Higher urine 1-hydroxy pyrene glucuronide (1-OHPG) is associated with tobacco smoke exposure and drinking maté in healthy subjects from Rio Grande do Sul, Brazil
© Fagundes et al; licensee BioMed Central Ltd. 2006
Received: 19 November 2005
Accepted: 26 May 2006
Published: 26 May 2006
The highest rates of esophageal squamous cell carcinoma (ESCC) in Brazil occur in Rio Grande do Sul, the most southern state, which has incidence rates of 20.4/100,000/year for men and 6.5/100,000/year for women. Exposure to carcinogenic polycyclic aromatic hydrocarbons (PAHs) through tobacco smoke and other sources may increase the risk of ESCC. The aims of the current study were to investigate the degree and sources of PAH exposure of the inhabitants of this region of southern Brazil.
Two hundred healthy adults (half smokers, half non smokers, half male and half female) were recruited, given a standardized questionnaire, and asked to provide a urine sample for measurement of 1-hydroxypyrene glucuronide (1-OHPG), a PAH metabolite). Urine 1-OHPG concentrations were measured using immunoaffinity chromatography and synchronous fluorescence spectroscopy and urine cotinine was measured using a dipstick test. We examined factors associated with 1-OHPG concentration using Wilcoxon tests and multiple linear regression.
Urine 1-hydroxypyrene glucuronide (1-OHPG) was successfully measured on 199 subjects. The median (interquartile range) of urine 1-OHPG in the 199 participants was 2.09 pmol/mL (0.51, 5.84). Tobacco smoke exposure and maté drinking were statistically significantly associated with higher urine 1-OHPG concentrations in the multivariate linear regression model.
Tobacco smoke and maté both contribute to high levels of benzo[a]pyrene exposure in the people of southern Brazil. This high PAH exposure may contribute to the high rates of ESCC observed in this population. The increased urine 1-OHPG concentrations associated with maté suggest that contaminants, not just thermal injury, may help explain the increased risk of ESCC previously reported for maté consumption.
Esophageal cancer is a common and usually fatal cancer that is characterized by great variation in rates among different populations. In South America, there is a geographic area of high esophageal squamous cell carcinoma (ESCC) incidence that encompasses southern Brazil, northeastern Argentina, Uruguay, and Paraguay, with age-standardized incident rates of approximately 20/100,000/year . Inhabitants of this area share a similar environment and similar habits and culture. Two habits they have in common that may contribute to the high ESCC rates are the high consumption of grilled red meat called churrasco and a daily consumption of large volumes of a beverage known as maté. Churrasco is barbequed meat grilled directly over a wood fire and is a potential source of polycyclic aromatic hydrocarbons (PAH), heterocyclic amines, and other contaminants which may be associated with cancer in humans [2, 3]. Mate is an infusion of the herb Ilex paraguayensis that is prepared in a gourd and is often drunk very hot through a metal straw, which delivers the liquid directly to the oropharynx and esophagus. Most epidemiologic studies that examined maté drinking have found significant associations with ESCC [4–10], but one reported no association .
In low-risk areas, most ESCC is attributable to alcohol, tobacco, and poor diet , but the etiologic agents in high-risk populations remain unclear. PAHs, such as benzo[a]pyrene, and nitrosamines from tobacco smoke and from other sources may act as esophageal carcinogens [2, 13–17]. Animal studies have demonstrated a dose-response relationship between benzo[a]pyrene food levels and the incidence of esophageal cancer in mice .
Studies in the very high risk population of Linxian, China, where consumption of tobacco and alcohol is low, suggest that the inhabitants there may be exposed to high-levels of carcinogenic PAHs from the coal and wood used for cooking and heating in unvented stoves [19–21]. This hypothesis is also supported by the finding of high levels of benzo[a]pyrene in uncooked food samples , histological changes suggestive of PAH exposure in esophagectomy specimens , and high 1-hydroxypyrene glucuronide (1-OHPG) concentrations in urine samples from the inhabitants of this region . 1-OHPG is a stable PAH metabolite that is excreted in the urine and is an index biomarker that reflects recent exposure to mixed PAHs [22–27]. In northeastern Iran, another area with very high rates of ESCC and little consumption of tobacco and alcohol, the population also has high urine 1-OHPG concentration, consistent with exposure to very high levels of PAHs .
To determine the degree and source of PAH exposure in inhabitants of southern Brazil we collected questionnaire data and determined urine concentrations of 1-OHPG from inhabitants of Rio Grande do Sul, the most southern state in Brazil.
