Hair trace metal concentration of pregnant women at term in comparison with blood and milk levels

Eur. J. Obstet. Gynecol. Reprod. Biol., 23 (1986) 349-357

349

Elsevier

EIO 00403

Hair trace metal concentration of pregnant women at term in comparison with blood and milk levels RadzisIaw Sikorski ‘, Teodor Juszkiewicz 2, Tomasz Paszkowski l, Tomasz Radomahki ‘, J6zef Szkoda 2 and Pawed Milart l ’ Clinic of Gynecology, Institute of Obstetrics and Gynecology Academy of Medicine, Lublin, Poland, and ’ Department

of Pharmacology and Toxicology, Veterinary Research Institute, Pulawy, Poland

Accepted for publication 8 September 1986

Summary

The concentrations of Fe, Cu, Zn and Pb were determined by means of AAS in the scalp hair (SH) and pubic hair (PH) as well as in maternal blood (MB), breast milk (BM) and cord blood (CB) taken from 104 pregnant women and their neonates. SH values of all studied metals correlated with strong statistical significance with their PH levels. Significant correlations were found between maternal Fe-PH and neonatal body length (r = 0.271) as well as between Fe-SH and placental weight (r = 0.168). The age of examined women was found to correlate negatively with Cu-SH values as well as with Zn-SH (P < 0.05). The inverse relationship between Zn-PH and the parity of examined women was established to be statistically significant (P < 0.001). The usefulness of hair analyses in assessing trace metal status of a pregnant woman is discussed. Iron; Copper; Zinc; Lead; Pregnancy; Hair; Blood; Milk

Introduction

The importance of both toxic and essential trace metals for the course and outcome of pregnancy in humans is commonly understood although many questions concerning the biological cycle of these elements in the matemo-fetal system remain unanswered. Since pregnancy constitutes a period of greater need for nutrients, resulting from both maternal and fetal requirements, efforts have been focused on obtaining a

Correspondence: Prof. Dr. Radzislaw Sikorski, Clinic of Gynecology, Academy of Medicine, Jaczewskiego 8, 20-090 Lublin, Poland.

0028s2243/86/$03.50 0 1986 Elsevier Science Publishers B.V. (Biomedical Division)

350

representative picture of the trace elements status in pregnant women. It has been demonstrated experimentally and clinically that the essential metals may seriously affect prenatal development if present in abnormal concentrations or rates [l]. Among heavy metals of known or suspected prenatal toxicity, lead occupies an important position as a persistent environmental pollutant able to penetrate into fetal tissues through the placental barrier [2]. Many toxico-clinical studies of the last two decades have shown the considerable diagnostic usefulness of human hair in assessing trace metal levels [3-lo]. As the trace metal atom incorporated in the hair structure is fixed, this biological material provides a record of past as well as present element status. Segmental hair analyses may therefore be used for evaluating the trace metal condition of a woman prior to conception as well as demonstrating retrospectively the mineral status during pregnancy. Unlike blood, hair is easy to store and transport. Moreover its collection can be considered a fully noninvasive method of sampling, which makes hair an especially attractive biopsy material in perinatology. The very scanty data available on pubic hair analyses for trace metal content indicate that this type of hair might be even more appropriate as an indicator medium than scalp hair [3,7,10]. Trace metal content in the neonatal scalp hair well reflects the transplacental intake by a developing fetus [3,6,10]. Although the above-mentioned advantages make hair an attractive material for trace metal testing, several critical remarks have been made [ll]. The most commonly emphasized criticism regarding the value of hair analyses concerns the lack of standardized washing procedures, segmental differences along the hair, regional dissimilarities across the scalp and the possible effect of external contamination on the determined element levels. The aim of the present study was to evaluate the usefulness of scalp and pubic hair for assessing the trace metals status of pregnant women at term. The correlations were tested between iron, copper, zinc and lead levels in scalp hair (SH) and pubic hair (PH). Hair metal levels were compared with their concentrations in maternal blood (MB), breast milk (BM) and cord blood (CB). Moreover, the relationships were studied between hair metal levels and selected individual characteristics and some perinatological factors such as neonatal birth weight, body length and placental weight. Material and methods

