Health Changes in Fishermen 2 Years After Clean-up of the Prestige Oil Spill Gema Rodríguez-Trigo, MD; Jan-Paul Zock, PhD; Francisco Pozo-Rodríguez, MD; Federico P. Gómez, MD; Gemma Monyarch, MSc; Laura Bouso, MSc; M. Dolors Coll, PhD; Héctor Verea, MD; Josep M. Antó, MD; Carme Fuster, PhD; and Joan Albert Barberà, MD, for the SEPAR (Sociedad Española de Neumología y Cirugía Torácica)-Prestige Study Group*

Background: In 2002, the oil tanker Prestige spilled more than 67 000 tons of bunker oil, heavily contaminating the coast of northwestern Spain.

Objective: To assess respiratory effects and chromosomal damage in clean-up workers of the oil spill 2 years after the exposure.

Design: Cross-sectional study.

Setting: Fishermen cooperatives in coastal villages.

Participants: Local fishermen who were highly exposed (n � 501) or not exposed (n � 177) to oil 2 years after the spill.

Measurements: Respiratory symptoms; forced spirometry; metha- choline challenge; markers of oxidative stress (8-isoprostane), air- way inflammation (interleukins, tumor necrosis factor-�, and interferon-�), and growth factor activity in exhaled breath conden- sate; and chromosomal lesions and structural alterations in circulat- ing lymphocytes.

Results: Compared with nonexposed participants, persons exposed to oil were at increased risk for lower respiratory tract symptoms (risk difference, 8.0 [95% CI, 1.1 to 14.8]). Lung function did not significantly differ between the groups. Among nonsmoking partic- ipants, exposed individuals had higher exhaled 8-isoprostane levels than nonexposed individuals (geometric mean ratio, 2.5 [CI, 1.7 to 3.7]), and exposed individuals with lower respiratory tract symp- toms had higher 8-isoprostane levels than those of exposed indi-

viduals without symptoms. Exposed nonsmoking participants also had higher levels of exhaled vascular endothelial growth factor (risk difference, 44.8 [CI, 27.9 to 61.6]) and basic fibroblast growth factor (risk difference, 16.0 [CI, 3.5 to 28.6]). A higher proportion of exposed participants had structural chromosomal alterations (risk difference, 27.4 [CI, 10.0 to 44.8]), predominantly unbalanced al- terations. The risk for elevated levels of exhaled 8-isoprostane, vascular endothelial growth factor, and basic fibroblast growth fac- tor and structural chromosomal alterations seemed to increase with intensity of exposure to clean-up work.

Limitations: The clinical significance of exhaled biomarkers and chromosomal findings are uncertain. The association between oil exposure and the observed changes may not be causal. The find- ings may not apply to spills involving other types of oil or to different populations of oil spill workers.

Conclusion: Participation in clean-up of a major oil spill was asso- ciated with persistent respiratory symptoms, elevated markers of airway injury in breath condensate, and chromosomal damage.

Primary Funding Source: Instituto de Salud Carlos III.

Ann Intern Med. 2010;153:489-498. www.annals.org For author affiliations, see end of text. * For members of the SEPAR-Prestige Study Group, see the Appendix (avail- able at www.annals.org). This article was published at www.annals.org on 24 August 2010.

More than 2.2 billion metric tons of oil is shipped bytanker every year around the world by a fleet com- prising more than 11 000 oil tankers. Between 1974 and 2008, more than 9000 tanker incidents were reported, among which 348 resulted in spills of more than 700 tons of oil (1). Oil spills cause great public concern, especially among people living in affected coastal areas, and large numbers of volunteers are mobilized to clean up the oil sediment.

Consequences of oil spills are usually evaluated in terms of environmental damage, effects on marine species, and economic losses, but relatively little is known about the effects of oil exposure on humans. Acute exposure to aromatic hydrocarbons, which are common constituents of oil, are known to cause respiratory symptoms (2). Certain volatile organic oil compounds, in particular benzene, are carcinogenic in humans and have been associated with he- matologic cancer (3). Exposure to polycyclic aromatic hy- drocarbons can damage the skin and mucous membranes and have been implicated in the pathogenesis of skin tumors (4).

In November 2002, the oil tanker Prestige foundered and spilled more than 67 000 tons of bunker oil, heavily contaminating the coast of Galicia in northwestern Spain. The spilled oil contained aromatic hydrocarbons (includ- ing benzene), saturated hydrocarbons, heavy metals, resins, and asphaltenes (5). More than 300 000 volunteers partic- ipated in clean-up activities; among them, local fishermen

See also:

Print Editors’ Notes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 490 Glossary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 490 Editorial comment. . . . . . . . . . . . . . . . . . . . . . . . . . 540 Summary for Patients. . . . . . . . . . . . . . . . . . . . . . . I-28

Web-Only Appendix Appendix Tables Appendix Figures Conversion of graphics into slides

Annals of Internal Medicine Original Research

Downloaded From: http://annals.org/ on 07/01/2016

were a large and highly exposed group. Studies of persons who participated in clean-up activities either as volunteers or as paid workers during the active period of clean-up showed that exposure to oil was associated with genomic damage (6 – 8), and a questionnaire that we distributed to fishermen showed increased rates of respiratory symptoms 1 to 2 years after participating in clean-up (9).

We sought to follow up those observations in this study of longer term health effects of the Prestige oil spill. Specifically, we evaluated changes in lung function; as- sessed respiratory markers of oxidative stress and airway inflammation in exhaled breath condensate (EBC) (10); and assessed chromosomal damage, a biomarker of in- creased risk for cancer (11, 12), 22 to 27 months after the spill in fishermen who had been highly exposed to oil dur- ing the clean-up work. We hypothesized that exposure to spilled oil would be associated with persistent abnormali- ties in lung function, inflammatory and oxidative changes in the airways, and evidence for genotoxicity similar to those reported in other occupational exposures to oil and its components (13–16).

METHODS Design and Participants

Study participants were fishermen who had taken part in a previous questionnaire survey that included qualita- tive and quantitative information about participation in clean-up activities (9). Using this self-reported informa- tion, we distinguished exposed from nonexposed individu- als (Figure). Exposed individuals (n � 1119) were mem- bers of fishermen cooperatives in heavily affected areas of the Atlantic coast who had participated at least 15 days in

clean-up activities, for 4 or more hours per day, including November and December 2002, when exposure presum- ably was greatest. Nonexposed fishermen (n � 577) were members of cooperatives in areas of the Cantabrian coast (which was less affected by the oil spill) who did not par- ticipate in clean-up activities for reasons other than those related to health. Among the 598 (53%) exposed and 205 (35%) nonexposed fishermen who agreed to participate in the study, 97 exposed and 28 nonexposed individuals re- ported inconsistencies in details of clean-up work in a sub- sequent interview and were excluded from this analysis,

Glossary

Aberrant metaphase: Metaphase with some chromosomal lesion or alteration.

Acentric chromosome/acentric fragments: Chromosome with no centromere.

Balanced chromosomal alterations: Exchange of segments between chromosomes so that no genetic material is lost or gained.

Banded chromosome: Chromosome that is clearly distinguishable from the others by showing darker or lighter regions obtained with banding techniques.

Banded metaphases: All chromosomes of the cell phase are banded. Banding techniques: Technical procedures that produce banding patterns on

metaphase chromosomes. Chromatid: One of 2 replicated arms of a chromosome. Chromatid break: Discontinuity of a single chromatid in which there is a

clear misalignment of 1 of the chromatids. Chromatid gap: Nonstaining region of a single chromatid in which there is

minimal misalignment of the chromatid. Chromosomal alteration: Change of chromosome number or structure. Chromosomal damage: Chromosomes with some lesion or structural

alteration. Chromosomal lesions: Chromosomes with gaps or breaks. Chromosomal unbalances: Chromosomes with loss or gain of genetic

material. Chromosome: Structure in which genes are located within the cell,

consisting of a highly compacted stretch of DNA with associated proteins. Chromosome break: Discontinuity at the same locus in both chromatids of a

single chromosome. Chromosome gap: Nonstaining region at the same locus in both chromatids

of a single chromosome in which there is minimal misalignment of the chromatids.

Chromosome preparation: Extension of a cell suspension on a slide. Cytogenetic: Pertaining to chromosomes. Deletion: Mutation due to loss of large chromosomal region. Destain: To remove the color from a chromosome preparation. G-banding: Chromosome staining by Giemsa resulting in characteristic

patterns of light and dark bands along the chromosome. Genotoxic effects: DNA damage produced by a toxic agent. Karyotype: Chromosome complement of a cell or organism; often

represented by an arrangement of metaphase chromosomes according to their lengths and the positions of their centromeres.

