A pilot study on the feasibility of European harmonized human biomonitoring: Strategies towards a common approach, challenges and opportunities.

In 2004 the European Commission and Member States initiated activities towards a harmonized approach for Human Biomonitoring surveys throughout Europe. The main objective was to sustain environmental health policy by building a coherent and sustainable framework and by increasing the comparability of data across countries. A pilot study to test common guidelines for setting up surveys was considered a key step in this process. Through a bottom-up approach that included all stakeholders, a joint study protocol was elaborated. From September 2011 till February 2012, 17 European countries collected data from 1844 mother-child pairs in the frame of DEMOnstration of a study to COordinate and Perform Human Biomonitoring on a European Scale (DEMOCOPHES).(1) Mercury in hair and urinary cadmium and cotinine were selected as biomarkers of exposure covered by sufficient analytical experience. Phthalate metabolites and Bisphenol A in urine were added to take into account increasing public and political awareness for emerging types of contaminants and to test less advanced markers/markers covered by less analytical experience. Extensive efforts towards chemo-analytical comparability were included. The pilot study showed that common approaches can be found in a context of considerable differences with respect to experience and expertize, socio-cultural background, economic situation and national priorities. It also evidenced that comparable Human Biomonitoring results can be obtained in such context. A European network was built, exchanging information, expertize and experiences, and providing training on all aspects of a survey. A key challenge was finding the right balance between a rigid structure allowing maximal comparability and a flexible approach increasing feasibility and capacity building. Next steps in European harmonization in Human Biomonitoring surveys include the establishment of a joint process for prioritization of substances to cover and biomarkers to develop, linking biomonitoring surveys with health examination surveys and with research, and coping with the diverse implementations of EU regulations and international guidelines with respect to ethics and privacy.

Interpreting biomarker data from the COPHES/DEMOCOPHES twin projects: Using external exposure data to understand biomarker differences among countries.

In 2011 and 2012, the COPHES/DEMOCOPHES twin projects performed the first ever harmonized human biomonitoring survey in 17 European countries. In more than 1800 mother-child pairs, individual lifestyle data were collected and cadmium, cotinine and certain phthalate metabolites were measured in urine. Total mercury was determined in hair samples. While the main goal of the COPHES/DEMOCOPHES twin projects was to develop and test harmonized protocols and procedures, the goal of the current paper is to investigate whether the observed differences in biomarker values among the countries implementing DEMOCOPHES can be interpreted using information from external databases on environmental quality and lifestyle. In general, 13 countries having implemented DEMOCOPHES provided high-quality data from external sources that were relevant for interpretation purposes. However, some data were not available for reporting or were not in line with predefined specifications. Therefore, only part of the external information could be included in the statistical analyses. Nonetheless, there was a highly significant correlation between national levels of fish consumption and mercury in hair, the strength of antismoking legislation was significantly related to urinary cotinine levels, and we were able to show indications that also urinary cadmium levels were associated with environmental quality and food quality. These results again show the potential of biomonitoring data to provide added value for (the evaluation of) evidence-informed policy making.

Gender differences in cadmium and cotinine levels in prepubertal children.

Susceptibility to environmental stressors has been described for fetal and early childhood development. However, the possible susceptibility of the prepubertal period, characterized by the orchestration of the organism towards sexual maturation and adulthood has been poorly investigated and exposure data are scarce. In the current study levels of cadmium (Cd), cotinine and creatinine in urine were analyzed in a subsample 216 children from 12 European countries within the DEMOCOPHES project. The children were divided into six age-sex groups: boys (6-8 years, 9-10 years and 11 years old), and girls (6-7 years, 8-9 years, 10-11 years). The number of subjects per group was between 23 and 53. The cut off values were set at 0.1 µg/L for Cd, and 0.8 µg/L for cotinine defined according to the highest limit of quantification. The levels of Cd and cotinine were adjusted for creatinine level. In the total subsample group, the median level of Cd was 0.180 µg/L (range 0.10-0.69 µg/L), and for cotinine the median wet weight value was 1.50 µg/L (range 0.80-39.91 µg/L). There was no significant difference in creatinine and cotinine levels between genders and age groups. There was a significant correlation between levels of cadmium and creatinine in all children of both genders. This shows that even at such low levels the possible effect of cadmium on kidney function was present and measurable. An increase in Cd levels was evident with age. Cadmium levels were significantly different between 6-7 year old girls, 11 year old boys and 10-11 year old girls. As there was a balanced distribution in the number of subjects from countries included in the study, bias due to data clustering was not probable. The impact of low Cd levels on kidney function and gender differences in Cd levels needs further investigation.

Biomarkers of low-level mercury exposure through fish consumption in pregnant and lactating Slovenian women.