One hundred and four normal pregnant women receiving perinatal care in the State Clinical Hospital No. 4 in Lublin were randomly recruited for this study. The age of the examined women ranged from 15 to 42 years (median 26 years). Nineteen women (18.3%) were from rural areas and the remaining 85 (81.7%) lived in towns. The duration of the pregnancy at the time of sampling ranged from 37 to 42 weeks (median 40 weeks). The mean parity of the examined gravidae amounted to 1.9. All were apparently healthy; none was receiving any medication; none had been using oral contraceptives or IUD prior to pregnancy. One hundred and four neonates (50 boys and 54 girls) were delivered to the examined women. Their birth weight ranged from 2400 to 4350 g (median 3320 g); the body length ranged from 42 to 60 cm (median 53 cm); placental weight ranged

3.51

from 400 to 1020 g (median 515 g). Maternal venous blood (MB) was aspirated from the cubital vein on admission to the delivery room and the cord blood (CB) was obtained during the routine delivery procedures. The approximately lo-ml blood samples were placed in acid-rinsed glass tubes and frozen immediately after sampling. 50.0 to 100.0 ml of breast milk (BM) were collected on the third or fourth day of puerperium in dark-glass containers with the addition of 1.0 ml of 1% formalin solution. Hair samples weighing approximately 0.3 g were taken from the suboccipital area as well as from the pubic region and were placed in paper envelopes. The hair was washed prior to the analysis with the following sequence of liquids, redistilled water - four times in acetone - redistilled water, and dried at room temperature. Lead was measured by flame atomic absorption spectrometry following chelation and organic extraction, whereas iron, copper and zinc levels were determined by AAS directly in the mineralized samples. A Varian 1250 spectrometer was used. The details of the analytical procedure are described elsewhere [12,13]. In total 2044 determinations were made on 104 SH samples, 95 PH samples (in 9 cases PH was not available on admission to the clinic), 104 MB samples, 104 CB samples and 104 BM samples. Personal characteristics as well as clinical data concerning the examined women and their newborn babies were obtained from clinical documentation. Since a skewing to the right of the obtained results was observed, log-transformed values and nonparametric methods of analysis were used. Descriptive statistics such as range, median, both arithmetic and geometric means, standard error and the coefficient of variation were used in order to demonstrate the determined metal concentrations. In addition, frequency distribution histograms were made for each metal concentration in both types of hair. The interrelationships between the examined trace metals in the analyzed materials as well as between hair metal levels and neonatal birth weight, body length and placental weight were characterized by means of the Spearman rank correlation coefficient. The hypotheses that age, parity, current smoking and the place of residence of the examined women do not affect the determined hair metal levels were tested using Chi-square tests. Results

The results of laboratory analyses are demonstrated in Table I. In all studied metals except lead the dispersion of the determined metal concentrations was greater in SH than in PH (Table I, Fig. 1). Geometric means of essential metal levels in SH were higher than in PH, whereas mean Pb-SH levels and mean Pb-PH levels appeared to be almost equal. The differences between copper, zinc and iron contents in SH and PH were statistically significant (p < 0.05). Table II shows significant metal-versus-metal relationships established in our study. SH levels of all studied metals were strongly correlated with their PH concentrations (p < 0.001). As shown in Fig. 2, lead, copper and zinc levels determined in the hair of both types exceeded many times those measured in CB, MB and BM. Only iron concentrations were found to be much greater in both MB and CB than in maternal hair.