Leishman stain: Stain used in nonbanding technique. Marker chromosomes: Structurally abnormal chromosome that cannot be

identified or characterized by conventional banding cytogenetics. Metaphase: Stage of cell division when chromosomes are aligned at the

center of the cell before separation. Ring chromosome: Chromosome abnormality in which a ring forms after

breakage of both the long and the short arms. Structural chromosomal alteration: Significant change of chromosome

structure. Translocation: Exchange of segments between chromosomes. Unbalanced chromosomal alterations: Exchange of segments between

chromosomes, with loss or gain of genetic material. Uniform stain: Chromosomes stained by methods that do not produce

bands.

Context

Oil spills are ecological disasters, but their health effects on humans are not well known.

Contribution

This study found that Spanish fishermen who participated in the clean-up of a coastal oil spill had a higher preva- lence of respiratory symptoms, higher levels of markers suggestive of airway injury in exhaled breath condensate, and chromosomal alterations in lymphocytes than did those who did not participate in clean-up activities.

Caution

The clinical significance of the marker and chromosomal findings is not known. The study does not prove that oil exposure caused the abnormalities.

Implication

Participation in clean-up of a major oil spill seemed to have adverse health effects. The clinical significance of the findings is not known.

—The Editors

Original Research Health Changes in Fishermen 2 Years After Oil Spill Clean-up

490 19 October 2010 Annals of Internal Medicine Volume 153 • Number 8 www.annals.org

Downloaded From: http://annals.org/ on 07/01/2016

Figure. Study flow diagram.

Determination of chromosomal damage in lymphocytes (n = 91)¶

Final exposed study population (n = 501)‡

Inflammatory markers in exhaled breath

condensate (n = 77 [49])||

Questionnaire survey (9): 38 fishermen cooperatives

(n = 6780)

Lifetime nonsmokers (n = 230)§

Excluded smokers and former smokers (n = 271)

Excluded smokers and former smokers (n = 90)

Excluded (n = 97) Did not meet inclusion criteria according to the face-to-face interview

Highly exposed to clean-up work (n = 1330)*

Excluded (n = 2414) Participated <15 d or <4 h/d or not during November– December 2002

Participated in clean-up work (n = 3744)

Excluded (n = 1861) Not participants in clean-up work

15 most-affected cooperatives invited to participate

(n = 1119)

More-affected areas: 22 fishermen cooperatives

(n = 5605)

Agreed to participate (n = 598)

Excluded (n = 211) Belonged to 7 less-affected fishermen cooperatives

Excluded asthmatics (n = 15) Randomly selected subsample matched to nonexposed group by sex

Excluded neoplasms or not fertile (n = 41) Randomly selected subsample

Determination of chromosomal damage in lymphocytes (n = 46)¶

Final nonexposed study population (n = 177)‡

Inflammatory markers in exhaled breath

condensate (n = 79 [50])||

Lifetime nonsmokers (n = 87)§

Excluded (n = 28) Did not meet inclusion criteria according to the face-to-face interview

Not exposed to clean-up work (n = 577)†

Excluded (n = 61) Did not participate for health-related reasons

Did not participate in clean-up work (n = 638)

Excluded (n = 537) Participants in clean-up work

All 16 less-affected cooperatives invited to participate

(n = 577)

Less-affected areas: 16 fishermen cooperatives

(n = 1175)

Agreed to participate (n = 205)

Excluded asthmatics (n = 8)

Excluded neoplasms or not fertile (n = 12) Randomly selected subsample

* Participated in clean-up activities for at least 15 days, for 4 or more hours per day on average, including November and December 2002. † Did not participate in clean-up activities for non– health-related reasons. ‡ Met the inclusion criteria both at the questionnaire survey and at the face-to-face interview. § Participants reported never having smoked both at the questionnaire survey and at the face-to-face interview. || 8-Isoprostane was measured in all participants indicated; the numbers in brackets indicates those for whom additional analyses of cytokines and growth factors were done. ¶ Participants reported having children (which proved their fertility) and had no history of malignant neoplasms.

Original ResearchHealth Changes in Fishermen 2 Years After Oil Spill Clean-up

www.annals.org 19 October 2010 Annals of Internal Medicine Volume 153 • Number 8 491

Downloaded From: http://annals.org/ on 07/01/2016

leaving 501 exposed and 177 nonexposed persons in the final study population (Figure).

The study was performed between September 2004 and February 2005, 22 to 27 months after the spill and almost 2 years after most of the exposed participants came into contact with the oil (Appendix Figure 1, available at www.annals.org, shows the timing of events). A face-to- face interview was performed and outcome measures were obtained on the same day at the fishermen cooperative in a mobile unit that traveled to participants’ coastal villages. Because the coastal area affected by the oil spill was known, nurses obtaining the measures were not blinded to expo- sure status. The project was approved by the Ethics Com- mittee on Clinical Research of Galicia, and all participants provided written informed consent.

Interview and Clinical Testing All participants completed a second interviewer-led

questionnaire on respiratory symptoms and medication use, smoking habits, participation in clean-up activities, and characteristics of these activities. Items in this ques- tionnaire were the same as those in the previous survey (9). Participants underwent spirometry testing for FEV1 and FVC measurement and methacholine challenge; bronchial hyperresponsiveness was defined as a 20% decrease in FEV1 associated with a methacholine dose of 2 mg or less. Participants also had serum total IgE measurement and skin-prick testing for 19 common and occupational aller- gens to help distinguish intrinsic (atopic) from extrinsic (environmental) causes of symptoms. Atopy was defined as a positive reaction to at least 1 of the tested allergens (Ap- pendix Table 1, available at www.annals.org).

Assessment of Biomarkers in EBC We used EBC to assess respiratory biomarkers of oxi-

dative stress and inflammation. Samples were obtained by using an EcoScreen condenser (Jaeger, Würzburg, Ger- many) following current recommendations (17), through breathing at normal frequency and tidal volume until a total expired volume of 180 L was achieved. After collec- tion, the condensing device was centrifuged at 4 °C, and the resultant EBC volume was distributed in 1-mL aliquots and rapidly frozen in liquid nitrogen. All samples were lyophilized and stored at �80 °C before analysis.

We measured 8-isoprostane in a subsample of the pop- ulation by using an enzyme immunoassay after resuspen- sion with 400 �L of assay buffer. The subgroup comprised all 79 nonexposed individuals who were nonasthmatic and lifetime nonsmokers (75% women) and 77 exposed indi- viduals randomly selected from 230 exposed individuals who also were nonasthmatic and lifetime nonsmokers, matched by sex to the nonexposed group (Figure). 8-Isoprostane is a well-known stable product of local oxida- tive stress (18). We excluded smokers because of associations between smoking and markers of oxidative stress and inflam- mation (19).

Among 49 exposed participants and 50 nonexposed participants for whom sufficient EBC samples remained after 8-isoprostane testing, we also measured 10 cytokines and growth factors (interleukin-1�, 2, 4, 6, and 8; tumor necrosis factor-�; interferon-�; vascular endothelial growth factor [VEGF]; monocyte chemotactic protein-1; and basic fibroblast growth factor [bFGF]) that are representative of Th1/Th2 inflammation (17, 20) and airway remodeling (21). Investigators who obtained these measurements were blinded to exposure status and used the Cytometric Bead Arrays Flex System (BD Biosciences, Erembodegem, Bel- gium) after resuspension with 100 �L of human soluble protein buffer.

Assessment of Chromosomal Damage We assessed chromosomal damage in circulating lym-

phocytes. This measure of harm is often used in environ- mental studies and is an early marker of genotoxicity that has been associated with an increased risk for cancer (22– 24). The assessment was conducted in a preselected sub- sample of lifetime nonsmoking participants without a his- tory of cancer. We again excluded smokers because of associations between smoking and chromosomal damage; we also excluded participants who did not have children because infertile individuals may have an impaired ability to produce gametes. Thus, 91 exposed and 46 nonexposed participants were included (Figure). Assessment of chro- mosomal damage included evaluation of chromosomal lesions (chromatid gaps and breaks) and structural chromo- somal alterations (deletions, acentric fragments, translocations, and marker chromosomes) (25).

Peripheral blood samples were obtained, and within 48 hours, lymphocytes were cultured for 72 hours in RPMI-1640 medium (GIBCO Invitrogen Cell Culture, Invitrogen, Carlsbad, California), according to standard procedures. All cultures were performed in duplicate. Chromosome preparations were uniformly stained with Leishman stain (1:4 in Leishman buffer) in order to detect gaps and breaks. For each participant, at least 100 ran- domly selected metaphases were investigated. In addition, the same preparations were destained and reexamined after G-banding to further detect break-points involved in chro- mosomal lesions and to characterize structural chromo- somal alterations. At least 25 banded metaphases were karyotyped in each participant. The analyses were carried out independently by 2 trained evaluators who were blinded to exposure status. All aberrant metaphases were checked by 2 observers, and agreement was reached in cases of discordance. We verified that the number of evaluated metaphases did not differ between exposed and nonex- posed participants.