Participants were volunteers from Santa Maria, a city in the central region of Rio Grande do Sul. People attending the outpatient unit of the University Hospital with minor injuries, minor skin diseases, to donate blood, women visiting the gynecologist for annual screening, patients with dyspepsia after a normal upper gastrointestinal endoscopy, and patient's companions were considered eligible. After a brief description of the study purpose and requirements they were invited to participate. Greater than 90% of the invited individuals chose to participate. Subjects were recruited so that among the 200 healthy people half were male and half were female and within each of these groups half were current smokers and half were non-smokers. Informed consent was obtained from each participant. The study was approved by the Ethical Committee on Research of the Health Sciences Center of the University of Santa Maria, RS, Brazil, and the analysis of anonymized data and samples was exempted from review by the Institutional Review Board of the National Cancer Institute, Bethesda, MD.
All subjects were interviewed face-to-face using a pre-tested, standardized questionnaire, administered by specially trained interviewers. The questionnaire included: basic demographic variables and residence characteristics; habits of tobacco smoking (age started, age stopped, typical number of cigarettes per day, type of tobacco, and passive smoke exposure); history of alcohol drinking (type of alcoholic beverage, amount of each beverage consumed, duration of consumption); history of maté drinking (amount usually consumed/day); frequency of and fuel used to make barbeque and other cooking; preferred doneness and amount of barbeque typically eaten; and home heating fuel system and its smokiness.
Each participant was recruited in the morning asked to provide a 10 ml urine sample. The urine samples were collected in a sterile container, frozen at -80°C, and shipped on dry ice to the National Cancer Institute. Urine samples were assayed in the laboratory of Dr. Strickland at the Johns Hopkins University. Urine 1-OHPG concentrations were measured using immunoaffinity chromatography and synchronous fluorescence spectroscopy as previously described [16, 25]. NicAlert Strips (Jant Pharmaceutical Corp., Encino, CA) were used to measure urine cotinine equivalents as directed by the manufacturer. This test produces categorical results ranging from zero (<1–10 ng/mL cotinine equivalents) to six (>2000 ng/mL). Because only a small number of subjects had urine cotinine results in each of categories two, three, and four, we collapsed these three groups into a single category.
Urine 1-OHPG concentrations were examined graphically and found to be skewed with a mode at the limit of detection which included 37/199 (19%) subjects. Log10 transformation produced a normal curve outside the mode. Urine 1-OHPG was also represented as quintiles for some analyses. Univariate associations with 1-OHPG were examined by forming exposure data into quantiles and comparing them with the Wilcoxon rank sum test or the Kruskal-Wallis test. Age and maté were divided into empirical quartiles. Multivariate associations were examined using log-transformed urine concentrations in linear regression models. The final model was built by adding all variables, with selected members of a class (e.g. only one of the variables associated with barbeque preparation), and deleting those that were not significant, based on the F-test, and whose removal did not change the estimates for the remaining variables. Interactions between sex or tobacco smoking and other variables were explored. When tested independently (data not shown) we found a significant interaction between barbeque preparation and smoke exposure, so it was retained in the final model. A borderline significant interaction between smoke exposure and maté was also retained. The p-values for the interactions were inflated slightly in the final model. To better show the effects of these interaction we plotted the data for the four groups. Data for plotting was jittered to improve clarity. All analyses were carried out using SAS version 9 (SAS Institute, Cary, NC). All p-values come from two-sided tests.
Characteristics and univariate comparisons of urine 1-hydroxypyrene glucuronide (1-OHPG) concentration in 199 healthy subjects from Rio Grande de Sul
Category N (%)
1-OHPG (pmol/ml) median (IQR)
Age quartile 1
Age quartile 2
Age quartile 3
Age quartile 4
Sex, N (%)
Residence, N (%)
Ever drink maté, N (%)
Maté quartile 1
Maté quartile 2
Maté quartile 3
Maté quartile 4
Current tobacco smoker, N (%)
Ever regular tobacco smoker, N (%)
Urine cotinine, N (%)
0 (<1–10 ng/mL)
Ever prepare BBQ, N (%)
Prepared BBQ in the last week, N (%)
Prepare BBQ at least once a week, N (%)
Ever eat BBQ well done, N (%)
Ever drink beer, N (%)
Ever drink wine, N (%)
Ever drink cachaca, N (%)
We used tabular analysis to look for associations between these factors and tobacco smoking, the most likely contributor to urine 1-OHPG. We found associations between tobacco smoking and each of the factors. By increasing order of age group, we found a prevalence of smoking of 52%, 63%, 58%, and 17%, respectively, (3 df chi-square < 0.0001). Fifty-three percent of maté consumers, but only 29% of non-consumers reported current smoking (chi-square P = 0.0031). We expected and found that most subjects with urine cotinine values of 5 or 6 reported current use of tobacco. Fifty-three percent of non-smokers had cotinine values greater than category zero, but only 1 had a value of 5 or 6. Significantly more cachaça drinkers (71%) reported current tobacco smoking than did non-drinkers (44%) (chi-square P = 0.0042).
