Objective. Heavy metals are potentially toxic substances, especially for the susceptible infant. Exposure to mercury (Hg) and lead (Pb) may result in neurotoxic and nephrotoxic impairment and in anemia. Previous data on breast milk Pb and Hg contents are sparse or missing for the Austrian population. No evaluations of the influence of mothers’ lifestyles on Pb and Hg levels in breast milk are available.
Methods. Five- to 10-mL individual samples of breast milk were provided from healthy mothers in Vienna (urban; n = 59), Linz (industrial; n = 47), and Tulln (rural; n = 59). A questionnaire about area of residence, maternal nutrition, smoking habits, and dental fillings was filled out by the lactating mothers. Milk samples and infant formulas were lyophilized, wet-ashed with nitric acid (65%), and analyzed with atomic absorption spectrophotometry. Spiked skim milk powder was used as reference material. Statistical analysis included the Kruskal-Wallis test and multiple robust regression analysis.
Results. Breast milk showed low Hg and Pb concentrations (Hg: 1.59 ± 1.21 1g/l, n = 116; Pb: 1.63 ± 1.66 6g/l, n = 138). Eight percent of the breast milk samples marginally exceeded the screening level of 3.5 μg/L for Hg. Austrian Pb values declined strongly during the last 20 years. Bivariate comparison revealed that the factors significantly related to metal levels in breast milk were area of residence (Hg, Pb), prematurity (Hg), consumption of fish (Pb) and cereals (Hg), vitamin supplementation (Hg), and smoking (Pb). The Hg and Pb contents of cow milk and infant formulas were far below respective guideline values.
Conclusions. Neither Hg nor Pb concentrations exceeded critical levels. There are no reports on infants harmed by the intake of milk from unexposed mothers. We conclude that even theoretical risks from current Hg or Pb levels for the breastfed infant of a healthy mother can be ruled out.
Breast milk contains both essential and nonessential trace elements. Lead (Pb) and mercury (Hg) are nonessential, potentially toxic heavy metals with hematotoxic, neurotoxic, and nephrotoxic properties even at very low concentrations.2 The human infant is especially susceptible to toxicity as a result of rapid growth, immaturity of kidneys and liver, and the vulnerability of the myelinizing central nervous system during the first year of life. Infants and young children may absorb as much as 50% of dietary Pb, compared with only 10% for adults.3 Additional evidence shows increased Pb absorption when diets are deficient in calcium, phosphate, selenium, or zinc.3
Although metals and other pollutants are excreted into breast milk in accordance with the environmental contamination and diet of the mothers,4 the advantages of breastfeeding outweigh the risks under normal conditions. Thus, breastfeeding should be encouraged.1,5 Nevertheless, accurate data should be collected about pollutants in breast milk to counsel mothers optimally.
Austrian data on Pb and Hg contents in breast milk are sparse, and recommendations or safety limits are missing. Breast milk Pb has been measured in Tyrol by Lechner et al6 and in Vienna by Maruna et al,7 Haschke and Steffan,8 and Plöckinger et al.9 Lechner et al6 observed higher mean Pb concentrations in areas with high traffic density. In Vienna, a mean value of 50 μg/L milk Pb was measured in 1981, with the value already dropping to 36 μg/L in 1993. Plöckinger et al9 reported that Hg in milk was below limits of detection.
Current Austrian data, particularly for Hg, are missing. Furthermore, the influence of a mother’s lifestyle on milk metal concentrations has not yet been investigated. For the present study, lactating women in Vienna, Linz, and Tulln—cities of different population size, area, and metal environmental pollution—were recruited for breast milk samples. Mothers filled out a questionnaire with respect to environmental factors, maternal nutrition, smoking habits, and amalgam fillings. In addition, cow milk and infant formulas were analyzed. The goals of this study on breast milk were 1) to measure Hg concentrations for the first time in Austria, 2) to measure the possible changes in Pb levels since previous measurements in 1981 and 1993, 3) to evaluate factors that influence Pb and Hg levels, and 4) to compare the above levels with those in infant formulas and cow milk.
Between February and December 1999, breast milk samples of healthy lactating women who lived in the cities of Vienna, Tulln, and Linz were measured. Sampling sites were selected on the basis of different Hg and Pb levels expected in breast milk. In Vienna (population: 1 606 300), diffuse metal sources include atmospheric emissions produced by domestic fuel, traffic, and incinerators. The metal emissions in Linz (population: 208 193) mainly stem from the local metal-processing industry. Among the investigated sites, the small town of Tulln (population: 14 600) can be regarded as rural and therefore least affected by heavy metal contamination.