Statistical Analysis Differences in characteristics between participants and

nonparticipants and between exposed and nonexposed par- ticipants were evaluated by using chi-square tests for cate- gorical variables and t tests for continuous variables. Dif-

Original Research Health Changes in Fishermen 2 Years After Oil Spill Clean-up

492 19 October 2010 Annals of Internal Medicine Volume 153 • Number 8 www.annals.org

Downloaded From: http://annals.org/ on 07/01/2016

ferences in categorical health outcomes between exposed and nonexposed persons were expressed as adjusted abso- lute risk differences, estimated from multivariable general- ized linear models with the binomial family with identity link and regular variance estimates based on the expected information matrix, controlling for sex and smoking where applicable. If the models did not converge, sex as a covari- ate was removed. Differences in percentage of predicted lung function (26) between the 2 groups were evaluated by using multivariable linear regression analyses adjusted for pack-years smoked. Because the concentration of 8-isoprostane in EBC approximated a log-normal distribu- tion, log-transformed values of this variable were used throughout in analyses. Thus, group means were expressed as geometric means, and differences in 8-isoprostane level between groups were expressed as adjusted geometric mean ratios obtained from multivariable linear regression analysis of the log-transformed concentration adjusted for sex. Lev- els of each of the 10 markers measured in EBC were di- chotomized by using the lower limit of detection provided by the manufacturer as the cutoff. Potential dependence of the presence of chromosomal lesions and structural alter- ations between metaphases within individuals was evalu- ated by using the correlation matrix from generalized esti- mating equation analysis. Because no dependence could be demonstrated (correlation coefficient �0.01), associations between exposure to clean-up work and chromosomal damage at the individual level were determined by using generalized linear models as described. Dose–response re- lationships were investigated for major study outcomes in analyses by using 3 increasing categories of exposure inten- sity. The P value for linear trend was obtained from ad-

justed regression models that included the respective expo- sure index as a continuous variable. Analyses were done by using Stata SE, version 10.0 (StataCorp, College Station, Texas).

Role of the Funding Source The study was supported by grants from Instituto de

Salud Carlos III/European Regional Development Fund, Sociedad Española de Neumologı́a y Cirugı́a Torácica (SEPAR), and Centro de Investigación en Red de Enfer- medades Respiratorias. The sponsors had no role in study design, data collection, data analysis, or data interpretation or in the writing of the report.

RESULTS Study participants were more likely than nonpartici-

pants in both the exposed and nonexposed groups to be female and never-smokers (P � 0.001); age distribution was similar. Nonexposed participants reported lower respi- ratory tract symptoms in the questionnaire survey (9) more often than nonparticipants (30% vs. 19%; P � 0.003), whereas the prevalence of lower respiratory tract symptoms was similar in exposed participants and nonexposed partic- ipants (37% vs. 38%; P � 0.80).

A higher proportion of participants exposed to clean-up work were men and were younger than those who were not exposed (Table 1). Female participants were 8 years older than men on average. Exposed and nonexposed persons did not significantly differ in smoking history, al- though a higher proportion of men than women were cur- rent smokers (39.8% vs. 12.8%) and former smokers (28.7% vs. 9.5%). Geometric mean serum total IgE levels

Table 1. Participant Characteristics

Characteristic All Participants (n � 678) Lifelong Nonsmokers (n � 317)

Exposed (n � 501)

Nonexposed (n � 177)

P Value* Exposed (n � 230)

Nonexposed (n � 87)

P Value*

Women, n (%) 141 (28.1) 80 (45.2) �0.001 109 (47.4) 64 (73.6) �0.001

Mean age (SD), y All participants 44.7 (11.4) 47.3 (10.6) 0.007 46.5 (11.6) 51.4 (9.4) 0.001 Women 51.0 (8.9) 51.3 (9.3) 0.85 53.0 (8.3) 53.9 (6.7) 0.48 Men 42.1 (11.3) 44.0 (10.6) 0.15 40.6 (11.1) 44.5 (12.2) 0.13

Smoking status, n (%) 0.49 Former 118 (23.6) 34 (19.2) – – – Current smokers 153 (30.5) 56 (31.6) – – –

Ever had asthma, n (%) 34 (6.8) 10 (5.6) 0.59 15 (6.5) 6 (6.9) 0.90

Participation in clean-up work Median days of clean-up work (range) 87 (15–429) 0 – 90 (15–429) 0 – Median hours per day of clean-up work (range) 6 (4–18) 0 – 6 (4–14) 0 – Median types of clean-up activity (range), n† 5 (1–10) 0 – 5 (1–9) 0 – Used facemask often or always, n (%) 166 (33.1) 0 – 81 (35.2) 0 –

* Obtained from chi-square tests for categorical variables and t tests for continuous variables. † The most common clean-up activities were gathering oil from coastal rocks (83%), transporting the gathered oil (81%), gathering oil from beaches (73%), gathering oil from the sea (67%), cleaning work clothes or boots that were used during the gathering of oil (45%), and cleaning boats used for gathering oil (43%).

Original ResearchHealth Changes in Fishermen 2 Years After Oil Spill Clean-up

www.annals.org 19 October 2010 Annals of Internal Medicine Volume 153 • Number 8 493

Downloaded From: http://annals.org/ on 07/01/2016

were 32 IU/mL in exposed participants and 24 IU/mL in nonexposed participants (P � 0.021); atopy (a positive re- action to at least 1 skin prick antigen) was present in 22.7% and 17.0%, respectively (P � 0.12). After sex, age, and smoking status were controlled for, the difference in serum total IgE level between exposed and nonexposed participants lost statistical significance (P � 0.086).

Participants exposed to oil had an increased risk for lower respiratory tract symptoms (adjusted risk difference, 8.0 [95% CI, 1.1 to 14.8]), a finding also evident in the subsample of nonsmokers (Table 2). No statistically signif- icant differences were found between exposed and nonex- posed participants in nasal symptoms, medication use, or lung function. In sensitivity analyses of all associations re- ported in Table 2 that were restricted to nonasthmatics, associations between exposure status and respiratory health outcomes remained similar in magnitude and direction (data not shown).

Exposed participants had statistically significantly higher concentrations of 8-isoprostane in EBC than non- exposed participants (geometric mean ratio, 2.5 [CI, 1.7 to 3.7]) (Table 3), and a higher proportion of exposed par- ticipants had measurable VEGF levels (risk difference, 44.8 [CI, 27.9 to 61.6]) and bFGF (risk difference, 16.0 [CI, 3.5 to 28.6]); differences between groups in other biomar- kers were not statistically significant. Exposed partici- pants with lower respiratory tract symptoms had higher 8-isoprostane levels in EBC than did those without symp- toms (geometric mean, 34 vs. 11 pg/mL; adjusted geomet- ric mean ratio, 3.2 [CI, 1.5 to 6.8]), a difference not found in nonexposed participants (geometric mean, 7.3 vs. 5.4 pg/mL; geometric mean ratio, 1.3 [CI, 0.7 to 2.4]) (Ap- pendix Figure 2, available at www.annals.org).

There were no statistically significant differences be- tween exposed and nonexposed participants in chromo- somal lesions, whereas a higher proportion of exposed par- ticipants had structural chromosomal alterations (adjusted risk difference, 27.4 [CI, 10.0 to 44.8]) (Table 4), primar- ily chromosomal imbalances (translocations, acentric frag- ments, deletions, and markers). Twelve exposed partici- pants and 1 nonexposed participant had metaphases with multiple structural chromosomal alterations when evalu- ated by using uniform stain.

The risk for several study outcomes increased with the degree of exposure to clean-up work (Appendix Table 2, available at www.annals.org). A statistically significant lin- ear trend was seen for 8-isoprostane concentration and measurable VEGF or bFGF levels in EBC, as well as struc- tural chromosomal alterations in lymphocytes, when the number of hours per day and the number of different clean-up activities was evaluated.

DISCUSSION This study of health effects among fishermen who par-

ticipated in clean-up of the Prestige oil spill about 2 years earlier confirms previously reported findings of an increase in respiratory symptoms and newly demonstrates increased 8-isoprostane levels and growth factor activity in EBC and more structural chromosomal alterations in circulating lymphocytes among fishermen exposed to the oil.

