ment 382 (2007) 14–21www.elsevier.com/locate/scitotenv

Science of the Total Environ

Pesticide exposure of two age groups of womenand its relationship with their diet

Ana Rivas, Isabel Cerrillo, Alicia Granada,Miguel Mariscal-Arcas, Fatima Olea-Serrano ⁎

Department of Nutrition and Food Science, University of Granada, Spain

Received 25 January 2007; received in revised form 20 March 2007; accepted 22 March 2007Available online 2 May 2007

Abstract

The widespread presence of organochlorine (OC) pesticides in human samples may be explained by the environmental exposureof the population. Foods are considered a constant source of exposure, despite compliance with maximum permitted residue levels.This study aimed to examine the relationship between nutritional habits of women in Southeast Spain and their serumconcentrations of OCs. A semi-quantitative questionnaire was used to estimate the frequency of consumption of foods by two agegroups of women, pre-menopausal (Pre-M) and post-menopausal (Post-M), and their serum pesticide levels were measured by gaschromatography (GC) with electron capture detector and confirmed by GC and mass spectrometry. The Pre-M group showedsignificantly higher serum concentrations of all OCs studied with the exception of DDE. The groups significantly differed inconsumption of all food groups with the exception of fruit. In the Pre-M group, the mean serum p,p-DDT concentration wassignificantly associated with milk/yoghurt (pb0.045) and red meat (pb0.023), serum o,p-DDT with red meat (pb0.049), serumaldrin with eggs (pb0.038) and poultry (pb0.024), and serum DDE with eggs (pb0.025). In the Post-M group, serum lindane wasassociated with fresh and cured cheese (pb0.001), red meat (pb0.001) and white and oily fish (pb0.001), and both serum DDEand dieldrin were associated with fresh cheese, cured cheese, red meat, and white and oily fish (pb0.001). These results confirmfoods as a source of human exposure to persistent organic molecules. Consideration should be given to the reduction of permittedresidue levels to minimize this threat to human and animal health.© 2007 Elsevier B.V. All rights reserved.

Keywords: Pesticides; Diet; Women

1. Introduction

The presence of persistent organochlorine (OC)pesticides in human serum may be explained byenvironmental exposure of the population. Foods areconsidered to represent a constant source of exposure,

⁎ Corresponding author. Department de Nutrition and Food Science,School of Pharmacy Campus de Cartuja University of Granada, 18071Granada, Spain. Tel.: +34 958 24 28 41; fax: +34 958 24 95 77.

E-mail address: folea@ugr.es (F. Olea-Serrano).

0048-9697/$ - see front matter © 2007 Elsevier B.V. All rights reserved.doi:10.1016/j.scitotenv.2007.03.032

despite compliance with the maximum permittedresidue levels (MRLs). Only 60% of all fruit orvegetables consumed by European citizens are freefrom pesticide residues; 36% show residues below theMRL, and 4% are contaminated by residues above theMRL. On the other hand, the use of extremelyhazardous substances has been progressively bannedin agriculture and some positive results of this policy arenow emerging. According to a recent study in Barcelona(Spain), OC pesticides, the most persistent substances,

15A. Rivas et al. / Science of the Total Environment 382 (2007) 14–21

appeared to have completely disappeared from somefoodstuffs, such as fruits, although they continued to bedetected in other types of food, such as vegetables andmilk (Vicente et al., 2004). In a recent analysis of OCresidues in U.S. foods, Pesticide Action Network foundthat even chemicals that had been banned for decadeswere consistently detected in food samples tested by theU.S. Food and Drug Administration (http://www. panna.org). This can be explained in part by the long life ofmany OCs in the environment (dieldrin and DDTbreakdown products, for example, can remain in soil fordecades) and in part by the importation or wind andwater transportation of pesticides still used in othercountries. Inhalation and dermal contact are additionalexposure routes for individuals working directly withpesticides (e.g., farm workers, seed treatment facilityworkers, etc). Air samples collected in Strasbourg(France) in 2001 and 2003 were found to contain α-HCH and γ-HCH and, at lower concentrations, aldrinand dieldrin (Scheyer et al., 2005). Infants can alsoingest OCs in breast milk from chemicals accumulatedin the fatty tissue of the mother during her life (Schaferand Kegly, 2002; Glynn et al., 2003; Botella et al., 2004;Minh et al., 2004; De Saeger et al., 2005.).

