Climate change, Environmental contaminants and Reproductive health

Projektinformationen

CLEAR

ID Finanzhilfevereinbarung: 226217

Projekt abgeschlossen

Startdatum 1 Mai 2009

Enddatum 31 Oktober 2013

Finanziert unter

• Specific Programme "Cooperation": Environment (including Climate Change)

Gesamtkosten € 3 181 078,80

EU-Beitrag € 2 377 603,60

Koordiniert durch

AARHUS UNIVERSITETSHOSPITAL
Denmark

Final Report Summary - CLEAR (Climate change, Environmental contaminants and Reproductive health)

Executive Summary:
The main aim of the CLEAR project was to evaluate how climate change may alter human exposure to environmental contaminants with the potential to disturb human reproductive health and development in a cohort including about 1400 pregnant women and 600 spouses from Greenland, Poland and Ukraine.

Male and female exposure to most of the measured persistent organic pollutants (POPs) were particularly high among Inuits from Greenland. Also Bisphenol A and mercury were detected in higher concentrations in Greenland, but lead, cadmium and phthalates showed more evenly distribution across countries, indicating that the overall load of chemical exposure was highest among the Inuits from Greenland.

Modelling of climate induced change in exposure was performed. Only small changes in exposure to POPs are expected due to changes in temperature, precipitation, sea ice cover, primary productivity and resultant organic carbon dynamics in the Arctic Ocean. However, dietary transitions which may partly be a consequence of changes in available local food items due to climate change induced habitat changes, are expected to play a major role in changes in exposures.

Studies of male reproductive toxicity in the cohort indicated a few strong associations. Most noteworthy was an inverse association between phthalate metabolites and male testosterone level. In addition exposure to PFOS was indicated to be associated to more abnormal sperm morphology. Also, female reproductive function measured as time to pregnancy and menstrual cycle characteristics was indicated to be adversely affected at higher level of exposure to perfluorinated compounds (PFCs). Finally organochlorine exposures were indicated to have minor if any effects on growth, motor and behavioural development, but PFCs was indicated to be associated more ADHD like behaviour.

Some of the potential associations between environmental exposures and reproductive outcomes may be masked by different susceptibility of genetically differing individuals. Some indications of gene-environment interactions were detected in relation to androgen receptor polymorphisms in the CAG repeat length and phthalate exposure as well as in the aryl hydrocarbon receptor pathway and POP exposure.

Finally, the project evaluated sperm DNA global methylation level as a possible mechanism by which environmental exposure may affect male fertility.
Geographical differences in methylation level were indicated. The study did not indicate major consistent associations across study populations between methylation level and the measured environmental contaminants.

In conclusion, the study demonstrates that climate change will not directly cause major changes in environmental POP exposure levels in the Arctic and, furthermore, the present study does not indicate that adult exposure to current levels of environmental contaminants have marked influence on male and female fertility or child growth and development. We emphasize that this study has examined male reproductive health in much more detail than female reproductive health and child development and – not least important – that pronounced long-term effects of fetal exposure on adult reproductive health has not been addressed.

Project Context and Objectives:
The overall objective of this multi-disciplinary research proposal is to investigate the possible impact of global climate change on reproductive health in the Arctic and in three local European populations. The key questions to be addressed are, first, how may climate change impact on human exposure to widespread environmental contaminants and, second, how may contaminants impact on occurrence of reproductive disorders as sensitive indicators of health?
To reach this goal the proposal has the following specific objectives:
1. To identify and describe mechanisms by which the changing climate may affect the exposure of Arctic and other human populations to contaminants through change in chemical use and emissions, delivery to the ecosystems as well as processing within the physical environment and human food chain. Under a number of assumptions this will result in estimates of the magnitude of changing exposure according to different scenarios of climate change. This work relies on modelling of existing data.
2. To expand the existing knowledge database on human contaminant exposure in the Arctic and selected European countries by measurements of biopersistent and non-persistent compounds in serum samples, namely
(i) polybrominated diphenylethers (PBDEs)
(ii) perfluorinated surfactants (PFOS and PFOA)
(iii) phthalates (5-oxy-MEHP as a marker of diethylhexyl phthalate)
(iv) Metals (lead and mercury)
This work relies on more than 3,500 biobanked serum specimens from the INUENDO cohorts established by project participant based on an EU FP5 grant.
3. To increase the limited knowledge on links between parental blood levels of environmental contaminants and reproductive health outcomes in terms of:
(i) functional and biological measures of fertility
(ii) child development [growth, developmental milestones, attention-deficit hyperactivity (ADHD) and obesity in children 5-9 years old]
This work relies on a large existing parent-child-cohort where a follow-up survey provides new crucial data that are fed into risk assessment.
4. To investigate mechanisms related to effects of contaminants on reproductive health as genomic methylation status in the germ cell line and impact of polymorphisms in strategic genes related to androgen, oestrogen and aryl hydrocarbon pathways. These studies will help disentangle causal from non-causal associations in epidemiological studies and thus qualify the risk assessment. This work relies on semen and blood samples already collected in INUENDO.
5. To integrate data on relative climate induced changes in contaminant mobility and distribution, external and internal exposure of humans and links between contaminant exposure and health surveys into a risk assessment and risk evaluation providing insight into possible future risk scenarios related to global climate change. Reviews of critical reproductive effects, exposure-response levels and benchmark doses in animal models provide part of the basis for risk assessments.
This proposal is for several reasons focused upon reproductive function and child development as indicator of human health. First, possible effects of environmental exposures are manifested within months (fertility and adverse pregnancy outcomes) or few years (child developmental outcomes), which means that epidemiological findings are supposed to reflect current exposures. Second, reproductive processes in both men and women are considered highly sensitive to actions of environmental xenobiotics. Third, methodologies have been developed to investigate reproductive health in human populations and, fourth, a large cohort including Arctic as well as three European populations has been established and is suitable for the purpose of this study.

The project draws upon a network of experts in climate modelling and experimental, epidemiological and risk assessment methodologies that partly were developed in a recent cost shared European project (the INUENDO project; www.inuendo.dk). The cohorts include more than 2,500 families with a balanced distribution of Inuits from almost all parts of Greenland, with a high intake of seafood, a mixed rural and urban Ukranian population with high environmental exposure to pesticide residues and a Polish reference population. More than 3,500 biobanked blood and semen samples are available for this project.

Project Results:
Description of main results


WP2 – Climate change and contaminant mobility and distribution

Deliverables
D2.1 Article identifying and describing mechanisms by which climate changes may affect contaminant exposure in the Arctic

D2.2 Article estimating changes in the extent of contaminant delivery to the Arctic ecosystem

D2.3 Article on estimated changes in the processing of contaminants within the physical Arctic environment

D2.4 Article estimating changes in the processing of contaminants within the Arctic human food chain

D2.5 Input to risk assessment WP7 in terms of data on relative change in the exposure of Northern populations to contaminants

Objectives & Accomplishments

The main objective of WP2 was to describe mechanisms by which a changing climate may affect the exposure to contaminants of people living in the Arctic considering i) changes in chemical use and emissions, ii) changes in the extent of contaminant delivery to the Arctic ecosystem (e.g. Long-range Transport Potential, LRTP), iii) changes in the processing of contaminants within the physical Arctic environment, iv) changes in the processing of contaminants within the Arctic food chain and v) changes in exposure due to changes in the life-style of Northern populations (e.g. increased consumption levels of imported food). Where possible, the potential influence of climate change-induced alterations in chemical fate, transport and exposure were quantified using various modelling tools. As discussed in a Critical Review article published by WP2 on interactions between global climate change and contaminants in the North [1], the potential impact of some climate change scenarios can more readily be assessed than others. For example, scenarios describing temperature and precipitation changes can be formulated directly from available projections whereas changes in chemical use and emissions or alterations to food web dynamics in the Arctic (e.g. via entry of new/invasive species) are more problematic in terms of realistic scenario development.