Compared with unexposed fishermen, a greater pro- portion of persons who voluntarily participated in clean-up activities of the oil spill had lower respiratory tract symp- toms 2 years later. Consistent with the results of our first cross-sectional survey (9), the increase in respiratory symp- toms was not explained by an increased prevalence of

Table 2. Associations Between Exposure to Clean-up Work and Respiratory Outcomes

Variable All Participants (n � 678) Lifelong Nonsmokers (n � 317)

Exposed (n � 501)

Nonexposed (n � 177)

Risk Difference (95% CI)*

Exposed (n � 230)

Nonexposed (n � 87)

Risk Difference (95% CI)†

Lower respiratory tract symptoms, n (%)‡ All types of symptoms 132 (26.6) 37 (20.9) 8.0 (1.1 to 14.8) 51 (22.3) 14 (16.1) 8.8 (0.1 to 17.4) Asthma-like symptoms 101 (20.2) 32 (18.1) 4.1 (�2.4 to 10.5) 41 (17.8) 13 (14.9) 5.0 (�3.1 to 13.1) Bronchitis-like symptoms 64 (12.9) 17 (9.6) 4.2 (�0.4 to 8.7)§ 19 (8.3) 4 (4.6) 3.7 (�2.0 to 9.4)||

Nasal symptoms, n (%) 128 (25.7) 42 (23.9) 1.4 (�5.9 to 8.7) 55 (23.9) 22 (25.6) �0.9 (�11.5 to 9.7) Use of inhaled medication, n (%) 41 (8.2) 9 (5.1) 2.9 (�1.1 to 7.0) 17 (7.4) 7 (8.0) �1.0 (�7.4 to 5.4) Use of oral medication, n (%) 28 (5.6) 11 (6.2) �0.7 (�4.8 to 3.5) 12 (5.2) 3 (3.4) 1.8 (�2.9 to 6.6) Mean FEV1 (SD), % predicted¶ 101.3 (16.5) 101.5 (17.0) �0.2 (�2.9 to 2.5)** 103.6 (15.5) 105.0 (17.4) �1.2 (�5.2 to 2.8)** Mean FVC (SD), % predicted¶ 98.7 (14.1) 99.7 (14.7) �0.9 (�3.3 to 1.5)** 98.7 (14.2) 100.3 (16.7) �1.6 (�5.3 to 2.1)** Bronchial hyperresponsiveness, n (%)†† 79 (18.2) 24 (15.5) 3.4 (�2.1 to 8.9) 30 (15.2) 7 (9.1) 6.1 (�2.1 to 14.2)||

* Adjusted for sex and smoking status. † Adjusted for sex. ‡ Wheeze with breathlessness, wheeze apart from colds, or nocturnal attacks of shortness of breath (asthma-like symptoms) and chronic cough or chronic phlegm (bronchitis-like symptoms). § Adjusted for smoking status only (the fully adjusted model did not converge). || Unadjusted estimate (the fully adjusted model did not converge). ¶ Based on 670 measurements for all participants and 312 measurements for lifelong nonsmokers. ** Difference (regression coefficient) in predicted lung function, adjusted for pack-years of smoking in analysis of all participants. †† Methacholine dose of 2 mg or less causing a 20% decrease in FEV1. Based on 589 measurements for all participants and 275 measurements for lifelong nonsmokers.

Original Research Health Changes in Fishermen 2 Years After Oil Spill Clean-up

494 19 October 2010 Annals of Internal Medicine Volume 153 • Number 8 www.annals.org

Downloaded From: http://annals.org/ on 07/01/2016

chronic respiratory diseases in the exposed group. We could not demonstrate reductions in lung function values associated with exposure to clean-up work, consistent with a previous study that failed to demonstrate changes in air- flow among residents of a coastal area affected by the Braer oil spill (27, 28) and inconsistent with another study show- ing lower lung function values in workers involved in the Tasman Spirit oil spill clean-up (29). A subgroup of life- time nonsmokers had a non–statistically significant higher risk for bronchial hyperresponsiveness in response to methacholine challenge, suggesting that compounds of spilled oil might act as respiratory irritants that increase bronchial reactivity, at least in persons not chronically ex- posed to cigarette smoke. This finding is consistent with previously demonstrated associations between various aro- matic hydrocarbons present in spilled oil and airway in- flammation and hyperresponsiveness (30, 31).

The previous respiratory symptom findings are rein- forced by the new observation of increased 8-isoprostane levels and growth factor activity in EBC of a subsample of nonasthmatic, nonsmoking exposed participants; the in- creased 8-isoprostane levels in this subgroup seem to be related to some measures of intensity of exposure. High levels of 8-isoprostane in EBC are considered to reflect local oxidative stress (32), and increased concentrations have been reported in a variety of inflammatory airway diseases (18, 19, 33, 34), in cigarette smokers (19), and in some occupational exposures (35). Of note, 8-isoprostane levels were associated with the presence of lower respiratory tract symptoms in exposed participants. These findings suggest that among the fishermen who participated in clean-up activities after the Prestige oil spill, exposure to oil

products over days to a few months might have contrib- uted to respiratory oxidative changes that were still mea- surable 2 years later. The association between exhaled 8-isoprostane levels and respiratory symptoms also suggests that oil-induced oxidative stress underlies the increased re- spiratory morbidity in exposed fishermen. Although we ac- knowledge that the clinical meaning of the measures is unclear, our findings suggest a potential role for measuring EBC biomarkers in epidemiologic studies.

The growth factors bFGF and VEGF were also in- creased in EBC in an even smaller subgroup of nonasth- matic, nonsmoking exposed participants. Basic fibroblast growth factor is involved in angiogenesis and stimulates the proliferation, migration, and differentiation of epithelial cells and fibroblasts. In asthma, epithelial injury by inflam- mation products or environmental agents can stimulate the release of a range of growth factors, including bFGF, that are active on fibroblasts and smooth-muscle cells, leading to airway remodeling (21). Accordingly, we speculate that increased bFGF levels in EBC might reflect an ongoing process of airway-wall remodeling. Regarding VEGF, little is known about its origins in biological lung fluids. It has been associated with asthma (36) and lung cancer (37), suggesting stimulated angiogenesis. In our study, the pres- ence of VEGF in EBC might be related to airway remod- eling; the origin and pathobiological significance of this finding merit further investigation.

We also detected an increased risk for structural chro- mosomal alterations in circulating lymphocytes among ex- posed workers, which had a dose-dependent relationship with some measure of intensity of exposure; frequency of chromosomal alterations in nonexposed persons was within

Table 3. Respiratory Biomarkers in Exhaled Breath Condensate of Exposed and Nonexposed Participants Without Asthma Who Are Lifelong Nonsmokers

Biomarker Geometric Mean Level (95% CI), pg/mL

Geometric Mean Ratio (95% CI)*

Exposed (n � 77)

Nonexposed (n � 79)

8-Isoprostane 14 5.6 2.5 (1.7 to 3.7)

Biomarker Lower Limit of Detection, pg/mL

Participants With Levels Above the Limit of Detection, n (%)

Risk Difference (95% CI)*

Exposed (n � 49)

Nonexposed (n � 50)

Interleukin-1� 2.3 2 (4) 2 (4) 0.1 (�6.8 to 6.9) Interleukin-2 11.2 7 (14) 4 (8) 4.0 (�6.9 to 14.8) Interleukin-4 1.4 8 (16) 8 (16) 1.5 (�10.7 to 13.8) Interleukin-6 1.9 0 (0) 0 (0) NE Interleukin-8 1.2 6 (12) 3 (6) 3.5 (�5.3 to 12.2) Tumor necrosis factor-� 0.7 2 (4) 3 (6) �0.5 (�7.5 to 6.5) Interferon-� 1.8 13 (27) 8 (16) 11.4 (�2.9 to 25.7) Vascular endothelial growth factor 4.5 29 (59) 7 (14) 44.8 (27.9 to 61.6) Monocyte chemotactic protein-1 1.3 15 (31) 10 (20) 9.0 (�6.5 to 24.6) Basic fibroblast growth factor 3.4 11 (22) 3 (6) 16.0 (3.5 to 28.6)

NE � not estimable. * Adjusted for sex.

Original ResearchHealth Changes in Fishermen 2 Years After Oil Spill Clean-up

www.annals.org 19 October 2010 Annals of Internal Medicine Volume 153 • Number 8 495

Downloaded From: http://annals.org/ on 07/01/2016

the normal range (22). Chromosomal damage in circulat- ing lymphocytes is an early marker of genotoxicity associ- ated with increased risk for cancer (22–24). Genotoxicity studies, particularly those focusing on structural chromo- somal alterations, are standard ways to assess risk for cancer in potentially toxic occupational or environmental expo- sures (12, 23). Genotoxic studies in volunteers exposed to oil spills are scarce (38). It is well-known that chromo- somal damage is increased in cigarette smokers (39) and is associated with occupational exposure to benzene (13–16, 40) or coal combustion products (41), both of which con- tain aromatic hydrocarbons that are an important constit- uent of the oil spilled by the Prestige tanker. Genome dam- age related to early exposure has been shown in clean-up workers of the Prestige oil spill during the clean-up (6, 8); the chromosomal alterations we analyzed provide addi- tional information on more persistent adverse health effects (11, 12).

We did not anticipate finding increased chromosomal damage, in particular unbalanced alterations, 2 years after exposure. Little information exists regarding the cytoge- netic effects induced by acute exposure to chemical clasto- gens (42). Persistent unstable chromosome alterations in

lymphocytes 20 years after acute benzene exposure have been reported (42). The precise pathways involved in chro- mosomal damage remain to be elucidated, but possible mechanistic explanations for the findings include persis- tence of genotoxic effects of acute exposure (43); persis- tence of chemical compounds leading to continuous expo- sure; genotoxic effects on bone marrow progenitor cells; and dysfunction in DNA repair proteins or alterations in epigenetic factors induced by exposure to oil.

