% above SR7501, 22.9% witheither unknown or irregular income and 10.7% refused to answerthe question. The geographical distribution of the current dwellingwas represented by dividing Al-Kharj into five regions: the West-ern Region reported the highest rate, with 37.3% of the participatingwomen, followed by the Southern Region with 20.7%, the EasternRegion with 16.9%, the Northern Region with 15.7%, and the CentralRegion with 9.4%. None of the mothers in the study smoked. How-ever, 26% had either a husband or at least one household memberwho smoked (cigarettes, sheesha or/and muaasal).
Exposure profile and potential confounding variables
Descriptive statistics of the levels of lead, cadmium and mer-cury in maternal and umbilical cord bloods and in placentaltissues are summarized in Table 2. The median levels of thesemetals in maternal blood had the following descending order:lead (2.540 �g/dl) > mercury (1.949 �g/l) > cadmium (0.983 �g/l).The median levels of metals in umbilical cord blood were:mercury (2.876 �g/l) > lead (2.057 �g/dl) > cadmium (0.704 �g/l).The median concentrations of the three metals in placental tis-sue had the following descending series: lead (0.450 �g/g drywt.) > cadmium (0.035 �g/g dry wt.) > mercury (0.031 �g/g dry wt.).The status of exposure for each metal has been reported previouslyin detail (Al-Saleh et al., 2011). However, we will briefly present inthis section the possible confounding variables and the profile ofeach metal.
Lead. The mean lead levels in maternal and umbilical cordblood at delivery were 2.897 ± 1.851 �g/dl, with a range of0.073–25.955 �g/dl, and 2.551 ± 2.592 �g/dl, with a range of0.154–56.511 �g/dl, respectively. Mean lead levels in maternalblood were significantly higher than the levels in umbilical cordblood (t = −9.52, p < 0.001). Maternal blood lead levels were signif-icantly correlated with umbilical cord blood lead levels (r = 0.456,p < 0.001). Lead was detected in all samples of maternal and umbil-ical cord blood, with only seven mothers and six newborns havinglead levels below the MDL (0.42 �g/dl). The mean concentration oflead in placenta was 0.579 ± 2.176 �g/g dry wt. (range, 0–78.0 �g/gdry wt). Lead was detected in 96% of the placental tissues, and morethan 75.4% (N = 1189) of the women had a level higher than the MDLfor lead (0.25 �g/g dry wt.). As shown in Table 3, lead in samplesof umbilical cord and maternal blood was significantly and posi-tively correlated with maternal age, parity and urinary cotinine. Thefirst and second maternal trimester BMIs were only correlated withumbilical cord blood lead levels. ANOVA analyses revealed that leadlevels in both umbilical cord and maternal blood were higher inmothers who were either illiterate or attending literacy class thanthe levels in mothers who attended primary, intermediate, or sec-ondary or college/university, with p < 0.01 for all. The levels of leadin umbilical cord and maternal blood were significantly lower inmothers with total family incomes >SR7500 than in those earn-ing < SR5000, with p-values of 0.003 and 0.015 respectively. Thelevels of lead in umbilical cord blood from mothers with familyincomes >SR7500 were lower than those reported from motherswith an irregular income (p = 0.026). The analyses also indicatedthat mothers living in the Western Region of Al-Kharj had signif-
and maternal exposure to heavy metals (lead, cadmium and mercury)oi.org/10.1016/j.ijheh.2013.04.009
icantly lower levels of blood lead than those living in the Central(p = 0.02) or Northern Regions (p < 0.001). Placental lead levels weresignificantly higher in mothers living in areas of heavy traffic com-pared to those living in residential areas (p < 0.001) but not to those
Please cite this article in press as: Al-Saleh, I., et al., Birth outcome measures and maternal exposure to heavy metals (lead, cadmium and mercury)in Saudi Arabian population. Int. J. Hyg. Environ. Health (2013), http://dx.doi.org/10.1016/j.ijheh.2013.04.009
ARTICLE IN PRESSG Model
IJHEH-12693; No. of Pages 14
4 I. Al-Saleh et al. / International Journal of Hygiene and Environmental Health xxx (2013) xxx– xxx
Table 1General characteristics of the studied population.
Continuous variables N Mean ± SD Median Min–Max
Maternal age (yr) 1574 28.47 ± 6.029 28 16–50Maternal height (cm) 1576 158.97 ± 5.787 159 132.0–189.0First maternal trimester body weight (kg) 1291 68.24 ± 14.636 67 31.50–160.0Second maternal trimester body weight (kg) 1406 71.820 ± 14.544 70 39.0–162.0Third maternal trimester body weight (kg) 1573 73.705 ± 14.677 72 36.0–168.0Total maternal weight gain (kg/week) for the
second and third trimester1283 0.461 ± 0.427 0.423 −1.62–5.37
Table 3The bivariate relationships between exposure to heavy metals (lead, cadmium and mercury) in cord and maternal blood as well as placental tissues of mothers and a number of confounding variables (demographic, socioeconomicand lifestyle).
a Result was tested by Pearson’s correlation analysis.b Result was tested by ANOVA.c Result was tested by Student-t test. Values between parentheses are the level of significance (p).
iving in other areas such as farms or deserts (p = 0.908). Student’s-tests showed that the blood lead levels were higher in mothers of
ale than of female newborns.Cadmium. The mean cadmium levels in maternal and umbil-
cal cord blood were 0.986 ± 0.313 �g/l (0.233–3.157 �g/l) and.780 ± 0.623 �g/l (0.245–15.325 �g/l), respectively. Cadmium wasetected in all samples, with approximately 94.8% and 97.9% of theamples of umbilical cord and maternal blood respectively, abovehe MDL of cadmium (0.42 �g/l). Even though cadmium levelsere significantly higher in samples of maternal than in umbil-
cal cord blood (t = −21.42, p < 0.001), no relationship was foundetween them (r = 0.042, p = 0.1). The overall mean + SD for placen-al cadmium levels was 0.045 ± 0.116 �g/g dry wt., with a rangef 0–4.363 �g/g dry wt. Cadmium was detected in 99% of placen-al samples. About 70.2% (N = 1107) of all cadmium measurementsere above the MDL of 0.025 �g/g dry wt. Cadmium in umbili-
al cord blood and placental tissues was associated with maternalge and parity (Table 3). Cadmium levels in maternal blood andlacental tissues were positively correlated with the duration of
iving in the district. Mothers living in areas of heavy traffic hadigher blood cadmium levels compare to those living in other areasuch as residential areas, farms or deserts (p = 0.026). Mothers whoefused to report their total family income had the highest bloodadmium levels, which were significantly different than the lev-ls in mothers from other classes: <SR5000 (p = 0.001), >SR7500p = 0.023) or irregular (p < 0.001) but not from mothers with a totalamily income of SR5001-7500 (p = 0.373). Even though placentaladmium levels were highest in mothers with a total family incomeSR7500, they were not significantly different from the levels inothers from other classes (SR < 5000, SR5001–7500, irregular or
efused). Mothers who refused to report family income, though,ad higher placental cadmium levels than those with a total family
ncome of SR5001–7500 (p = 0.036).Mercury. The maternal mercury blood levels ranged from 0 to
06.41 �g/l, with a mean of 3.005 ± 6.319 �g/l. The correspond-ng levels in their newborns ranged from 0 to 26.532 �g/l with a
ean of 3.354 ± 2.673 �g/l. In 96 newborns (6.1%) and 196 moth-rs (12.5%), mercury levels were below the MDL of 0.25 �g/l.nlike lead and cadmium, mercury was significantly higher inmbilical cord blood than in maternal blood, with a t-value of.036 (p < 0.001). These two parameters were significantly corre-
ated (r = 0.202, p < 0.001). The average placental mercury level was.064 ± 0.403 �g/g dry wt. (range, 0–13.003 �g/g dry wt). Approxi-ately 93% of tested placental tissues contained mercury. Placentalercury levels in 47.8% (N = 750) of the women were above the MDL
f 0.033 �g/g dry wt. Mercury in cord blood was correlated withoth the second and third maternal trimester BMIs and with urinaryotinine. On the other hand, mercury in maternal blood and placen-al tissues was associated with the third maternal trimester BMI.
