Article Text
PDF
Abstract
Objectives Fishermen who had participated in clean-up activities of the Prestige oil spill showed an excess risk of respiratory symptoms 1–2 years later, but the long-term persistence of these health effects is unclear. The aim of this study was to evaluate the persistence of these respiratory symptoms 5 years after clean-up work.
Methods Subgroups of 501 fishermen who had been exposed to clean-up work and 177 non-exposed individuals were re-interviewed by telephone in 2008, including the same symptom questions as in the initial survey. Associations between participation in clean-up work and respiratory symptoms were assessed using log-binomial and multinomial regression analyses adjusting for sex, age and smoking.
Results Information from 466 exposed (93%) and 156 non-exposed (88%) fishermen was obtained. The prevalence of lower respiratory tract symptoms (including wheeze, shortness of breath, cough and phlegm) had slightly decreased in both groups, but remained higher among the exposed (RR 1.4, 95% CI 1.1 to 1.9). The risk of having persistent respiratory symptoms (reported both at baseline and at follow-up) increased with the degree of exposure: RR ratio 1.7 (95% CI 0.9 to 3.1) and 3.3 (95% CI 1.8 to 6.2) for moderately and highly exposed, respectively, when compared with those without any symptoms. Findings for nasal symptoms and for respiratory medication usage were similar.
Conclusions Participation in clean-up activities of oil spills may result in respiratory symptoms that persist up to 5 years after exposure. Guidelines for preventive measures and a continued surveillance of clean-up workers of oil spills are necessary.
- Oil spill
- emergency responders
- environmental disaster
- respiratory
- longitudinal
- allergy
- hygiene/occupational hygiene
- epidemiology
- lung function
- exposure monitoring
- exposure assessment
- smoking
- air pollution
- occupational asthma
- asthma
Statistics from Altmetric.com
- Oil spill
- emergency responders
- environmental disaster
- respiratory
- longitudinal
- allergy
- hygiene/occupational hygiene
- epidemiology
- lung function
- exposure monitoring
- exposure assessment
- smoking
- air pollution
- occupational asthma
- asthma
What is known about this subject
Exposure to oil spills has been associated with adverse health effects, but to date, no study has evaluated long-term respiratory health effects of participation in clean-up activities of a major oil spill.
What this paper adds
This longitudinal study among 466 fishermen who had participated in clean-up work of the Prestige oil spill and 156 non-exposed fishermen shows that involvement in clean-up work of oil spills may result in upper and lower respiratory tract symptoms that persist up to 5 years after exposure.
Guidelines for preventive measures and a continued surveillance of clean-up workers of oil spills are necessary.
Introduction
Major marine oil spills are environmental disasters that typically mobilise large numbers of clean-up workers. There is growing concern about the chemical exposures that clean-up activities involve and their potential health effects.1 The wreckage of the oil tanker Prestige in November 2002 heavily contaminated the coast of Galicia, Spain. Hundreds of thousands of emergency responders participated in clean-up operations, including many local fishermen who were among the highest exposed populations. A questionnaire survey among 6780 fishermen in 2004 showed increased prevalences of respiratory symptoms more than 1 year after having participated in clean-up work.2 The risk increased with the degree of exposure to clean-up work and was related to various types of clean-up activities. The excess risk of respiratory symptoms was less apparent when more time had elapsed between the clean-up activities and the interview, which was suggestive of a potential reversibility of the adverse respiratory health effects over time. However, it was stressed that the latter hypothesis could only be formally tested using an appropriate longitudinal design.
A more detailed respiratory health assessment performed shortly after the questionnaire survey included 501 fishermen who had been highly exposed to clean-up work and 177 non-exposed individuals.3 While effects on conventional spirometric indices of lung function were not apparent, there was evidence that the exposed had increased non-specific bronchial responsiveness and increased levels of 8-isoprostane and growth factors in exhaled breath condensate. The latter suggested a potential involvement of oxidative stress as a mechanism of airway damage and a lasting airway remodelling process.