352 TABLE I The concentration of trace metals in scalp and pubic hair of pregnant women at term. n, number of samples; SE, standard error; CV, variation coefficient; Fe-SH, scalp hair iron vahres; Fe-PH pubic hair iron values; Cu-SH, scalp hair copper values; Cu-PH, pubic hair copper values; Zn-SH, scalp hair zinc values; Zn-PH pubic hair zinc values; Pb-SH, scalp hair lead values; Pb-PH, pubic hair lead values. Metal/material Fe-SH Fe-PH Cu-SH Cu-PH Zn-SH Zn-,PH Pb-SH Pb-SH

n

104 95 104 95 104 95 104 95

PbSH

Range ‘7.2 -452.5 1.6 - 63.63 0.592- 19.330 0.520- 15.288 2.805-639.94 44.1 -492.8 0.003- 10.982 0.340- 20.230

Arithmetic mean

SE

30.100 18.936 8.526 6.195 230.915 162.911 2.962 24.008

3.993 1.206 0.431 0.318 10.485 7.034 0.204 20.934

cvm

Median mean

134.8 61.7 51.5 45.1 46.31 42.1 70.4 854.3

20.700 16.430 7.800 6.552 209.000 152.008 2.251 2.141

23.483 16.966 1.342 5.951 209.628 148.312 2.262 2.264

PbPH

lnnn&2

4

6

8

10PPrn

Fig. 1. The frequency distribution of trace metal concentrations in scalp and pubic hair of pregnant women at term. FeSH, s&p hair iron values; FePH, pubic hair iron values; CuSH, scalp hair copper values; CuPH, pubic hair copper values; ZnSH, scalp hair zinc values; ZnPH, pubic hair zinc values; PbSH, scalp hair lead values; PbPH, pubic hair lead values.

353 TABLE

II

Interrelationships between the examined trace metals in the hair samples of pregnant women at term Fe-SH, scalp hair iron values; Fe-PH, pubic hair iron values; CU-SH, scalp hair copper values; Cu-PH, pubic hair copper values; Zn-SH, scalp hair zinc values; Zn-PH, pubic hair zinc values; Pb-SH, scalp hair lead values; Pb-PH, pubic hair lead values. Metal pair examined

Spearman coefficient

Same metal in heterogeneous Fe-SH vs. Fe-PH Cu-SH vs. Cu-PH Zn-SH vs. Zn-PH Pb-SH vs. Pb-PH Different Fe-SH Fe-PH Cu-SH Cu-PH Pb-SH Pb-PH

Different metals in heterogeneous Cu-SH vs. Zn-PH Cu-PH vs. Zn-SH Pb-PH vs. Zn-SH

IRON

p value

materials

metals in homogeneous vs. CuSH vs. ZnPH vs. Zn-SH vs. Zn-PH vs. Fe-SH vs. Fe-PH

LEAD

correlation

0.32148 0.25809 0.58433 0.37222

< < < <

0.01 0.01 0.001 0.001

0.18748 0.19138 0.30659 0.40571 0.17816 0.24345

< < < 4 < <

0.05 0.05 0.001 0.001 0.05 0.001

0.18605 0.34842 0.19027

< 0.05 < 0.001 i 0.05

material

materials

COPPEI

ZINC

Fig. 2. Trace metal concentrations in scalp hair (SH), pubic hair (PH), maternal blood (MB), cord blood (CB) and breast milk (BM); geometric means. All values are expressed as ppm.

354

Zn PH /pm/ 150 0

-

1000

1243

Is.1

1=21

500

PARA 1

PARA 2

PAF~A 3 OR MORE

Fig. 3. Pubic hair zinc levels (geometric means) in the examined woman in relation to parity.