others living in the Central Region of Al-Kharj had the highestevels of mercury in their blood, but these levels were significantlyifferent only from those in mothers living in the Northern Regionp = 0.041). The levels of mercury in placental tissues were highern mothers living in areas of heavy traffic compared to those livingn residential areas (p = 0.025). The levels of mercury in umbilicalord and maternal blood were significantly higher in mothers with
college diploma and/or university degree than in those who wereither illiterate or attended literacy classes with p-values of 0.032nd 0.04 respectively. Results are shown in Table 3.
isk factors associated with the measures of birth outcome
Please cite this article in press as: Al-Saleh, I., et al., Birth outcome measuresin Saudi Arabian population. Int. J. Hyg. Environ. Health (2013), http://dx.d
Uni- and multivariate logistic regression analyses determinedhe crude (unadjusted) and adjusted ORs for the associationsetween the various measures of birth outcome and each metal
n umbilical cord blood, maternal blood and placental tissues.
PRESSnd Environmental Health xxx (2013) xxx– xxx
The observed trends in the prevalence of participants with themeasures of birth outcome below the 10th percentile were:placental weight (17.3%) > Apgar 5-minute score (17%) > Apgar1-min score (14.4%) > crown-heel length (13.9%) > placenta thick-ness (12.6%) > birth height (11.2%) > cord length (10.4%) > birthweight (10.3% > head circumference (10.2%) > ponderal index(10%) > SGA (10%) > cephalization index (9.9%).
Univaiate logistic regression analysis. As shown in Table 4, cad-mium levels in umbilical cord blood were significantly higher innewborns with <10th percentile head circumferences (p = 0.005),while placental mercury levels were significantly higher in new-borns with ≥10th percentile head circumferences (p = 0.047). Thecadmium levels in umbilical cord and maternal blood were sig-nificantly higher in newborns with <10th percentile crown-heellengths with p-values of 0.001 and 0.027, respectively. Newbornswith <10th percentile Apgar 5-minute scores had significantlyhigher cord blood cadmium (p = 0.011). The same pattern wasseen with mercury levels in maternal blood (p = 0.044). In contrast,newborns with <10th percentile Apgar 5-minute scores had signifi-cantly lower placental mercury levels (p = 0.018) compared to thosewith ≥10th percentile scores. Newborns with <10th percentilebirth weights had higher cord blood cadmium levels (p = 0.002).Maternal blood mercury levels were significantly higher in moth-ers with <10th percentile placental weights (p = 0.024). The levels ofmaternal blood lead, maternal blood cadmium and placental cad-mium were significantly higher in mothers with <10th percentileplacental thicknesses with p-values of 0.022, 0.029 and 0.024,respectively. Mercury levels in umbilical cord blood and cadmiumlevels in placental tissues were respectively lower in mothers with<10th percentile placenta thicknesses (p = 0.029) and in newbornswith <10th percentile cord lengths (p = 0.005). Lead levels in umbil-ical cord blood were significantly higher in newborns with a ≥10thpercentile ponderal index (p = 0.022). The risk of SGA births (<10thpercentile) was significantly increased with higher cadmium levelsin umbilical cord blood (p = 0.009).
Birth anthropometric measures below <10th percentile sig-nificantly increased with maternal age such as head cir-cumference (OR = 0.949, 95% CI: 0.922–0.977, p < 0.001), birthweight (OR = 0.966, 95% CI: 0.939–0.994, p = 0.017), birth height(OR = 0.962, 95% CI: 0.936–0.989, p = 0.006), and placental weight(OR = 0.969, 95% CI: 0.948–0.991, p = 0.006). On the other hand,older mothers had a higher risk of giving birth to newbornswith a <10th percentile cephalization index (OR = 1.071, 95%CI: 1.043–1.10, p < 0.001). Parity was only associated with therisk of <10th percentile cephalization index (OR = 1.129, 95% CI:1.057–1.207, p < 0.001). Maternal BMIs during the third trimesterwere significantly lower for newborns with <10th percentiles ofhead circumference (OR = 0.066, 95% CI: 0.025–0.174, p < 0.001),crown-heel length (OR = 0.303, 95% CI: 0.135–0.683, p = 0.004),birth weight (OR = 0.057, 95% CI: 0.022–0.15, p < 0.001), birthheight (OR = 0.063, 95% CI: 0.025–0.16, p < 0.001), placental weight(OR = 0.085, 95% CI: 0.039–0.183, p < 0.001), placental thickness(OR = 0.327, 95% CI: 0.141–0.758, p = 0.009), cord length (OR = 0.197,95% CI: 0.078–0.499, p = 0.001), ponderal index (OR = 0.302, 95%CI: 0.119–0.767, p = 0.012) and SGA (OR = 0.084, 95% CI: 0.032–0.22,p < 0.001). In contrast, maternal BMIs during the third trimesterwere significantly higher for newborns with a <10th percentilecephalization index (OR = 24.024, 95% CI: 9.481–60.872, p < 0.001).Male newborns had an increased risk of <10th percentile cephal-ization index (�2 = 8.392, p = 0.004) and SGA birth (�2 = 7.925,p = 0.007), while female newborns had a higher risk of <10th per-centile head circumferences (�2 = 6.627, p = 0.01). Newborns in
and maternal exposure to heavy metals (lead, cadmium and mercury)oi.org/10.1016/j.ijheh.2013.04.009
the Western Region had a higher risk of <10th percentile headcircumferences (�2 = 19.46, p = 0.001), birth weights (�2 = 13.559,p = 0.009), birth heights (�2 = 10.572, p = 0.032) and placental thick-nesses (�2 = 15.149, p = 0.004). Mothers who were either illiterate
Please cite this article in press as: Al-Saleh, I., et al., Birth outcome measures and maternal exposure to heavy metals (lead, cadmium and mercury)in Saudi Arabian population. Int. J. Hyg. Environ. Health (2013), http://dx.doi.org/10.1016/j.ijheh.2013.04.009
ARTICLE IN PRESSG Model
IJHEH-12693; No. of Pages 14
I. Al-Saleh et al. / International Journal of Hygiene and Environmental Health xxx (2013) xxx– xxx 7
Table 4Arthmatic mean, n, unadjusted ORs, 95% CI, and p-value for the association between birth outcome measures falling < or ≥10th percentile and heavy metals in cord blood,maternal blood and placental tissues.