To date, we are not aware of longitudinal studies that have evaluated long-term respiratory health effects of exposure to oil spills.1 4 Here, we report the results of a follow-up study among clean-up workers of the Prestige oil spill and non-exposed fishermen more than 5 years after the event. We hypothesised that exposure to clean-up work involving inhalation and/or dermal exposure to oil or degradation products would be associated with long-term respiratory health effects reflected by persistent respiratory symptoms and increased medication usage.
Methods
Study design and participants
A follow-up study of fishermen was conducted in 2008, that is, 4 years after the baseline survey2 and more than 5 years after exposure to the clean-up of the Prestige oil spill (figure 1). The selection of the study participants was based on the information obtained through the baseline questionnaire survey2 and has been detailed elsewhere.3 Briefly, exposed individuals (n=501) were members of fishermen cooperatives in heavily affected areas on the Atlantic coast who had participated at least 15 days in clean-up activities, during ≥4 h/day, including the period November to December 2002 when exposure presumably was at its highest level. Non-exposed fishermen (n=177) were members of cooperatives in areas on the Cantabrian coast, less affected by the oil spill and who had not participated in any clean-up activity for other than health reasons. All participants who complied with these criteria according to both the baseline questionnaire2 and a face-to-face interview3 were eligible to be included. The study was approved by an institutional committee on ethical practice and participants provided informed consent.
Flow diagram showing the design of the different stages of the study.
Telephone interview
An explanatory letter was sent in May to June 2008 to the 678 eligible participants and to their cooperatives. A questionnaire was compiled using a selection of the items included in the baseline study2: upper and lower respiratory tract symptoms and inhaled and oral medication usage in the previous 12 months, respiratory and allergic conditions and questions about anxiety and beliefs about the effects of the oil spill on the participant's own health. Contact information was updated wherever necessary and feasible, and questionnaires were administered by telephone by four trained fieldworkers in the period June to July 2008.
Consistent with the baseline survey,2 lower respiratory tract symptoms were defined as an affirmative answer to at least one out of five questions (wheeze with breathlessness in the last 12 months, wheeze apart from colds in the last 12 months, nocturnal attacks of shortness of breath in the last 12 months, cough for as much as 3 months each year and/or phlegm for as much as 3 months each year). Nasal symptoms were defined as a problem with sneezing or a runny or a blocked nose when not having a cold or the flu in the last 12 months. Smoking status (never, former or current) was based on the information provided at the baseline interviews2 3 and at the follow-up interview. The information on details of clean-up work provided at baseline was used to categorise clean-up workers according to an increasing degree of exposure. We used four characteristics of clean-up exposure that appeared to be associated with respiratory symptoms in the baseline study.2 We first defined the highest exposed category aiming at a high specificity as those who had performed 35 or more days of clean-up work, on average during ≥5 h/day, having performed at least five types of clean-up activities and who never or only sometimes had been using a face mask. The moderately exposed category consisted of clean-up workers who did not comply with these stringent criteria.
Analysis
Taking advantage of the longitudinal data, two types of analysis were performed. First, cross-sectional associations between exposure to clean-up work and dichotomous respiratory health outcomes both at baseline and at follow-up were assessed using multivariable binomial regression models with a log link (providing relative risks) adjusted for sex, age and smoking status. Second, in order to take into account individual changes in health outcomes, multinomial regression analysis (providing RR ratios (RRRs)) with additional adjustment for the time elapsed between baseline and follow-up studies was applied using variables with four categories based on presence or absence of the outcome at baseline and at follow-up. In order to study dose–response relationships, the p value for linear trend was obtained by the Wald χ2 test from the adjusted regression models using the degree of exposure as ‘continuous’ variable (0, 1, 2). Potential modification of the associations by demographic and respiratory health characteristic at baseline was evaluated by including a multiplicative interaction term in the adjusted regression model. Statistical analyses were done with Stata V.10 (StataCorp).
