1 23

Environmental Science and PollutionResearch ISSN 0944-1344 Environ Sci Pollut ResDOI 10.1007/s11356-012-1320-3

BIOAMBIENT.ES study protocol: rationaleand design of a cross-sectional humanbiomonitoring survey in Spain

Beatriz Pérez-Gómez, Roberto Pastor-Barriuso, Marta Cervantes-Amat, MartaEsteban, Montserrat Ruiz-Moraga,Nuria Aragonés, et al.

1 23

Your article is protected by copyright and

all rights are held exclusively by Springer-

Verlag Berlin Heidelberg. This e-offprint is

for personal use only and shall not be self-

archived in electronic repositories. If you

wish to self-archive your work, please use the

accepted author’s version for posting to your

own website or your institution’s repository.

You may further deposit the accepted author’s

version on a funder’s repository at a funder’s

request, provided it is not made publicly

available until 12 months after publication.

SHORT RESEARCH AND DISCUSSION ARTICLE

BIOAMBIENT.ES study protocol: rationale and designof a cross-sectional human biomonitoring surveyin Spain

Beatriz Pérez-Gómez & Roberto Pastor-Barriuso & Marta Cervantes-Amat &Marta Esteban & Montserrat Ruiz-Moraga & Nuria Aragonés & Marina Pollán &

Carmen Navarro & Eva Calvo & Javier Román & Gonzalo López-Abente &

Argelia Castaño & on behalf of BIOAMBIENT.ES

Received: 7 August 2012 /Accepted: 9 November 2012# Springer-Verlag Berlin Heidelberg 2012

Abstract Human biomonitoring (HBM), defined as the mea-surement of concentrations of chemicals or of their metabo-lites in human biological matrices, is considered the method ofchoice for determining internal exposures in individuals.HBM is part of environmental exposure surveillance systemsin several countries. In 2007, the Ministry of Agriculture,Food and Environment promoted BIOAMBIENT.ES project,a national-level HBM study on environmental pollutants car-ried out in Spain to estimate levels of heavy metals, persistentorganic pollutants, and other substances on the Spanish activeworkforce. BIOAMBIENT.ES is a nationwide cross-sectional

study, with a stratified cluster sampling designed to cover allgeographical areas, sex and occupational sectors, and aimed toobtain a representative sample of the Spanish workforce. Par-ticipants were recruited among people residing in Spain for5 years or more, which underwent their annual occupationalmedical check-up in the health facilities of the Societies forPrevention of IBERMUTUAMUR, MUTUALIA, MC-PREVENCIÓN, MUGATRA, UNIMAT PREVENCIÓN,and PREVIMAC (March 2009–July 2010). A total of 1,892subjects fulfilled the criteria for inclusion, donated biologicalsamples (1,880 blood, 1,770 urine, and 577 hair) and complet-ed a short self-administrated epidemiological questionnaire onenvironmental and lifestyle-related exposures. Additionally,clinical information from participant’s health exams wasobtained. This project will provide a first overview of thebody burden of selected pollutants in a representative sampleof the Spanish-occupied population. This information will beuseful to establish reference values of the studied populationand, eventually, to evaluate temporal trends and the effec-tiveness of environmental and health policies.

Keywords Human biomonitoring . Pollutants . Spain .

Heavymetals . Study protocol . Biomarkers

Background

Human biomonitoring (HBM) has become the gold standardin pollutant exposure estimations in humans (Centers forDisease Control and Prevention 2012). HBM can be definedas the measurement of concentrations of chemicals or theirmetabolites in human biological media such as blood, urine,hair, or breast milk. This term may also refer to chemicaland biological parameters which allow inferences about the

Responsible editor: Philippe Garrigues

Other members of BIOAMBIENT.ES: Jose A. Jimenez , Ana I. Cañas,Olga Huetos, Mónica Bartolomé,Carolina Rodríguez, Gema Díaz, AnaLópez-Herranz, Juan Mayor, José Manuel Aleixandre, Miquel Vila,Carlos-Félix Méndez, Mª Victoria Cortés, Mª Ángeles Molina.

B. Pérez-Gómez (*) : R. Pastor-Barriuso :M. Cervantes-Amat :N. Aragonés :M. Pollán :G. López-AbenteEnvironmental and Cancer Epidemiology Unit,National Centre for Epidemiology, Carlos III Institute of Health,Monforte de Lemos 5,28029 Madrid, Spaine-mail: bperez@isciii.es

B. Pérez-Gómez :R. Pastor-Barriuso :M. Cervantes-Amat :N. Aragonés :M. Pollán :G. López-AbenteCIBERESP, Carlos III Institute of Health, Madrid, Spain

M. Esteban :C. Navarro :A. CastañoEnvironmental Toxicology Area,National Centre on Environmental Health, Carlos IIIInstitute of Health, Carretera Majadahonda- Pozuelokm. 2( Majadahonda( 28220 Madrid, Spain

M. Ruiz-Moraga : E. Calvo : J. RománIbermutuamur, C/Ramírez de Arellano 27, 4ª planta,28043 Madrid, Spain

Environ Sci Pollut ResDOI 10.1007/s11356-012-1320-3

Author's personal copy

pollutants’ biological effects. HBM allows to estimate thenet amount of these chemicals present in human beings as aresult of exposures from different sources, as well as indi-vidual variations on metabolism due to genetic or lifestyledifferences. HBM is considered the method of choice fordetermining internal exposures in the population, populationgroups, or individuals, thereby supplying a basis for esti-mating health risks and, if necessary, for risk management(Angerer et al. 2007). This approach has classically been anessential component of the surveillance strategies in occu-pational health (Manno et al. 2010). Nowadays, however,HBM has achieved a wider scope and is considered afundamental tool in environmental health research (EuropeanUnion 2005; Paustenbach and Galbraith 2006; Porta et al.2008). Environmental exposures may affect the whole popu-lation during long periods of life and, though these exposuresare generally low—and indeed much lower than those foundin specific work environments—their possible health impactcan be more relevant from a public health perspective. For thisreason, during the last decades, some countries have imple-mented surveillance systems using biomarkers. Their aim is toprovide accurate estimates of the level of exposure to envi-ronmental chemicals in the general population at nationallevel by measuring the concentrations of these chemicals inhuman blood, urine, or hair. These initiatives, which werepioneered by GerES (Seifert et al. 2000), and the NationalHealth and Nutrition Examination Surveys in the USA (Cen-ters for Disease Control and Prevention 2011), have also beencarried out in other parts of Europe (National Institute ofPublic Health 2011; Fréry et al. 2010; Flemish Centre ofExpertise of Environmental Health 2007), as well as in othercountries (Kim and Lee 2010; Health 2010).