Hg and Pb levels were measured in milk samples from a total of 165 women: 59 from Vienna, 59 from Tulln, and 47 from Linz. The mean age of the mothers was 29 ± 5 years (± standard deviation); 40% of them were multipara. The mean age of the infants was 6.6 ± 6 days. Birth weight and length were 3121 ± 797 g and 48.6 ± 4.4 cm, respectively. Nineteen percent of the infants were premature, representing the patients of the neonatal intensive care units of the participating hospitals. The study was approved by the ethical committee of the Kinderklinik Glanzing. Written informed consent was obtained from the mothers.
Individual samples of 5 to 10 mL of breast milk, cow milk, and infant formulas (prepared with tap water) were collected and stored in acid-prewashed polyethylene tubes. Because of organizational reasons and feasibility of sampling, breast milk samples were mostly taken between the 2nd and 14th days postpartum. Samples collected after the sixth week postpartum were excluded, because milk contamination decreases with the duration and amount of breastfeeding, as were milk samples from mothers of twins or triplets.
Pretreatment of Samples
Breast milk, cow milk, and infant formula samples were stored at −20° until lyophilization (−48°C) and subsequent homogenization with a glass rod. A total of 200 to 500 mg of each milk sample was digested with 5 mL of nitric acid (65%, suprapure) in closed Teflon vessels (Milestone Laboratory Systems, Monroe, CT) for 10 minutes under constant temperature (175°C). Pure HNO3 samples were used as controls. Spiked skim milk powder (Institute of Reference Materials and Measurements, Geel, Belgium) was used as reference material and treated as described for other milk samples. After cooling, samples were rinsed with 7 mL of Milli-Q-water and stored. The tap water used for the preparation of infant formulas was acidified (pH <2), stored in a freezer (−20°C), and thawed before measurement.
Pb and Hg Measurements
Pb was determined in the graphite furnace of a Hitachi Z 8200 (Berkshire, United Kingdom) Polarized Zeeman Atomic Absorption Spectrophotometer (AAS); Hg was determined by cold-vapor-AAS associated with a hydride formation system (HFS-3, Hitachi) and an amalgamation trap made of gold wire (Uwe Binninger Analytik, Schwäbisch-Gmünd, Germany).
The detection limit (3 standard deviations for the blanks) was 0.14 μg/L for Hg and 0.10 μg/L for Pb. All samples were measured in duplicate; when the coefficient of variation exceeded 10%, the result was rejected and the measurement was repeated. The accuracy of Pb and Hg measurements was checked on external reference samples. Our results for Pb (0.98 mg/L) and Hg (9.7 g/L) contents of the reference material spiked skim milk powder (n = 19) were not significantly different from certified values (Pb: 1.00 mg/L; Hg: 9.4 μg/L). A representative number of samples were co-analyzed. Quality control of Hg measurement was performed by the Arbeitsmedizinischer Dienst in Linz. The Institut für Analytische Chemie (Universität Wien) analyzed Pb in the milk samples. These results were within the variation limits of Hg and Pb determined in the laboratory of the Institut für Medizinische Biologie.
Mothers filled out a questionnaire on residential area, nutrition, smoking habits, and maternal dental fillings. The questions on diet included consumption of vegetables, fruits (especially berries), fish, shellfish, innards (especially pork or bovine liver), meat, mushrooms, cereals, and red wine.
Pb and Hg data (μg/g dry weight) were divided by the factor 7.52 to calculate wet weight (μg/L). The Kruskal-Wallis test was performed for the bivariate relationship between environmental and dietary determinants and milk metal contents using the SPSS 8.0 program (SPSS Inc, Chicago, IL). Relevant determinants showing a significant bivariate relationship with milk metal concentrations were included in a multivariate robust regression model, whereby the multivariate analysis was done with the S-PLUS 4.5 program (Math Soft, Inc, Cambridge, MA). All statistical tests were performed at the significance level of P ≤ .05.
The mean Hg concentration of breast milk (n = 116) in the 3 Austrian cities was 1.59 ± 1.21 μg/L (Fig 1); the mean Pb content of breast milk was 1.63 ± 1.66 μg/L. Hg contents of breast milk and cow milk (1.12 ± 0.56 μg/L; n = 11) were significantly higher than in the infant formulas (0.52 ± 0.35 μg/L; n = 7). The Pb content of tap water was 3.4 μg/L. Infant formulas (1.81 ± 1.04 μg/L; n = 7) showed marginally higher Pb concentrations than the milk samples; mean Pb content of cow milk was 0.9 ± 0.78 (n = 12).