Multivariate associations1 between characteristics and urine 1-hydroxypyrene glucuronide (1-OHPG) concentration in the healthy subjects from Rio Grande de Sul
95% Confidence interval
Ever drink cachaHa
Age quartile 1
Age quartile 2
Age quartile 3
Age quartile 4
Interaction (Smoke * Maté)
Prepare barbeque weekly
Interaction (Smoke * BBQ)
We wanted to assure that our results were not due to the shape of the distribution of urine 1-OHPG concentrations, so we created an alternative urine 1-OHPG scale by categorizing subjects into five quintiles where the first quintile was subjects at or near the limit of detection and the remaining subjects were assigned to the remaining values. Replacing the log-transformed urine 1-OHPG concentration value with the quintile category and fitting the same final model produced very similar results, reassuring us that the results were not sensitive to the shape of the distribution (data not shown).
According to the Brazilian Ministry of Health, southern Brazil has incidence rates for ESCC of approximately 20.4/100,000/year for men and 6.5/100,000 for women. These rates are much higher than those observed in most western countries. Several epidemiologic studies have examined potential etiologic factors other than tobacco and alcohol in this region of South America that may contribute to the high rates of ESCC. Some evidence suggests that high consumption of churrasco and hot maté could be additional important risk factors [4–10].
The median urine 1-OHPG level of the inhabitants of southern Brazil who were examined in this study (2.09 pmol/ml) was similar to those found in two other high ESCC-risk areas, namely Linxian, China (2.06 pmol/ml)  and northeastern Iran (4.2 pmol/ml). All of these concentrations are much higher than those reported for non-smoking US residents (0.23 pmol/ml). As expected, tobacco smoking in our population had a significant association with urine 1-OHPG. Non-smoke exposed subjects who regularly prepare barbeque also had elevated urine 1-OHPG concentrations, presumably from increased smoke exposure during this activity. Surprisingly, we also found that any maté consumption significantly increased urine 1-OHPG concentrations and that there was a step-wise increase in 1-OHPG concentration with the volume of maté consumed.
Brazil is a country with recognized regional socio-economical and cultural differences. Rio Grande do Sul State has an economy based largely on agriculture and cattle production which has led to high consumption of red meat, due to relatively low prices and the availability of this product, and a preference for barbequed meat. The churrasco maker is exposed to coal or wood smoke when preparing the meat. We did not see a significant association between urine 1-OHPG concentration and the amount of barbeque consumed, but exposure to other potentially hazardous compounds that may be present in the barbequed meat, such as heterocyclic amines, should also be investigated .
Several epidemiologic studies in this region of South America have shown an association between maté consumption and risk of esophageal cancer [4–10]. Possible reasons for this association include ingestion of carcinogens present in the unprocessed leaves of Ilex paraguayensis, ingestion of carcinogens produced or added as contaminants during the processing of the leaves, and thermal injury to the esophageal mucosa caused by drinking maté tea at very hot temperatures. Many people in this region drink large amounts of maté at very high temperatures. In our study population, we saw a median intake (interquartile range) of 500 (100 – 1000) mls/day. Previous studies in southern Brazil have reported mean maté consumptions of 1200 and 1800 ml/day and mean temperatures measured just before consumption of 63.4 and 69.5°C [31, 32]. Several studies have reported that only the temperature at which maté was drunk was significantly associated with ESCC risk, while the amount of maté consumed and the temperature at which it was extracted were inconsequential [7, 11]. Other studies, however, reported that high temperature and a high volume of consumption were both important, and were independently associated with significantly increased risk of ESCC [4, 8]. Yet another study reported that duration and amount of maté consumption was consistently associated with cancer risk, but temperature was not . We did not collect information on mate temperature in this study because we had no a priori reason to suspect that consumption temperature would affect urine 1-OHPG concentration. We also thought that without objective temperature measurements, questionnaire data concerning mate temperature would not be sufficiently reliable for meaningful evaluations.