In Spain, environmental surveillance has mainly relied onmeasures of selected pollutants in air, water, food, and soil. Inthe last decade, however, several initiatives have tried to incor-porate human samples in the framework of environmentalsurveillance aiming to estimate pollution levels around pointsources (Gonzalez et al. 2000), to study vulnerable groups—children, pregnant women—(Llop et al. 2011; Aragones et al.2008), or to provide reference values of some toxicants for thegeneral population at a regional level: heavy metals in Madrid(140 triads of father–mother–newborn) (Aragones et al. 2008)and Huelva (857 adults (Aguilera et al. 2008) and 227children (Aguilera et al. 2010)); persistent organic pol-lutants (POPs) in Madrid (Aragones et al. 2008), Catalonia/Barcelona (919 adults) (Porta et al. 2010; Porta et al. 2012),and Canary Islands (682 adults) (Zumbado et al. 2005).

In this context, in 2007 the Ministry of Agriculture, Foodand Environment promoted the design of a surveillance systemto enhance current knowledge of the distribution of environ-mental pollutants in Spain.Within this system, they intended toincorporate, among other strategies, a HBM-based assessmentof exposure in the general population. This task was entrusted

to the Environmental Toxicology Area (TA) of the NationalCentre on Environmental Health of the Carlos III Institute ofHealth (ISCIII) (Ministerio de Agricultura Alimentación yMedio Ambiente 2007). To execute this assignment, a collab-orative multidisciplinary research group was established in-cluding TA, the Environmental and Cancer EpidemiologyUnit of the National Centre of Epidemiology (also fromISCIII) and the coordinators of the Prevention Societies ofthe Mutual Insurance Companies that make up CorporationMutua, the biggest Spanish Consort of Mutual InsuranceCompanies (Castaño et al. 2009). The objectives of this groupwere to design and, afterwards, to carry out a national HBMstudy to estimate the levels of selected pollutants in a repre-sentative sample of Spanish adults. In this project, we focusedon the occupied population, which represented in 2009 a 63 %of adult population in Spain. This paper describes the mainaspects of BIOAMBIENT.ES study design and protocol. Atpresent, as the field work has already been completed, we willalso summarize the basic field work results.

Aims and objectives

The aim of this project was to study exposure to heavymetals, cotinine, and POPs in a representative sample ofthe Spanish workforce, providing an estimation of the pres-ence of these biomarkers in blood, serum, urine, and hair aswell as by sex, age, economic sector, and geographical area.

Main features of the study design

Study population and sampling

To achieve the aforementioned objective, we designed a nation-wide cross-sectional epidemiological study. Participants wererecruited among occupied people older than 16 years, residentsin Spain for 5 years or more, which underwent their annualoccupational medical check-up in the health facilities of theSocieties for Prevention of IBERMUTUAMUR,MUTUALIA,MC-PREVENCIÓN, MUGATRA, UNIMAT PREVENCIÓN,and PREVIMAC during March 2009 to July 2010. Participantswere selected through a stratified cluster sampling which cov-ered all geographical areas, sex, and economic sectors, aimed toobtain a representative sample of the Spanish workforce.

In order to guarantee the nationwide representativenessof the sample, 12 geographic zones were defined by joiningclose regions, obtaining the following strata: Northwest I(Galicia), Northwest II (Asturias and Cantabria), North(Basque Country), Northeast I (Navarre, La Rioja, andAragon), Northeast II (Catalonia), Central I (Castile andLeón), Central II (Madrid), Central III (Castile-La Manchaand Extremadura), East (Valencian Community and Balearic

Environ Sci Pollut Res

Author's personal copy

Islands), South I (Andalusia and Ceuta), South II (Murcia),and Canary Islands. A total of 38 Prevention Health Centerswere selected across the above mentioned geographical strata.A minimum of two examination centers were selected withineach geographical area, with the remaining 14 centers beingdistributed proportionally to the working population in eacharea, according to data from the Spanish Active PopulationSurvey 2007 (Instituto Nacional de Estadística 2012).

At a first sampling stage, the corresponding centers withineach geographical area were randomly selected with probabil-ities proportional to their annual number of physical examina-tions, according with the participating Societies for Preventionregisters. Similar numbers of male and female participantswere to be recruited in each sampling unit to be able to providepollutant levels for men and women separately. In addition, inorder to stratify by economical sector, two groups were definedtaking into account the National Classification of EconomicActivities for 2009 (Instituto Nacional de Estadística 2011),which is the national translation of the statistical classificationof economic activities in the European Community (NACERev. 2) (European Commission 2012c). The first group includ-ed “Service activities” and the second “farming, industry, andconstruction activities.” Therefore, in the second samplingstage, each selected examination center had to recruit 25workers in the service sector (10 men and 15 women) and25 workers of other economic sectors—farming, industry, orconstruction—(15 men and 10 women). Also, to ensure aproper seasonal distribution of the sample, the 38 collaborat-ing centers were assigned to four annual periods: January–March, April–June, July–September, and October–December.Participants had to be consecutively selected, beginning at therecruitment period randomly assigned to each center.