Variables ORs (95% CI), p
Head circumference Crown-heel length Apgar 1-minute scores
r attending literacy class had a higher risk of newborns with10th percentile cephailzation index (�2 = 11.157, p = 0.011). Fam-lies with low income < SR5000 had a higher risk of SGA births�2 = 10.006, p = 0.04).
Adjusted multivariate logistic models. We further evaluated theignificant association between metals and various measures ofirth outcome in the bivariate logistic after adjusting for con-ounding variables. As shown in Table 5, – model I, only placental
ercury levels remained significantly correlated with the risk ofewborns with ≥10th percentile head circumferences (p = 0.017).his relationship remained significant after adjusting for gesta-ional age (p = 0.026) or excluding preterm births (p = 0.021) ashown in Table 5, models II and III. The AUC of placental mer-ury was 0.686 (95% CI, 0.626–0.746) which was larger than 0.5p < 0.001). The AUC did not change after adjusting for gestationalge or excluding preterm births. On the other hand, the associationetween <10th percentile head circumferences and cadmium levels
n umbilical cord blood disappeared (p = 0.230) after adjusting foronfounders.
The risk of newborns with <10th percentile crown-heel lengthsemained significantly associated with cadmium levels in umbil-cal cord blood (p = 0.034) in the adjusted model I (Table 5). Theame relationship was observed with maternal blood cadmiump = 0.007). The AUC was 0.586 (95% CI, 0.533–0.639, p = 0.003)or cadmium levels in umbilical cord blood and 0.589 (95% CI,.531–0.646, p = 0.002) in maternal blood. After controlling for ges-ational age or excluding preterm births, the association betweenhe risk of <10th percentile crown-heel lengths and cadmium levelsn umbilical cord blood disappeared. On the other hand, the asso-iation remained significant with maternal blood cadmium afterdjusting for gestational age (p = 0.01) and after excluding pretermirths (p = 0.042), with no change in the AUC. Results are presented
n Table 5, models II and III.The increase in the risk of <10th percentile Apgar 5-minute
cores with cadmium levels in umbilical cord blood remained sig-ificant in the adjusted model (p = 0.004). The AUC of the modelas 0.637 (95% CI, 0.581–0.693, p < 0.001). The risk of ≥10th per-
entile Apgar 5-minute scores remained significant with placental
Please cite this article in press as: Al-Saleh, I., et al., Birth outcome measuresin Saudi Arabian population. Int. J. Hyg. Environ. Health (2013), http://dx.d
ercury levels (p = 0.01). The model’s AUC was 0.573 (95% CI,.519–0.627, p = 0.012). After adjusting for gestational age (Table 5,odel II), the relationship remained significant with cadmium lev-
ls in umbilical cord blood (p = 0.025) and with placental mercury
(p = 0.015). After excluding preterm births, however, a marginallysignificant relationship was seen between <10th percentile Apgar5-minute scores and cadmium levels in umbilical cord (p = 0.069).A similar pattern was also seen between placental mercury andthe risk of ≥10th percentile Apgar 5-minute scores after excludingpreterm births (p = 0.076).
The risk of <10th percentile birth weights remained signifi-cantly associated with cadmium levels in umbilical cord (p = 0.015)after adjusting for confounders. The model’s AUC was 0.652 (95%CI, 0.58–0.723, p < 0.001). The trend, however, was weaker and nolonger significant after adjusting for gestational age (p = 0.188) orexcluding preterm births (p = 0.124).
The association between <10th percentile placental weights andmercury levels in maternal blood remained marginally significant(p = 0.059), with an AUC of 0.639 (95% CI: 0.598–0.681 p < 0.001).Additional adjustment for gestational age or excluding pretermbirths changed the estimates little, with p-values of 0.057 and0.053, respectively, as displayed in Table 5, models II and III.
The risk of placental thickness below the 10th percentileremained significant with maternal blood lead (p = 0.011), mater-nal blood cadmium (p = 0.022) and placental cadmium (p = 0.032). Aborderline significant relationship was observed between mercuryin umbilical cord blood and ≥10th percentile placental thicknesses(p = 0.05). The model’s AUCs were 0.625 (95% CI: 0.575–0.675) formaternal blood lead, 0.618 (95% CI: 0.573–0.662) for maternalblood cadmium, 0.618, (95% CI: 0.57–0.667) for placental cadmiumand 0.607 (95% CI: 0.560–0.654) for mercury in umbilical cordblood, with p < 0.001 for all. The relationships remained unchangedafter adjusting for gestational age or excluding preterm births, asshown in Table 5, models II and III.
The risk of cord lengths above the 10th percentile remainedsignificant with placental cadmium (p = 0.012), with an AUC of0.59 (95% CI: 0.532–0.648, p = 0.004). The risk of ≥10th per-centile cord lengths remained unchanged after controlling forgestational age (p = 0.025) but not after excluding preterm births(p = 0.532).
The relationship between umbilical cord blood lead and the riskof a ≥10th percentile ponderal index disappeared (p = 0.216) after
and maternal exposure to heavy metals (lead, cadmium and mercury)oi.org/10.1016/j.ijheh.2013.04.009
controlling for confounders, as shown in Table 5 – Model I.After adjusting for confounding variables, cadmium in umbili-
cal cord blood remained significantly associated with SGA births(p = 0.049). The AUC was 0.659 (95% CI: 0.6–0.718, p < 0.001). A
Table 5Adjusted ORs (and corresponding 95% CI, p-value) for predicting the association between birth outcome measures at the 10th percentile and heavy metals in cord blood, maternal blood and placental tissues after controlling forcofounders (Model I); plus adjusting for gestational age (Model II); and excluding preterm births (Model III).
Lead (�g/dl) Cadmium (�g/l) Mercury (�g/l) Placenta (�g/g dry wt.)