A key concept articulated in the Critical Review published by WP2 [1] is that of “compensatory behaviour”. In short, there are some projected alterations to the physical environmental related to global climate change that are already known to be antagonistic and hence can be expected to exert a dampening effect on the magnitude of increase or decrease in exposure that might be expected by examining processes in isolation. In some cases, changes to a single parameter can act both positively and negatively on exposure. For example, warmer temperatures in source regions can lead to enhanced emissions/revolatilization of contaminants from reservoirs but would also be associated with enhanced degradation in the atmosphere and environment in general. The fact that the Arctic region is projected to experience greater warming than any other region is also relevant in the context of chemical exposure for Northern populations. Warmer ambient temperatures are associated with reduced deposition to the surface compartments that are most important for far-field exposure pathways (e.g. via the marine food web). These types of compensatory behaviours largely drive the model results summarized in the Critical Review paper [1] and those conducted for the CLEAR project [2,3] which indicate a limited influence (less than a factor of two) of the physical changes considered to date. Other researchers have arrived at similar findings using different modelling tools and scenarios [4–6].

Another key aspect of the studies published by WP2 is the need to consider the potential implications of global climate change on chemical exposure from a multimedia perspective. For many chemicals of concern (e.g. polychlorinated biphenyls, PCBs), the atmosphere is most sensitive to climate-induced changes (i.e. exhibits the largest ‘negative’ response) [2–4]. However, levels of contaminants in the atmosphere are typically the least relevant for human exposure in the far-field context (i.e. excluding the indoor environment). Indeed, the use of atmospheric concentrations as an indicator of overall contamination is premised on the assumption that there have not been nor will there be any significant changes to the physical environment that may alter the fate and distribution of the chemicals of interest. Work published by WP2 [1–3] clearly demonstrates that projected changes to atmospheric levels in the Arctic are not necessarily indicative of changes in exposure-relevant media (e.g. surface ocean waters) with respect to the magnitude and direction of change. Accordingly, sampling campaigns seeking to elucidate the impact of global climate change on chemical exposure must be more holistic in nature (i.e. can not be limited to air sampling) in order to address divergences in response.

While physical alterations to the environment related to global climate change do not appear to exert a substantial influence on chemical exposure, work published by WP2 illustrates how the ongoing dietary transition could be far more important [7]. Consumption of marine mammals by Northern populations is a key pathway of exposure to hydrophobic contaminants. However, many communities are shifting towards greater consumption of imported (store-bought) foods at the expense of locally-harvested ‘country foods’ like ringed seal (skin, blubber). This dietary transition is important for two reasons. First, the extent to which Northern populations shift towards imported food determines the extent to which they become disconnected from chemical exposure pathways in the Arctic and hence the influence of climate-induced changes to this environment. Second, the shift from consuming high trophic level/marine organisms to low trophic level/terrestrial organisms is associated with a reduction in exposure far in excess of any increase in exposure quantified for changes to the physical environment.

The publications generated by WP2 during the CLEAR project have contributed to the general consensus emerging that indirect changes linked to global climate change (e.g. human behavioural responses) are likely to be far more influential on human exposure than the direct changes linked to global climate change (i.e. alterations to the physical environment). While additional work is required to better understand the potential impacts of changes to key aspects of the Arctic environment (particularly the cryosphere), the modeling studies published to date by WP2 and other authors [1–7] have allowed some of the issues surrounding global climate change and contaminant exposure to be put in a proper context and provided insights useful to designing monitoring campaigns and focusing research on key issues.

Literature Cited

1. Armitage, J.M.; Quinn, C.L. Wania, F. Global climate change and contaminants--an overview of opportunities and priorities for modelling the potential implications for long-term human exposure to organic compounds in the Arctic. J. Environ. Monitor. 2011, 13, 1532 – 1546.
2. Gouin, T.; Armitage, J.M.; Cousins, I.T.; Muir, D.C.G.; Ng, C.A.; Reid, L.; Tao, S. Influence of global climate change on chemical fate and bioaccumulation: The role of multimedia models. Environ. Toxicol. Chem. 2013, 32, 20 – 31.
3. Armitage, J.M.; Wania, F. Exploring the potential influence of climate change and particulate organic carbon scenarios on the fate of neutral organic contaminants in the Arctic environment. Environ. Sci. Process Impacts 2013, 15, 2263 – 2272.
4. Lamon, L.; von Waldow, H.; MacLeod, M.; Scheringer, M.; Marcomini, A.; Hungerbühler, K. Modeling the global levels and distribution of polychlorinated biphenyls in air under a climate change scenario. Environ. Sci. Technol. 2009, 43, 5818 – 5824.
5. Wöhrnschimmel, H.; MacLeod, M.; Hungerbühler, K. Emissions, fate and transport of Persistent Organic Pollutants to the Arctic in a changing global climate. Environ. Sci. Technol. 2013, 47, 2323 – 2330.
6. Kong, D.; MacLeod, M.; Li, Z.; Cousins, I.T. Effects of input uncertainty and variability on the modelled environmental fate of organic pollutants under global climate change scenarios. Chemosphere 2013, 93, 2086 – 2093.
7. Quinn, C.L.; Armitage, J.M.; Breivik, K.; Wania, F. A methodology for evaluating the influence of diets and intergenerational dietary transitions on historic and future human exposure to persistent organic pollutants in the Arctic. Environ. Internat. 2012, 49, 83 – 91.

WP3 Exposure assessment

The aims mentioned in the application was to develop and document a sensitive method to analyse mono-(2-ethyl-5-oxohexyl) phthalate (5-oxo-MEHP) in serum, to analyse biobanked serum samples for contaminants as mentioned below, to characterize the exposure profiles and describe correlations between the contaminants.

Furthermore, in the application it was mentioned that analysis of 5-oxo-MEHP, perfluorooctane sulfonic acid (PFOS), perfluorooctanoic acid (PFOA), hexachorobenzene (HCB), lead and mercury should be performed in the biobanked serum from 1000 subjects. The 1000 samples should be 600 paternal and 400 maternal serum samples from couples where the male partner had provided a semen sample.

The deliverables included in WP 3 are the following

D3.1 A novel method to analyse the concentration of 5-oxo-MEHP in human biobanked serum samples is expected to be documented and validated.

D3.2 Levels of HCB.

D3.3 Profile of PCB in pooled serum samples from different countries

D3.4 Database with analyses of PBDE

D3.5 Levels of PFOS and PFOA.

D3.6 Levels of 5-oxo-MEHP.

D3.7 Levels of lead and mercury.

All of the deliverables in the project have been carried out and even extended to include more than was mentioned in the original application:

We have developed and validated a highly sensitive analytical method for simultaneous analysis of six phthalate metabolites, eight perfluorinated chemicals, bisphenol A and cotinine in aliquots of only 0.1 ml serum.

The phthalate metabolites analysed by the method were mono-(2-ethyl-5-hydroxylhexyl) phthalate (5-oh-MEHP), 5-oxo-MEHP and mono-(2-ethyl-5-carboxypentyl) (5-cx-MEPP), all metabolites of di-ethylhexylphthalate (DEHP). Furthermore, the metabolites of di-isononylphthalate (DiNP) mono-(4-methyl-7-hydroxyloctyl)phthalate (7-oh-MMeOP), mono-(4-methyl-7-oxo octyl)phthalate (7-oxo-MMeOP) and mono-(4-methyl-7-carboxyheptyl)phthalate (7-cx-MMeHP) were analysed by the method.