The presence and persistence of OCs in our bodies isa constant, whatever the source of exposure. They havebeen universally reported in human fatty tissue andserum, although most studies are based on serum levelsgiven the ease of access to this medium. Numerousstudies have related the presence of OCs in humansamples to the consumption of certain foods, especiallyfish (Cole et al., 2002, Chun and Kang, 2003; Jiang etal., 2005). The present study was designed to examinethe relationship between nutritional habits of women inSoutheast Spain and concentrations of OCs in theirserum.

2. Materials and methods

2.1. Study population

Two age groups of women, women of reproductiveage (pre-menopausal, Pre-M) and post-menopausalwomen (Post-M), were randomly selected from a largersample of female patients recruited to investigate therelationship between OC exposure risk and onset ofhormone-dependent diseases (Botella et al., 2004;Ibarluzea et al., 2004; Cerrillo et al., 2005). All studysubjects lived in the same region of southern Spain(Granada and Almeria provinces). The Pre-M group wasformed by 200 women who had just given birth (agerange, 17–42 yr) and the Post-M group by 200 women

undergoing surgery for various diseases (age range, 50–75 yr). All participants signed their informed consent.

2.2. Nutritional survey

Structured face-to-face interviews were conducted atthe hospitals by trained interviewers before the surgeryor delivery in order to gather data on sociodemographiccharacteristics, reproductive history, diet, and tobaccoand alcohol consumption. The frequency of consump-tion of foods was studied by means of a semi-quantitative questionnaire, classifying the frequency ofconsumption of foods as follows: never; 1–3 times/month; 1–3 times/week; 4–6 times/week; and (only forsome foods) every day. The results were expressed astimes/week or g/week. The questionnaire design tookaccount of the eating habits of the Spanish population.The mean portions consumed by the study populationwere estimated using habitual domestic measurements:spoons, glasses, cups, etc. (SENC 2001; Carvajal andSánchez-Muniz, 2003).

2.3. Biological samples

Approximately 10 mL of whole blood was collectedduring surgery or at delivery in glass vials with noanticoagulant or serum separator. Serum samples werestored at − 70 °C until laboratory analysis.

3. Methods

3.1. Analysis of OCs in serum

3.1.1. ReagentsAll solvents used were of high purity grade for

HPLC: methanol, 2-isopropanol and hexane (Panreac,Barcelona, Spain). The chemicals used (and suppliers)were: aldrin, dieldrin, lindane, p,p-DDT, and o,p-DDT(Sulpeco, Bellefonte, PA); p,p-DDE, and p-p′dichlor-obenzophenone (Dr Ehrenstorfer Lab., Ausburg, Ger-many); and p,p-DDE (Chem Service, West Chester, PA).

3.2. Apparatus

3.2.1. Chromatographic conditionsThe chromatograph apparatus comprised a Varian-

3350 ECD Electron Capture Detector (63Ni) (WalnutCreek, CA), a split/splitless injector operated in splitlessmode with Millennium Chromatography ManagerSoftware, and a Varian CP8944 chromatographiccolumn (VF-5ms; 30 m×0.25 mm i.d. ×0.25 μm filmthickness).

16 A. Rivas et al. / Science of the Total Environment 382 (2007) 14–21

3.2.2. Working conditionsInjector and detector temperatures were 250° and

300°, respectively. The temperature column wasprogrammed from 130 °C (1 min) to 150° at 20 °C/min, 150 °C to 200° at 10 °C/min, and 200 °C to 260° at20 °C/min (20 min). Carrier and auxiliary gas wasnitrogen at a flow rate of 30 mL/min and 40 mL/min,respectively.

3.2.3. GC — mass spectrometry (GC/MS)Saturn 2100T Varian equipment was used with

Varian injector 1177 and CP5860 WCOT fused silicacolumn (30 m×0.25 mm). The computer controlling thesystem contained a specifically created library for thetarget analytes under our experimental conditions.Working conditions: injector temperature 250 °C; initialcolumn temperature 50 °C (2 min), 30 °C/min to 185 °C(5.5 min), 2 °C/min to 250 °C (32.5 min), and 30 °C to300 °C (6.67 min). The carrier gas used was helium(purity 99.999%) at flow rate of 1 mL/min. Manifold,transfer-line, and trap temperatures were 50, 230, and200 °C, respectively.

3.2.4. Serum analysisHalf of the same volume of methanol was added to

2–4 mL of serum, and the solution was shaken for5 min. The extraction of OCs was performed using10 mL ethyl ether/hexane (1:1 v/v) centrifuged for15 min at 3000 rpm. The organic phase was obtainedand the extraction procedure was repeated twice more.Organic phases were evaporated to 1 mL under a streamof nitrogen. To this residue, 0.5 ml of concentratedsulfuric acid was added and centrifuged for 10 min at3000 rpm. Acid residue was extracted twice more withthe addition of 1 ml hexane. The three organic phaseswere collected and dried in a flow of nitrogen. Dryresidue was dissolved in 1 mL of hexane and cleanedup.