Results
Telephone interviews were successfully completed by 466 exposed (participation rate 93%) and 156 non-exposed individuals (88%). The most important reasons for loss to follow-up were the lack of updated contact information (63%) and the failure to contact the participants by phone (20%) in spite of multiple attempts (online data supplement table 1). There were no significant differences in sex, age or baseline respiratory health status between responders and non-responders (online data supplement table 2). Among the non-exposed, but not among the exposed, responders to the follow-up interview were more likely to be never-smokers and less likely to be current smokers.
Consistently with the baseline survey, exposed participants were more likely to be male and were on average 3 years younger than the non-exposed (table 1). No significant differences in smoking status and lifetime history of respiratory and allergic conditions were found between the exposed and non-exposed groups. Almost half of the exposed individuals (n=216) were classified as highly exposed with an average duration of participation in clean-up work of more than 7 months (online data supplement table 3).
Demographic and respiratory health characteristics of the study population
The prevalence of lower respiratory tract symptoms had decreased during the follow-up period, both among the exposed and the non-exposed (table 2). The prevalence of nasal symptoms had decreased among the exposed but had increased among the non-exposed. More medication usage was reported at follow-up, both by exposed and non-exposed individuals, being more pronounced for inhaled medication. In spite of the changes over time, the prevalence of each of the four respiratory health outcomes at follow-up was still higher among the exposed. The risk of having lower respiratory tract symptoms at follow-up increased with the degree of exposure (online data supplement table 4). Such a dose-dependent association was less apparent for the other health outcomes under study.
Cross-sectional associations between exposure to clean-up work and respiratory health status at baseline and at follow-up
There was considerable variation in reported symptoms and medication usage during the follow-up period in both groups (table 3). The strongest associations with exposure to clean-up work were found for persistent respiratory symptoms or medication usage (ie, reported on both occasions). The associations between exposure and the report of a health outcome only once were overall weaker and indicated a larger remission of nasal symptoms and inhaled medication usage among the exposed. The risks of persistent respiratory symptoms and medication usage increased with the degree of exposure (table 4) and were three to four times higher among the highly exposed as compared with the non-exposed.
Associations between exposure to clean-up work and individual changes in respiratory health status between baseline and follow-up
Associations between the degree of exposure to clean-up work and persistent respiratory symptoms and medication usage
The associations between exposure to clean-up work and respiratory symptoms at follow-up were studied in subgroups defined according to demographic and respiratory health characteristic at baseline (online data supplement table 5). The effect on both types of symptoms was modified by smoking status; significantly increased risks were found among current smokers, while no increased risks were found among never-smokers. Nevertheless, the risk of persistent lower respiratory tract symptoms among never-smokers was associated with exposure to clean-up work and increased with the degree of exposure; RRR 1.4 (95% CI 0.6 to 3.1) and 2.3 (95% CI 1.0 to 5.6) for the moderately and highly exposed, respectively (data not presented). There were indications that the association with lower respiratory tract symptoms was stronger in men and in the subgroup of participants with non-specific bronchial hyperresponsiveness but the interaction terms did not reach statistical significance (p>0.1). Similar trends were observed when stratified analyses were performed using persistent symptoms, in spite of wide CIs due to small numbers (data not presented).
Anxiety about health effects of the oil spill was reported by 26% of the participants, which was slightly lower than 33% reported at the baseline study. The belief that the oil spill had affected their health had increased during the follow-up period from 14% to 22%. Both issues were more often reported by exposed individuals (online data supplement table 6) and by those with respiratory symptoms (data not presented). We stratified the analyses between exposure to clean-up work and symptoms by any reported anxiety and/or health belief, either at baseline or at follow-up. The RR of having lower respiratory tract symptoms at follow-up was 1.3 (95% CI 0.9 to 1.8) and 1.3 (95% CI 0.8 to 2.0) for the groups with and without anxiety or health belief, respectively. The corresponding RRRs of persistent lower respiratory tract symptoms were 1.9 (95% CI 0.8 to 4.4) and 1.7 (95% CI 0.7 to 4.3), respectively.