In general, hair levels of essential metals were not well correlated with their concentrations in other studied materials. Only Fe-PH versus Fe-MB (r = 0.238) and Cu-SH versus Cu-BM (r = 0.196) relationships were found to be significant (p < 0.05). Hair lead levels appeared to be better correlated with MB, CB and BM values than the essential metal levels were found to be. Pb-SH values correlated significantly with Pb-MB (r = 0.184, p < 0.05), Pb-CB (r = 0.319) and Pb-BM (r = 0.279 p < 0.01) while Pb-PH showed good correlation with Pb-MB (r = 0.210, p -z0.05)and Pb-CB (r = 0.372, p < 0.001). The inverse relationship between the age of examined women and hair metal levels was statistically significant in the cases of Cu-SH and Zn-PH (p < 0.05). The examined women were grouped according to parity as follows: para 1 (n = 43), para 2 (n = 31) and para 3 or more (n = 21). As shown in Fig. 3, as parity of the examined women increased their pubic hair zinc levels decreased. The relationship between the parity and ZnPH was established to be strongly significant (p < 0.001). Neonatal body length measured at birth was correlated significantly with Fe-PH (r = 0.271) and Fe-SH correlated with placental weight (r = 0.168). It should be emphasized that Fe-MB and Fe-CB levels did not correlate significantly with any of the studied fetoplacental parameters. In the examined group of women 59 (56.7%) were current smokers and the remaining 45 (43.3%) were nonsmokers. No significant relationship was observed in our study between current smoking and trace metal levels in the hair of examined pregnant women. Hair metal levels were not correlated in our study with urban or rural residence of examined gravidae except for Pb-PH, established to be significantly higher in rural women (p -c0.05).

355

Discussion The results of our study seem to confirm that hair analysis may become a valuable research tool in monitoring the mineral status within certain populations of pregnant women, although its diagnostic reliability with respect to individual cases is limited because of the great person-to-person variability of hair metal concentrations. It has been suggested by some authors that pubic hair represents a better indicator medium for trace metal analyses than scalp hair [3,7,10], since the latter may be exposed to external sources of determined elements and exhibits regional differences across the scalp. In our study scalp hair levels exceeded pubic hair values and the dispersion of the determined concentrations was evidently greater in scalp than in pubic hair. These differences together with the excellent correlation established by us between scalp hair metal ievels and their pubic hair values indicate that pubic hair may successfully replace scalp hair in trace element testing. Good correlations found in our study between the hair content of iron and copper, iron and zinc as well as copper and zinc confirm the essentiality of these metals and seem to reflect known interactions between them. However, the positive correlation observed in both scalp and pubic hair between lead and iron levels is rather surprising. The same correlation was reported by Baumslag et al. in neonatal hair [3]. These findings show that known or suspected interrelationships between toxic and essential metals in the ecological niche of the materno-fetal system need to be further explored using ‘unconventional’ biopsy materials also. The lack of correlation established in most cases between hair levels of essential metals and their milk and blood values indicates that the hair deposits of these elements should not be considered as variables directly dependent on the metal concentrations in other tissue compartments. It has been found by Hambidge et al. that during pregnancy the amount of iron incorporated into scalp hair slightly decreases [5]. In our study, iron content, as opposed to copper and zinc levels, was smaller in the hair than in maternal blood. Moreover, the pubic hair iron levels were negatively correlated with iron concentration in maternal blood, whereas no such correlations were found in the cases of copper and zinc. This observation seems to indicate that exceptionally profound and complex changes concerning iron metabolism in pregnancy may also specifically affect the mineral processes during hair formation. Hair lead has been reported to be a valuable indicator of the body burden with this metal, which suggests a remarkable toxicological value of hair analysis for measuring long-term lead exposure [3,6]. Relatively good hair versus blood and hair versus milk correlations for lead established in our study seem to confirm this opinion. Only hair iron appeared to be significantly correlated with the clinical parameters such as neonatal body length and placental weight. No such correlations were found when maternal and cord blood iron levels were analyzed. This observation is not surprising, since hair levels provide a lasting record of metal levels over the previous months, including the whole period of pregnancy, while blood determinations offer information only on immediate levels.