The clinical implications of chromosomal damage in individuals exposed to oil while cleaning up spills are un- known. Chromosomal damage is a characteristic feature of cancer cells and is crucial for tumor pathogenesis (44). An increased frequency of chromosomal alterations in circulat- ing lymphocytes has been associated with augmented can- cer risk (45). The distribution of chromosomal damage is assumed to be similar in different tissues (11). Accordingly, we speculate that chromosomal damage detected in circu- lating lymphocytes might reflect a more general increased risk for cancer. Because the possibility of a higher risk for cancer in exposed workers cannot be excluded, a surveil- lance program in the target population would be appropri- ate. Follow-up studies to evaluate persistent respiratory

Table 4. Associations Between Exposure to Clean-up Work and Chromosomal Lesions and Structural Alterations in Lymphocytes of Lifelong Nonsmoking Participants*

Chromosomal Lesions† Exposed Participants Nonexposed Participants Risk Difference (95% CI)‡

Total (n � 91), n (%)

Metaphases With Chromosomal Lesions (n � 9520), n

Total (n � 46), n (%)

Metaphases With Chromosomal Lesions (n � 4859), n

Absent 42 (46) 0 26 (57) 0 0.0 (reference) Present 49 (54) 84 20 (43) 32 10.8 (�7.4 to 28.9) Gaps 31 (34) 44 15 (33) 18 5.4 (�14.0 to 24.7) Chromatid gap 27 (30) 38 11 (24) 13 9.9 (�9.4 to 29.2) Chromosome gap 6 (7) 7 5 (11) 5 �4.4 (�20.8 to 12.1) Breaks 31 (34) 45 13 (28) 14 10.9 (�8.2 to 29.9) Chromatid break 17 (19) 21 8 (17) 8 5.7 (�12.9 to 24.2) Chromosome break 18 (20) 27 7 (15) 7 9.9 (�8.9 to 28.7)

Structural Chromosomal Alterations§

Exposed Participants Nonexposed Participants Risk Difference (95% CI)‡

Total (n � 91), n (%)

Metaphases With Structural Chromosomal Alterations (n � 2448), n

Total (n � 46), n (%)

Metaphases With Structural Chromosomal Alterations (n � 1285), n

Absent 27 (30) 0 25 (54) 0 0.0 (reference) Present 64 (70) 136 21 (46) 27 27.4 (10 to 44.8) Balanced abnormalities 11 (12) 12 7 (15) 7 7.1 (�13.3 to 27.4)|| Unbalanced abnormalities 62 (68) 125 17 (37) 20 31.3 (13.4 to 49.1) Deletions 29 (32) 32 8 (17) 9 27.6 (7.3 to 48.0) Translocations 24 (26) 29 4 (9) 4 33.3 (14.7 to 51.8)|| Rings 8 (9) 8 0 (0) 0 NE Acentric fragments 18 (20) 25 0 (0) 0 NE Markers 30 (33) 37 9 (20) 9 27.8 (8.3 to 47.2)

NE � not estimable. * Participants reported never having smoked both at the questionnaire survey and at the face-to-face interview, had children (proven fertility), and had no history of cancer. † Detected by using uniform stain. ‡ Adjusted for sex. § Detected by using G-banded karyotypes. || Unadjusted estimate (fully adjusted model did not converge).

Original Research Health Changes in Fishermen 2 Years After Oil Spill Clean-up

496 19 October 2010 Annals of Internal Medicine Volume 153 • Number 8 www.annals.org

Downloaded From: http://annals.org/ on 07/01/2016

health effects, chromosomal damage, and the development of cancer in these individuals for longer periods are cur- rently under way.

Our study has limitations. First, its design cannot es- tablish that any of the observed associations are causal, and the findings cannot be extrapolated to other populations of clean-up workers or to the general population living in the area of the oil spill. In addition, the findings cannot be extrapolated to spills of other types of oil. For example, the Prestige tanker contained bunker C oil, whereas oil spilled in the 2010 U.S. Deepwater Horizon disaster is crude oil. The main components of the oil spills are presumably sim- ilar, but proportions of those components (for example, volatile organic compounds, polycyclic aromatic hydrocar- bons, hydrogen sulfide gas, and heavy metals) are probably different, as might be dispersants to break up the oil slick and the proportion of oil that evaporates and could be inhaled by humans. Findings from this study therefore cannot predict what effects individuals exposed to other oil spills, such as that in the Gulf of Mexico might experience.

In summary, 2 years after participating in clean-up efforts of the Prestige oil spill, exposed fishermen had in- creased prevalence rates of respiratory symptoms and bi- omarkers of pulmonary oxidative stress and growth factor activity, suggesting persistent airway injury. In addition, they had more structural chromosomal abnormalities in circulating lymphocytes. Our findings indicate that expo- sure to oil sediments, even for short periods, may have detrimental health effects. To fully understand the impor- tance and nature of these effects, further longitudinal and mechanistic research in similar episodes is warranted. Be- cause, unfortunately, oil spills will most likely occur again, it is crucial that the authorities responsible for organizing clean-up operations take appropriate measures to guarantee the health protection of persons involved in the clean-up activities and to establish registries to systematically assess possible adverse health outcomes in exposed workers over time.

From Complexo Hospitalario Universitario A Coruña, A Coruña, Spain; Hospital Clı́nico San Carlos and University Hospital 12 de Octubre, Madrid, Spain; Centre for Research in Environmental Epidemiology (CREAL), Municipal Institute of Medical Research (IMIM-Hospital del Mar), Centro de Investigación Biomédica en Red de Epidemiologı́a y Salud Pública (CIBERESP), Hospital Clı́nic-Institut d’Investigacions Biomèdiques August Pi i Sunyer (IDIBAPS), Autonomous University of Barcelona, and Pompeu Fabra University, Barcelona, Spain; and Centro de Investigación Biomédica en Red de Enfermedades Respiratorias (CIBERES), Bunyola, Mallorca, Spain.

Acknowledgment: The authors thank the participating fishermen, the staff of their cooperatives, and the efforts made by A. Devesa; the SEPAR-Prestige data collection team (Y. Torralba, A. Souto, M. Rodrı́guez-Valcárcel, and L. Vázquez-Rey) for their outstanding work; the staff of Complexo Hospitalario Universitario A Coruña (S. Lamela, E. Rodrı́guez, M. Garea, and M. Saleta) for their support; D. Macfarlane for assistance with data management and reviewing the manuscript; A. Espinosa for her contribution to the statistical analyses; A. Serrano-

Mollar and G. Gay for analysis of EBC; A. López-Rodrı́guez and C. Rodrı́guez-Escudero, formerly of Servizo Galego de Saúde, for their valu- able collaboration; and J.L. Alvarez-Sala and J. Ancochea, past presidents of SEPAR, for their initiative and continuous support.

Grant Support: By Instituto de Salud Carlos III/European Regional Development Fund (grants PI03/1685 and PI05/0548), SEPAR, and Centro de Investigación en Red de Enfermedades Respiratorias. The Instituto de Salud Carlos III/European Regional Development Fund also provided grants to Dr. Rodrı́guez-Trigo (grant BAE 06/90018) and Dr. Zock (grant Miguel Servet 01-3058).

Potential Conflicts of Interest: Disclosures can be viewed at www.acponline .org/authors/icmje/ConflictOfInterestForms.do?msNum�M10-1030.

Reproducible Research Statement: Study protocol and statistical code: Available from Dr. Zock (e-mail, [email protected]). Data set: Not available.

Requests for Single Reprints: Joan Albert Barberà, MD, Servei de Pneumologia, Hospital Clı́nic, Villarroel 170, 08036 Barcelona, Spain; e-mail, [email protected].

Current author addresses and author contributions are available at www .annals.org.