The perfluorinated chemicals analysed by the method were perfluorohexane sulfonic acid (PFHxS), perfluoroheptanoic acid (PFHpA), PFOA, perfluorononanoic acid (PFNA), PFOS, perfluorodecanoic acid (PFDA), perfluoroundecanoic acid (PFUnDA) and perfluorododecanoic acid (PFDoDA).

With this method we have analysed the phthalate metabolites, the perfluorinated chemicals, bisphenol A and cotinine in the blood sera from 602 male partners that provided a semen sample. Furthermore, we analysed serum HCB from 588, blood lead and cadmium from 505 and blood mercury from 531 of these men using already developed and validated methods.

In addition, we have analysed the phthalate metabolites, the perfluorinated chemicals, bisphenol A and cotinine in sera from all 1441 women providing serum samples. Also, we analysed HCB in serum from 380 and lead, cadmium and mercury in whole blood from 281 of these women.

Significant results

Quantitative analytical methods
A paper describing the analytical method for determination of PFCs, phthalate metabolites, bisphenol A and cotinine has been submitted for publication.

Exposure and determinants
A paper describing exposure and determinants of exposure of perfluorinated compounds in men from the three countries has been published:

Lindh CH, Rylander L, Toft G, Axmon A, Rignell-Hydbom A, Giwercman A, Pedersen HS, Góalczyk K, Ludwicki JK, Zvyezday V, Vermeulen R, Lenters V, Heederik D, Bonde JP, Jönsson BAG. Blood serum concentrations of perfluorinated compounds in men from Greenlandic Inuit and European populations. Chemosphere. 2012; 88: 1269-1275

This paper identifies markedly different exposure profile between countries with the highest exposures for most PFCs found on Greenland. The levels of PFOS were among the highest ever reported in a general population. However, except for intake of seafood, tea and age being determinants of PFOS exposure on Greenland, the study was not able to identify strong determinants of exposure to perfluorinated compounds.

Similar results were obtained for the women where the exposures of the women to the PFCs were highest on Greenland for all PFCs except for PFOA, PFHxS and PFHpA where the highest levels were found in Poland. The PFC with the highest concentration in all three populations was PFOS. The other PFCs were found in lower levels.

The mean concentration of bisphenol A was much higher on Greenland than in Poland and Ukraine. The highest levels were found in the areas on Greenland that have changed from a traditional Inuit life-style to a more westernised one. The results have been presented on a scientific meeting as an oral presentation:

Jönsson BAG, Rylander L, Lind CH, Lenters V, Rignell-Hydbom A, Giwercman A, Pedersen HS, Toft G, Bonde JP. Higher blood serum concentrations of bisphenol A in Greenlandic Inuit compared to men from two other European populations. ISEE 2013, 19th-23rd August 2013, Basel, Switzerland.

A paper describing levels and determinants of the phthalate metabolites is in preparation. The exposure to DEHP and DiNP, according to the levels of their metabolites in serum was rather similar on Greenland and in Poland and Ukraine. The exposure to DEHP was higher than for DiNP. Interestingly, the levels of phthalate metabolites were higher than in Swedish pregnant women 20-30 years ago.

The cotinine levels were higher in Greenland than in Ukraine. The lowest cotinine levels were found in Poland.

Also, the levels of mercury were high on Greenland while the levels of lead and cadmium were similar to other countries. This has been published in a paper on levels of mercury in blood and semen quality and reproductive hormones:

Mocevic E, Specht IO, Marott JL, Giwercman A, Jönsson BAG, Toft G, Lundh T, Bonde JP. Environmental mercury exposure, semen quality and reproductive hormones in Greenlandic Inuit and European men: a cross-sectional study. Asian Journal of Andrology, 2013; 15: 97-104.

The levels and determinants of flame retardants have been described in a publication:

Lenters V, Thomsen C, Smit LAM, Jönsson BAG, Pedersen HS, Ludwicki JK, et al. Serum concentrations of polybrominated diphenyl ethers (PBDEs) and a polybrominated biphenyl (PBB) in men from Greenland, Poland and Ukraine. Environ Int 2013;61:8–16.

At least one flame retardant (PBDE or PBB) was detected in 298 of the 300 participants’ serum samples. Concentrations of individual congeners were 2.7 to 15 fold higher in Greenlandic relative to Polish and Ukrainian men. Within Greenland, concentrations were higher in men from the South-East than in men from the West coast (including Nuuk). It is unclear what proportion of Greenlanders’ exposure is due to emissions from imported flame retardant-impregnated products and their disposal via burning, relative to consumption of contaminated fish and high trophic level animals (e.g. seals and polar bears).

25 PCBs were measured in a subset of 45 serum samples from the CLEAR men (July 2011). Correlations between PCB-153, which was measured in the complete cohort during the baseline (INUENDO) study (Jönsson et al. 2005), and other PCB congeners were characterized, thereby evaluating the validity of congener 153 as a biomarker for exposure to PCBs. Many PCB congeners, particularly higher-chlorinated congeners, are strongly correlated (r >0.85). PBDE and PCB congener profiles analyses were subcontracted by Utrecht University to the Norwegian Institute of Public Health.
WP4 Male and female fertility


All data on contaminant exposure have been merged for males and females separately, and a database including imputed values on levels below level of detection has been generated by partner 8. The contaminants measured were hexachlorbenzene (HCB), eight brominated flame retardants, eight perfluorinated chemical (PFCs), six metabolites of the phthalates from diethylhexyl phthalate and diisononyl phthalate, bisphenol A, the smoking biomarker cotinine, vitamin D and the heavy metals lead, mercury and cadmium. We ended up with a considerably more extensive database on exposures compared to what has been described in the proposal including a total of 29 chemicals assessed compared to 7 mentioned in the proposal, and furthermore most of these chemicals (PFCs, Phthalates, bisphenol A, cotinine and vitamin D) were measured on all 1441 women from the cohort with samples in our biobank instead of the proposed 400 samples due to additional funding from other sources.

For practical reasons we keep data from the previously collected reproductive outcome database from the INUENDO project and the exposure database as separate files and separated by sex, but all files are available in SAS and STATA format and the merging procedure have been checked.