In most studies from this region, mate consumption is considered to be an independent risk factor for esophageal cancer. The underlying mechanism, whether thermal or chemical, remain unclear. Fonseca et al. reported that extracts of unprocessed Ilex paraguayensis are mutagenic in bacterial assays and can cause chromosomal aberrations in human peripheral lymphocytes treated ex vivo . The processing of this herb for maté involves roasting the leaves over an open fire, which can lead to the formation or addition of PAHs or other contaminants. A single study of processed mate purchased in Germany reported that the leaves contained up to 461 μg/kg benzo[a]pyrene, but there were relatively low concentrations of this PAH in the prepared beverage . Differences in maté brand and the details of tea preparation might change the amount of benzo[a]pyrene in the tea. The finding of benzo[a]pyrene in maté implies, however, that elevated 1-OHPG concentrations in maté drinkers may come directly from the maté, and need not be attributed to uncontrolled confounders such as smoking. Our findings raise the possibility that PAH exposure from consumption of maté may be one cause for the previously reported association between maté drinking and ESCC risk.
We conclude that people in southern Brazil are exposed to high levels of PAH from tobacco smoke and maté drinking, as well as barbeque preparation and other unmeasured sources, and that this exposure may contribute to the high rates of ESCC observed in this area. Additional studies are needed to characterize this exposure more fully and to determine if it is etiologically associated with the high esophageal cancer rates found in this area.
This research was supported in part by the Intramural Research Program of the NIH, National Cancer Institute, Division of Cancer Epidemiology and Genetics and by NIH grant P01-ES06052 to Dr. Strickland.
- Parkin DM, Pisani P, Ferlay J: Estimates of the worldwide incidence of 25 major cancers in 1990. Int J Cancer. 1999, 80: 827-841. 10.1002/(SICI)1097-0215(19990315)80:6<827::AID-IJC6>3.0.CO;2-P.View ArticlePubMedGoogle Scholar
- Nadon L, Siemiatycki J, Dewar R, Krewski D, Gerin M: Cancer risk due to occupational exposure to polycyclic aromatic hydrocarbons. Am J Ind Med. 1995, 28: 303-324.View ArticlePubMedGoogle Scholar
- Ward MH, Sinha R, Heineman EF, Rothman N, Markin R, Weisenburger DD, Correa P, Zahm SH: Risk of adenocarcinoma of the stomach and esophagus with meat cooking method and doneness preference. Int J Cancer. 1997, 71: 14-19. 10.1002/(SICI)1097-0215(19970328)71:1<14::AID-IJC4>3.0.CO;2-6.View ArticlePubMedGoogle Scholar
- Castellsague X, Munoz N, De Stefani E, Victora CG, Castelletto R, Rolon PA: Influence of maté drinking, hot beverages and diet on esophageal cancer risk in South America. Int J Cancer. 2000, 88: 658-664. 10.1002/1097-0215(20001115)88:4<658::AID-IJC22>3.0.CO;2-T.View ArticlePubMedGoogle Scholar
- De Stefani E, Munoz N, Esteve J, Vasallo A, Victora CG, Teuchmann S: Maté drinking, alcohol, tobacco, diet, and esophageal cancer in Uruguay. Cancer Res. 1990, 50: 426-431.PubMedGoogle Scholar
- De Stefani E, Deneo-Pellegrini H, Ronco AL, Boffetta P, Brennan P, Munoz N, Castellsague X, Correa P, Mendilaharsu M: Food groups and risk of squamous cell carcinoma of the oesophagus: a case-control study in Uruguay. Br J Cancer. 2003, 89: 1209-1214. 10.1038/sj.bjc.6601239.View ArticlePubMedPubMed CentralGoogle Scholar
- Rolon PA, Castellsague X, Benz M, Munoz N: Hot and cold maté drinking and esophageal cancer in Paraguay. Cancer Epidemiol Biomarkers Prev. 1995, 4: 595-605.PubMedGoogle Scholar