Models are adjusted for the following confounders:a Maternal age, parity, mother’s third trimester BMI, urinary cotinine, geographical distribution of current dwelling, newborm mother’s highest education, total family income.b Maternal age, parity, mother’s third trimester BMI, urinary cotinine, mother’s highest education, total family income.c Maternal age, parity, mother’s third trimester BMI, geographical distribution of current dwelling, newborn’s gender.d Maternal age, parity, mother’s third trimester BMI, illness during pregnancy, newborn’s gender.e Maternal age, parity.f Maternal age, parity, mother’s third trimester BMI, geographical distribution of current dwelling.g Maternal age, parity, mother’s third trimester BMI, total family income, newborn’s gender.h Maternal third trimester BMI, illness during pregnancy, duration of living in the district, area of current dwelling, total family income.i Maternal third trimester BMI, urinary cotinine, mother’s highest education, geographical distribution of current dwelling.k Maternal third trimester BMI, mother’s highest education, geographical distribution of current dwelling.l Maternal third trimester BMI, maternal age, illness during pregnancy, mother’s highest education, geographical distribution of current dwelling.
m Maternal third trimester BMI, maternal age, parity, duration of living in the district, total family income.n Maternal third trimester BMI, maternal age, parity, duration of living in the district, total family income.o Maternal age, maternal third trimester BMI, newborn’s gender, location of current dwelling.p Maternal third trimester BMI, location of current dwelling.
imilar pattern was observed after adjusting for gestational agep = 0.04) but not after excluding preterm births (p = 0.158).
iscussion
The present study expands upon our previous paper thatevealed Saudi mothers and their newborns were substantiallyxposed to toxic metals that might jeopardize the health of bothAl-Saleh et al., 2011) by examining their influence on severalirth anthropometric measures. We will start first briefly recap-ing on the prevalence of heavy metals in our population. Both
ead and cadmium were detected in all samples of umbilical cordnd maternal blood and in 96% and 99.2%, respectively of placentalamples. Mercury was detected in 96%, 93.5% and 92.6% of sam-les of umbilical cord blood, maternal blood and placental tissue,espectively. Only few participants had metal levels outside thecceptable threshold limits. Only 6.4% and 7.1% of cord and mater-al blood samples, respectively, had lead levels above the newlyevised threshold limit of the United States Centers for Diseaseontrol and Prevention (CDC) of 5 �g/dl (CDC, 2012).
The American Occupational Safety and Health AdministrationOSHA, 2003) established an occupational threshold limit for cad-
ium in blood of 5 �g/l. Five newborns in this study had cadmiumlood levels above this limit. This occupational threshold limit,owever, is not applicable to the general population. Mijal andolzman (2010) reported a geometric mean of 0.29 �g/l (95%I: 0.28–0.3), and the 75th percentile blood concentration was.35 �g/l (0.33–0.36), based on a survey of 1594 nonsmokingomen of childbearing age conducted between 1999 and 2006. Our
tudy found a substantially higher geometric mean of 0.935 �g/l95% CI: 0.971–1.002), and 25% of the mothers had blood cadmiumevels above the 75th percentile of 1.207 �g/l. Such large discrep-ncies between our results and the US levels indicate a potentialource of exposure to cadmium in Saudi Arabia that requires inves-igation, particularly because none of our participants smoked.imilarly, the overall geometric mean for umbilical cord bloodadmium was 0.716 �g/l (95% CI: 0.971–1.002), and 24.8% of theeasurements were above the 75th percentile of 0.853 �g/l. These
evels, which are higher than those reported by other researchers,an be a risk factor for subsequent developmental impairment innfants. Tian et al. (2009) found that the concentration of cadmiumn umbilical cord blood ≥0.6 �g/l influenced fetal growth and theevelopment of IQ.
Mercury levels in 11.2% and 13% of samples of maternal andmbilical cord blood, respectively, were ≥5.8 �g/l, which is thePA reference dose (US EPA, 2012). Some believe, however, thathis concentration should be lowered to 3.5 �g/l during pregnancy,ased on the fact that mercury is usually 70% higher in umbili-al cord blood than in maternal blood. Furthermore, an associationetween delays in neurological development and prenatal mercuryxposures below this lower threshold has been reported (Mahaffey,005). In our study, 39.9% newborns and 25.1% of mothers hadlood mercury levels ≥3.5 �g/l. This maternal percentage is quiteigh compared to the 2% of pregnant with these levels found byiranda et al. (2011). Furthermore, the 75th percentile of mercury
n umbilical cord and maternal blood in this study were 4.6 �g/lnd 3.507 �g/l, respectively. These values are much higher than the5th percentile of blood mercury reported by the CDC (2004) in USomen of childbearing-age (1.81 �g/l) and in children (0.6 �g/l).
Lead, cadmium and mercury were detected in 96%, 99% and3%, respectively, of placental tissue samples, confirming their
Please cite this article in press as: Al-Saleh, I., et al., Birth outcome measuresin Saudi Arabian population. Int. J. Hyg. Environ. Health (2013), http://dx.d
ransplacental passage. Because no threshold levels have yet beenstablished for these metals in the placenta, the 75th percentiles ofead (0.63 �g/g dry wt.), cadmium (0.048 �g/g dry wt.) and mercury0.06 �g/g dry wt) were used for comparison with other studies.
PRESSnd Environmental Health xxx (2013) xxx– xxx 11
Of the tested placental tissues, 26.1%, 25.2% and 24.8% had ≥ their75th percentile levels of lead, cadmium and mercury, respectively.These values are generally higher than those reported in other stud-ies (Amaya et al., 2013) and are likely to affect fetal growth anddevelopment.
Among the three tested heavy metals, cadmium in the variouscompartments appeared to affect metal measures of birth outcome.Newborns with Apgar 5-minute scores below the 10th percentile(0–8) had higher levels of cadmium in umbilical cord blood. It isnot clear whether exposure to cadmium in utero could be deemedas an influential factor on these scores. Low Apgar 5-minute scoresmight contribute to poor prognoses. Ehrenstein et al. (2009) foundthat Apgar 5-minute scores under 7 were consistently associatedwith the prevalence of neurologic disability and with low cog-nitive function in early adulthood. A recent study indicated thatlow Apgar 5-minute scores lower than 5 were associated witha higher risk of childhood cancer (Li et al., 2012). Among new-borns reported in our study, 24 (1.5%) and 91 (5.8%) had Apgar5-minute scores ≤5 and 7, respectively. Even though the accuracy ofthe Apgar 5-minute scores has been questioned (O’Donnell et al.,2006), it remains a widely used tool for assessing the vitality ofnewborns at birth around the world (Casey et al., 2001). Despiteadjusting for gestational age or excluding preterm births, our studysuggests that a true association between cadmium exposure inutero and lower Apgar 5-minute scores might lead to a series ofadverse developments in childhood. Only one study has reported apotential association between cadmium in cord blood (0.66 �g/l,range: 0.2–1.5 �g/l) and Apgar 5-minute scores (Mokhtar et al.,2002). The levels of umbilical cord blood cadmium in our study(0.78 �g/l, range: 0.245–15.325 �g/l) were higher than those ofMokhtar et al.‘s study. The relevance of our findings requires fur-ther evaluation, because low Apgar 5-minute scores have importantimplications for adverse consequences in the short and long term.
We have also shown that umbilical cord blood cadmium lev-els were associated with an increased risk of <10th percentile birthweights. This association disappeared, though, when including onlypreterm births or adjusting for gestational age. Similar findingswere reported by Lin et al. (2011) who found that prenatal cad-mium exposure was associated with lower birth weights. Theirmedian value of umbilical cord blood cadmium (0.31 �g/l) waslower than ours (0.704 �g/l). These authors also observed smallerhead circumferences, which may have a detrimental effect on childgrowth in the first three years of life. An association between umbil-ical cord blood cadmium and the risk of <10th percentile headcircumferences was seen in our study (p = 0.005) but the associa-tion disappeared when the analysis was adjusted for confounders.The developmental effects of cadmium exposure during early liferequires further exploration. A recent study by Kim et al. (2012)suggested that a dose-dependent interaction between prenatalexposures to lead and cadmium might affect cognitive develop-ment at the age of six months.