Drafting of papers on the main associations between contaminant exposures and male and female reproductive outcomes were initiated already during the first reporting period. The following manuscripts have been published/drafted as indicated below

Published papers
Toft G, Lenters V, Vermeulen R, Heederik D, Thomsen C, Becher G, Giwercman A, Bizzaro D, Manicardi GC, Spanò M, Rylander L, Pedersen HS, Struciński P, Zviezdai V, Bonde JP. Exposure to polybrominated diphenyl ethers and male reproductive function in Greenland, Poland and Ukraine. Reprod. Toxicol. 43 (2014) 1– 7

Mocevic E, Specht IO, Marott JL, Giwercman A, Jönsson BA, Toft G, Lundh T, Bonde JP. Environmental mercury exposure, semen quality and reproductive hormones in Greenlandic Inuit and European men: a cross-sectional study.
Asian J Androl. 2013 Jan;15(1):97-104.
O Specht I, Spanò M, S Hougaard K, C Manicardi G, Bizzaro D, Toft G, Giwercman A, E Bonde JP. Relationship between apoptotic markers in semen from fertile men
and demographic, hormonal and seminal characteristics. Asian J Androl. 2012, 14(6):890-6
Kvist L, Giwercman YL, Jönsson BA, Lindh CH, Bonde JP, Toft G, Strucinski P,
Pedersen HS, Zvyezday V, Giwercman A. Serum levels of perfluorinated compounds and sperm Y:X chromosome ratio in two European populations and in Inuit from Greenland. Reprod Toxicol. 2012, 34, 644; 650 .
Toft G, Jonsson BAG, Lindh CH, Giwercman A, Spano M, Heederik D, Lenters V Vermeulen R, Rylander L, Pedersen HS, Ludwicki JK, Zviezdai V, Bonde JP. Exposure to perfluorinated compounds and human semen quality in Arctic and European populations. Hum. Reprod. 2012; 27; 2532 -3540.
Specht IO, Hougaard KS, Spanò M, Bizzaro D, Manicardi GC, Lindh CH, Toft G, Jönsson BA, Giwercman A, Bonde JP. Sperm DNA integrity in relation to exposure to environmental perfluoroalkyl substances - A study of spouses of pregnant women in three geographical regions. Reprod Toxicol. 2012 Jul;33(4):577-83.
Lyngsø J, Ramlau-Hansen CH, Høyer BB, Støvring H, Bonde JP, Jönsson BAG, Lindh CH, Pedersen HS, Ludwicki JK, Zviezdai V, Toft G. Menstrual cycle characteristics in fertile women from Greenland, Poland and Ukraine exposed to perfluorinated chemicals. A cross-sectional study. Hum. Reprod. Oct 25[Epub ahead of press].
Manuscripts in draft
Jørgensen KT, Specht IO, Lenters V, Rylander L, Jönsson BA, Giwercman A, Heederik D, Toft GV, Bonde JP. Perfluorinated compounds and time-to-pregnancy in couples from Greenland, Poland and Ukraine.

Specht IO, Toft G, Hougaard KS, Lindh CH, Lenters V, Jönsson BAG, Heederik D, Giwercman A, Bonde JP. Associations between serum phthalates and biomarkers of reproductive function in 589 adult men.

Lenters V, Portengen L, Smit LAM, Jönsson BAG, Giwercman A, Rylander L, Lindh CH, Spanò M, Pedersen HS, Ludwicki JK, Chumak L, Piersma AH, Toft G, Bonde JP, Heederik D, Vermeulen R. Global screening of environmental contaminants and reproductive function in Greenlandic and European men using partial least squares regression.

Significant results
A few noteworthy associations appeared in an overall assessment of 15 exposures and 22 male outcomes of reproductive function assessed by PLSR regression to take into account multiple exposures and outcomes. The most consistent effects across countries included inverse associations between several phthalate metabolites and testosterone. More detailed epidemiological studies of male reproductive function in relation to PFCs, phthalates and metals confirms the overall tendency of some associations between phthalate exposure and male reproductive hormones, but also few consistent associations to other outcomes including a negative association between PFOS and sperm morphology, but no consistent association between PFCs or metals and other male reproductive outcomes.

Associations between PFCs and female reproductive outcomes including time to pregnancy (TTP) and menstrual cycle characteristics suggested increased TTP at the highest PFOS exposure and increased risk of long menstrual cycles at high PFOA exposure.



WP5 Growth and development


During the course of the project data on childhood growth and development were collected from 573 children from Greenland, 156 children from Poland and 502 children from Ukraine.

All participants filled in infromed consent and the data collection was approved by the local ethical comittees.

Data evaluations on childhood growth and development were mainly performed at AUH in collaboration with partner 8 for advanced statistical analyses and with partners 4,5 and 6 for evaluation of cohort specific issues related to data interpretation.

The following tasks for WP5 during the third reporting period included drafting of papers on obesity, growth, developmental milestones and ADHD like symptoms in relation to intrauterine exposures (D5.5-5.8)

The following manuscripts have been drafted/submitted for publication.

Høyer BB, Ramlau-Hansen CH, Pedersen HS, Goralczyk K, Chumak L, Jönsson BAG, Bonde JP, Toft G. Motor Development Following inUtero Exposure to p,p´-DDE and CB-153: a follow-up study of children aged 6-9 years in Greenland, Ukraine and Poland.

Høyer BB, Ramlau-Hansen CH, Henriksen TB, Pedersen HS, Góralczyk K, Zviezdai V, Jönsson BAG, Heederik D, Lenters V, Vermeulen R, Bonde JP, Toft G. Body mass index in young school age children in relation to organochlorine compounds in early life: a prospective study.

Høyer BB, Ramlau-Hansen CH, Pedersen HS, Goralczyk K, Zviezdai V, Jönsson BAG, Bonde JP, Toft G. Pregnancy concentrations of perfluorinated compounds and behaviour and motor development at age 5 to 9 years.

Christensen LH, Høyer BB, Pedersen HS, Zviezdai V, Jönsson BAG , Bonde JP , Toft G. Prenatal smoking exposure, measured as maternal serum cotinine, and children’s motor development: a follow-up study


The results indicated that organochlorine or PFC exposure was not significantly associated to child motor development but smoking measured as cotinine exposure was associated to impaired motor development among the oldest children in Greenland. In addition, indications of lower BMI at high p,p’-DDE exposure in Ukraine was found. Regarding child behaviour, our data indicated that higher PFC exposure may be related to behavioural problems measured by the SDQ in Greenland and Ukraine.


WP6 Gene-environment interactions and epigenetics

The main aim of the gene environment interaction studies in the CLEAR project was:
To assess – in vivo - the impact of polymorphisms in strategic genes related to androgen, estrogen and aryl hydrocarbon pathway on the adverse effect of xenobiotics on reproductive function.
Corresponding to following deliverables:
D6.2 Database with results from analysis of genetic polymorphisms.

D6.4 Article evaluating statistical interaction between genotypes and the effect of contaminant exposure on reproductive outcomes.

In summary:
- Following genetic polymorphisms have been determined and included in the CLEAR database:
* Androgen receptor CAG and GGN repeats;
* Arylhydrocarbon receptor (AhR) and AhR repressor (AhRR);
* 5 alpha reductase type II
* Estrogen receptor 1 and 2
- Both deliverables has been submitted. In summary, an interaction between the androgen receptor CAG repeat length and phthalate exposure in relation to serum testosterone levels was found. Furthermore, an interaction between exposure to persistent organohalogen pollutants and the AhR genotype as well as the AhRR genotype, in relation to Inhibin B levels and sperm DNA Fragmentation Index, respectively, was seen.
These findings are reported in two original papers being currently under preparation:
Brokken L.J.S. Lenters L., Rylander L., Spanò M, Pedersen H.S. Ludwicki J.K. Zviezdai V., Jönsson B.A. Heederik D., Bonde J.P. Toft G., Lundberg Giwercman Y., Giwercman A. Androgen receptor glutamate repeat length modifies the effect of phthalate exposure on serum testosterone levels in Greenlandic Inuit. Under preparation.
Brokken L.J.S. , Spanò M, Lundberg Giwercman Y., Lenters L., Rylander L., Pedersen H.S. Ludwicki J.K. Zviezdai V., Jönsson B.A. Heederik D., Bonde J.P. Toft G., Giwercman A. Polymorphisms in aryl hydrocarbon receptor related genes affect the relationship between DNA fragmentation index and Inhibin B and exposure to environmental pollutants. Under preparation.
and also in a review paper:
Brokken L.J.S. and Lundberg Giwercman Y. Gene-environment interactions in male reproductive health: special reference to the aryl hydrocarbon receptor signalling pathway. Asian Journal of Andrology, In press.
As additional task outside the project proposal, Anti-Müllerian hormone (AMH) analysis has been performed on serum from almost 600 females from Ukraine, Poland and Greenland from whom exposure data and information on time to pregnancy is available. The following manuscript is under consideration for publication:

Bungum L, Bungum M, Toft G, Axmon A, Bonde JP, Pedersen HS, Ludwicki JK, Zviezdai V, Spano M and Giwercman A. Anti Müllerian Hormone and time to pregnancy in a fertile population.