In order to assess the mercury exposure of pregnant and lactating women in Slovenia, levels of total mercury (THg) and methylmercury (MeHg) were determined in hair, cord blood and breast milk. In addition, the frequency of fish consumption was estimated, because fish is generally the main pathway for human exposure to MeHg. Hair samples were collected from 574 women participating in this study, while cord blood and breast milk samples were collected from 446 and 284 women, respectively. As expected, the levels of THg in hair (median (Med)=297 ng/g, 10th percentile (P10)=73 ng/g, 90th percentile (P90)=781 ng/g), cord blood (Med=1.5 ng/g, P10=0.5 ng/g, P90=4.2 ng/g) and breast milk (Med=0.2 ng/g, P10=0.06 ng/g, P90=0.6 ng/g) were low, due to low consumption of fish (X=25 g/day). A significant linear correlation was found between levels of lnTHg in hair and lnTHg in cord blood (r=0.87, 95% confidence interval (CI): 0.84-0.89), between levels of lnTHg in hair and lnMeHg in cord blood (r=0.94, 95% CI: 0.90-0.96) and between lnTHg levels in cord blood and lnTHg levels in breast milk (r=0.36, 95% CI: 0.25-0.47). Spearman's rank correlations between the frequency of fish consumption and THg in hair (rs=0.35, 95% CI: 0.28-0.42), and between the frequency of fish consumption and THg in cord blood (rs=0.43, 95% CI: 0.36-0.51) or MeHg in cord blood (rs=0.31, 95% CI: 0.06-0.52) were weak. This could be due to the approximate information on fish consumption obtained from the questionnaires, the high variability of MeHg concentrations in fish and a relatively high proportion of inorganic mercury in the biomarkers which originates from sources other than fish. In conclusion, THg levels in cord blood, THg levels in hair and MeHg levels in cord blood are suitable biomarkers of low-level Hg exposure through fish consumption. Compared to cord blood, hair samples are easy to collect, store and analyse.

Relationship between the prenatal exposure to low-level of mercury and the size of a newborn's cerebellum.

Exposure to methylmercury at any stage of central nervous system development could induce alterations and result in severe congenital abnormalities. Total mercury level in maternal hair during pregnancy correlates well with blood levels of methylmercury and with total mercury levels in fetal brain. A prospective study has been conducted and a total of 137 childbearing women living at the coastal region with term, normal pregnancies were included and their newborns evaluated by ultrasonography. Mothers and their newborns are divided in two groups according to their hair mercury levels; examined group with high body levels of mercury (≥ 1 µg/g) and control group with low body levels of mercury (<1 µg/g). Neurosonographic examination was conducted to all newborns. Two dimensions of cerebellum in the sagital-medial plane have been measured: maximum height and width starting from the roof of the fourth chamber. Majority of mothers had hair mercury levels lower than 1 µg/g (N = 107). Mean value was 0.88 µg/g (SD 1.24), ranging from 0.02 to 8.71 µg/g. There was no significant difference between the two groups when it comes to the width of cerebellum (Mann-Whitney test: Z = 1471; p = 0.141). However, comparison related to the length of cerebellum shows statistically significant smaller cerebellum in newborns whose mother had hair mercury levels higher than 1 µg/g (Mann-Whitney test: Z = 2329; p = 0.019). Our results lead to a conclusion that prenatal exposure to, what we consider to be, low-levels of methylmercury does influence fetal brain development detected as decreased size of newborn's cerebellum. From a clinical point of view, a question related to the influence of prenatal low-level methylmercury exposure on fetal neurodevelopment remains open. Our further objectives are to direct the research towards performing detailed neuropshychological tests on children at the age of 18 months. Such tests could indicate the presence of subtle neurological or neuropsychological deficits.

A study of low level selenium determination by hydride generation atomic fluorescence spectrometry in water soluble protein and peptide fractions.

Development of a method for very low level selenium determination in water soluble protein and peptide fractions, obtained after various separation procedures, is presented. A hydride generation atomic fluorescence spectrometry (HG-AFS) detection system was optimised and the influence of Cu(II), Sb(V), As(III) and HNO3 interferences in the measurement of Se by HG-AFS was investigated. A destruction procedure using HNO3 and H2O2 was also optimised and the average recovery of the digestion of a solution of selenomethioneine was 92 +/- 4% (n=14). Combination of this digestion with the detection system gave reliable results. Accuracy was tested by comparison with two independent methods. A very low detection limit (DL) of 0.2 ng/g of measuring solution was achieved. The whole procedure from weighing to measuring was performed in the same Teflon tube. The addition of HNO3 to the fractions before long term storage at -20 degrees C was necessary to prevent adsorption on the test tubes. Selenium was measured in water soluble protein and peptide fractions obtained after extraction, and Sephadex G-75 chromatography performed on liver samples from: i) hens exposed to As2O3, ii) hens fed with a high fat feed and iii) the certified reference material dogfish liver (CRM DOLT-2). Because of the very low DL we were able to observe the Se distribution in chromatographic fractions of samples of organisms which were not exposed to excess amounts of Se. The presence of selenium associated with metallothioneins was observed.

Effect of arsenic trioxide on metallothionein and its conversion to different arsenic metabolites in hen liver.

The metabolism of arsenic, its affinity to metallothionein (MT), its influence on selenium levels, and its biotransformation to different metabolites in the liver tissue of laying hens exposed to arsenic trioxide (As2O3) was investigated. The experiment was performed with two groups of hens fed for 19 d with either a standard diet or with the same diet enriched in arsenic (30 microg/g). The major findings were as follows: 1. After 19 d exposure, about 65% of the total liver As was found in the water-soluble phase (100,000g centrifuged supernatant). In liver supernatant, As binding was found mostly in the range of very low-molecular-weight proteins (Mr < 10,000). Although after exposure the amount of MT-like proteins increased, the As bound to it was only in trace amounts. The protein was identified by convential procedures as Zn,Cu-thionein with traces of selenium and arsenic. 2. Arsenic exposure resulted in almost unchanged Se levels regarding its tissue concentrations and distribution between supernatant and pellet, where about 10% of total Se was found in the supernatant. On the contrary, As exposure did affect Cd levels. Tissue Cd concentration was slightly diminished, but the percentage of tissue Cd found in the water-soluble phase was increased from 20% to 40%. 3. In methanol extracts of tissue and supernatant of the As-exposed group, only two arsenic compounds were detected, As(III) and dimethylarsinic acid (DMA), the latter prevailing.