We found an association between higher levels of cadmiumin maternal blood and lower crown-heel lengths. This associa-tion remained significant after controlling for gestational durationor excluding preterm births. Previous studies have found associa-tions between lower crown-heel lengths and exposure to smoking(Lindley et al., 2000) and organoclorinated pesticides (Dewan et al.,2013). The mean cadmium level in maternal blood in our study was0.986 �g/l, which was higher than the average levels observed insome other countries, as previously reported (Al-Saleh et al., 2011).Although all the women were nonsmokers, 48.6% had cadmiumlevels >1 �g/l higher than the international standards for male and
and maternal exposure to heavy metals (lead, cadmium and mercury)oi.org/10.1016/j.ijheh.2013.04.009
female non-smokers set by the Commission of the German FederalEnvironment Agency (Wilhelm et al., 2004). Cadmium in maternalblood and placental tissues was associated with reduction in theplacenta thickness, which remained significant even after adjusting
2 I. Al-Saleh et al. / International Journal of Hyg
or gestational age or excluding preterm births, suggesting possiblendependent and opposing effects of cadmium on these measures.t has been shown that placenta thickness <2.5 cm has been associ-ted with retarded intrauterine fetal growth (Nagi, 2011). Placentalccumulation of cadmium can have an indirect effect on fetalrowth by altering placental function (Menai et al., 2012). A recentxperimental study observed a decreased incidence of pregnancy,elayed maternal weight gain, altered placental weight, decreasedetal length, and delayed neonatal growth (Blum et al., 2012).
Newborns with ≥10th percentile cord lengths had significantlyigher placental cadmium levels. This relationship disappearednly after excluding preterm births. We cannot account for thisncrease. Cord length is usually an index of fetal activity (Ente andenzer, 1991), and abnormally short or long cords are usually com-ared to an average cord length of 50–60 cm The mean length of thembilical cord in our study was 51.97 ± 9.3 cm (13–104 cm) and 56ewborns (3.6%) had umbilical cords longer than 70 cm. Baergent al. (2001) observed an increased risk of brain-imaging abnor-alities and/or abnormal neurological follow-ups with higher cord
engths. We cannot rule out the possibility, however, that our find-ngs may be biased by unmeasured confounders or by residualonfounding.
The above results indicate a partial retention of cadmium inhe placenta, which confirms other findings (Esteban-Vasallo et al.,012). Placental concentrations of metallothionein increased withhe accumulation of cadmium, but at a slower rate that does notompletely prevent its passage (Nakamura et al., 2012). The reten-ion of cadmium in the placenta or a transfer to the fetus canoth have adverse effects on fetal development and/or pregnancyutcome. For example, a recent experimental study by Ji et al.2011) found that exposure to cadmium during late pregnancympaired testicular steroidogenesis in male offspring. Maternallood, though, only had traces of cadmium compared to theigher levels in the placenta, suggesting an indirect effect on fetalestes. Cadmium can disrupt endocrine function, causing variouseproductive problems (Takiguchi and Yoshihara, 2006). Cadmiumay interfere with the production of placental progesterone that
an impair steroidogenesis and consequently affect fetal growthnd development (Stasenko et al., 2010). Few studies, though,ave reported inverse relationships between cadmium exposure
n non-smoking mothers and various birth outcomes such as headircumference (Kippler et al., 2012), birth height (Zhang et al., 2004)nd birth weight (Salpietro et al., 2002; Kippler et al., 2012).
Both our crude and adjusted logistic regression models providedvidence of increasing odds of SGA birth associated with higher cad-ium levels in umbilical cord blood (p = 0.049) but not with higher
admium in maternal blood. This association was marginally sig-ificant after adjusting for confounders and gestational age but notfter excluding preterm births with p-values of 0.084, 0.068 and.102 respectively (data not shown). Controlling for gestationalge gave the same results, but the effect disappeared when wexcluded preterm births. Murphy et al. (2010) observed that theost sensitive window for human development affected by antie-
trogenic polychlorinated biphenyls occurred before conceptionnd during early pregnancy. The authors suggested that a reliancen maternal blood samples could underscore fetal exposure duringmbryogenesis. Furthermore, with only 10% SGA births, low powerikely decreased the precision of our estimates. To our knowledge,uch associations between cadmium and SGA risk have not beeneported previously.
Most of the relationships between prenatal cadmium exposurend several birth outcomes were independent of gestational age or
Please cite this article in press as: Al-Saleh, I., et al., Birth outcome measuresin Saudi Arabian population. Int. J. Hyg. Environ. Health (2013), http://dx.d
reterm birth. Also, some birth measures appeared to be more sen-itive than other possibly indicating different effects of cadmiumt different developmental periods. The relevance of our findings,owever, requires evaluation.
PRESSnd Environmental Health xxx (2013) xxx– xxx
We found no associations between lead in maternal blood,umbilical cord blood and placental tissues with birth outcomes,other than with placental thickness. Similarly, no associationshave been reported in other studies (Rahman and Hakeem, 2003;Rahman et al., 2012). In contrast, relationships between lead lev-els in maternal or umbilical cord blood and various fetal growthparameters have been found (Bellinger, 2005; Gundacker et al.,2010). The timing and dose of lead exposure can play an importantrole in fetal growth (Cantonwine et al., 2010). Jelliffe-Pawlowskiet al. (2006) found that lead levels in maternal blood ≥10 �g/dl to beassociated with significant decreases in total duration of gestationand an increased risk of preterm and SGA birth. The lower mean lev-els of blood lead in our study population may explain this apparentlack of an inverse association of lead levels in umbilical cord and/ormaternal blood with fetal growth. Likewise to maternal cadmium,the multiple logistic regression analysis showed that an increasein maternal blood lead levels caused reduced placental thickness.Controlling for gestational age or excluding preterm births did notaffect this relationship, which again suggests that lead exposureis likely to influence birth outcome, but via an indirect route ofa altering placental thickness. Abnormalities in placental growthmay precede fetal complications (Meèëjus, 2005). Mital et al. (2002)suggested that placental thickness can be a useful indicator of thegestational age of the fetus.
In contrast, mercury in cord blood was marginally higher inmothers with ≥10th percentile placenta thicknesses. This findingalso confirmed that the presence of independent effects of ges-tational age or preterm births. Few studies have linked placentathickness >4 cm with various poor birth outcome (Pinette et al.,1998; Lee et al., 2012). We do not know if these associations are dueto chance or to unadjusted confounders such as the amniotic fluidmeasurements; excess fluid may have a compressive effect on pla-cental thickness and may distort the results (Lee et al., 2012). Afteran adjustment for confounding variables, maternal blood mercurywas marginally higher in mothers with <10th percentile placentaweights, which persisted after controlling for gestational age orexcluding preterm births. Husslein et al. (2012) found a high risk ofemergency delivery (C.S. or vacuum-assisted delivery) when pla-cental weight was higher than the 10th percentile. An associationwith mercury exposure has not been reported. Ma et al. (2006),however, observed a reduction in placenta weight in rats exposedto lead during different gestational periods, due to its effect onthe trophoblast that led to interference in nutrition and oxygentransfer.