DNA methylation studies.
The main aim of the DNA methylation studies in the CLEAR project was:
“to measure global genomic methylation in 200 human semen samples”.

A total of 269 individuals, 75 from Greenland, 97 from Warsaw, and 97 from Kharkiv was successfully analyzed by an improved flow cytometric (FCM) method to measure the sperm DNA global methylation level (DGML). These results were integrated in an annotated database of the results of sperm DGML measurements of coded samples in terms of arbitrary units of fluorescence. We used a reference sample to evaluate the overall methodology performances. The variability of the methodology, expressed by the coefficient of variation (CV) of the several measurements carried out on this reference sample for every FCM session needed to measure all the samples, resulted <15%. Measuring the sperm DGML is a potentially important test to screen deleterious agents, which are not specific for any particular DNA sequence. However, our approach, like all the other global methylation assays as well, can only define an average genomic methylation level, so that subtle changes involving in specific regions could escape and remain undetected. Aware of the limitations and drawbacks of this approach, for a large subset of samples we have integrated the FCM DGML data with the results of a more specific genome-wide assessment of the sperm DGML level. The strategy was to measure the methylation level of some representative genome-wide DNA repetitive sequences, an approach already described to evaluate the DGML in human peripheral blood lymphocytes (but not in sperm cells) and already recently deployed in epidemiological surveys. In this way the information is “global”, in the sense that it is representative of the whole genome being these repetitive motifs scattered throughout the genome, and, at the same time, is “positional”, as it is derived from some specific sequences for which the average methylation status is known. In summary, we have taken into consideration the methylation level of interspersed and repetitive transposonic sequences either long, such as 6 kb long LINE-1 (Long Interspersed Nucleotide Element-1), or short, such as the 300 bp long Alu, which is a member of the SINEs (Short Interspersed Nuclear Elements). Additionally, we have also considered the methylation status of the 171 bp long non-transposonic interspersed repetitive sequence Satbelonging to the tandem repeats Satellite family. It is estimated that almost 0.5x106 LINE-1 elements and ~1.4 million Alu repetitive elements are scattered in the whole human genome and are generally highly methylated. In this way, we have covered more than 30% of the whole genome and the methylation level of these sequences can be considered a good proxy of the DGML. These activities have been carried out in collaboration with Dr. A. Budillon (Dept. Experimental Oncology, National Cancer Institute - G. Pascale, Naples, Italy) by pyrosequencing methods after treatment of the DNA extracted from the sperm nuclei with sodium bisulfite.

As far as LINE-1 analysis is concerned, we measured 224 samples, 90 from Greenland, 65 from Warsaw (Poland) and 69 from the Kharkiv district (Ukraine). For the Satα analysis, we collected data from 213 samples, 81 from Greenland, 63 from Warsaw and 69 from the Kharkiv district. Alu methylation was evaluated on 215 samples, 84 from Greenland, 64 from Warsaw and 67 from the Kharkiv district. As far the FCM DMGL is concerned, two investigators, blind to all the information regarding the samples except their code, performed the FCM analyses. 47 out of 316 samples could not be measured because of an insufficient cell number at the end of manipulation procedures. At the end, as already reported, we measured 269 samples, 75 from Greenland, 97 from Warsaw (Poland) and 97 from the Kharkiv district (Ukraine).

In our opinion, this integrated approach, the first being attempted on a relatively large scale on human sperm from the general population, could provide a better description of the sperm DGML thus increasing the chances to pick up environmental epigenetic effects, if any, on the human male reproductive system.

As determinants of the methylation level in young healthy fertile individuals are still unknown, as a first step, in order to investigate and clarify the role of potential confounding factors on the sperm global methylation level, we studied the impact of demographic and lifestyle factors, and of hormonal and seminal correlates in our population of fertile men from the 3 geographic areas (Greenland, Ukraine, Poland). General linear regression models were used to evaluate associations between Alu, LINE-1 and Satα methylation levels (expressed as %5-mC) together with the average fluorescence intensity from FCM measurements (expressed as channel number) and subjects’ characteristics, i.e. country (Greenland, Warsaw, Kharkiv), age, body mass index (BMI kg/m2, continuous), smoking (blood cotinine level), alcohol drinking (drinks/week) and abstinence time (ln transformed abstinence time, in days). We also assessed the associations between methylation level and semen quality parameters, i.e. sperm concentration (ln transformed millions/mL), normal morphology (% normal sperm), motility (% immotile sperm) and sperm chromatin structure parameters assessed by the SCSA, i.e. DNA Fragmentation Index (%DFI) and High DNA Stainable (%HDS) cells, as well as associations between seminal biochemical parameters of epididymal and sex accessory gland function, i.e. neutral α-glucosidase (NAG), fructose, zinc, and prostate specific antigen (PSA); and reproductive hormone concentration, i.e. testosterone, sex hormone binding globulin (SHBG), estradiol, follicle-stimulating hormone (FSH), luteinizing hormone (LH), and inhibin B plasma levels.

As far as the FCM DGML is concerned, the average values (expressed as the channel number of the fluorescence intensity distribution ± SE) were 251.1±7.6 for all the study population, the highest value was observed in Kharkiv and the lowest was reported for the Greenland Inuits. The results were not significantly different across the 3 geographical areas.

As far as the methylation levels obtained after pyrosequencing (expressed as %5-mC ± SE) are concerned, for LINE-1 we got an overall mean of 77.4±0.4%, the highest value reported in Greenland, followed by Kharkiv and Warsaw. Greenland, Kharkiv and Warsaw data were statistically different from each other with a LINE-1 methylation 3.3 (CI 1.6 to 5.1)% higher in Greenland compared to Warsaw and 2.3 (CI 0.4 to 4.3)% higher in Kharkiv compared to Warsaw. For Alu, the overall mean was 22.6±0.3%, the highest value observed in Kharkiv followed by Warsaw, and the lowest in Greenland. The percentage of Alu methylation in samples from Greenland was statistically significantly 2.7 (CI 1.2 to 4.2)% lower than for Warsaw, whereas the two European towns were not diverse. For Satα, an overall mean of 46.6±0.9% methylation was obtained, the highest value reported in Greenland, 6.9 (CI 2.1 to 11.7)% higher than Warsaw. Levels in Warsaw and Kharkiv were similar. Thus, consistent statistically significant differences among the 3 countries were observed for Satα, Alu and LINE-1 methylation levels, but not for the FCM DGML parameter. Satα is negatively associated with Alu (r=-0.22 p=0.002) and positively with LINE-1 (r=0.29 p<0.0001). LINE-1 and Alu showed a non-significant inverse association. FCM DGML showed no associations with the methylation level of all the other repetitive sequences.

Associations with POPs (PCB and DDE)

Alu and LINE-1 methylation level was not associated to POPs body burden either considering the population as a whole or stratified into each single country. Satmethylation level was associated only with DDE in Warsaw. Finally, FCM DMGL resulted negatively associated with PCB-153 and DDE in Kharkiv and when populations were combined. Therefore, these results appeared not consistent either for sperm DGML assays or across populations.