As a result of this selection procedure, the overall theo-retical sample was planned to comprise 1,900 participants.Small geographical areas, nonservice economic sectors, andwomen were over-sampled to improve precision estimationin these subgroups.

Biological specimens and selected pollutants

The biological matrices selected were blood, urine, andhair. Biomarkers were decided by a steering committee(MAGRAMA and TA), following priorities included in theNational Implementation Plan for the Stockholm conventionon POPs (Ministerio de Agricultura Alimentación y MedioAmbiente 2012), the EU Mercury strategy (EuropeanCommission 2012b), and the recommendations of the consul-tative forum on Environment and Health established under theScale strategy (European Commission 2012a). The attentionwas focused on those chemical that persist in the environment,bio-accumulate trough the food chain, and of public healthconcern due to their largely recognized toxicity and riskassessments available (i.e., threshold/HBM values) (Ewers et

al. 1999). Final compounds selection included certain heavymetals (mercury, cadmium, and lead), some POPs—poly-chlorinated biphenyl (PCBs), polybrominated diphenyl ether(PBDEs), organochlorine pesticides, and polycyclic aromatichydrocarbons (Table 1). These pollutants have been usuallyincluded in the main biomonitoring studies such as NHANES,GerES, or the French National Biomonitoring Study (Beckeret al. 2002; Centers for Disease Control and Prevention 2011;Fréry et al. 2010). Cotinine has also being selected, as a meansto evaluate exposure to tobacco smoke.

Epidemiological questionnaire and clinical information

A short self-administered epidemiological questionnairewas designed to collect basic individual information onsociodemographic data and living and environmental con-ditions—including data about house heating, source ofdrinking water, and distance of usual residence to mayortraffic ways; specific questions on tobacco active and pas-sive exposure, both at home and in occupational and leisuretime settings; a short food frequency questionnaire to regis-ter usual diet (food groups) and alcohol consumption, withspecial attention to fish consumption; and some questionson recent diseases and use of medicines. The questionnairealso included some specific items on factors associated withthe studied toxicants, such as the presence of dental fillings,recent changes of weight, and breastfeeding history. A clin-ical form was also designed to be filled in by the doctorsperforming the health exams of the included workers togather basic clinical information on all the participants in auniform way. In addition, participants were asked to grant

Table 1 Biological matrices and pollutants analyzed in BIOAMBIENT

Biologicalmatrices

Pollutants analyzed

Urine Cadmium, mercury, lead, cotinine, and metabolitesof PAHs (1-OH-pyrene, (1, 2, 3, 4, and 9)-hydroxyphenanthrenes, and 3-hydroxy(a)benzopyrene)

Blood Cadmium, mercury, and lead

Serum PCBs (PCB 28, PCB 52, PCB 101, PCB 138, PCB153, and PCB 180) and PBDEs (PBDE 85, PBDE99, PBDE 100, PBDE 119, PBDE 126, PBDE196, PBDE 197, PBDE 28, PBDE 47, PBDE 99,PBDE 100, PBDE 153, PBDE 154, PBDE 183,and PBDE 209)

Organochlorine pesticides (2,4′-DDT, 2,4′-DDE,4,4,′-DDT, 4,4′ DDE, HCB, (a, b, and g), HCH,aldrin, dieldrin, endrin, heptachlor, and heptachlorepoxide)

Hair Mercury

PAHs polycyclic aromatic hydrocarbons, PCBs polychlorinated biphenyls,PBDEs polybrominated diphenyls ethers

Environ Sci Pollut Res

Author's personal copy

access to their clinical records in order to allow the researchgroup to have both the complete results of the occupationalhealth exam as well as to obtain the data needed to assesspossible occupational exposures.

Ethical approval

The project has been approved by both the Ethical Com-mittee and the law department of IBERMUTUAMUR.Participants were asked for their written informed consent.Their collaboration was voluntary and altruistic; only asmall token pen drive was given to the participants. BIO-AMBIENT.ES follows legal ethical principles and regula-tions of research involving individual information andbiological samples, including Organic Law 15/1999 onPersonal Data protection and its Regulations, the Law41/2002, on Autonomy of Patients regulation patient auton-omy and rights and obligations relating to health informationand documentation as well as law 14/1986, General ofHealth. Since the study collects biological human samples,the ethical principles of the Declaration of Helsinki will beobserved.

An information leaflet including individual levels of theresults available at this stage for each participant as well as

general information on each toxicant has been designed inorder to provide the volunteers feedback about the research.Additionally, BIOAMBIENT has established a committee toanalyze each result that could be considered as anomalouscase by case, in order to confirm the result and take theappropriate measurements that were due.

Field work results

Center participation

The final location of the centers that collaborated withBIOAMBIENT is depicted in Fig. 1. The overall centerparticipation rate was 86.8 %, as only 5 of the 38 Pre-vention Health Centers initially selected for this cross-sectional study had to be replaced, which belonged tothe following regions: one to the Northeast I, two inCentral III, one in South I, and one in Canary Islands.We replaced these five centers either for logistic limita-tions—such as the availability of sample material (centri-fuge and freezer)—and staff restrictions. Therefore, thesecenters were considered unable to collaborate in the proj-ect, and they were randomly replaced by other centers

La CoruñaLugo

Almería

Zaragoza

León

Talavera de la Reina

Cartagena

LogroñoVigo

Granada

Barcelona

Ceuta

Valencia

Badajoz

Córdoba

Málaga

Santander

Salamanca

Valladolid Granollers

Girona

San Sebastián

Onda

Rubí

Tortosa

Petrer

Espinardo

Pamplona

Guadalajara

Torrejón de Ardoz

Arganda del Rey

Madrid

Erandio

ZamudioOviedo

Alcorcón

Santa Cruz de Tenerife

NORTHWEST I

NORTHWEST II NORTH

NORTHEAST I

NORTHEAST II

CENTRAL I

CENTRAL IIIEAST

SOUTH I

SOUTH II

CANARY ISLANDS

January-MarchApril-JuneJuly-SeptemberOctober-December

Seasonal sample distribution

CENTRAL II

Fig. 1 Geographical distribution of the sampling points and trimester of samples recruitment of the BIOAMBIENT.ES participants

Environ Sci Pollut Res

Author's personal copy

located in the same geographical area which had not beenselected at the initial sampling, with the single exceptionof the center in Canary Islands, where substitution was notpossible; in this case, the center was replaced by one inthe mainland.