Discussion
This longitudinal study performed more than 5 years after the Prestige oil spill shows that fishermen who had participated in the clean-up operations had an increased prevalence of upper and lower respiratory tract symptoms and medication usage when compared with fishermen who had not participated in any clean-up work. The excess risks were particularly apparent when respiratory symptoms were reported both at the baseline and the follow-up studies and increased with the degree of exposure to clean-up work. Our findings suggest that intensive participation for several months in clean-up work of oil spills without appropriate respiratory protection may result in persistent respiratory health effects on the long term.
In the last 2 decades, potential health effects of eight major oil tanker spills have been evaluated through epidemiological studies on residents, clean-up workers or both and have been summarised in recent reviews.1 4 Although most of the studies showed an increased risk of respiratory symptoms in exposed individuals, there is little information about the persistence of these effects. Medium-term respiratory health effects related to the Braer oil spill (Shetland Islands, Scotland, UK, 1993) were evaluated in a 5-month follow-up study among residents.5 The prevalences of wheeze and breathlessness on exertion were higher among the exposed residents as compared with a non-exposed control population. Although clean-up workers were not studied and the follow-up was performed in less than a year after the spill, findings regarding respiratory health effects were in line with our study.
As far as we know, our study is the first assessment of long-term respiratory health effects of exposure to a major oil spill. A particular strength of this study was the longitudinal design including three contacts with the exposed and non-exposed fishermen over a 4-year period. The methodology of assessing information on respiratory symptoms and medication usage was identical in the first (2004) and the third (2008) study and followed an internationally standardised protocol with validated questionnaires.6 Additional strengths included a relatively small loss to follow-up with a similar magnitude among exposed and non-exposed and an in-depth clinical and functional characterisation of the study participants at baseline. However, there are a number of potential limitations that need to be considered. First, although the overall participation rate was more than 90%, a selective response could have introduced bias in the associations between exposure to clean-up work and respiratory symptoms at follow-up. Nevertheless, the vast majority of loss to follow-up was not due to refusal of study participation but because eligible individuals could not be contacted. In addition, there were no significant differences in reported respiratory symptoms at baseline between responders and non-responders, neither for exposed nor for non-exposed. Thus, it is unlikely that loss follow-up introduced an important bias.
Second, the information on respiratory health outcome was obtained by self-report. Although the excess risk of symptoms at baseline was parallelled by differences in objectively measured respiratory health indices,3 it is not impossible that those who participated in clean-up work were more likely to report respiratory symptoms 5 years later, particularly if they were anxious about potential effects of the oil spill on their health. A stratified analysis showed that the excess risk of respiratory symptoms was very similar among those who had and those who had not reported any anxiety or belief that the oil spill had affected their health. Thus, a differential reporting of symptoms and hence a perceptual bias7 is unlikely to explain the major part of the excess risk.
The reports of experienced respiratory symptoms showed a considerable variability over time. This is a common observation in longitudinal studies using a repeated assessment of respiratory symptoms and is typically related to the intermittent nature of the underlying respiratory condition (true variation) as well as to measurement error.8 In our study, more than a quarter of all participants reported lower respiratory tract symptoms only at one of the two occasions. In order to study persistence, we compared participants reporting respiratory symptoms both at baseline and at follow-up with those who reported no symptoms at both occasions. An additional advantage of this approach was the reduction of bias due to variability over time, either or not due to measurement error.
The decrease in prevalence of lower respiratory tract symptoms among the exposed in itself could be suggestive of reversibility of the adverse respiratory health effects. However, this relatively small decrease among clean-up workers was parallelled by a decrease among non-exposed fishermen. This remission of symptoms may be partly attributed to the fact that medically unexplained physical symptoms as part of a post-traumatic stress disorder9 were more common shortly after the oil spill than several years later. This hypothesis was supported by the observed decrease of reported anxiety in both groups. Although they did not participate in clean-up operations, the non-exposed fishermen were also affected by the oil spill and suffered psychosocial effects10 that could have influenced their reports of respiratory symptoms.