Even though 48% of placentas had levels of mercury above theMDL, mercury levels were significantly higher in newborns with≥10th percentile head circumferences. This association remainedsimilar even after adjusting for gestational age or excludingpreterm births. Normal head circumference is usually in the rangeof 33–35 cm (25th to 75th percentile). Head circumferences <33and >35 cm were found in 14.6% and 23.9% of the newbornsrespectively. Head circumferences above the 75th percentile wereassociated more with impaired adaptive behaviors and less withimpairment of IQ and motor and verbal-language development(Sacco et al., 2007). Some researchers have found evidence thatautistic infants at the age of four months had higher head circum-ferences, weights and lengths than normal (Torrey et al., 2004) andhave related these observations to abnormalities in metabolism,growth factors, hormone levels or general development. McKean-Cowdin (2006) suggested that higher in head circumference at birthcould be a risk factor for brain cancer during childhood and rec-ommended further study. On the other hand, another study found
and maternal exposure to heavy metals (lead, cadmium and mercury)oi.org/10.1016/j.ijheh.2013.04.009
no such association (Hobbs et al., 2007). Apgar 5-minute scoresabove the 10th percentile in newborns were also unexpectedlyassociated with placental mercury, which persisted after adjus-ting for gestational age. The possibility of residual confounding
hat was either not measured or was difficult to measure is againlausible in the case of head circumferences and Apgar 5-minutecores.
In the present study, the AUC for the prediction of the aboveodels was larger than 0.5 and smaller than 0.7, which represents aodest predictive capability for the measures of birth outcome. We
elieve that prediction could be improved by including additionalisk factors. Lindell et al. (2013) recently found that predictionsor large-gestational-age neonates at term using fetal weights esti-
ated during routine third-trimester ultrasounds improved theUC after including maternal characteristics.
Our study has many strengths: (1) the large sample sizetrengthened the analysis of the effects of heavy metal exposure onealth after adjusting for several confounders, (2) the populationas homogeneous for socioeconomic and prenatal-care profiles,
educing the potential for uncontrolled confounding, and (3) heavyetals were measured in umbilical cord and maternal blood and
articularly in the placenta, providing a more accurate picture ofhe redistribution of metals between the mother and the fetus. Thistudy, however, does have limitations: (1) the heavy metals wereeasured only at a single time point (at delivery), and extent tohich these single measurements can reflect exposures at criti-
al times during pregnancy is still debated (Murphy et al., 2010),2) the low numbers of newborns of low birth weight and SGAisk reduced the precision of our statistical estimates due to lowower, mainly related to the cross-sectional design of the study thatnrolled any eligible mother without considering the characteris-ics of newborns, (3) genetic information that could be associatedith birth outcome was not included (Rossner et al., 2011), and
4) the measures of birth outcome were recorded according toospital protocol, and our findings may have been subject to non-ifferential measurement error.
onclusion
Despite its partial passage through the placenta, cadmium hadhe most prominent effect on several measures of birth outcome.rown-heel lengths, Apgar 5-minute scores, birth weights and SGAselow the 10th percentile were influenced by cadmium levels in thembilical cord. On the other hand, crown-heel length and placen-al thickness were affected by cadmium levels in maternal blood.s placental cadmium levels increased, placental thickness signif-
cantly decreased and an cord length significantly increased. Onlyead levels in maternal blood influenced placental thickness. Thebsence of some effects in our study may be due to either lowaternal lead exposure or exposure measurements taken out-
ide the window of fetal vulnerability. Even though mercury wasfficiently transferred transplacentally to the fetus, mercury inoth umbilical cord and maternal blood was marginally associ-ted with placental thickness and placental weight, respectively.onversely, placental mercury significantly influenced head cir-umference, Apgar 5-minute scores and cord length. The predictionf these measures of birth outcome by the AUCs of metals wereenerally modest (larger than 0.5 and smaller than 0.7). The studyhowed that the retention of metals in the placenta or the trans-er of metals to the fetus can both have adverse effects on fetalevelopment and/or pregnancy outcome. The independence of ges-ational age or preterm births on the observed effect of metals onome measures of birth outcome, suggested detrimental effects ofxposure on fetal development. The magnitude of the estimatedffects might not necessarily be of clinical significance for infants
Please cite this article in press as: Al-Saleh, I., et al., Birth outcome measuresin Saudi Arabian population. Int. J. Hyg. Environ. Health (2013), http://dx.d
ut may have a considerable public-health relevance given the highrevalence of exposure to heavy metals. Further research should beonducted to confirm these findings and to evaluate their long-termisks, if any. In addition, the sources of exposure to these metals
PRESSnd Environmental Health xxx (2013) xxx– xxx 13
remain to be determined in this population. Finally and impotantly,the findings of this study are likely to pertain only to nonsmokingwomen living in the Al-Kharj area that has history of occupationalexposure. The results thus cannot be generalized to women livingin other areas but may reflect typical exposure in rural settings.The intensity of exposure is obviously expected to be far greaterin urban settings due mainly to the emissions of motor vehiclesand/or industrial activities, which may have larger adverse effectson pregnancy outcome.
Conflict of interest
The authors do declare that there is no conflict of interest.
Acknowledgments
The investigators would like to thank King Abdulaziz City forScience and Technology for funding this study ARP-23-7. We wouldlike to thank all the women who participated in this study and thestaff of King Khalid Hospital in Al-Kharj.
References
Al-Saleh, I., Shinwari, N., Mashhour, A., Mohamed Gel, D., Rabah, A., 2011. Heavymetals (lead, cadmium and mercury) in maternal, cord blood and placenta ofhealthy women. Int. J. Hyg. Environ. Health 214, 79–101.
Amaya, E., Gil, F., Freire, C., Olmedo, P., Fernández-Rodríguez, M., Fernández, M.F.,Olea, N., 2013. Placental concentrations of heavy metals in a mother–childcohort. Environ. Res. 120, 63–70.
Baergen, R.N., Malicki, D., Behling, C., Benirschke, K., 2001. Morbidity, mortality,and placental pathology in excessively long umbilical cords: retrospective study.Pediatr. Dev. Pathol. 4, 144–153.
Bellinger, D.C., 2005. Teratogen update: lead and pregnancy. Birth Defects Res. A:Clin. Mol. Teratol. 73, 409–420.
Cantonwine, D., Hu, H., Sanchez, B.N., Lamadrid-Figueroa, H., Smith, D., Ettinger, A.S.,Mercado-Garcia, A., Hernandez-Avila, M., Wright, R.O., Tellez-Rojo, M.M., 2010.Critical windows of fetal lead exposure: adverse impacts on length of gestationand risk of premature delivery. J. Occup. Environ. Med. 52, 1106–1111.
Caserta, D., Mantovani, A., Ciardo, F., Fazi, A., Baldi, M., Sessa, M.T., la Rocca, C., Ronchi,A., Moscarini, M., Minoia, C., 2011. Heavy metals in human amniotic fluid: a pilotstudy. Prenat. Diagn. 31, 792–796.