Participants’ recruitment, biological sampling, and transport

The personnel in the selected Prevention Health Centersreceived specific training in both the recruitment processand the handling of the corresponding biological samples.Doctors and nurses were in charge of providing informationon the objectives of the project to those attending the occu-pational health exam. The working routine of participantsdid not suffer any change on behalf of the study. Afterchecking that the inclusion criteria were met, they invitedthe participants to collaborate in BIOAMBIENT.ES andcollected the informed consent form. Each recruited subjectwas then given the self-administered epidemiological ques-tionnaire and was asked to donate blood and urine samples,in addition to those obtained for the health exam. Urinecontainers prewashed with 10 % HNO3 and Milli-Q waterwere given to participants to collect first morning urine. Hairsamples were considered as optional. Fasting whole bloodfor metals analysis was collected in special sodium heparintubes for trace element determination (Cañas et al. 2010).

Usually, the blood was centrifuged at the same healthexamination center to obtain the serum; in some centers,however, samples had to be transported to an auxiliarylaboratory to be processed. All biological samples werestored at 4 °C in ad hoc designed isotherm containers, whichwere afterwards shipped, as well as those participants’

corresponding documentation, to the TA laboratories (Estebanet al. 2012).

BIOAMBIENT.ES participants: the final sample

Figure 2 summarizes the main results of BIOAMBIENT.ESfield work. A total of 1,936 people were recruited to partic-ipate in this project. After additional quality check-ups, 44subjects were excluded due to incomplete information onstratification variables or because they finally did not meetall inclusion criteria. Therefore, the main sample, availablefor future analyses in this project, is made up by 1,892subjects. Logistic problems have produced small differencesbetween the initially designed sample and the one finallyobtained, being the most noteworthy the lower number ofparticipants from nonservice economic sectors, which hasattenuated the planned over-sampling in this group. Table 2presents the final distribution of BIOAMBIENT.ES partic-ipants as well as the structure of Spanish working popula-tion in 2009, allowing the comparison between themaccording to the stratification variables. This informationwill be the base of the weights that will be used in futureanalyses. Another feature which was taken into account inthe initial design was seasonality. Figure 1 includes alsothe distribution of sampling in each region by trimester(January–March, April–June, July–September, and October–December); in most of them, the recruitment period has beendistributed along the whole year. This way, the national-levelaverage measures of pollutant levels will not be affected bypossible variations in the concentration of toxicants due to thisfactor.

1936 Subjects recruited

1880 subjects 1770 subjects 577 subjects

Exclusion criteria: - 20 Residence in Spain <5 years Stratification variables:- 6 No data on sex- 18 No data on occupational sector sector

33 Specimen not provided11 Transport/Logistic problems78 <0.3 or >3 g creatinine/L

12 Insufficient volumes 1297 Specimen not provided18 Insufficient samples

Blood Urine Hair

1892 Bioambient participants

Fig. 2 Flowchart ofBIOAMBIENT.ESparticipation

Environ Sci Pollut Res

Author's personal copy

Regarding biological specimens, 100 % of the recruitedparticipants donated blood, although in 12 cases the amountavailable was considered insufficient to perform the analy-ses. Most of the participants also supplied a urine specimen(98.3 %); some of them—44 urine samples—had to bediscarded due to insufficient amount of substrate or becausethey were not in optimal condition due to logistic problems.We also excluded in the analyses 78 urine samples withcreatinine levels below 0.3 or over 3 g creatinine/l urine,following the guidelines of biomonitoring studies recom-mended by the World Health Organization (1996). Theclassification of hair specimen as optional led several cen-ters to ignore it; thus, hair sample was provided only by a31.4 % of the participants. In 19 of them, the hair obtainedwas also insufficient to carry out any chemical determina-tion. In summary, a total of 1,880 blood samples, 1,770urine samples, and 577 hair samples are available for thelaboratory analyses. The main characteristics of BIOAM-BIENT.ES participants are shown in Table 3, both globallyand by biological matrix available.

The transport of samples was organized in each preven-tion health centre in an individualized way to assure that

refrigerated samples were received in TA laboratories withinthe first 96 h after collection. When arrived, a quality checkwas performed including sample integrity (no spills), totalamount of specimen (both for hair, urine, blood, and serum)and adequate documentation of the sample (informed con-sent, self-administered questionnaire, and a form with thedate, time, and any remarkable observation during sam-pling). Afterwards, serum samples were centrifuged againand aliquoted in glass vials while urine and whole bloodaliquots were stored in polypropylene tubes. Tubes for metalanalysis aliquots were pre-washed with 10 % HNO3 andMilli-Q water. All the aliquots were preserved at −20 °Cuntil laboratory analysis.