The remarkably strong dose-dependent association between participation in clean-up work and the persistence of respiratory symptoms 5 years later is suggestive of a lasting and perhaps irreversible damage of the airways. The observation from the baseline survey that the excess risk of respiratory symptoms became less apparent when more than 2 years had elapsed2 could therefore not be extrapolated to longer time periods. It needs to be taken into account that the definition of exposure in the present analysis was different; the baseline analysis was based on any participation in clean-up work—albeit 1 day—while the follow-up study included fishermen with an intensive exposure to clean-up work for several months. In particular, the highest risks were found among the highest exposed, who were characterised by a participation of more than 35 days (most of them more than half a year) and lack of use of appropriate respiratory protective devices.
The mechanisms underlying these long-term respiratory health effects are unclear. The detailed respiratory health assessment at baseline3 suggested increased bronchial responsiveness and a potential involvement of increased oxidative stress and, possibly related, elevated growth factor activity suggestive of remodelling behind the appearance of respiratory symptoms. We found indications that the excess risk of respiratory symptoms 5 years after exposure was more pronounced among those with bronchial hyperresponsiveness at baseline, but due to limited statistical power, we could not formally demonstrate effect modification. In a more general sense, establishing mechanisms of longer-term respiratory health effects of environmental disasters is challenging. The exposure is often characterised by relatively short-term inhalations of high levels of various irritants, and the suspected effects typically include a disruption and injury of the bronchial epithelium.11 The latter may give rise to inflammatory reactions mediated by cytokines and leukotrienes and to an increased secretion of growth factors, among others. A well-documented pattern of such effects is the WTC cough syndrome, a lasting respiratory disorder including symptoms and bronchial hyperresponsiveness among firefighters who were involved in the emergency response following the collapse of the WTC in New York on 11 September 2001.12 Thus, we speculate that exposure to hydrocarbons and other volatile chemicals from the oil or the dispersants for several months may have induced a disruption of the epithelial structure, inducing lasting airway inflammation and bronchial hyperresponsiveness. More detailed follow-up investigations are needed in order to properly evaluate effect mechanisms of long-term respiratory health effects of exposure to oil spills. In addition, a continuous surveillance and follow-up of clean-up workers can be recommended since the clinical evolution of the observed effects is uncertain.
Oil spills are relatively common but unpredictable and are likely to occur again in any part of the world. The experience from studies related to the Prestige and other oil spills is being incorporated in a large-scale epidemiological study on long-term health effects of the Deepwater Horizon oil spill that occurred in 2010 in the Gulf of Mexico.1 13 We conclude that participation in clean-up activities of oil spills may result in persistent respiratory symptoms up to 5 years after exposure. Guidelines for preventive measures and a continued surveillance of clean-up workers of oil spills are necessary.
Acknowledgments
The authors wish to thank María Isabel Hernández Díaz, Tamara Gómez Pérez and Silvia Suárez Varela (University Hospital A Coruña) for performing the telephone interviews and Dave Macfarlane (IMIM) for assistance in data management.
References
Supplementary materials
Supplementary Data
This web only file has been produced by the BMJ Publishing Group from an electronic file supplied by the author(s) and has not been edited for content.
Files in this Data Supplement:
- Download Supplementary Data (PDF) - Manuscript file of format pdf
- Download Supplementary Data (PDF) - Manuscript file of format pdf
Footnotes
Funding This study was financially supported by the Instituto de Salud Carlos III/FEDER (grants FIS PI03/1685 and PI07/0086), the Spanish Respiratory Society (SEPAR), the Galician Health Service (SERGAS), the CIBER Enfermedades Respiratorias (CIBERES) and the CIBER Epidemiología y Salud Pública (CIBERESP).
Competing interests None.
Patient consent Obtained.
Ethics approval This study was conducted with the approval of the Ethics Committee on Clinical Research of Galicia.
Provenance and peer review Not commissioned; externally peer reviewed.