356

Zinc imbalance appears to be a remarkable embryo-fetotoxic factor of well documented pathological effects on the course and outcome of human pregnancy [l].. Since zinc is predominantly intracellular, its plasma levels do not always accurately reflect zinc status in the whole body mass. The analysis of zinc levels in the hair has been found useful in the diagnosis of diverse zinc status in man [14,15]. Baumslag et al. have found that parity affects scalp hair zinc levels in the mother [3]. In our study the pubic hair zinc values versus parity relationship was established to be strongly tignificant. Since the age of the examined women was not correlated with zinc values in pubic hair, the strongly negative correlation between parity and pubic hair zinc concentrations merits special emphasis. Relatively high breast milk zinc levels may contribute to the zinc loss due to multiple maternity. In the light of the results obtained the question reappears as to whether women who undergo many deliveries constitute a group at increased risk of zinc deficiency, and therefore zinc supplementation should be considered in these cases. The results obtained by Grasmick et al. [16] suggest that smoking may increase blood lead levels. Hair lead values were not significantly correlated in our study with current smoking in the examined pregnant women, nor was smoking correlated with the hair levels of the essential metals studied. The hair levels of the essential metals determined by us in rural women did not differ significantly from those of urban women. The significant increase in pubic hair lead levels observed.by us in rural women is surprising, since it is thought that lead exposure in urbanized areas exceeds that in rural territories [17]. Trace metal storage in the hair has its own time-scale and biodynamics not comparable to any other storage or excretory process known in human physiology. Therefore hair analysis may provide a representative and clinically useful record of the trace element status of a patient only when standardized methods and normal values are established for this particular tissue. The results obtained in the present study seem to prove that the benefits of hair analysis are worth these standardizing efforts. References 1 Jameson S. Effe& of zinc deficiency on human reproduction. Acta Med Stand 1976; suppl 593: 5-76. 2 Clarkson TW, Nordberg GF, Sager PR, editors. An overview of the reproductive and developmental toxicity of metals. In: Reproductive and Developmental Toxicity of Metals. New York, Plenum Press; 1983: 3-25. 3 Baumslag N, Yeager D, Levin L, et al. Trace metal content of maternal and neonate hair. Arch Environ Health 1974; 29: 186-191. 4 Dang HS, Jaiswal DD, Mehta U, et al. Trace element changes in hair during pregnancy: preliminary study. Sci Total Environ 1983; 31: 187-192. 5 Hambidge KM, Drogemueller W. Changes in plasma and hair concentrations of zinc, copper, chromium, manganese during pregnancy. Obstet Gynecoll974; 44: 666-672. 6 Hue1 G, Bondene C, Ibrahim MA. Cadmium and lead content of maternal and newborn hair. Arch Environ Health 1981; 35: 221-227. 7 Juszkiewicz T, Szprengier T, Radomafrski T. Mercury content in human pubic and scalp hairs. Natural Sci 1981; 3: 53-62. 8 Laker M. On determining trace element levels in man: the uses of blood and hair. Lancet 1982; ii: 260-262.

357 9 Maugh TH. Hair: A diagnostic tool to complement blood serum and urine. Science 1978; 202: 1271-1273. 10 Sikorski R, Paszkowski T, Szprengier-Juszkiewicz T. Mercury in neonatal hair. Sci Total Environ (in the press). 11 Lazar P. Hair analysis: what does it tell us. JAMA 1974; 229: 1908-1909. 12 Yeager DW, Cholak J, Henderson EW. Determination of lead in biological and related material by atomic absorption spectrophotometry. Environ Sci Technol 1971; 10: 1020-1022. 13 imudzki, J. Oznaczanie zawartosci kadmu w materiale biologicznym metoda absorpcyjnej spektrometrii atomowej. Bromat Chem Toksykol 1980; 1: 77-81. 14 Amador M, Pena M, Garcia-Miranda A. Low hair zinc concentrations in acrodermatitis enteropatica. Lancet 1975; i: 1379. 15 Bergmann KF, Mukosch G, Tews KH. Abnormalities of hair zinc concentrations in mothers of newborn infants with spina bifida. Am J Clin Nutr 1980; 33: 2145-2150. 16 Grasmick C, Hue1 G, Moreau T, et al. The combined effect of tobacco and alcohol consumption on the level of lead and cadmium in blood. Sci Total Environ 1985; 41: 207-217. 17 Tsuchiya K. Lead. In: Friberg L, Nordberg GF, Vouk VB, eds. Handbook on the Toxicology of Metals. Amsterdam, EIsevier/North-Holland Biomedical Press: 1979: 451-484.