Associations with Perfluorinated compounds (PFCs)

As far as PFOS and sperm DGML end points are concerned, when considering the populations as a whole, no changes in any of the methylation parameters as a function of PFOS exposure were observed. A negative association emerged between PFOS and FCM DGML in Warsaw where, for each unit increase in ln transformed PFOS concentration, a decrease of 108 FCM DMGL units (95% CI -192 to -25) was observed. In Kharkiv a positive association with Satα was detected (each unit increase in the ln transformed PFOS units was associated with an increase in Satα of 8 percentage points).

As far as PFOA and sperm DGML end points are concerned, also in this case, no changes in all the methylation parameters were detected as a function of PFOA exposure when considering the populations as a whole. A positive association with LINE-1 was detected in Kharkiv where, for each unit increase in ln transformed PFOA concentration, LINE-1 increased by 2.6 percentage points. No other statistically significant associations were observed.

Considering PFHxS and sperm DGML end points, as observed for the previous pollutants, no changes in all the methylation parameters were detected as a function of exposure for PFHxS when considering the populations as a whole. A sporadic negative association emerged with FCM DGML in Greenland only, where for each unit increase in ln transformed PFHxS concentration, FCM DMGL decreased by 75 units. No other associations were observed.

Finally, considering PFNA and sperm DGML end points, a negative association with FCM DGML was observed when considering the whole population with a FCM DGLM decrease of 39 units (95% CI -73 to -5) for each unit increase in ln transformed PFNA concentration. But, when analysing the populations separately, this association remained significant in Warsaw only.

The combined analysis across populations showed several statistically significant associations in the crude analysis that were not confirmed in the adjusted multivariate linear regression analysis, in particular, all the significant associations emerging from the combined populations.

Associations with the metals Hg, Cd and Pb
The only statistically significant association emerged between LINE-1 methylation level and Cd concentration in Greenland and in the whole population. The sign of the association is positive, with an increase of 1.08 units (in the populations combined) and 1.01 units (in Greenland) for each unit increase in ln transformed Cd concentration. No other associations could be observed.

Associations with Bisphenol A
The only statistically significant association was a positive association with Sat methylation level in Greenland (Sat increase by 1.6 units for each unit increase in ln transformed Bisphenol A concentration). No other associations could be observed.

Finally, a preliminary inspection of the crude associations between the sperm DMGL end points and blood concentration of measured phthalates metabolites was informative of the lack of consistent associations and it was decided not to enter into more detailed analyses.

This study represents one of the first and largest efforts ever attempted using sperm DGML proxies in a reproductive epidemiological study on the effects of environmental pollutants on male reproduction integrity. Firstly, we found that the only main determinant of human sperm DGML was the geographical location (i.e. Greenland vs. European towns). This was clearly evident for the methylation assessment carried out using the genome-wide repetitive elements, either transposonic or not, Alu, LINE-1 and Satα. Generally, information available for each participant at the individual level could not clearly explain these differences. We speculate that unexplored environmental exposures, gene-environment interactions, or other unmeasured lifestyle or personal factors may play a role behind the observed differences between Inuit and European men.

PUBLICATIONS

Consales C, Leter G, Bonde JPE, Toft G, Eleuteri P, Moccia T, Budillon A, Jönsson BAG, Giwercman A, Pedersen HS, Ludwicki JK, Zviezdai V, Heederik D, Spanò M. Determinants of global methylation levels in sperm DNA of fertile men from three different geographical regions. Submitted to Human Reproduction, September 2013

Leter G, Consales C, Eleuteri P, Uccelli R, Specht IO, Toft G, Moccia T, Budillon A, Jönsson BAG, Lindh CH, Giwercman A, Pedersen HS, Ludwicki JK, Zviezdai V, Heederik D, Bonde JPE, Spanò M. Exposure to perfluoroalkylsubstances (PFASs) and sperm DNA global methylation in Arctic and European populations. Submitted to Environmental and Molecular Mutagenesis, September 2013



WP7 Causal inference, risk assessment and dissemination
The primary objective of Work Package 7 was to perform a risk assessment of the potential impact of environmental contaminants on reproductive health. This risk assessment was to consider how climate change will influence future exposure profiles, and how forecasted environmental contaminant levels will impact reproductive health.
The human health risk assessment framework involves the following steps: 1) hazard identification, 2) hazard characterization, 3) exposure assessment, and 4) risk characterization (NRC, 1983, 2009). How Partner 8 (Utrecht University) addressed these risk assessment steps is described herein.

Hazard identification and hazard characterization
We sought to identify which environmental contaminants had an effect on reproductive health outcomes (hazard identification), and to ascertain the magnitude of the effect (exposure-response or hazard characterization). This investigation was restricted to the set of environmental contaminants which were measured in the biological samples (refer to Work Package 3). These represent high priority legacy and emerging pollutants (European Commission, 2000), which were or are high production volume chemicals to which virtually all humans are exposed, and are suspected to interfere with the complex endocrine signalling pathways and homeostasis; also known as endocrine-disrupting chemicals.
We faced several methodological challenges in testing the significance of associations between exposures and outcomes. First, the high number of comparisons (hundreds), which would lead to a large number of false positives if the significance threshold, typically set at p-value<0.05 was not adjusted for multiple testing. We therefore applied relatively established multiple testing strategies, such as controlling the false discovery rate. Second, the exposure data exhibited a complex correlation structure (Figure 1). Conventional ordinary least squares regression models break down with multiple correlated independent variables, termed multicollinearity, producing inflated variance estimates and sometimes inverted coefficient estimates. We therefore decided to use statistical tools which have more commonly been applied in genomics and metabolomics, and which can accommodate correlated variables: dimension reduction and variable selection approaches. It has recently been advocated that these statistical approaches are appropriate and useful for the analysis of environmental epidemiological data (Chadeau-Hyam et al., 2013). The recent emergence of the exposome concept—the totality of an individual’s exposures over at lifetime (Wild, 2005, 2012)—and environment-wide associations studies (Patel et al., 2010) require that statistical tools evolve to match the increasing quantity and complexity of available data.

To provide further theoretical support for the application of these statistical modelling approaches, we conduced simulation exercises, evaluating the sensitivity, specificity and false discovery rate of a dozen approaches. The number of exposures, the magnitude of effect estimates, the total number of observations (subjects), and the correlations between exposures were varied across simulation runs. A multivariate dimension reduction method, partial least squares regression, and a penalisation method, elastic net regression, both proved to efficiently identify true positives in datasets with a similar structure as the CLEAR data: with <100 exposures and moderate correlations between exposures (Lenters et al. unpublished).
Thus, we employed a systematic and data-driven approach, and made use of advanced bioinformatics methods to identify the most statistically robust exposure-response relationships. In assessing the 440 exposure-response relationships between the 20 biomarkers of exposure (phthalates, perfluorinated compounds, metals and organochlorines) and the 22 biomarkers of male reproductive function, we identified 12 significant associations (Figure 2). Four of these associations were significant in both the univariate (single-exposure) and multivariate (multiple-exposure) patial least squares (PLS) models, and were consistent in direction across the three study populations: namely, positive associations between organochlorines and sex hormone-binding globulin; and negative associations between phthalate metabolites and testosterone and neutral α-glucosidase, a marker of epididymal function (Figure 3).