Casey, B.M., McIntire, D.D., Leveno, K.J., 2001. The continuing value of the Apgarscore for the assessment of newborn infants. N. Engl. J. Med. 344, 467–471.
Centers for Disease Control and Prevention (CDC), 2004. Blood mercury levelsin young children and childbearing-aged women—United States, 1999-2002.MMWR Morb. Mortal. Wkly Rep. 53, 1018–1020.
CDC Centers for Disease Control and Prevention. Low level lead exposure harmschildren: a renewed call for primary prevention. Atlanta, GA: US Departmentof Health and Human Services, CDC; 2012. Available at http://www.cdc.gov/nceh/lead/acclpp/final document 030712.pdf (accessed 09.02.13).
Dewan, P., Jain, V., Gupta, P., Banerjee, B.D., 2013. Organochlorine pesticide residuesin maternal blood, cord blood, placenta, and breastmilk and their relation to birthsize. Chemosphere 90, 1704–1710.
Ehrenstein, V., Pedersen, L., Grijota, M., Nielsen, G.L., Rothman, K.J., Sørensen, H.T.,2009. Association of Apgar score at five minutes with long-term neurologicdisability and cognitive function in a prevalence study of Danish conscripts.BMC Pregnancy Childbirth 9, 14.
Ente, G., Penzer, P.H., 1991. The umbilical cord: normal parameters. J. R. Soc. Health111, 138–140.
US EPA Environmental Protection Agency, 2012. US Environmental Protec-tion Agency. Mercury. Human exposure. Available http://www.epa.gov/mercury/exposure.htm (accessed 03.03.13).
Esteban-Vasallo, M.D., Aragonés, N., Pollan, M., López-Abente, G., Perez-Gomez, B.,2012. Mercury, cadmium and lead levels in human placenta: a systematicreview. Environ. Health Perspect. 120, 1369–1377.
Gundacker, C., Frohlich, S., Graf-Rohrmeister, K., Eibenberger, B., Jessenig, V., Gicic,D., Prinz, S., Wittmann, K.J., Zeisler, H., Vallant, B., Pollak, A., Husslein, P.,2010. Perinatal lead and mercury exposure in Austria. Sci. Total Environ. 408,5744–5749.
Helsel, D.R., 2005. Nondetections and Data Analysis: Statistics for Censored Envi-ronmental Data. John Wiley and Sons, Hoboken, NJ.
and maternal exposure to heavy metals (lead, cadmium and mercury)oi.org/10.1016/j.ijheh.2013.04.009
Hobbs, K., Kennedy, A., Dubray, M., Bigler, E.D., Petersen, P.B., McMahon, W., Lainhart,J.E., 2007. A retrospective fetal ultrasound study of brain size in autism. Biol.Psychiatry 62, 1048–1055.
Husslein, H., Moswitzer, B., Leipold, H., Moertl, M., Worda, C., 2012. Low placentalweight and risk for fetal distress at birth. J. Perinat. Med. 40, 693–695.
4 I. Al-Saleh et al. / International Journal of Hyg
ijima, K., Otake, T., Yoshinaga, J., Ikegami, M., Suzuki, E., Naruse, H., Yamanaka,T., Shibuya, N., Yasumizu, T., Kato, N., 2007. Cadmium, lead, and selenium incord blood and thyroid hormone status of newborns. Biol. Trace Elem. Res. 119,10–18.
yengar, G.V., Rapp, A., 2001. Human placenta as a ‘dual’ biomarker for monitoringfetal and maternal environment with special reference to potentially toxic traceelements, Part 3: toxic trace elements in placenta and placenta as a biomarkerfor these elements. Sci. Total Environ. 280, 221–238.
elliffe-Pawlowski, L.L., Miles, S.Q., Courtney, J.G., Materna, B., Charlton, V., 2006.Effect of magnitude and timing of maternal pregnancy blood lead (Pb) levels onbirth outcomes. J. Perinatol. 26, 154–162.
i, Y.L., Wang, H., Liu, P., Zhao, X.F., Zhang, Y., Wang, Q., Zhang, H., Zhang, C., Duan,Z.H., Meng, C., Xu, D.X., 2011. Effects of maternal cadmium exposure during latepregnant period on testicular steroidogenesis in male offspring. Toxicol. Lett.205, 69–78.
hanjani, N., Sim, M.R., 2006. Maternal contamination with dichlorodiphenyl-trichloroethane and reproductive outcomes in an Australian population.Environ. Res. 101, 373–379.
im, R., Aro, A., Rotnitzky, A., Amarasiriwardena, C., Hu, H.K., 1995. X-ray fluores-cence measurements of bone lead concentration: the analysis of low-level data.Phys. Med. Biol. 40, 1475–1485.
im, Y., Ha, E.H., Park, H., Ha, M., Kim, Y., Hong, Y.C., Kim, E.J., Kim, B.N., 2012. Prena-tal lead and cadmium co-exposure and infant neurodevelopment at 6 monthsof age: The Mothers and Children’s Environmental Health (MOCEH) study. Neu-rotoxicology 35C, 15–22.
ippler, M., Hoque, A.M., Raqib, R., Ohrvik, H., Ekstrom, E.C., Vahter, M., 2010.Accumulation of cadmium in human placenta interacts with the transport ofmicronutrients to the fetus. Toxicol. Lett. 192, 162–168.
ippler, M., Tofail, F., Gardner, R., Rahman, A., Hamadani, J.D., Bottai, M., Vahter,M., 2012. Maternal cadmium exposure during pregnancy and size at birth: aprospective cohort study. Environ. Health Perspect. 120, 284–289.
ee, A.J., Bethune, M., Hiscock, R.J., 2012. Placental thickness in the second trimester:a pilot study to determine the normal range. J. Ultrasound Med. 31, 213–218.
i, J., Cnattingus, S., Gissler, M., Vestergaard, M., Obel, C., Ahrensberg, J., Olsen,J., 2012. The 5-minute Apgar score as a predictor of childhood cancer: apopulation-based cohort study in five million children. BMJ Open. 2 (August(4)), doi:pii:e001095.10.1136/bmjopen-2012-001095.
in, C.M., Doyle, P., Wang, D., Hwang, Y.H., Chen, P.C., 2011. Does prenatal cadmiumexposure affect fetal and child growth? Occup. Environ. Med. 68, 641–646.
indell, G., Marsál, K., Källén, K., 2013. Predicting risk for large-for-gestationalage neonates at term: a population-based Bayesian theorem study. UltrasoundObstet. Gynecol. (March), doi:10.1002/uog.11218 [Epub ahead of print] PubMedPMID: 23505150.
indley, A.A., Becker, S., Gray, R.H., Herman, A.A., 2000. Effect of continuing or stop-ping smoking during pregnancy on infant birth weight, crown-heel length, headcircumference, ponderal index, and brain:body weight ratio. Am. J. Epidemiol.152, 219–225.
a, H.Y., Li, H., Wang, J.C., Liu, X.Q., Xu, F.S., Tan, J.S., 2006. Effects of lead exposureto rat placenta and pubs during different gestation periods. Zhonghua Yu FangXue Za Zhi 40, 101–104.
ahaffey, K.R., 2005. Mercury exposure: medical and public health issues. Trans.Am. Clin. Climatol. Assoc. 116, 127–153.
cKean-Cowdin, R., 2006. Head circumference at birth linked to cancer in child-hood. Lancet Oncol. 7, 4–5.
eèëjus, G., 2005. Influence of placental size and gross abnormalities on intrauterinegrowth retardation in high-risk pregnancies. Acta Med. Lituanica 12, 14–19.
enai, M., Heude, B., Slama, R., Forhan, A., Sahuquillo, J., Charles, M.A., Yazbeck,C., 2012. Association between maternal blood cadmium during pregnancy andbirth weight and the risk of fetal growth restriction: the EDEN mother–childcohort study. Reprod. Toxicol. 34, 622–627.