BIOAMBIENT.ES after field work: analyses planned

Chemical analysis

Metal determinations in blood and urine using inductivelycoupled plasma/mass spectrometry (Perkin-Elmer ELANDRC-e) are being carried out in clean room facilities with

Table 2 BIOAMBIENT study participants and employed Spanish population by geographical area, economic sector, and sex

Geographical area(autonomouscommunity)

No. ofselectedcenters

No. of study participants (%) Employed Spanish populationa (thousands, %)

Services Other economicsectorsb

Services Other economic sectorsb

Men Women Men Women Men Women Men Women

Northwest I (Galicia) 3 33 (1.7) 49 (2.6) 42 (2.2) 23 (1.2) 326.5 (1.7) 423.6 (2.3) 302.3 (1.6) 99.0 (0.5)

Northwest II (Asturiasand Cantabria)

2 26 (1.4) 29 (1.5) 23 (1.2) 21 (1.1) 203.5 (1.1) 266.1 (1.4) 166.2 (0.9) 31.7 (0.2)

North (Basque Country) 3 48 (2.5) 63 (3.3) 26 (1.4) 12 (0.6) 265.4 (1.4) 365.1 (2.0) 249.3 (1.3) 52.3 (0.3)

Northeast I (Navarre,La Rioja, and Aragon)

3 52 (2.7) 61 (3.2) 27 (1.4) 6 (0.3) 267.3 (1.4) 357.3 (1.9) 296.6 (1.6) 68.2 (0.4)

Northeast II (Catalonia) 5 29 (1.5) 58 (3.1) 94 (5.0) 67 (3.5) 978.1 (5.2) 1,202.5 (6.4) 767.6 (4.1) 240.7 (1.3)

Central I (Castile andLeon)

3 40 (2.1) 54 (2.9) 38 (2.0) 22 (1.2) 300.4 (1.6) 365.2 (2.0) 297.1 (1.6) 54.8 (0.3)

Central II (Madrid) 4 64 (3.4) 76 (4.0) 28 (1.5) 31 (1.6) 1,137.3 (6.1) 1,236.6 (6.6) 438.2 (2.3) 105.6 (0.6)

Central III (Castile-LaMancha andExtremadura)

3 53 (2.8) 37 (2.0) 27 (1.4) 34 (1.8) 365.3 (2.0) 397.4 (2.1) 355.9 (1.9) 62.4 (0.3)

East (Valencia andBalearic Islands)

4 68 (3.6) 83 (4.4) 42 (2.2) 13 (0.7) 788.8 (4.2) 978.2 (5.2) 614.3 (3.3) 119.3 (0.6)

South I (Andalusia andCeuta)

5 89 (4.7) 106 (5.6) 34 (1.8) 13 (0.7) 1,079.9 (5.8) 1,088.7 (5.8) 648.1 (3.5) 133.0 (0.7)

South II (Murcia) 2 31 (1.6) 35 (1.8) 31 (1.6) 4 (0.2) 175.4 (0.9) 202.9 (1.1) 155.1 (0.8) 42.3 (0.2)

Canary Islands 1 24 (1.3) 26 (1.4) –c –c 328.0 (1.8) 320.1 (1.7) –c –c

Total 38 557 (29.4) 677 (35.8) 412 (21.8) 246 (13.0) 6,215.9 (33.2) 7,203.7 (38.5) 4,290.7 (22.9) 1,009.3 (5.4)

Definition of economic sectors according to the Spanish Classification of Economic Activities for 2009 (Instituto Nacional de Estadística 2011)a Employed Spanish population estimates in 2009 obtained from the Spanish Active Population Survey (Instituto Nacional de Estadística 2011)b Agriculture, industry, and construction sectorsc No participants were selected from economic sectors other than services in Canary Islands

Environ Sci Pollut Res

Author's personal copy

ISO6 air quality, differential pressure rooms and con-trolled temperature, which enable analyses of trace ele-ment concentrations. Lyphochek® Urine Metals Controland Seronorm™ Trace Elements Whole Blood (Sero,Norway) are used for internal quality control (QC) forurine and blood samples, respectively. The limits ofquantification (LOQ) for metals are: 0.1 μg/l in bloodand 0.05 μg/l in urine. For mercury determination inhair, only the first 3 cm closest to the scalp were used.Samples were cut into 2 mm pieces and analyzed with a directmercury analyzer DMA-80 (Milestone) which combines thetechniques of thermal decomposition, catalytic conversion,amalgamation, and atomic absorption spectrophotometry at254 nm. Certified reference materials IAEA-086 and NIES 13are used for calibration and internal QCs. Mercury measure-ments have been accredited by ISO17025 and by the National

Entity of Accreditation (ENAC, Accreditation No 223/LE460).

PCBs and PBDEs are analyzed in 0.5 ml human serum.After a solid phase extraction (SPE), both groups of com-pounds are quantified byGC-MS-NCIwith a DB-5ms column(Agilent, SA). For internal standard we used 13C-PCB118 and13C-PBDE138.

The presence of polycyclic aromatic hydrocarbon metabo-lites (1-hydroxypyrene) is quantified in 10 ml of human urineby means of enzymatic hydrolysis followed by a SPE. 1-hydroxypyrene is analyzed by liquid chromatography (HPLCAgilent 1200) and detected with a Fluorescence DetectorAgilent with a linearity range from 0.05 to 2.0 μl/l.

Additional determinations needed to evaluate the bio-markers selected included: (a) analysis of urinary creatinine,performed by Jaffé method using a commercial kit

Table 3 Frequency distribution of general characteristics in BIOAMBIENT participants

Total Blood Urine Hair

N % N % N % N %

Total 1,892 100.0 1,880 99.4 1,770 93.6 577 30.5

Sex

Men 969 51.2 962 51.2 897 50.7 250 43.3

Women 923 48.8 918 48.8 873 49.3 327 56.7

Age groups

16–29 years 373 19.9 372 19.9 342 19.4 93 16.1

30–39 years 766 40.8 764 40.9 715 40.6 234 40.6

40–49 years 469 25.0 466 25.0 449 25.5 150 26.0

≥50 years 270 14.4 266 14.2 255 14.5 100 17.3

Geographical area (autonomous communities)

Northwest I (Galicia) 147 7.8 147 7.8 134 7.6 70 12.1

Northwast II (Asturias and Cantabria) 99 5.2 99 5.3 96 5.4 12 2.1

Northeast I (Basque Country) 149 7.9 149 7.9 134 7.6 75 13.0

Northeast II (Navarre, La Rioja, and Aragon) 146 7.7 145 7.7 137 7.7 37 6.4

Central I (Castile and León) 248 13.1 247 13.1 235 13.3 16 2.8

Central II (Madrid) 154 8.1 153 8.1 148 8.4 6 1.0

Central III (Castile-La Mancha and Extremadura) 199 10.5 199 10.6 190 10.7 63 10.9