3.2.5. Clean-upIn all extractions, organic extracts obtained were

purified with silica Sep-Pak (Wat 051900) cartridgepretreated with 2 mL hexane. Extracts were eluted with10 mL hexane and then with 10 ml hexane: methanol:isopropanol (45:40:15; v/v/v). Both eluates werecollected and dried in a stream of nitrogen.

Dry residue was dissolved in 1 mL hexane, labeledwith p-p′dichlorobenzophenone internal standard, andanalyzed using GC/ECD. The results were confirmed bymeans of GC/MS. (Rivas et al., 2001; Botella et al.,2004; Moreno Frias et al., 2004; Cerrillo et al., 2005;Carreño et al., 2006).

3.2.6. Statistical analysisStandard descriptive statistics (frequency, mean±

standard deviation) were calculated for all variables.Quantitative variables were correlated by using thePearson or Spearman test, and Student's t test andLevene test and non-parametric tests were used tocompare pesticide concentrations and nutritional vari-ables. The chi-square test was also used to analyzequalitative variables. pb0.05 was considered significantin all tests. All statistical analyses were performed usingSPSS version 12 statistical software.

4. Results

Anthropometric and sociological characteristics ofthe study population are summarized in Table 1. Besidesage and menopausal status, the two groups significantlydiffered in mean BMI and number of children, with ahigher BMI and more children for the post-menopausalwomen. No difference in socioeconomic characteristicswas observed between the groups.

4.1. Analysis of persistent OCs in serum

There was a higher concentration and frequency ofall studied OCs in the Pre-M group than in the Post-Mgroup. Because the pesticide concentrations did nothave a normal distribution, the Mann–Whitney U testwas used to compare samples. A significant differencewas found between the groups in all of the OCcompounds analyzed (pb0.001). The results (in ng/mL of serum) are listed in Table 2.

4.2. Analysis of nutritional questionnaire results

The frequencies of consumption of foods collectedby the semi-quantitative questionnaire are shown inTables 3 and 4, comparing the frequencies between Pre-M and Post-M women. Intake of foods (in g/week) wascalculated from the reported frequency and weight (in g)of each portion, as described by the women or in somecases according to the usual average portion in Spain(Table 3). The comparison of means (Table 3) showedsignificant differences between the study groups for allfood groups with the exception of fruit (Levene test,pb0.130; t test, pb0.455).

However, no significant difference was found betweenthe groups in the frequency of consumption of foods(Table 4). Interestingly, ecological food was consumed atsome time by 32% of the Pre-M group but was neverconsumed by any of the Post-M group. Only one subject(in Pre-M group) described herself as vegetarian.

Table 1Anthropometric and sociological characteristics of the studypopulation, reproductive age women (Pre-M) and post-menopausalwomen (Post-M)

Mean(SD)

Minimum Maximum t test p

AgePre-M 32.9

(5.39)17 44 102.57 0.001

Post-M 61.66(6.40)

50.1 75.4

Age of menarchePre-M 12.61

(1.58)8 14 1.576 0.210

Post-M 12.89(1.72)

9 19

Height (m)Pre-M 1.63

(0.57)1.45 1.79 3.104 0.079

Post-M 1.58(0.063)

1.40 1.75

Weight (kg)Pre-M 61.62

(10.51)40.00 112.50 1.503 0.221

Post-M 71.90(11.63)

45.00 110.00

BMI (kg/m2)Pre-M 23.26

(3.89)15.24 39.86 12.278 0.001

Post-M 28.73(5.03)

17.0 44.09

Number of childrenPre-M 1.77

(1.02)1 9 85.07 0.001

Post-M 3.63(1.90)

1 11

Educational level frequency (%)Noschooling

Primaryschool

Secondary/technical school

Universitydegree

p

Pre-M 7 (3.5) 93 (46.5) 64 (32.0) 36 (18) 0.700Post-M 45 (23.5) 71 (37.2) 54 (28.3) 21 (11)

pb0.01 was considered significant.