In separate analysis of 1310 mother-child pairs, conducted in collaboration with Partner 2, DEHP, PFOA, PCB-153 and p,p′-DDE were significantly associated with decreased birth weight among full-term infants (born ≥37 weeks gestation) in single-exposure models. In penalised elastic net regression models, the associations for PFOA and p,p′-DDE remained the most robust (Figure 3).

Exposure assessment
Cross-sectional blood samples were collected from the adult CLEAR participants at baseline. This allowed for characterization of the internal levels, at or close to the target organs; representing the most accurate method on the hierarchy of exposure assessment methods (Baker and Nieuwenhuijsen, 2008). Partner 2 (Lund University) analysed the concentrations of 15 to 20 environmental contaminants in biobanked serum and whole blood samples from the adult male (n~600) and female (n~1400) participants of reproductive age [levels presented in Appendix A; refer also to Work Package 3 and Lindh et al., 2012].
In addition, brominated flame retardants, namely six polybrominated diphenyl ethers (PBDEs) and one polybrominated biphenyl (BB-153), were measured in 300 male serum samples. These chemical analyses were subcontracted to the Division of Environmental Medicine at the Norwegian Institute of Public Health by Partner 8. In a publication (Lenters et al., 2013), the levels and determinants of flame retardants were described. At least one flame retardant (PBDE or PBB) was detected in 298 of the 300 participants’ serum samples. Concentrations of individual congeners were 2.7 to 15 fold higher in Greenlandic relative to Polish and Ukrainian men. Within Greenland, concentrations were higher in men from the South-East than in men from the West coast (including Nuuk). It is unclear what proportion of Greenlanders’ exposure is due to emissions from imported BFR-impregnated products and their disposal via burning, relative to consumption of contaminated fish and high trophic level animals (e.g. seals and polar bears). In Figure 4, we compare levels of the most abundant BDEs (47 and 153) and the sum of the most abundant PBDEs of the Penta-BDE commercial mixture (Σtri–heptaPBDE) observed in the current study with levels reported for other populations (n > 50 individual samples), not occupationally exposed or exposed to site-specific sources (e.g. living nearby an incinerator). The concentrations in Greenlandic men were higher than in European and Asian populations, but generally lower than those reported for North American populations. Relatively high PBDE concentrations in US populations have been attributed to more stringent furniture flammability standards (Frederiksen et al., 2009).

To our knowledge, this is the first description of blood levels of brominated flame retardants in the Greenlandic, Polish or Ukrainian populations. However, almost simultaneously the paper on PBDEs in cord blood of Polish women has been published in 2013 (Hernik et al., 2013).
In addition, Partner 8 used a physiologically-based pharmacokinetic or toxicokinetic (PBPK/PBTK) model to more accurately characterize the post-natal exposure of the CLEAR children. A PBTK model mathematically accounts for absorption, distribution, metabolism and excretion of compounds within an individual. The toxicokinetic model (see Figure 5) was recently developed and has been validated in similar study populations—Inuit from Canada and a Slovak cohort—and the predictive ability of the toxicokinetic model was moderate to high. The R2 of measured child blood levels versus estimated levels was 0.81-0.83 for the Inuit cohort (n=156) and 0.44-0.59 for the Slovak cohort (n=754) for both PCB-153 and p,p’-DDE (Verner et al., 2013). Duration of exclusive breastfeeding, birth weight and maternal blood levels were the most important predictors of child exposure.

For the CLEAR project, Partner 8 modelled the cumulative exposure to age 12 months to PCB-153 and p,p’-DDE. We focused on modelling these contaminants because they have well-characterized half-lives, are lipophilic and bioaccumulate, and were considered top priority contaminants for the epidemiological analyses of child health risks. Mirroring measured maternal levels, estimated child PCB-153 levels were highest in those from Greenland, followed by those from Ukraine and Poland. p,p’-DDE levels were more similar across the three cohorts. The exposure rankings, which ultimately influence epidemiological assessments of exposure-response relationships, changed substantially: the correlation between measured maternal levels and the estimated child cumulative exposures was 0.76 for PCB-153 and 0.61 for p,p’-DDE. These PBTK modelled exposure estimates represent more accurate estimates of exposure than simply relying on the measured maternal serum levels, and were incorporated in the epidemiologic study on the effects of organochlorines on child body mass index (Work Package 5; Høyer et al. unpublished).

Risk characterization
In this ultimate step of risk assessment, the probability and magnitude of the risk is characterized. A qualitative rather than a quantitative risk characterization was performed (Partner 1 and 8), and was summarized in several review papers (Lenters et al. unpublished; Specht et al. unpublished; Vested et al. unpublished). These reviews addressed the statistical robustness of exposure-response associations, and also the clinical relevance of outcomes, biological plausibility, causal inference and potential sources of bias, and the weight of the evidence based on systematic literature reviews.
Based on the extensive modelling conducted by Partner 3 for Work Package 2, it was deemed that the uncertainties surrounding the forecasted estimates of contaminant levels based on future global climate change scenarios were too great to consider this (quantitatively) in a risk assessment; and furthermore, for the Inuit, the influence of dietary transitions from traditional high trophic level foods (e.g. seal) to store-bought foods was a much more important determinant of future exposures than climate change.

Conclusion
Work performed for Work Package 7 helped to advance the science of modelling exposure-response relationships when multiple, correlated exposures (exposure mixtures) are considered. Furthermore, this work identified several statistically robust and mechanistically plausible associations between phthalates, some PFCs and organochlorines and (bio)markers of reproductive health.






References
Baker DB, and Nieuwenhuijsen MJ, eds. Environmental Epidemiology: Study methods and application. Oxford University Press: Oxford; 2008.
Chadeau-Hyam M, Campanella G, Jombart T, Bottolo L, Portengen L, Vineis P, Liquet B, Vermeulen RCH. Deciphering the complex: Methodological overview of statistical models to derive OMICS-based biomarkers. Environ Mol Mutagen 2013;54:542–57.
European Commission. (2000). Towards the establishment of a priority list of substances for further evaluation of their role in endocrine disruption–preparation of a candidate list of substances as a basis for priority setting. Final Report. BKH Consulting Engineers in association with TNO Nutrition and Food Research, produced for European Commission DG ENV (available from http://europa. eu. int/comm/environment/docum/01262_en. htm# bkh).
Frederiksen M, Vorkamp K, Thomsen M, Knudsen LE. Human internal and external exposure to PBDEs--a review of levels and sources. Int J Hyg Environ Health 2009;212:109–34.
HernikA.: Góralczyk K., Struciński P., Czaja K., Korcz W., Minorczyk M., Ludwicvki J.K.. Chemosphere 2013;93: 526-531.
Høyer BB, Ramlau-Hansen CH, Henriksen HS, Pedersen HS, et al. Body mass index in young school age children in relation to organochlorine compounds in early life: a prospective study. Unpublished.
Lenters V, Thomsen C, Smit LAM, Jönsson BAG, Pedersen HS, Ludwicki JK, et al. Serum concentrations of polybrominated diphenyl ethers (PBDEs) and a polybrominated biphenyl (PBB) in men from Greenland, Poland and Ukraine. Environ Int 2013;61:8–16.
Lenters V, Portengen L, Smit LAM, Jönsson BAG, Giwercman A, Rylander L, Lindh CH, Spanò M, Pedersen HS, Ludwicki JK, Chumak L, Piersma AH, Toft G, Bonde JP, Heederik D, Vermeulen R. Global screening of environmental contaminants and reproductive function in Greenlandic and European men using partial least squares regression. Unpublished.
Lenters V, Portengen L, Rignell-Hydbom A, Jönsson BAG, Lindh CH, Piersma AH, Toft G, Bonde JP, Heederik D, Rylander L, Vermeulen R. Sparse penalized regression of prenatal exposure to phthalates, perfluorinated compounds and organochlorines and birth outcomes. Unpublished.
Lenters V, Vermeulen R, Portengen L, et al. Performance of dimension reduction and variable selection methods for exposure-response analysis of highly correlated exposures. Unpublished.
Lindh CH, Rylander L, Toft G, Axmon A, Rignell-Hydbom A, Giwercman A, et al. Blood serum concentrations of perfluorinated compounds in men from Greenlandic Inuit and European populations. Chemosphere 2012;88:1269–75.
NRC (National Research Council). Risk assessment in the federal government: Managing the process. National Academies Press: Washington, DC; 1983.
NRC (National Research Council). Science and decisions: Advancing risk assessment. National Academies Press: Washington, DC; 2009.
Patel CJ, Bhattacharya J, Butte AJ. An Environment-Wide Association Study (EWAS) on type 2 diabetes mellitus. PLoS One 2010;5:e10746.
Verner M-A, Sonneborn D, Lancz K, Muckle G, Ayotte P, Dewailly É, et al. Toxicokinetic modeling of persistent organic pollutant levels in blood from birth to 45 months of age in longitudinal birth cohort studies. Environ Health Perspect 2013;121:131–7.
Wild CP. Complementing the genome with an “exposome”: the outstanding challenge of environmental exposure measurement in molecular epidemiology. Cancer Epidemiol Biomarkers Prev 2005;14:1847–50.
Wild CP. The exposome: from concept to utility. Int J Epidemiol 2012;41:24–32.