East I (Catalonia) 151 8.0 150 8.0 144 8.1 69 12.0

East II (Valencian Community) 206 10.9 205 10.9 191 10.8 39 6.8

South I (Andalusia and Ceuta) 242 12.8 239 12.7 218 12.3 134 23.2

South II (Murcia) 101 5.3 97 5.2 95 5.4 10 1.7

Canary Islands 50 2.6 50 2.7 48 2.7 46 8.0

Economic sector

Services sector 1,234 65.2 1,225 65.2 1,151 65.0 450 78.0

Other sectors (farming, construction, and industry) 658 34.8 655 34.8 619 35.0 127 22.0

Sampling trimester

January–March 352 19.0 349 18.9 337 19.4 87 15.2

April–June 564 30.4 562 30.5 524 30.2 155 27.1

July–September 376 20.3 373 20.2 353 20.3 88 15.4

October–December 563 30.4 561 30.4 524 30.2 242 42.3

Environ Sci Pollut Res

Author's personal copy

(Spinreact, Spain). Assayed urine control levels 2 and 3(Randox) have been used as internal QCs; (b) serum totalcholesterol and triglycerides, measured colorimetrically us-ing commercial kits (Spinreact, Spain). Clinical chemistrycalibration serum levels 2 and 3 (Randox) were used forcalibration and Spìntrol H Normal and Pathologic(Spinreact) as internal QCs; and (c) total lipids estimatedaccording to Bernet et al. (2007).

External quality assurance has also been done by participa-tion in the Quebec multi-element external quality assessmentscheme and Mercury-in-Hair Interlaboratory Comparison Pro-gram for metals and G-EQUAS and AMAP ring exercises forthe other biomarkers.

Statistical analysis plan and statistical considerations

For each studied pollutant, the analysis plan includes theestimation of the main descriptive statistics: arithmeticmean, geometric mean, confidence intervals (95 %), andpercentiles (10, 25, 50, 75, and 90). Geometric mean willbe prioritized since it is the best estimate of the averagelevels for data that do not fit a normal distribution and havea long dispersion in the right end of the distribution—asoccurs in most of the studied pollutants. For all descriptiveanalyses, those subjects with pollutant levels below the limitof quantification, which corresponds to the lowest concen-tration that can be determined with acceptable level offidelity and accuracy for a particular toxic compound, willbe considered to have a concentration equal to the LOQdivided by the square root of two. In the case of thosetoxicants in which there might be levels over the limit ofquantification, i.e., cotinine, this upper limit will be replacedby its value multiplied by square root of two. For determi-nations in urine the levels of pollutants will be presented inmicrograms per gram corrected by urinary creatinine, tocorrect the possible effect of dilution. Serum POPs will beexpressed on lipid weight.

The complex design of the study will always be takeninto account in all analyses, to obtain results and conclu-sions that can be extrapolated to a certain extent the overallSpanish employed population. Strata (geographic zones),primary sampling units (health exam centers), and popula-tion weights will be used to obtain unbiased point estimatesand robust standard errors; each participant will be assignedthe inverse of his probability of selection relative to thedistribution of occupied people in Spain by autonomouscommunity, sex, and economic activity sector, in accor-dance with the information provided by the last SpanishActive Population Survey (Instituto Nacional de Estadística2009). A specific set of weights will be calculated for thoseanalyses focused on pollutants measured on hair, due to thedifferent distribution of these samples.

In a second step, multivariate regression models will befitted for each pollutant. They will include the stratificationvariables as well as age, sampling trimester, and thoseessential explanatory factors defined for each biomarker(i.e., BMI for PCB, or tobacco consumption for cadmiumor cotinine). In those cases in which there are subjects withlevels below or over the limit of detection/quantification,Tobit or censored regression models will be used, as theyallow incorporating this information adequately (Bleda andTobías 2002).

Discussion

BIOAMBIENT is the first national-level HBM study onenvironmental pollutants carried out in Spain. This projecthas been designed with the aim of providing a representativeimage of the levels of certain toxicants in biological matri-ces of the Spanish working population, with participantsrecruited along the whole country, including the CanaryIslands and the Autonomous city of Ceuta. Its objective isnot to evaluate occupational exposures; it intends to give anapproximate estimation of pollutant levels in Spanish adults.This initiative, which has been promoted and financed bySpanish National Environmental Authorities, will informthe future assessments and public health measures relatedto the surveillance and control of the selected pollutants.This project may help to monitor the level of exposure tothese chemicals and to evaluate the effectiveness of futureinterventions destined to control it.

The selection of the sampling frame for this project hasclear advantages, as well as significant limitations. We de-cided to focus the study on those workers who underwenttheir annual occupational medical check-up in the healthfacilities of the Societies for Prevention of the Mutual In-surance Companies that make up Corporation Mutua, Soci-eties for Prevention aim to control work-related risks as wellas to perform the annual preventive health exams thatemployers must offer to their employees. A total amountof 113 Prevention Health centers were available for theproject. Those centers are distributed across the whole coun-try; they provide their services to more than 436,000 com-panies in Spain in all activity sectors, with 3,600,000workers employed within a large spectrum of occupationsand occupational categories, and perform more than650,000 occupational health exams per year. This highnumber of annual surveys, as well as its wide geographicalcoverage could allow us to obtain a fairly representativesample of the Spanish workforce. In addition, since bothmen and women from services and other economic sectorsare sufficiently represented in the sample, it will be possibleto obtain reliable sex- and sector-specific exposure esti-mates, as well as to compute unbiased national-wise

Environ Sci Pollut Res

Author's personal copy

estimates for the entire Spanish workforce by using standardinverse probability-of-selection weighting methods (Horvitzand Thompson 1952).