17A. Rivas et al. / Science of the Total Environment 382 (2007) 14–21

A large majority (84.5%) of the population only usedolive oil, 11.5% consumed olive oil and other oils, whereasonly 4% never consumed olive oil. An equally largemajority (88%) never drank beer, with 12% drinking oneglass/day and only one woman (Post-M group) drinkingthree glasses/day; a similar consumption pattern wasobserved for wine, while consumption of spirits was amaximum of two glasses/week by 2% of the Pre-M groupand of three glasses/week by 6% of the Post-M group.

4.3. Relationship between diet and concentrations ofOCs in serum

In the Pre-M group, the serum concentration of p,p-DDT was significantly associated with milk/yoghurt(pb0.045) and red meat (pb0.023), serum o,p-DDTwith red meat (pb0.049), serum aldrin with eggs(pb0.038) and poultry (pb0.024), and serum DDEwith eggs (pb0.025). In the Post-M group, serumlindane was associated with fresh and cured cheese(pb0.001), red meat (pb0.001) and white and oily fish(pb0.001), and both serum DDE and dieldrin wereassociated with fresh cheese, cured cheese, red meat,and white and oily fish (pb0.001). Correlations betweenconsumption of foods and serum OC values in the wholeseries are listed in Table 5.

5. Discussion

Human impregnation by persistent OC molecules isuniversal, and they can be detected in any biologicalsample. Lipophilic OCs are highly resistant to bothabiotic and biotic degradation. They can be transportedby air and water, and trace amounts are found all overthe world, even where PCBs have never been producedor used (AMAP, 1998). The present study analyzed andcompared the presence of OC residues in sera of pre-menopausal (n=200) and post-menopausal women(n=200) in Southern Spain, a Mediterranean region ofintensive agriculture. There were no significant differ-ences in the socioeconomic or cultural characteristics ofthe groups, which only differed in age (32.9±5.39 vs.61.6±6.4 yr, pb0.001) and BMI. Our finding of apositive association of age and BMI with serum OCconcentrations is consistent with previous reports(Ahlborg et al., 1995; Deutch and Hansen, 2000) andis biologically plausible, since OCs accumulate in fattissue over time. The pre-menopausal women showedsignificantly higher serum concentrations of all OCsstudied with the exception of DDE. The higher DDElevels in the post-menopausal group may reflecthistorical exposure, since this pesticide was authorizedand widely used during a large part of their lives(Akkina et al., 2004).

The high frequency of OC molecules in serum maybe due to their current utilization and their low butconstant presence as a persistent residue in foods and theenvironment, since dietary exposure and gastrointestinalabsorption are considered the most important routes inhumans (NRC, 1999). Thus, the same OC residues arefound in foods in both Europe and the USA. Twenty-two OCs were detected by a study that followed up

Table 2Pesticides in serum of Pre-M group and Post-M group, expressed in ng/mL of serum

Pre-M (n=200) Post-M (n=200)

No. of samples quantified Minimum Maximum Mean (SD) No. of samples quantified Minimum Maximum Mean (SD)

Lindane 142 0.02 61.86 1.96(5.28) 105 0.063 11.399 1.42(1.98)Aldrin 107 0.05 36.01 3.00(3.87) 109 0.212 14.155 2.22(2.43)Dieldrin 51 0.48 28.33 4.21(5.14) 84 0.214 8.818 1.30(1.43)DDE 162 0.81 27.36 4.04(3.74) 188 0.436 49.022 7.01(8.02)o,p′-DDT 57 0.17 30.24 4.03(3.86) 48 0.200 10.119 1.36(1.60)p,p′-DDT 122 0.10 56.02 8.56(10.71) 73 0.262 20.908 4.07(4.10)

18 A. Rivas et al. / Science of the Total Environment 382 (2007) 14–21

Spanish foods for more than 15 years. Thus, aldrin,dieldrin, endosulfan, and DDTs were detected in 2.8%of vegetables, 3.10% of meats, and 2% of dairyproducts, but no OC residues were found in fruit(Vicente et al., 2004). This last finding is consistent withthe absence of any correlation between fruit consump-tion and the presence of OCs in the present serumsamples. In Spain, as in other countries, residues ofDDE, PCB, hexachlorobenzene, and lindane isomerssuch as β-hexachlorocyclohexane are frequently foundin samples of meat, fish, milk, butter, cheese, and cereals(Hernández et al., 1994; Lazaro et al., 1996; Porta et al.,2002). Schafer and Kegley (2002) estimated that in mostfoods, e.g., vegetables, meat, and dairy products, thereare usually at least five OCs in a single food item, mostcommonly DDTs and dieldrin. Similar findings werereported in Yukon, Canada, where concentrations of 61pesticides were analyzed in 132 foods; DDT and DDEwere found in 25% of samples (Rawn et al., 2004), and asimilar prevalence was found for PCBs and dioxins(Schepens et al., 2001). Numerous recent studies haveestimated the presence of persistent OCs in food and