References 1. The International Tanker Owners Pollution Federation. Oil Tanker Spill Statistics: 2008. London: The International Tanker Owners Pollution Federa- tion; 2009. 2. Lyons RA, Temple JM, Evans D, Fone DL, Palmer SR. Acute health effects of the Sea Empress oil spill. J Epidemiol Community Health. 1999;53:306-10. [PMID: 10396538] 3. International Agency for Research on Cancer. IARC Monographs on the Overall Evaluations of Carcinogenicity: An Updating of IARC Monographs Vol- umes 1 to 42. IARC Monographs on the Evaluation of Carcinogenic Risk to Humans (Supplement 7). Lyon: International Agency for Research on Cancer; 1998. 4. Agency for Toxic Substances and Disease Registry. Toxicological profile for fuel oils. Atlanta: U.S. Department of Health and Human Services, Public Health Service; 1995. 5. Rodrı́guez-Trigo G, Zock JP, Isidro Montes I. [Health effects of exposure to oil spills]. Arch Bronconeumol. 2007;43:628-35. [PMID: 17983548] 6. Laffon B, Fraga-Iriso R, Pérez-Cadahı́a B, Méndez J. Genotoxicity associated to exposure to Prestige oil during autopsies and cleaning of oil-contaminated birds. Food Chem Toxicol. 2006;44:1714-23. [PMID: 16814914] 7. Pérez-Cadahı́a B, Lafuente A, Cabaleiro T, Pásaro E, Méndez J, Laffon B. Initial study on the effects of Prestige oil on human health. Environ Int. 2007; 33:176-85. [PMID: 17055056] 8. Pérez-Cadahı́a B, Laffon B, Valdiglesias V, Pásaro E, Méndez J. Cytogenetic effects induced by Prestige oil on human populations: the role of polymorphisms in genes involved in metabolism and DNA repair. Mutat Res. 2008;653:117-23. [PMID: 18495522] 9. Zock JP, Rodriguez-Trigo G, Pozo-Rodriguez F, Barbera JA, Bouso L, Tor- ralba Y, et al; SEPAR-Prestige Study Group. Prolonged respiratory symptoms in clean-up workers of the prestige oil spill. Am J Respir Crit Care Med. 2007;176: 610-6. [PMID: 17556713] 10. Antczak A, Górski P. Markers of pulmonary diseases in exhaled breath con- densate. Int J Occup Med Environ Health. 2002;15:317-23. [PMID: 12608619] 11. Norppa H, Bonassi S, Hansteen IL, Hagmar L, Strömberg U, Rössner P, et al. Chromosomal aberrations and SCEs as biomarkers of cancer risk. Mutat Res. 2006;600:37-45. [PMID: 16814813] 12. Au WW. Usefulness of biomarkers in population studies: from exposure to susceptibility and to prediction of cancer. Int J Hyg Environ Health. 2007;210: 239-46. [PMID: 17174154]

Original ResearchHealth Changes in Fishermen 2 Years After Oil Spill Clean-up

www.annals.org 19 October 2010 Annals of Internal Medicine Volume 153 • Number 8 497

Downloaded From: http://annals.org/ on 07/01/2016

13. Holecková B, Piesová E, Sivikova K, Dianovsky J. Chromosomal aberrations in humans induced by benzene. Ann Agric Environ Med. 2004;11:175-9. [PMID: 15627321] 14. Tompa A, Jakab MG, Major J. Risk management among benzene-exposed oil refinery workers. Int J Hyg Environ Health. 2005;208:509-16. [PMID: 16325561] 15. Roma-Torres J, Teixeira JP, Silva S, Laffon B, Cunha LM, Méndez J, et al. Evaluation of genotoxicity in a group of workers from a petroleum refinery aro- matics plant. Mutat Res. 2006;604:19-27. [PMID: 16431152] 16. McHale CM, Lan Q, Corso C, Li G, Zhang L, Vermeulen R, et al. Chromosome translocations in workers exposed to benzene. J Natl Cancer Inst Monogr. 2008:74-7. [PMID: 18648008] 17. Horvath I, Hunt J, Barnes PJ, Alving K, Antczak A, Baraldi E, et al; ATS/ERS Task Force on Exhaled Breath Condensate. Exhaled breath conden- sate: methodological recommendations and unresolved questions. Eur Respir J. 2005;26:523-48. [PMID: 16135737] 18. Montuschi P, Corradi M, Ciabattoni G, Nightingale J, Kharitonov SA, Barnes PJ. Increased 8-isoprostane, a marker of oxidative stress, in exhaled con- densate of asthma patients. Am J Respir Crit Care Med. 1999;160:216-20. [PMID: 10390403] 19. Montuschi P, Collins JV, Ciabattoni G, Lazzeri N, Corradi M, Kharitonov SA, et al. Exhaled 8-isoprostane as an in vivo biomarker of lung oxidative stress in patients with COPD and healthy smokers. Am J Respir Crit Care Med. 2000; 162:1175-7. [PMID: 10988150] 20. Idali F, Wikén M, Wahlström J, Mellstedt H, Eklund A, Rabbani H, et al. Reduced Th1 response in the lungs of HLA-DRB1*0301 patients with pulmo- nary sarcoidosis. Eur Respir J. 2006;27:451-9. [PMID: 16507843] 21. Fedorov IA, Wilson SJ, Davies DE, Holgate ST. Epithelial stress and struc- tural remodelling in childhood asthma. Thorax. 2005;60:389-94. [PMID: 15860714] 22. Hagmar L, Strömberg U, Bonassi S, Hansteen IL, Knudsen LE, Lindholm C, et al. Impact of types of lymphocyte chromosomal aberrations on human cancer risk: results from Nordic and Italian cohorts. Cancer Res. 2004;64:2258- 63. [PMID: 15026371] 23. Norppa H. Cytogenetic biomarkers and genetic polymorphisms. Toxicol Lett. 2004;149:309-34. [PMID: 15093278] 24. Rossner P, Boffetta P, Ceppi M, Bonassi S, Smerhovsky Z, Landa K, et al. Chromosomal aberrations in lymphocytes of healthy subjects and risk of cancer. Environ Health Perspect. 2005;113:517-20. [PMID: 15866756] 25. Shaffer LG, Slovak ML, Campbell LJ. ISCN 2009. An International System for Human Cytogenetic Nomenclature: Recommendations of the International Standing Committee on Human Cytogenetic Nomenclature. Basel: Karger; 2009. 26. Roca J, Sanchis J, Agusti-Vidal A, Segarra F, Navajas D, Rodriguez-Roisin R, et al. Spirometric reference values from a Mediterranean population. Bull Eur Physiopathol Respir. 1986;22:217-24. [PMID: 3730638] 27. Campbell D, Cox D, Crum J, Foster K, Christie P, Brewster D. Initial effects of the grounding of the tanker Braer on health in Shetland. The Shetland Health Study Group. BMJ. 1993;307:1251-5. [PMID: 8281057] 28. Campbell D, Cox D, Crum J, Foster K, Riley A. Later effects of grounding of tanker Braer on health in Shetland. BMJ. 1994;309:773-4. [PMID: 7950562] 29. Meo SA, Al-Drees AM, Meo IM, Al-Saadi MM, Azeem MA. Lung function in subjects exposed to crude oil spill into sea water. Mar Pollut Bull. 2008;56:88-

94. [PMID: 18031764] 30. Jang AS, Choi IS, Koh YI, Park CS. Volatile organic compounds contribute to airway hyperresponsiveness. Korean J Intern Med. 2007;22:8-12. [PMID: 17427638] 31. Podechard N, Lecureur V, Le Ferrec E, Guenon I, Sparfel L, Gilot D, et al. Interleukin-8 induction by the environmental contaminant benzo(a)pyrene is aryl hydrocarbon receptor-dependent and leads to lung inflammation. Toxicol Lett. 2008;177:130-7. [PMID: 18289803] 32. Koutsokera A, Loukides S, Gourgoulianis KI, Kostikas K. Biomarkers in the exhaled breath condensate of healthy adults: mapping the path towards reference values. Curr Med Chem. 2008;15:620-30. [PMID: 18336277] 33. Barreto M, Villa MP, Olita C, Martella S, Ciabattoni G, Montuschi P. 8-Isoprostane in exhaled breath condensate and exercise-induced bronchocon- striction in asthmatic children and adolescents. Chest. 2009;135:66-73. [PMID: 18753466] 34. Biernacki WA, Kharitonov SA, Barnes PJ. Increased leukotriene B4 and 8-isoprostane in exhaled breath condensate of patients with exacerbations of COPD. Thorax. 2003;58:294-8. [PMID: 12668789] 35. Pelclová D, Fenclová Z, Kacer P, Kuzma M, Navrátil T, Lebedová J. Increased 8-isoprostane, a marker of oxidative stress in exhaled breath condensate in subjects with asbestos exposure. Ind Health. 2008;46:484-9. [PMID: 18840939] 36. Leung TF, Wong GW, Ko FW, Li CY, Yung E, Lam CW, et al. Analysis of growth factors and inflammatory cytokines in exhaled breath condensate from asthmatic children. Int Arch Allergy Immunol. 2005;137:66-72. [PMID: 15832052] 37. Beinert T, Binder D, Oehm C, Ziemer S, Priem F, Schweigert M, et al. Increased levels of vascular endothelial growth factor in bronchoalveolar lavage of patients with bronchial carcinoma effect of tumour activity and oxidative stress due to radio-chemotherapy? Eur J Med Res. 1999;4:328-34. [PMID: 10471544] 38. Cole J, Beare DM, Waugh AP, Capulas E, Aldridge KE, Arlett CF, et al. Biomonitoring of possible human exposure to environmental genotoxic chemi- cals: lessons from a study following the wreck of the oil tanker Braer. Environ Mol Mutagen. 1997;30:97-111. [PMID: 9329634] 39. Littlefield LG, Joiner EE. Analysis of chromosome aberrations in lympho- cytes of long-term heavy smokers. Mutat Res. 1986;170:145-50. [PMID: 3713724] 40. Zhang L, Eastmond DA, Smith MT. The nature of chromosomal aberra- tions detected in humans exposed to benzene. Crit Rev Toxicol. 2002;32:1-42. [PMID: 11846214] 41. Celik M, Donbak L, Unal F, Yüzbasioglu D, Aksoy H, Yilmaz S. Cytoge- netic damage in workers from a coal-fired power plant. Mutat Res. 2007;627: 158-63. [PMID: 17178253] 42. Forni A. Benzene-induced chromosome aberrations: a follow-up study. En- viron Health Perspect. 1996;104 Suppl 6:1309-12. [PMID: 9118911] 43. Sprent J, Tough DF. Lymphocyte life-span and memory. Science. 1994;265: 1395-400. [PMID: 8073282] 44. Fröhling S, Döhner H. Chromosomal abnormalities in cancer. N Engl J Med. 2008;359:722-34. [PMID: 18703475] 45. Bonassi S, Norppa H, Ceppi M, Strömberg U, Vermeulen R, Znaor A, et al. Chromosomal aberration frequency in lymphocytes predicts the risk of cancer: results from a pooled cohort study of 22 358 subjects in 11 countries. Carcinogenesis. 2008;29:1178-83. [PMID: 18356148]