The Kruskal-Wallis test evaluated differences between the mean values of the categories, such as metal contents in smokers versus nonsmokers. The metal contents in Vienna, Linz, and Tulln differed significantly: breast milk Hg levels were the highest in Linz and Vienna, whereas Pb contents were highest in Linz.
Increased Hg levels were also found in the breast milk of mothers <60 kg and in those who had premature infants. Consumption of fish seemed to influence Pb concentrations; the frequent consumption of cereals correlated with higher Hg contents. Vitamin supplements seemed to increase Hg levels. Smokers showed significantly higher Pb contents in breast milk than nonsmokers.
Multivariate robust regression models (Tables 1 and 2) included relevant environmental and dietary determinants that were significantly related to metal contents of breast milk in the bivariate analysis (Table 3). Positive regression coefficients were found for prematurity and the consumption of cereals versus milk Hg contents (Table 1). Both variables contributed significantly to Hg contents of breast milk. Relationships between Pb concentrations and the regarded determinants were not confirmed by the multivariate regression analysis (Table 2).
Current Hg and Pb Levels of Breast Milk in Austria
Previous authors have described different background values for Hg and Pb in breast milk. Jensen2 stated that average background levels of Hg in breast milk are <1 μg/L. Abadin et al1 concluded that, under normal conditions, this concentration is 1.4 to 1.7 μg/L. Women exposed environmentally or occupationally can have higher levels in their breast milk.
The mean breast milk Hg contents (1.59 μg/L) in the present study are within background levels cited above. A total of 7.8% of the samples showed values above 3.5 μg/L, which is the screening level given by Abadin et al1; the present maximum was 6.8 μg/L. Hg concentrations as high as 63 μg/L in breast milk from Minamata, Japan, and values up to 200 μg/L in Iraq2 caused severe neurobiological damage. Current values in Austria are far below such critical levels, and we can thus exclude adverse effects on breastfed infants. Compared with international data (Fig 2), the current Austrian values are within the European range.
In Austria, Pb contents in breast milk have decreased continuously since 1981 and 1993, most likely as a result of the mandatory shift from leaded to unleaded fuels in 1993. In the present study, the mean value was 1.63 μg/L; this is far below the average background levels in breast milk published by Jensen2 (5–20 μg/L) or Abadin et al1 (2–5 μg/L). It is also one of the lowest in Europe (Fig 3).
Factors Related to Metal Contents of Breast Milk in Austria
Area of Residence
The present study shows that mothers from Linz had significantly higher Hg and Pb contents in their breast milk compared with mothers from Tulln (Table 3). The metal-processing industry is located in Linz, and metal contamination of the environment was confirmed by several studies: Hg and Pb concentrations of soil17 or in the leaves of certain trees18 were markedly higher near the industrial area. Elevated Pb contents in breast milk from mothers living near metal-processing industry were described by Hallén et al16 and Frković et al.13
We measured the highest milk Hg concentrations in Vienna (Table 3). In 1994, a third of Vienna’s Hg emissions were produced by incinerators, followed by industrial emissions.19 Several authors have described higher metal contents in the milk of mothers who live in urban versus rural areas.14,15 Our study confirms this: the lowest values were found in the more rural area of Tulln (Table 3).
Maternal Body Weight
Mothers who weighed <60 kg showed higher Hg contents in the present study. Yoshinaga et al20 observed a significant influence of mother’s and infant’s body weight on the elementary composition of breast milk, although the biological attributes were not the major factors contributing to elemental variation.
Three strongly correlated factors—body length and weight of the child at birth, and prematurity—were significantly related to Hg concentrations: mothers of preterm infants showed significantly higher milk Hg contents (Table 3), which might be attributable to changing trace element metabolism during pregnancy and/or the lactation period. Perrone et al21 found that Pb contents declined with lactation time, both in pre- and full-term samples. Friel et al22 reported lower Pb concentrations in preterm milk, which is confirmed by our results: this milk had marginally lower contents (Table 3).
Relationships between maternal nutrition and milk metal contents have been studied thoroughly. Fish consumption was significantly related to milk Pb levels in the present study. In contrast, milk Hg contents were unaffected. Mothers who eat predominantly fish, whale, or reindeer meat accumulated considerable amounts of Hg in their breast milk.4,12,23–25 The levels were also influenced by the amount of cereals ingested (Table 3): wheat, in particular, may contain elevated Hg concentrations.26
Abadin et al1 discussed studies indicating that vitamin E decreases the toxic effect of Hg and that increased vitamin D and fat in the diet increase Pb absorption in rats. In the present study, we found a significant positive relation between vitamin supplementation and milk Hg (P < .05) but not Pb (P > .05) contents (Table 3). The reason for this remains unknown.