Table 3Intake (g/week) of each food in Pre-M and Post-M groups calculated from theor the usual average portion in Spain)

Foods a

g/weekPre-M group

Maximum Mean SD

Milk/yoghurt 625.00 457.70 238Fresh cheese 200.00 104.44 85Cured cheese 200.00 89.09 84Eggs 375.00 163.89 75Oily fish 300.00 133.33 71White fish 300.00 127.42 71Poultry 250.00 124.36 64Red meat 250.00 104.41 60Vegetables 500.00 384.59 173Salad 1000.00 943.93 202Fresh fruit 750.00 703.03 157a Classification of foods: Red meat: pork, beef, or veal; Poultry: fresh or

grapes, plums, peaches, melon, etc.; Eggs: in any preparation; Salad: lettuce, tbeans, etc.

their intake by different population groups (Bjerregaardet al., 2001; Deutch et al., 2004).

The approach of the present investigation was torelate serum levels of persistent OCs in these women totheir diet. Although the two groups showed a similarfrequency of consumption of food groups, theysignificantly differed (pb0.001) in the weekly amount(in g) of each food group consumed, with the exceptionof fruit (pb0.455).

Given their geographical and cultural setting, thesewomen could be expected to follow a Mediterraneandiet. Its positive qualities are increasingly recognized,and the Spanish government has backed recommenda-tions to disseminate scientific knowledge of this diet anddevelop strategies for its preservation (Observatory ofthe Mediterranean, 2005) The Mediterranean diet isgenerally considered to comprise the following: i)abundance of fruit and vegetables, bread, pasta, rice,pulses, and potatoes, with moderate daily consumptionof cheese and yoghurt (questionnaire variable: 4–6/week); ii) weekly consumption of a moderate amount of(preferably oily) fish, poultry, and eggs (questionnaire

frequency and weight (in g) of each portion (as reported by the women

Post-M group

Maximum Mean SD

.69 1000.00 922.50 245.45

.23 200 18.90 43.47

.21 200 54.30 61.02

.75 375.00 181.50 122.42

.94 300 101.70 114.83

.36 300 81.75 110.91

.34 250 126.25 105.55

.94 250 54.00 93.25

.94 750.00 585.75 237.33

.11 1000.00 870.50 284.40

.46 750.00 714.00 134.28

frozen; Fish: fresh, frozen or processed; Fresh fruit: apples, oranges,omatoes, peppers and other vegetables eaten raw; Vegetables: potatoes,

Table4

Frequency

ofconsum

ptionof

foodscomparedbetweenpre-menopausal(Pre-M

)andpost-m

enopausal(Post-M)wom

en

Pre-M

group

Yoghurt/m

ilkfrequency(%

)Fresh

cheese

frequency(%

)Cured

cheese

frequency(%

)Egg

sfrequency(%

)Salad

frequency(%

)Fresh

vegetables

Fruit

Pou

ltry

Red

meat

White

fish

Oily

fish

Never

17(8.6)

43(21.7)

60(30.3

5(2.5)

3(1.5)

14(7.1)

1(0.5)

5(2.5)

10(5.1)

8(4)

8(4)

1–3tim

es/m

onth

16(8.1)

26(13.1)

22(11.1)

7(3.6)

3(1.5)

13(6.6)

6(3)

5(2.5)

21(10.6)

12(6.1)

19(9.6)

1–3tim

es/wk

36(18.2)

47(23.7)

50(25.3)

166(84.3)

9(4.5)

50(25.3)

10(5.1)

150(75.8)

144(72.7)

152(76.8)

148(74.7)

4–6tim

es/wk

129(65.2)

82(41.4)

66(33.3)

19(9.6)

183(92.4)

120(60.6)

181(91.4)

38(19.2)

23(11.6)

268(13.1)

23(11.6)

Post-M

grou

p

Never

5(2.5)

149(75.5)

56(28)

3(1.5)

1(0.5)

00

60(30)

140(70)

111(55.5)

89(44.5)

1–3tim

es/m

onth

7(3.5)

19(9.5)

55(27.5)

42(21)

13(6.5)

12(6)

4(2)

4(2)

2(1)

5(2.5)

6(3)

1–3tim

es/wk

7(3.5)

25(12.5)

67(33.5)

104(52)

22(11.0)

55(27.5)

10(5)

59(29.5)

25(12.5)

50(259)

63(31.5)

4–6tim

es/wk

181(90.5)

7(3.5)

22(11)

51(25)

164(82.0)

133(66.5)

186(93)

77(38.5)

33(16.5)

34(17)

42(21)

p0.49

30.796

0.597

0.308

0.12

20.921

0.306

0.551

0.243

0.929

0.235

Nosign

ificantdifferencesbetweenthegrou

pswerefoun

d.pb0.05

was

considered

sign

ificant.