Original Research Health Changes in Fishermen 2 Years After Oil Spill Clean-up

498 19 October 2010 Annals of Internal Medicine Volume 153 • Number 8 www.annals.org

Downloaded From: http://annals.org/ on 07/01/2016

Current Author Addresses: Dr. Rodrı́guez-Trigo: Servicio de Neumo- logı́a, Hospital Clı́nico San Carlos, Profesor Martin Lago s/n, 28040 Madrid, Spain. Drs. Zock and Antó and Ms. Bouso: Centre for Research in Environ- mental Epidemiology (CREAL), Doctor Aiguader, 88, 08003 Barcelona, Spain. Dr. Pozo-Rodrı́guez: Servicio de Neumologı́a y Unidad de Epidemiolo- gı́a Clı́nica, Hospital 12 de Octubre, Avenida de Córdoba s/n, 28041 Madrid, Spain. Drs. Gómez and Barberà: Servei de Pneumologia, Hospital Clı́nic, Vil- larroel 170, 08036 Barcelona, Spain. Ms. Monyarch and Drs. Coll and Fuster: Unidad de Biologı́a Celular y Genética Médica, Facultad de Medicina, Universidad Autónoma de Bar- celona, Campus de Bellaterra, 08193 Cerdanyola del Vallés, Barcelona, Spain. Dr. Verea: Servicio de Neumoloxı́a, Complexo Hospitalario Universita- rio A Coruña, Xubias de Arriba 84, 15006 A Coruña, Spain.

Author Contributions: Conception and design: G. Rodrı́guez-Trigo, J.P. Zock, F. Pozo-Rodrı́guez, F.P. Gómez, H. Verea, J.M. Antó, C. Fuster, J.A. Barberà. Analysis and interpretation of the data: G. Rodrı́guez-Trigo, J.P. Zock, F. Pozo-Rodrı́guez, F.P. Gómez, G. Monyarch, M.D. Coll, H. Verea, J.M. Antó, C. Fuster, J.A. Barberà. Drafting of the article: G. Rodrı́guez-Trigo, J.P. Zock, F. Pozo- Rodrı́guez, F.P. Gómez, M.D. Coll, J.M. Antó, C. Fuster, J.A. Barberà. Critical revision of the article for important intellectual content: G. Rodrı́guez-Trigo, J.P. Zock, F. Pozo-Rodrı́guez, F.P. Gómez, M.D. Coll, H. Verea, J.M. Antó, C. Fuster, J.A. Barberà. Final approval of the article: G. Rodrı́guez-Trigo, J.P. Zock, F. Pozo- Rodrı́guez, F.P. Gómez, G. Monyarch, M.D. Coll, H. Verea, J.M. Antó, C. Fuster, J.A. Barberà. Provision of study materials or patients: G. Rodrı́guez-Trigo, J.P. Zock, F. Pozo-Rodrı́guez, F.P. Gómez, H. Verea. Statistical expertise: J.P. Zock, F. Pozo-Rodrı́guez. Obtaining of funding: G. Rodrı́guez-Trigo, J.P. Zock, F. Pozo- Rodrı́guez, H. Verea, J.M. Antó, C. Fuster, J.A. Barberà. Administrative, technical, or logistic support: G. Rodrı́guez-Trigo, F. Pozo-Rodrı́guez, G. Monyarch, H. Verea. Collection and assembly of data: G. Rodrı́guez-Trigo, F. Pozo- Rodrı́guez, F.P. Gómez, G. Monyarch, L. Bouso, H. Verea, C. Fuster.

APPENDIX: MEMBERS OF THE SEPAR-PRESTIGE STUDY GROUP Chairs

J.A. Barberà, F. Pozo-Rodrı́guez, H. Verea (Spanish Society of Pulmonology and Thoracic Surgery, SEPAR).

Investigators J.P. Zock, J.M. Antó, L. Bouso (Centre for Research in

Environmental Epidemiology and Municipal Institute of Medi- cal Research, Barcelona); G. Rodrı́guez-Trigo (Complexo Hospi- talario Universitario A Coruña, A Coruña, and Hospital Clı́nico San Carlos, Madrid); F.P. Gómez, Y. Torralba, F. Burgos (Hos- pital Clı́nic-IDIBAPS, Barcelona); C. Fuster, G. Monyarch (Autonomous University of Barcelona, Barcelona).

Collaborators L. Vázquez, L. Rodrı́guez-Valcárcel, A. Souto, M. Blanco

(Complexo Hospitalario Universitario A Coruña, A Coruña); A. Serrano-Mollar, O. Bulbena (Institute of Biomedical Investiga- tions of Barcelona [IIBB-CSIC], Barcelona); J. Tò (Hospital Clı́nic-IDIBAPS, Barcelona); M.D. Coll, J. Egozcue† (Autono- mous University of Barcelona, Barcelona); E. Toubes (Complexo Hospitalario Universitario de Ourense, Ourense); I. Isidro (Na- tional Silicosis Institute, Oviedo); A. Palacios (Complexo Hospi- talario Universitario de Santiago de Compostela, Santiago de Compostela); M. Suárez (Complexo Hospitalario Universitario de Vigo, Vigo).

†Deceased.

Appendix Figure 1. Timeline of the study relative to the oil spill and clean-up events.

Clean-up work

2003 2004 2005

Start of oil spill

Questionnaire survey (9)

Present study

November 2002

January 2004

September 2004

February 2005

Annals of Internal Medicine

W-174 19 October 2010 Annals of Internal Medicine Volume 153 • Number 8 www.annals.org

Downloaded From: http://annals.org/ on 07/01/2016

Appendix Table 1. Allergens Included in Skin-Prick Testing

House dust mites: Dermatophagoides pteronyssinus and D. farinae Cat: Felis domesticus Dog: Canis familiaris Grass pollen mix: Gramineae spp., Bermuda grass (Cynodon dactylon),

timothy grass (Phleum pratense), and spreading pellitory (Parietaria judaica)

Tree pollen: birch (Betula nigra) and European pear (Pyrus communis) Molds: Penicillium notatum, Cladosporium herbarum, Aspergillus

fumigatus, Mucor spp., and Alternaria alternata Soybean (Glycine max) and whitefish mix

Appendix Figure 2. 8-Isoprostane levels in exhaled breath condensate of exposed and nonexposed participants, by presence of lower respiratory tract symptoms.