In second place, occupational settings offer many impor-tant logistic advantages for recruitment in a biomonitoringstudy since blood samples are obtained simultaneously withthe scheduled medical exam, which minimizes both the costsand the disavowal usually associated with the donation of thiskind of invasive samples. Furthermore, this strategy incorpo-rates additional value to the project since it makes accessiblethe participant’s clinical records, as well as information onoccupational exposures, which will be available to be used inthe epidemiological analyses. Therefore, feasibility and effi-ciency reasons oriented the selection of the study populationfor this project. Field work results have proven true theseprevious assumptions, even though there have been smallmodifications along the recruitment process, both in the se-lected centers and in the participant’s distribution. In essence,BIOAMBIENT.ES sample corresponds with the planned oneand will allow fulfilling the stated objectives of the project.

Undoubtedly, an obvious consequence of this approach isthat it leaves out of the focus of the assessment children andother adults are not actively included in the labor market,such as unemployed, retired and housewives. Since weopted for efficiency, we assumed a significant limitation:part of the population is out of our scope. For this reason,future HBM projects should specifically be oriented towardsthese groups not included in the present project, in order toensure a complete representation of the exposure to thesepollutants in the entire Spanish population.

In conclusion, BIOAMBIENT.ES will provide an overviewof the body burden of selected pollutants in the occupiedSpanish adults. This is an important step to promote the incor-poration of human biomonitoring in Spain among the basictools in environmental surveillance, in order to assess andevaluate the level of internal exposures to widespread pollu-tants in the general population to environmental chemicals.

Acknowledgments This study was financed by the Spanish Ministryof Agriculture, Food and Environment, previously named Ministry ofthe Environment and Rural and Marine Affairs (SEG 1251/07).

Conflict of interest The author(s) declare that they have no compet-ing interests.

References

Aguilera I, Daponte A, Gil F, Hernandez AF, Godoy P, Pla A,Ramos JL (2008) Biomonitoring of urinary metals in a popula-tion living in the vicinity of industrial sources: a comparisonwith the general population of Andalusia, Spain. Sci TotalEnviron 407:669–678

Aguilera I, Daponte A, Gil F, Hernandez AF, Godoy P, Pla A, RamosJL, Daponte A, Aguilera I, Fernandez-Ajuria A, Toro S, Martin-Olmedo P, Lacasana M, Mayoral JM, Pla A, Gil F, Hernandez A,Villanueva E, Rodrigo L, de Santiago E, Lopez O, Ramos JL,Godoy P, Sanchez-Parra F (2010) Urinary levels of arsenic andheavy metals in children and adolescents living in the industri-alised area of Ria of Huelva (SW Spain). Environ Int 36:563–569

Angerer J, Ewers U, Wilhelm M (2007) Human biomonitoring: state ofthe art. Int J Hyg Environ Health 210:201–228. doi:10.1016/j.ijheh.2007.01.024

Aragones N, Perez-Gomez B, Astray J, Gil E, Perez-Meixeira AM, dePaz C, Iriso A, Cisneros M, de Santos A, Arias P, Sanz JC,Asensio A, Fernandez MA, Gonzalez MJ, de Leon A, Garcia-Sagredo JM, Pollan M, Lopez-Abente G, Frutos GJ, Martinez M(2008) Biomonitoring of exposure to environmental pollutants innewborns and their parents in Madrid, Spain (BioMadrid): studydesign and field work results. Gac Sanit 22:483–491

Becker K, Kaus S, Krause C, Lepom P, Schulz C, Seiwert M, Seifert B(2002) German Environmental Survey 1998 (GerES III): environ-mental pollutants in blood of the German population. Int J HygEnviron Health 205:297–308. doi:10.1078/1438-4639-00155

Bernert JT, TurnerWE, Patterson DG Jr, NeedhamLL (2007) Calculationof serum “total lipid” concentrations for the adjustment of persistentorganohalogen toxicant measurements in human samples. Chemo-sphere 68:824–831. doi:10.1016/j.chemosphere.2007.02.043

Bleda M, Tobías A (2002) Aplicación de los modelos de regresión tobiten la modelización de variables epidemiológicas censuradas. GacSanit 16:188–195

Cañas A, Castaño A, Esteban M, Navarro C, Jimenez J (2010) Selectionof sampling material for the analysis of heavy metals in blood forhuman biomonitoring studies. Toxicol Lett 196:S44. doi:10.1016/S0378-4274(10)01450-5

Castaño A, Esteban M, Roman J, Pérez B, Jimenez J, Pastor R, Calvo E,Díaz G, Ruiz M, Pollan M, Aragonés N, Navarro C, Cañas A,Huetos O, BartolomeM, Lopez-Abente G (2009) A national humanbiomonitoring program on pops and heavy metals in spain. Epide-miology 20:S243. doi:10.1097/01.ede.0000362814.59310.9b

Centers for Disease Control and Prevention (2011) National Health andNutrition Examination Survey. Available from http://www.cdc.gov/nchs/nhanes/about_nhanes.htm. Accessed 2 August 2012

Centers for Disease Control and Prevention (2012) Biomonitoringprogram. Available from www.cdc.gov/biomonitoring. Accessed2 August 2012

EstebanM, Ruiz-MoragaM, Pérez-Gómez B, Grupo BIOAMBIENT.ES,Castaño A (2012) Practical aspects of the pre-analytical phase inBIOAMBIENT.ES. Gac Sanit (in press)

European Commission (2012a) Consultative forum on Environmentand Health. Available from http://ec.europa.eu/health/healthy_environments/working_groups/index_en.htm. Accessed 2 August2012

European Commission (2012b) EU Mercury strategy. Available fromhttp://ec.europa.eu/environment/chemicals/mercury/index.htm.Accessed 2 August 2012

European Commission (2012c) Statistical classification of economicactivities. Available from http://epp.eurostat.ec.europa.eu/portal/page/portal/product_details/publication?p_product_code0KS-RA-07-015