Potential Impact:

CLEAR dissemination strategy

Website
A website with information about the project including a descriptions of the project in layman terms, and contact addresses for the partners in the project has been established: www.inuendo.dk/clear.
On the public part of the website we will include news about the project including newsletters published every 6-12 months.

A list of the scientific publication arising from the project will be made available including links to PubMed or other online sources for abstract/full paper when these are available.

On a password protected part of the website, we will put information about the project only relevant to members of the group including proposal, budget, agendas for upcoming meetings and meeting reports. Furthermore, unpublished paper drafts and power point presentations on results from the project may be uploaded on this site to keep the partners informed about the latest news in the project.

Finally templates for scientific and financial reporting will be uploaded, for easy access for all partners to these documents.

Brochure
A brochure describing the project and with contact information on partners was printed within the first 6 moths of the project after approval from all partners. The brochure was distributed to partners, colleagues and stakeholders by direct mailing to a list of these and by distribution at conferences where the project was presented during the lifetime of the project. Furthermore, an electronic version of the brochure will be made available on the website.

Newsletters
In order to inform the partners and other persons interested in the project about the progress of the project and new results – a newsletter was prepared every 6-12 months and distributed to a mailing list, open for anyone interested in the project. Furthermore the newsletters have been made available on the public part of the website.

Scientific publications and presentations
The results of the project have been and will be disseminated to the relevant parts of the scientific community at conferences and in peer reviewed scientific papers.
A list drafts/manuscripts in preparation including expected date of submission is attached in addition to the list of published papers in the next section. Several papers especially on the modelling of contaminant release will be made in collaboration with persons from the ArcRisk group, in order to utilise the expertise from both of these groups and ensure that the results will not overlap.
In all publications and presentations it has been mentioned in the acknowledgement that:
The study was part of the CLEAR Project: Climate, Environmental Contaminants and Reproductive Health (www.inuendo.dk/clear) supported by The European Commission 7th Framework Programme FP7-ENV-2008-1 Environment (including Climate Change) Grant no.: 226217. Furthermore, posters and PowerPoint presentations included the European emblem.


Press releases are provided to attract media and public attention on important new findings from the project after these have been published in peer reviewed scientific journals. Specifically, we submitted a press release in connection with the final meeting to increase public awareness about our project.


Summary report
In the end of the project a summary report was prepared. This report will be made available to national and international regulatory bodies in order to provide documentation necessary for implementing new regulations or changing existing regulations in the areas of climate change and reproductive toxicology.


International conference
The consortium organized an international conference in late 2013 that updated, reviewed and as far as possible establish international scientific consensus with respect to climate change, contaminant exposure and human health. The conference included participants from national regulatory bodies and other European research groups.


Future dissemination activities
The project has ended by October 31, 2013. We are still working on publications from the project and about 22 manuscripts are presently in draft or in preparation and expected to be published during 2014.
In addition upcoming results from the project will be presented at relevant international scientific meetings and meetings with other stakeholders and policy makers on future research and regulatory initiatives.

A table of future publications submitted or planned is attached



Contact details

Coordinator
Aarhus University Hospital (AUH-AS)

Associate Professor, Gunnar Toft
Department of Occupational Medicine
Aarhus University Hospital
Norrebrogade 44, build 2C
8000 Aarhus C
Denmark
Phone: [phone]
Fax: [phone]
E-mail: [e-mail]

Professor, Jens Peter Bonde
Department of Occupational and Environmental Medicine
Bispebjerg Hospital
Phone: [phone]
Fax: [phone]
E-mail: [e-mail]

Research secretary, Hanne Tulinius
Department of Occupational and Environmental Medicine
Bispebjerg Hospital
Bispebjerg Bakke
2400 København NV
Phone: [phone]
Fax: [phone]
E-mail: [e-mail]

Partners
Lund University (ULUND)
Professor Bo Jönsson
Department of Occupational and Environmental Medicine
Lund University Hospital
Sweden
E-mail:[e-mail]
Professor Aleksander Giwercman
Reproductive Medicine Center
Malmö University Hospital
Sweden
E-mail: [e-mail]


Agenzia nazionale per le nuove tecnologie, l’Energia e lo sviluppo economico sostenibile (ENEA)
Dr. Marcello Spano
Section of Toxicology and Biomedical Sciences
ENEA CR Casaccia
Via Anguillarese 301
I – 00123 Rome
Italy
Tel: [phone]
Fax: [phone]
E-mail: [e-mail]

National Institute of Public Health – National institute of Hygiene (PZH)
Professor Jan K. Ludwicki
Department of Environmental Toxicology
National Institute of Hygiene
00-791 Warsaw
P-Chocimska 24
Poland
Phone: [phone]
Fax: [phone]
e-mail: [e-mail]

Kharkiv National Medical University (KhNMU)
Professor Valentyna Zviezdai
Kharkiv National Medical University
Lenin Avenue 4,
Kharkiv
61022 Ukraine
Phone: [phone]
Fax: [phone]
e-mail: [e-mail]

Greenland Institute of Natural Resources (NI-GR)
Dr. Henning Sloth Pedersen
Centre for Arctic Environmental Medicine
Pinngortitaleriffik - Grønlands Naturinstitut
Postbox 570
3900 Nuuk Greenland
Phone: [phone]
Fax:[phone]
e-mail: [e-mail]

Universitet Utrecht (UU)
Professor Dick Heederik
Institute for Risk Assessment Sciences
E-mail:[e-mail]
Professor Aldert H. Piersma
National Institute for Public Health and the Environment RIVM
Phone [phone]
Fax [phone]
E-mail: [e-mail]

Governing Council of the University of Toronto (UTS)
Associate professor Frank Wania
University of Toronto Scarborough,
Department of Physical and Environmental Sciences,
1265 Military Trail
Toronto, ON, Canada M1C 1A4
Phone. [phone]
Fax. [phone]
E-mail: [e-mail]