We were unable to detect an influence of dental amalgam on milk Hg contents for statistical reasons (small number of mothers without fillings; n = 3). Some authors reported positive correlations between the number of amalgam fillings and Hg concentrations in body fluids.12,23,27,28 Klemann et al11 found no significant influence on the Hg contents of amniotic fluid and cord blood. Jones29 stressed that the amount of Hg released from such fillings is minimal and that the uptake of food-related organic Hg is 6 times higher; moreover, food-related Hg is significantly more toxic. Drexler and Schaller24 concluded that the additional exposure of breastfed infants from maternal amalgam fillings is of minor importance compared with maternal fish consumption.
Frković et al13 found that breast milk of nonsmokers had higher Pb contents compared with that of smokers. The present study yielded a contrasting result: current smokers showed significantly increased Pb levels (Table 3). Similar results were obtained by al-Saleh30 and Symanski and Hertz-Picciotto.31 An average increase of approximately 15% in cord blood Pb levels was estimated by Rhainds and Levallois32 for every 10 cigarettes smoked per day when mothers smoked and consumed alcohol at the same time.
Smoking had no influence on the milk Hg contents in the present study. Langworth et al27 observed only a poor correlation between smoking and Hg levels in blood and urine when they compared a group of chloralkali workers with a control group.
Comparison of Breast Milk Hg and Pb Contents With Cow Milk and Infant Formulas
Hg in Infant Formulas
In the present study, breast milk and cow milk showed significantly higher Hg contents than infant formulas. Drasch et al12 found that during the first days after delivery, some colostrum samples had higher values than formula milk samples. Later, the concentration in the breast milk was equal to or even lower than that in formula milk.
Pb in Infant Formulas
Pb levels in breast milk are normally lower than in milk-based infant formulas.2 Schumann,33 however, reported that heavy metal concentrations were usually in the same order of magnitude. Reconstitution of infant formulas with contaminated tap water can result in much higher Pb concentrations. The tap water used here had a Pb content of 3.4 μg/L, which is far below the Austrian safety limit of 50 μg/L.34 Nevertheless, tap water Pb clearly contributed to the relatively high Pb contents of infant formulas compared with breast milk or cow milk. In Graz (southeast Austria), for example, tap water Pb contributed to 45% of the Pb content in infant formulas.35
Cow Milk Pb and Hg
Hg (1.12 μg/L) and Pb (0.9 μg/L) contents of cow milk were also far below the current recommendations36: 10 μg/L (Hg) and 30 μg/L (Pb), respectively.
Breast milk is the optimal nutrition for the young infant. Pb and Hg levels in Austrian breast milk, cow milk, and infant formulas are relatively low and are by far below the currently recommended safety limits. As maternal and environmental conditions lead to significant differences in milk metal levels, however, all possible measures must be taken to reduce environmental contamination and contamination of infants. We conclude that infants and small children in Austria are not endangered by Pb or Hg, whether they are breastfed or fed infant formulas or cow milk.
We are grateful to the Austrian National Bank for supporting this study (project 7662): Human Milk Lead (Pb) and Mercury (Hg) Levels in Austria.
We thank the Arbeitsmedizinischer Dienst Linz, in particular Günter Kufner for the free quality control of breast milk samples; and Gabriele Jochinger (lactation counselor), Ingeborg Hanreich (dietitian), Peter Weiss (Umweltbundesamt Wien), and Johann Wimmer (OÖ. Umweltanwaltschaft Linz) for valuable information.