European Union (2005) Science, Children, Awareness, Legal instrument,Evaluation (SCALE initiative). Available from http://europa.eu/legislation_summaries/environment/general_provisions/l28133_es.htm. Accessed 7 August 2012

Ewers U, Krause C, Schulz C, Wilhelm M (1999) Reference valuesand human biological monitoring values for environmental tox-ins. Report on the work and recommendations of the Commissionon Human Biological Monitoring of the German Federal Envi-ronmental Agency. Int Arch Occup Environ Health 72:255–260

Environ Sci Pollut Res

Author's personal copy

Flemish Centre of Expertise of Environmental Health (2007) FlemishHuman Biomonitoring Program: monitoring and surveillance.Available from http://www.milieu-en-gezondheid.be/English/Luik21case%20studies.html. Accessed 2 August 2012

Fréry N, Saoudi A, Garnier R, Zeghnoun A, Falq G, Guldner L (2010)Exposure of the French population to environmental pollutants—environmental components of the French national survey on nu-trition and health—initial results. French institute for publichealth. Surveillance, Saint-Maurice (Fra)

Gonzalez CA, Kogevinas M, Gadea E, Huici A, Bosch A, Bleda MJ,Papke O (2000) Biomonitoring study of people living near orworking at a municipal solid-waste incinerator before and aftertwo years of operation. Arch Environ Health 55:259–267.doi:10.1080/00039890009603416

Health Canada (2010) Report on Human Biomonitoring of Environ-mental Chemicals in Canada. Results of the Canadian HealthMeasures Survey Cycle 1 (2007–2009). Health Canada, Ottawa

Horvitz DG, Thompson DJ (1952) A generalization of sampling with-out replacement from a finite universe. J Am Stat Assoc 47:663–685

Instituto Nacional de Estadística (2009) Spanish Active PopulationSurvey 2009. Avaialble from www.ine.es. Accessed 2 August2012

Instituto Nacional de Estadística (2011) National Classification of Eco-nomic Activities 2009. Available from http://www.ine.es/jaxi/menu.do?type0pcaxis&path0/t40/clasrev&file0inebase. Accessed2 August 2012

Instituto Nacional de Estadística (2012) Spanish Active PopulationSurvey 2007. Available from www.ine.es. Accessed 2 August2012

Kim NS, Lee BK (2010) Blood total mercury and fish consumption inthe Korean general population in KNHANES III, 2005. Sci TotalEnviron 408:4841–4847. doi:10.1016/j.scitotenv.2010.06.026

Llop S, Aguinagalde X, Vioque J, Ibarluzea J, Guxens M, Casas M,Murcia M, Ruiz M, Amurrio A, Rebagliato M, Marina LS,Fernandez-Somoano A, Tardon A, Ballester F (2011) Prenatalexposure to lead in Spain: cord blood levels and associatedfactors. Sci Total Environ 409:2298–2305. doi:10.1016/j.scitotenv.2011.02.004

Manno M, Viau C, Cocker J, Colosio C, Lowry L, Mutti A, NordbergM, Wang S (2010) Biomonitoring for occupational health riskassessment (BOHRA). Toxicol Lett 192:3–16. doi:10.1016/j.toxlet.2009.05.001

Ministerio de Agricultura Alimentación y Medio Ambiente (2007)Encomienda de gestión entre el ministerio de medio ambiente yel instituto de salud carlos III, para la investigación sobre lapresencia y vigilancia de contaminantes orgánicos persistentes yotras sustancias en humanos. Boletín Oficial del Estado 192:34495–34500

Ministerio de Agricultura Alimentación y Medio Ambiente (2012) Poli-cies Declaration for the National Implementation Plan (NIP) of theStockholm Convention and of Regulation (EC) No 850/2004.Avaialble from http://www.cnrcop.es/gc/assets/docs/pnestocolmofinal.pdf. Accessed 1 August 2012

National Institute of Public Health (2011) Environmental health mon-itoring system in the Czech Republic. Summary report 2009.http://www.szu.cz/uploads/documents/chzp/souhrnna_zprava/Szu_10.pdf. Accessed 1 August 2012

Paustenbach D, Galbraith D (2006) Biomonitoring: is body burdenrelevant to public health. Regul Toxicol Pharmacol 44:249–261.doi:10.1016/j.yrtph.2006.01.005

Porta M, Gasull M, Puigdomenech E, Gari M, Bosch DB, GuillenM, Lopez T, Bigas E, Pumarega J, Llebaria X, Grimalt JO,Tresserras R (2010) Distribution of blood concentrations ofpersistent organic pollutants in a representative sample of thepopulation of Catalonia. Environ Int 36:655–664. doi:10.1016/j.envint.2010.04.013

Porta M, Lopez T, Gasull M, Rodriguez-Sanz M, Gari M, Pumarega J,Borrell C, Grimalt JO (2012) Distribution of blood concentrationsof persistent organic pollutants in a representative sample of thepopulation of Barcelona in 2006, and comparison with levels in2002. Sci Total Environ 423:151–161

Porta M, Puigdomenech E, Ballester F, Selva J, Ribas-Fito N, Llop S,Lopez T (2008) Monitoring concentrations of persistent organicpollutants in the general population: the international experience.Environ Int 34:546–561. doi:10.1016/j.envint.2007.10.004

Seifert B, Becker K, Hoffmann K, Krause C, Schulz C (2000) TheGerman environmental survey 1990/1992 (GerES II): a represen-tative population study. J Expo Anal Environ Epidemiol 10:103–114. doi:10.1038/sj.jea.7500075

World Health Organization (1996) Biological monitoring of chemicalexposure in the workplace. WHO, Geneva

Zumbado M, Goethals M, Alvarez-Leon EE, Luzardo OP, Cabrera F,Serra-Majem L, Dominguez-Boada L (2005) Inadvertent expo-sure to organochlorine pesticides DDT and derivatives in peoplefrom the Canary Islands (Spain). Sci Total Environ 339:49–62

Environ Sci Pollut Res

Author's personal copy