19A. Rivas et al. / Science of the Total Environment 382 (2007) 14–21

variable: 1–3/week); iii) red meat a few times per month(questionnaire variable: 1–3/month); iv) use of olive oilas main fat for both frying and dressings; and v)consumption of seasonal and fresh foods in their naturalstate, eating nuts, honey and olives in moderation,drinking wine (preferably red) in moderation duringmeals, and using aromatic herbs as a healthy alternativeto salt (Tur et al., 2004).

In the Pre-M group, the consumption of yoghurt/milk, cheese, vegetables, and fruit was significantly(pb0.00) lower than these recommendations and theirintake of meat was significantly higher (pb0.004),while their consumption of eggs and fish was asrecommended. In the Post-M group, the consumptionof yogurt/milk, cheese, vegetables, and fruit wassignificantly (pb0.001) lower than recommendations,their intake of fish and poultry was significantly higher(pb0.001), while their consumption of eggs was asrecommended. Olive oil was the only fat used by 84.5%of the whole series, 11.5% consumed olive and otheroils, and only 4% did not use olive oil.

A correlation (Spearman's test) was found betweenthe weekly intake of foods (in grams) and the serumconcentration of OC pesticides in each group and in theoverall series. These results are in agreement withreports by various authors of a relationship betweenfood intake and the presence of persistent OCs inbiological samples (Schinas et al., 2000; Sandanger etal., 2003). The association observed between OCs andfoods of animal origin (Table 5) is consistent with otherrecently published findings. Thus, the intake by urbanChinese women of animal meat, especially animal fat,was reported to be a major contributor to their serumOCP levels (Lee et al., 2007). A high positivecorrelation was found by numerous authors betweenthe consumption of oily fish and high serum OC levelsin Swedish fishermen and their families (Axmon andRignell-Hydbom, 2006; Rylander et al., 2005).

In conclusion, foods can be considered a source ofhuman exposure to persistent organic molecules.Mechanisms to control this contamination are required,and norms need to be established by international bodiesin order to improve the effectiveness of follow-upsystems. Finally, consideration must be given to thereduction of permitted residue levels in order tominimize this threat to human and animal health.

Acknowledgements

We thank Richard Davies for the editorial assistance.This research was supported by grants from the SpanishMinistry of Health (FIS, 94/1551), Junta de Andalucia

Table 5Spearman correlation between organochlorines in the sera (ng/mL) of the study population and the amount of food consumed (g/week)

Global studypopulation

Yoghurt/milk

Freshcheese

Curedcheese

Eggs Salad Freshvegetables

Fruit Poultry Red meat Whitefish

Oily fish

Lindane ρ −0.063 0.144(⁎) 0.125(⁎) −0.090 0.009 −0.177(⁎⁎) −0.050 0.201(⁎⁎) 0.250(⁎⁎) 0.269(⁎⁎) 0.304(⁎⁎)n=273 p 0.300 0.018 0.039 0.140 0.886 0.003 0.409 0.001 0.001 0.001 0.001Aldrin ρ −0.229(⁎⁎) 0.112 −0.018 −0.136(⁎) 0.047 −0.190(⁎⁎) 0.054 0.043 0.087 0.041 0.060n=215 p 0.001 0.101 0.795 0.046 0.490 0.005 0.431 0.530 0.205 0.554 0.378DDE ρ 0.199(⁎⁎) −0.023 0.052 0.025 −0.078 0.080 0.024 0.100 −0.001 0.077 0.146(⁎⁎)n=348 p 0.001 0.669 0.333 0.648 0.144 0.138 0.657 0.062 0.987 0.150 0.006Dieldrin ρ −0.374(⁎⁎) 0.451(⁎⁎) 0.382(⁎⁎) −0.079 −0.020 −0.297(⁎⁎) −0.012 0.418(⁎⁎) 0.543(⁎⁎) 0.542(⁎⁎) 0.555(⁎⁎)n=135 p 0.001 0.001 0.001 0.361 0.815 0.001 0.888 0.001 0.001 0.001 0.001o,p-DDT ρ −0.157(⁎) 0.142 0.110 −0.101 −0.027 −0.146(⁎) −0.034 0.288(⁎⁎) 0.315(⁎⁎) 0.240(⁎⁎) 0.312(⁎⁎)n=105 p 0.031 0.050 0.130 0.167 0.708 0.044 0.641 0.001 0.001 0.001 0.001p,p-DDT ρ −0.056 0.160(⁎) −0.019 −0.054 0.044 −0.136(⁎) 0.028 −0.040 −0.020 0.024 0.037n=195 p 0.390 0.014 0.775 0.413 0.503 0.038 0.672 0.539 0.757 0.718 0.577