- ↵Abadin HG, Hibbs BF, Pohl HR. Breast-feeding exposure of infants to cadmium, lead, and mercury: a public health viewpoint. Toxicol Industrial Health.1997;134 :495– 517
- ↵Jensen AA. Levels and trends of environmental chemicals in human milk. In: Jensen AA, Slorach SA, eds. Chemical Contaminants in Human Milk. Boca Raton, FL: CRC Press; 1991:45–198
- ↵World Health Organization. Inorganic Lead. WHO Environmental Health Criteria Series. Geneva, Switzerland: World Health Organization; 1995;165:1–300
- ↵American Academy of Pediatrics. Breastfeeding and the use of human milk. Pediatrics.1997;100 :1035– 1039
- ↵Lechner W, Battista HJ, Dienstl F. Untersuchungen zum Bleigehalt in der Muttermilch in verkehrsreichen und verkehrsarmen Gegenden Tirols. Gynäk Rdsch.1980;20(suppl 2) :268– 270
- ↵Maruna H, Maruna RFL, Eisner R. Cadmium und Blei in Muttermilchproben. Ärztl Praxis.1976;1 :6– 7
- ↵Haschke F, Steffan I. Die Bleibelastung des jungen Säuglings mit der Nahrung in den Jahren 1980/81. Wr Klin Wochenschr.1981;93 :613– 616
- ↵Plöckinger B, Dadak C, Meisinger V. Blei, Quecksilber und Cadmium bei Neugeborenen und deren Müttern. Z Geburtsh Perinat.1993;197 :104– 107
- ↵Klemann D, Weinhold J, Strubelt O, Pentz R, Jungblut JR, Klink F. Effects of amalgam fillings on the mercury concentrations in amniotic fluid and breast milk. Dtschz Ahnärztl Z.1990;453 :142– 145
- ↵Guidi B, Ronchi S, Ori E, et al. Lead concentrations in breast milk of women living in urban areas compared with women living in rural areas. Pediatr Med Chir.1992;146 :611– 616
- ↵Weiss P, Riss A. Schadstoffe im Raum Linz. Monographien Band 20. Vienna, Austria: Umweltbundesamt Wien; 1992:32–63
- ↵Oberösterreichische Umweltanwaltschaft, Landesforstdirektion O.Ö. Project report. Biomonitoring Raum Linz; 1997
- ↵Umweltbundesamt Wien, Corinair 1994. Available at: www.ubavie.gv.at
- ↵Yoshinaga J, Li JZ, Suzuki T, et al. Trace elements in human transitory milk. variation caused by biological attributes of mother and infant. Biol Trace Elem Res.1991;312 :159– 170
- ↵Perrone L, Di Palma L, Di Toro R, Gialanella G, Moro R. Interaction of trace elements in a longitudinal study of human milk from full-term and preterm mothers. Biol Trace Elem Res.1994;413 :321– 330
- ↵Friel JK, Andrews WL, Jackson SE, et al. Elemental composition of human milk from mothers of premature and full-term infants during the first 3 months of lactation. Biol Trace Elem Res.1999;673 :225– 247
- ↵Oskarsson A, Schütz A, Skerfving S, Palminger Hallén I, Ohlin B, Lagerkvist BJ. Total and inorganic mercury in breast milk and blood in relation to fish consumption and amalgam fillings in lactating women. Arch Environ Health.1996;513 :234– 241
- ↵Klopov VP. Levels of heavy metals in women residing in the Russian Arctic. Int J Circumpolar Health.1998;571 :582– 585
- ↵Classen HG, Elias PS, Hammes WP. Toxikologisch-hygienische Beurteilung von Lebensmittelinhalts- und -zusatzstoffen sowie bedenklicher Verunreinigungen. Pareys Studientexte 54. Berlin, Germany: Verlag Paul Parey; 1987:228–238
- ↵Langworth S, Elinder CG, Gothe CJ, Vesterberg O. Biological monitoring of environmental and occupational exposure to mercury. Int Arch Occup Environ Health.1991;633 :161– 167
- ↵Vimy MJ, Hooper DE, King WW, Lorscheider FL. Mercury from maternal “silver’ tooth fillings in sheep and human breast milk. A Source of Neonatal Exposure. Biol Trace Elem Res.1997;562 :143– 152
- ↵Jones DW. Exposure or absorption and the crucial question of limits for mercury. J Can Dent Assoc.1999;651 :42– 46
- ↵al-Saleh IA. Lead exposure in Saudi Arabia and its relationship to smoking. Biometals.1995;83 :243– 245
- ↵Symanski E, Hertz-Picciotto I. Blood lead levels in relation to menopause, smoking, and pregnancy history. Am J Epidemiol.1995;141 :1047– 1058
- ↵Rhainds M, Levallois P. Effects of maternal cigarette smoking and alcohol consumption on blood lead levels of newborns. Am J Epidemiol.1997;1453 :250– 257
- ↵Schumann K. The toxicological estimation of the heavy metal content (Cd, Hg, Pb) in food for infants and small children. Z Ernährungswiss.1990;291 :54– 73
- ↵Pescheck R, Herlicska H. Schadstoffbelastung von Wasser und Abwasser in Osterreich. Monographien Band 24. Vienna, Austria: Umweltbundesamt Wien; 1990:112–115
- ↵Krachler M, Rossipal E, Irgolic KJ. Trace elements in formulas based on cow and soy milk and in Austrian cow milk determined by inductively coupled plasma mass spectrometry. Biol Trace Elem Res.1998;651 :53– 74
- ↵Richtwerte für Schadstoffe in Lebensmitteln. Bundesgesundheitsblatt.1996;39 :193– 194
- Copyright © 2002 by the American Academy of Pediatrics