Please cite this article in press as: Al-Saleh, I., et al., Birth outcome measuresin Saudi Arabian population. Int. J. Hyg. Environ. Health (2013), http://dx.d
ijal, R.S., Holzman, C.B., 2010. Blood cadmium levels in women of childbearingage vary by race/ethnicity. Environ. Res. 110, 505–512.
iranda, M.L., Edwards, S., Maxson, P.J., 2011. Mercury levels in an urban pregnantpopulation in Durham County, North Carolina. Int. J. Environ. Res. Public Health8, 698–712.
PRESSnd Environmental Health xxx (2013) xxx– xxx
Mital, P., Hooja, N., Mehndiratta, K., 2002. Placental thickness: a sonographic param-eter for estimating gestational age of the fetus. Ind. J. Radiol. Imag. 12, 553–554.
Mokhtar, G., Hossny, E., el-Awady, M., Zekry, M., 2002. In utero exposure to cadmiumpollution in Cairo and Giza governorates of Egypt. East Mediterr. Health J. 8,254–260.
Murphy, L.E., Gollenberg, A.L., Buck Louis, G.M., Kostyniak, P.J., Sundaram, R., 2010.Maternal serum preconception polychlorinated biphenyl concentrations and
infant birth weight. Environ. Health Perspect. 118, 297–302.Nagi, A.H., 2011. A monograph-placenta. Examination and pathology. Biomedica
27, 81–99.Nakamura, Y., Ohba, K., Suzuki, K., Ohta, H., 2012. Health effects of low-level cad-
mium intake and the role of metallothionein on cadmium transport from motherrats to fetus. J. Toxicol. Sci. 37, 149–156.
O’Donnell, C.P., Kamlin, C.O., Davis, P.G., Carlin, J.B., Morley, C.J., 2006. Interobservervariability of the 5-min Apgar score. J. Pediatr. 149, 486–489.
OSHA Occupational Safety and Health Administration DoL., 2003. Occupationalsafety and health standards: toxic and hazardous substances: cadmium. CodeFed. Regul. 29, 135–229, CFR 1910.1027.
Park, S.H., Goo, J.M., Jo, C.H., 2004. Receiver operating characteristic (ROC) curve:practical review for radiologists. Korean J. Radiol. 5, 11–18.
Rahman, A., Hakeem, A., 2003. Blood lead levels during pregnancy and pregnancyoutcome in Karachi women. J. Pak. Med. Assoc. 53, 529–533.
Rahman, A., Al-Rashidi, H.A., Khan, A.R., 2012. Association of maternal blood leadlevel during pregnancy with child blood lead level and pregnancy outcome inKuwait. Ecol. Food Nutr. 51, 40–57.
Rossner Jr., P., Tabashidze, N., Dostal, M., Novakova, Z., Chvatalova, I., Spatova, M.,Sram, R.J., 2011. Genetic, biochemical, and environmental factors associatedwith pregnancy outcomes in newborns from the Czech Republic. Environ. HealthPerspect. 119, 265–271.
Sacco, R., Militerni, R., Frolli, A., Bravaccio, C., Gritti, A., Elia, M., Curatolo, P., Manzi,B., Trillo, S., Lenti, C., Saccani, M., Schneider, C., Melmed, R., Reichelt, K.L., Pas-cucci, T., Puglisi-Allegra, S., Persico, A.M., 2007. Clinical, morphological, andbiochemical correlates of head circumference in autism. Biol. Psychiatry 62,1038–1047.
Salpietro, C.D., Gangemi, S., Minciullo, P.L., Briuglia, S., Merlino, M.V., Stelitano, A.,Cristani, M., Trombetta, D., Saija, A., 2002. Cadmium concentration in maternaland cord blood and infant birth weight: a study on healthy non-smoking women.J. Perinat. Med. 30, 395–399.
Shirai, S., Suzuki, Y., Yoshinaga, J., Mizumoto, Y., 2010. Maternal exposure to low-level heavy metals during pregnancy and birth size. J. Environ. Sci. Health A:Tox. Hazard. Subst. Environ. Eng. 45, 1468–1474.
Stasenko, S., Bradford, E.M., Piasek, M., Henson, M.C., Varnai, V.M., Jurasovic, J., Kusec,V., 2010. Metals in human placenta: focus on the effects of cadmium on steroidhormones and leptin. J. Appl. Toxicol. 30, 242–253.
Takiguchi, M., Yoshihara, S., 2006. New aspects of cadmium as endocrine disruptor.Environ. Sci. 13, 107–116.
Thompson, J., Bannigan, J., 2008. Cadmium: toxic effects on the reproductive systemand the embryo. Reprod. Toxicol. 25, 304–315.
Tian, L.L., Zhao, Y.C., Wang, X.C., Gu, J.L., Sun, Z.J., Zhang, Y.L., Wang, J.X., 2009. Effectsof gestational cadmium exposure on pregnancy outcome and development inthe offspring at age 4.5 years. Biol. Trace Elem. Res. 132, 51–59.
Torrey, E.F., Dhavale, D., Lawlor, J.P., Yolken, R.H., 2004. Autism and head circum-ference in the first year of life. Biol. Psychiatry 56, 892–894.
Wilhelm, M., Ewers, U., Schulz, C., 2004. Revised and new reference values for sometrace elements in blood and urine for human biomonitoring in environmentalmedicine. Int. J. Hyg. Environ. Health 207, 69–73.
Yoshida, M., 2002. Placental to fetal transfer of mercury and fetotoxicity. Tohoku J.Exp. Med. 196, 79–88.
and maternal exposure to heavy metals (lead, cadmium and mercury)oi.org/10.1016/j.ijheh.2013.04.009
Zhang, Y.L., Zhao, Y.C., Wang, J.X., Zhu, H.D., Liu, Q.F., Fan, Y.G., Wang, N.F., Zhao, J.H.,Liu, H.S., Ou-Yang, L., Liu, A.P., Fan, T.Q., 2004. Effect of environmental expo-sure to cadmium on pregnancy outcome and fetal growth: a study on healthypregnant women in China. J. Environ. Sci. Health A: Tox. Hazard. Subst. Environ.Eng. 39, 2507–2515.