8 -I

so p ro

st an

e Le

ve l, p g/ m L

Exposed Participants Nonexposed Participants

0.1

100

1000

10 000

No Symptoms (n = 59)

Symptoms (n = 18)

No Symptoms (n = 69)

Symptoms (n = 10)

Bars indicate the geometric mean of the 8-isoprostane level. Lower re- spiratory tract symptoms were wheeze with breathlessness, wheeze apart from colds, nocturnal attacks of shortness of breath, chronic cough, or chronic phlegm.

www.annals.org 19 October 2010 Annals of Internal Medicine Volume 153 • Number 8 W-175

Downloaded From: http://annals.org/ on 07/01/2016

A pp

en di

x T

ab le

2. A

ss o ci

at io

n s

B et

w ee

n th

e D

eg re

e o

f Ex

p o

su re

to C

le an

-u p

A ct

iv it

ie s

an d

Se le

ct ed

H ea

lt h

O u

tc o

m es

Ex p o su

re C

h ar

ac te

ri st

ic Lo

w er

R es

p ir

at o ry

T ra

ct Sy

m p to

m s

B ro

n ch

ia l

H yp

er re

sp o n si

ve n es

s* 8 -I

so p ro

st an

e Le

ve l†

D et

ec ta

b le

V EG

F o r

b FG

F† ‡

C h ro

m o so

m al

Le si

o n §

St ru

ct u ra

l C

h ro

m o so

m al

A lt

er at

io n s|

P ar

ti ci

p an

ts ,

n /N

(% )

R D

(9 5 %

C I)

|| P ar

ti ci

p an

ts ,

n /N

(% )

R D

(9 5 %

C I)

¶ P ar

ti ci

p an

ts ,

n G

M Le

ve l,

p g /m

L G

M R

(9 5 %

C I)

* *

P ar

ti ci

p an

ts ,

n /N

(% )

R D

(9 5 %

C I)

† †

P ar

ti ci

p an

ts ,

n /N

(% )

R D

(9 5 %

C I)

† †

P ar

ti ci

p an

ts ,

n /N

(% )

R D

(9 5 %

C I)

† †

N o n ex

p o se

d 1 7 7 /3

7 (2

0 .9

) 0 .0

(r ef

er en

ce )

7 7 /7

(9 .1

) 0 .0

(r ef

er en

ce )

7 9

5 .6

1 .0

(r ef

er en

ce )

5 0 /8

(1 6 )

0 .0

(r ef

er en

ce )

4 6 /2

0 (4

3 )

0 .0

(r ef

er en

ce )

4 6 /2

1 (4

6 )

0 .0

(r ef

er en

ce )

Ex p o se

d O

ve ra

ll d u ra

ti o n

1 5 –1

0 0

d 2 8 3 /7

8 (2

7 .6

) 8 .5

(0 .8

to 1 6 .1

) 1 1 4 /2

3 (2

0 .2

) 1 1 .1

(1 .3

to 2 0 .9

) 4 1

1 4

2 .5

(1 .6

to 4 .1

) 2 9 /1

5 (5

2 )

3 4 .9

(1 4 .1

to 5 5 .6

) 4 5 /2

5 (5

6 )

1 2 .5

(� 8 .4

to 3 3 .5

) 4 5 /3

5 (7

8 )

3 3 .7

(1 5 .0

to 5 2 .4

) �

1 0 0

d 2 1 3 /5

4 (2

5 .4

) 7 .3

(� 0 .7

to 1 5 .3

) 8 4 /7

(8 .3

) �

0 .8

(� 9 .5

to 8 .0

) 3 6

1 4

2 .4

(1 .5

to 4 .0

) 2 0 /1

4 (7

0 )

5 3 .7

(3 1 .6

to 7 5 .8

) 4 6 /2

4 (5

2 )

9 .1

(� 1 1 .7

to 2 9 .9

) 4 6 /2

9 (6

3 )

2 0 .2

(� 0 .4

to 4 0 .8

) P

va lu

e‡ ‡

0 .1

4 0 .4

3 �

0 .0

0 1

� 0 .0

0 1

0 .4

1 0 .0

6 4

D ai

ly d u ra

ti o n

4 –5

h 1 9 7 /5

5 (2

7 .9

) 9 .2

(0 .9

to 1 7 .5

) 9 1 /1

5 (1

6 .5

) 7 .4

(� 2 .6

to 1 7 .4

) 4 2

1 1

2 .0

(1 .3

to 3 .3

) 2 8 /2

1 (7

5 )

5 8 .3

(3 9 .9

to 7 6 .7

) 5 3 /2

8 (5

3 )

9 .7

(� 1 0 .1

to 2 9 .4

) 5 3 /3

5 (6

6 )

2 2 .8

(3 .2

to 4 2 .3

) 6 –8

h 2 9 9 /7

7 (2

5 .8

) 7 .1

(� 0 .5

to 1 4 .7

) 1 0 7 /1

5 (1

4 .0

) 4 .9

(� 4 .3

to 1 4 .1

) 3 5

1 8

3 .2

(1 .9

to 5 .2

) 2 1 /8

(3 8 )

1 8 .2

(� 5 .6

to 4 2 .1

) 3 8 /2

1 (5

5 )

1 2 .7

(� 9 .9

to 3 5 .3

) 3 8 /2

9 (7

6 )

3 3 .3

(1 3 .9

to 5 2 .7

) P

va lu

e‡ ‡

0 .1

3 0 .6

5 �

0 .0

0 1

0 .0

0 8

0 .2

6 0 .0

0 1

N u m

b er

o f

ty p es

o f

cl ea

n -u

p w

o rk

1 –4

2 0 0 /5

4 (2

7 .0

) 7 .3

(� 0 .9

to 1 5 .5

) 8 2 /1

5 (1

8 .3

) 9 .2

(� 1 .3

to 1 9 .7

) 3 9

1 2

2 .2

(1 .4

to 3 .6

) 2 4 /1

4 (5

8 )

4 1 .5

(1 9 .2

to 6 3 .7

) 4 0 /2

6 (6

5 )

2 1 .5

(0 .6

to 4 2 .4

) 4 0 /2

6 (6

5 )

2 1 .2

(0 .8

to 4 1 .6

) 5 –9

2 9 6 /7

8 (2

6 .4

) 8 .5

(0 .9

to 1 6 .0

) 1 1 6 /1

5 (1

2 .9

) 3 .8

(� 5 .0

to 1 2 .7

) 3 8

1 6

2 .8

(1 .7

to 4 .6

) 2 5 /1

5 (6

0 )

4 3 .9

(2 2 .5

to 6 5 .2

) 5 1 /2

3 (4

5 )

1 .6

(� 1 9 .0

to 2 2 .1

) 5 1 /3

8 (7

5 )

3 3 .6

(1 4 .5

to 5 2 .6

) P

va lu

e‡ ‡

0 .0

7 6

0 .8

7 �

0 .0

0 1

� 0 .0

0 1

0 .9

5 0 .0

0 2

U se

d fa

ce m

as k

d u ri n g

cl ea

n -u

p A

lw ay

s 1 6 4 /4

6 (2

8 .0

) 9 .7

(0 .9

to 1 8 .6

) 7 1 /1

1 (1

5 .5

) 6 .4

(� 4 .2

to 1 7 .0

) 2 5

1 5

2 .6

(1 .5

to 4 .6

) 1 5 /1

1 (7

3 )

5 5 .8

(3 1 .2

to 8 0 .4

) 2 9 /1

4 (4

8 )

5 .1

(� 1 8 .1

to 2 8 .2

) 2 9 /2

1 (7

2 )

2 8 .0

(6 .3

to 4 9 .6

) So

m et

im es

3 3 2 /8

6 (2

5 .9

) 7 .1

(� 0 .2

to 1 4 .4

) 1 2 7 /1

9 (1

5 .0

) 5 .9

(� 3 .1

to 1 4 .8

) 5 2

1 4

2 .5

(1 .6

to 3 .8

) 3 4 /1

8 (5

3 )

3 7 .1

(1 7 .8

to 5 6 .5

) 6 2 /3

5 (5

6 )

1 3 .9

(� 5 .8

to 3 3 .7

) 6 2 /4

3 (6

9 )

2 7 .1

(8 .3

to 4 5 .9

) P

va lu

e‡ ‡

0 .1

6 0 .3

8 �

0 .0

0 1

� 0 .0

0 1

0 .1

7 0 .0

0 8

bF G

F �

ba si

c fi

br ob

la st

gr ow

th fa

ct or

; G

M �

ge om

et ri

c m

ea n

; G

M R

� ge

om et

ri c

m ea

n ra

ti o;

R D

� ri

sk d

if fe

re n

ce ;

V E

G F

� va

sc u

la r

en d

ot h

el ia

l gr

ow th

fa ct

or .

* In

lif el

on g

n on

sm ok

er s.

† M

ea su

re d

in ex

h al

ed br

ea th

co n

d en

sa te

. ‡

V E

G F

le ve

l �

4. 5

p g/

m L

or bF

G F

le ve

l �

3. 4

p g/

m L

. §

D et

ec te

d in

ly m

p h

oc yt

es .

|| C

om p

ar ed

w it

h n

on ex

p os

ed p

ar ti

ci p

an ts

, ad

ju st

ed fo

r se

x an

d sm

ok in

g st

at u

s. ¶

C om

p ar

ed w

it h

n on

ex p

os ed

p ar

ti ci

p an

ts ,

u n

ad ju

st ed

. * *

F or

ex p

os ed

vs .

n on

ex p

os ed

p ar

ti ci

p an

ts ,

ad ju

st ed

fo r

se x.

† †

C om

p ar

ed w

it h

n on

ex p

os ed

p ar

ti ci

p an

ts ,

ad ju

st ed

fo r

se x.

‡ ‡

F or

lin ea

r tr

en d

W-176 19 October 2010 Annals of Internal Medicine Volume 153 • Number 8 www.annals.org

Downloaded From: http://annals.org/ on 07/01/2016 View publication statsView publication stats

https://www.researchgate.net/publication/45821271