p ⁎⁎The correlation is significant at 0.01;⁎ The correlation is significant at 0.05. ρ (rho) = Spearman correlation coefficient.

20 A. Rivas et al. / Science of the Total Environment 382 (2007) 14–21

(AGR141) and the European Union Commission(FOOD-CT-2003-506319).

References

Ahlborg UG, Lipworth L, Titus-Ernstoff L, Hsieh CC, Hanberg A,Baron J, et al. Organochlorine compounds in relation to breastcancer, endometrial cancer, and endometriosis: an assessment ofthe biological and epidemiological evidence. Crit Rev Toxicol1995;25(6):463–531.

Akkina JE, Reif JS, Keefe TJ, Bachand AM. Age at natural menopauseand exposure to organochlorine pesticides in hispanic women. JToxicol Environ Health Part A 2004;67:1407–22.

AMAP. Arctic pollution issues. Arctic Monitoring and AssessmentProgramme AMAP, Oslo, Norway; 1998.

Axmon A, Rignell-Hydbom A. Estimations of past male and femaleserum concentrations of biomarkers of persistent organochlorinepollutants and their impact on fecundability estimates. Environ Res2006;101:387–94.

Bjerregaard P, Dewailly E, Ayotte P, Pars T, Ferron L, Mulvad G.Exposure of Inuit in Greenland to organochlorines through themarine diet. J Toxicol Environ Health 2001;A 62:69–81.

Botella B, Crespo J, Ana Rivas A, Isabel Cerrillo I, Maria Fatima Olea-Serrano MF, Olea N. Exposure of women to organochlorinepesticides in Southern Spain. Environ Res 2004;96:34–40.

Carreño J, Rivas A, Granada A, Lopez-EspinosaMJ, Mariscal M, Olea N,et al. Exposure of youngmen to organochlorine pesticides in SouthernSpain. Environ Res 2006. doi:10.1016/j.envres.2006.06.007.

Carvajal A, Sánchez-Muniz FJ. Pesos de medidas caseras y racioneshabituales de consume. In: García-Arias MT, García-FernándezMC, editors. Nutrición y dietética, León: Universidad de León.ISBN:84-9773-023-2; 2003. p. 1a–130a.

Cerrillo I, Granada A, Lopez-EspinosaMJ, Olmos B, JiménezM, CañoA, et al. Endosulfan and its metabolites in fertile women, placenta,cord blood, and human milk. Environ Res 2005;98:233–9.

Chun OK, Kang HG. Estimation of risks of pesticide exposure, byfood intake, to Koreans. Food Chem Toxicol 2003;41:1063–76.

Cole DC, Sheeshka J, Murkin EJ, Kearney J, Scott F, Ferron LA, et al.Dietary intakes and plasma organochlorine contaminant levelsamong Great Lakes fish eaters. Arch Environ Health 2002;57:496–509.

De Saeger S, Sergeant H, Piette M, Bruneel N, Van de Voorde W, VanPeteghen C. Monitoring of polychlorinated biphenyls in Belgianhuman adipose tissue samples. Chemosphere 2005;58:953–60.

Deutch B, Hansen JC. High human plasma levels of organochlorinecompounds in Greenland. Regional differences and lifestyleeffects. Dan Med Bull 2000;47(2):132–7.

Deutch B, Pedersen HS, Hansen JC. Dietary composition in Greenland2000, plasma fatty acids and persistent organic pollutants. Sci TotalEnviron 2004;331:177–88.

Glynn AW, Fredrik Granath F, Aune M, Samuel Atuma S, DarnerudPO, Bjerselius R, et al. Organochlorines in Swedish women:determinants of serum concentrations. Environ Health Perspect2003;111:349–55.

Hernández LM, Fernández MA, Jiménez B, González MJ. Organo-chlorine pollutants in meats and cows milk from Madrid (Spain).Bull Environ Contam Toxicol 1994;52:246–53.

Ibarluzea JMJ, Fernández M, Santa-Marina L, Olea-Serrano MF,Rivas AM, Aurrekoetxea JJ, et al. Breast cancer risk and thecombined effect of environmental estrogens. CCC, Cancer CausesControl 2004;15:591–600.

Jiang QT, Lee TK, Chen K, Wong HL, Zheng JS, Giesy JP, et al.Human health risk assessment of organochlorines associated withfish consumption in a coastal city in China. Environ Pollut2005;136: 155–65.

Lázaro R, Herrera A, Ariño A, Conchello MP, Bayarri S. Organochlorinepesticide residues in total diet samples from Aragón (NortheasternSpain). J Agric Food Chem 1996;44:2742–7.

Lee Sang-Ah, Dai Qi, Zheng Wei, Gao Yu-Tang, Blair Aaron, TessariJohn D, et al. Association of serum concentration of organochlo-rine pesticides with dietary intake and other lifestyle factors amongurban Chinese women. Environ Int 2007;33:157–63.

Minh NH, Someya M, Minh TB, Kunisue T, Iwata H, Watanabe M,et al. Persistent organochlorine residues in human breast milkfrom Hanoi and Ho Chi Minh City, Vietnam: contamination,accumulation kinetics and risk assessment for infants. EnvironPollut 2004;129:431–41.

Moreno Frias M, Jiménez Torres M, Garrido Frenich A, MartinezVidal JL, Olea Serrano MF, Olea N. Determination of organo-chlorine compounds in human biological samples by GC/MS/MS.Biomed Chromatogr 2004;18:102–11.

National Research Council. Hormonally active agents in theenvironment. Washington, DC: National Academy Press; 1999.

21A. Rivas et al. / Science of the Total Environment 382 (2007) 14–21

Observatory of the Mediterranean. www.dietamediterranea.com; 2005.Porta M, Kogevinas M, Zumeta E. Concentrations of persistent toxic

compounds in the Spanish population: a puzzle without pieces andthe protection of public health. Gac Sanit 2002;16:257–66.

Rawn DF, Cao XL, Doucet J, Davies DJ, Sun WF, Dabeka RW, et al.Canadian Total Diet Study in 1998: pesticide levels in foods fromWhitehorse, Yukon, Canada, and corresponding dietary intakeestimates. Food Addit Contam 2004;21:232–50.

Rivas AM, Fernández MF, Cerrillo I, Ibarluzea J, Olea-Serrano MF,Pedraza V, et al. Human exposure to endocrine disruptors:standardization of a marker of estrogenic exposure in adiposetissue. APMIS 2001;109:1–13.

Rylander L, Rignell-Hydbom A, Hagmar L. A cross-sectional study ofthe association between persistent organochlorine pollutants anddiabetes. Environmental health: a global access science source;2005. 4:28. doi:10.1186/1476–069X-4.

Sandanger TM, Brustad M, Odland JO, Doudarev AA, Miretsky GI,ChaschinV, et al. Human plasma levels of POPs, and diet among nativepeople from Uelen, Chukotka. J Environ Monit 2003; 5:689–96.

Schafer KS, Kegley SE. Persistent toxic chemicals in the US foodsupply. J Epidemiol Community Health 2002;56:813–7.

Schepens PJ, Covaci A, Jorens PG, Hens L, Scharpe S, Van LarebekeN. Surprising findings following a Belgian food contaminationwith polychlorobiphenyls and dioxins. Environ Health Perspect2001;109:101–3.

Scheyer A, Gaeff C, Morville S, Mirabel P, Millet M. Analysis of someorganochlorine pesticides in an urban atmosphere (Strasbourg, eastof France). Chemosphere 2005;58:1517–24.

Schinas V, Leotsinidis M, Alexopoulos A, Tsapanos V, Kondakis XG.Organochlorine pesticide residues in human breast milk fromsouthwest Greece: associations with weekly food consumptionpatterns of mothers. Arch Environ Health 2000;55:411–7.

SENC. Sociedad Española de Nutrición Comunitaria. Guias alimen-tarias para la población española. Recomendaciones para una dietasaludable. Madrid: IMC&SENC; 2001.

Tur JA, Romaguera D, Pons A. Food consumption patterns in aMediterranean region: does the Mediterranean diet still exist? AnnNutr Metab 2004;48:193–201.

Vicente A, Arques JF, Villalbi JR, Centrich F, Serrahima E, Llebaria X,et al. Pesticides in the diet: adding pieces to the puzzle. Gac Sanit2004;18:425–30.