- Sewram V, De Stefani E, Brennan P, Boffetta P: Maté consumption and the risk of squamous cell esophageal cancer in Uruguay. Cancer Epidemiol Biomarkers Prev. 2003, 12: 508-513.PubMedGoogle Scholar
- Vassallo A, Correa P, De SE, Cendan M, Zavala D, Chen V, Carzoglio J, eo-Pellegrini H: Esophageal cancer in Uruguay: a case-control study. J Natl Cancer Inst. 1985, 75: 1005-1009.PubMedGoogle Scholar
- Victora CG, Munoz N, Day NE, Barcelos LB, Peccin DA, Braga NM: Hot beverages and oesophageal cancer in southern Brazil: a case-control study. Int J Cancer. 1987, 39: 710-716.View ArticlePubMedGoogle Scholar
- Castelletto R, Castellsague X, Munoz N, Iscovich J, Chopita N, Jmelnitsky A: Alcohol, tobacco, diet, maté drinking, and esophageal cancer in Argentina. Cancer Epidemiol Biomarkers Prev. 1994, 3: 557-564.PubMedGoogle Scholar
- Engel LS, Chow WH, Vaughan TL, Gammon MD, Risch HA, Stanford JL, Schoenberg JB, Mayne ST, Dubrow R, Rotterdam H, West AB, Blaser M, Blot WJ, Gail MH, Fraumeni JFJ: Population attributable risks of esophageal and gastric cancers. J Natl Cancer Inst. 2003, 95: 1404-1413.View ArticlePubMedGoogle Scholar
- Li JY: Epidemiology of esophageal cancer in China. Natl Cancer Inst Monogr. 1982, 62: 113-120.PubMedGoogle Scholar
- Roth MJ, Guo-Qing W, Lewin KJ, Ning L, Dawsey SM, Wesley MN, Giffen C, Yong-Qiang X, Maher MM, Taylor PR: Histopathologic changes seen in esophagectomy specimens from the high-risk region of Linxian, China: potential clues to an etiologic exposure?. Hum Pathol. 1998, 29: 1294-1298. 10.1016/S0046-8177(98)90260-X.View ArticlePubMedGoogle Scholar
- Roth MJ, Strickland KL, Wang GQ, Rothman N, Greenberg A, Dawsey SM: High levels of carcinogenic polycyclic aromatic hydrocarbons present within food from Linxian, China may contribute to that region's high incidence of oesophageal cancer. Eur J Cancer. 1998, 34: 757-758. 10.1016/S0959-8049(98)00262-7.View ArticlePubMedGoogle Scholar
- Roth MJ, Y-L Q, Rothman N, J. T, Dawsey SM, Wang GQ, Cho SH, Kang D, Taylor PR, Strickland PT: High urine 1-hydroxypyrene glucuronide concentrations in Linxina, China, an area of high-risk for squamous oesophageal cancer. Biomarkers. 2001, 6: 381-386. 10.1080/13547500110044780.View ArticleGoogle Scholar
- Yang CS: Research on esophageal cancer in China: a review. Cancer Res. 1980, 40: 2633-2644.PubMedGoogle Scholar
- Culp SJ, Gaylor DW, Sheldon WG, Goldstein LS, Beland FA: A comparison of the tumors induced by coal tar and benzo[a]pyrene in a 2-year bioassay. Carcinogenesis. 1998, 19: 117-124. 10.1093/carcin/19.1.117.View ArticlePubMedGoogle Scholar
- Finkelman RB, Belkin HE, Zheng B: Health impacts of domestic coal use in China. Proc Natl Acad Sci U S A. 1999, 96: 3427-3431. 10.1073/pnas.96.7.3427.View ArticlePubMedPubMed CentralGoogle Scholar
- Phillips DH: Polycyclic aromatic hydrocarbons in the diet. Mutat Res. 1999, 443: 139-147.View ArticlePubMedGoogle Scholar
- Siwinska E, Mielzynska D, Bubak A, Smolik E: The effect of coal stoves and environmental tobacco smoke on the level of urinary 1-hydroxypyrene. Mutat Res. 1999, 445: 147-153.View ArticlePubMedGoogle Scholar
- Bouchard M, Viau C: Urinary 1-hydroxypyrene as a biomarker of exposure to polycyclic aromatic hydrocarbons. Biomarkers. 1999, 4: 159-187. 10.1080/135475099230859.View ArticleGoogle Scholar
- Buckley TJ, Waldman JM, Dhara R, Greenberg A, Ouyang Z, Lioy PJ: An assessment of a urinary biomarker for total human environmental exposure to benzo[a]pyrene. Int Arch Occup Environ Health. 1995, 67: 257-266. 10.1007/BF00409408.View ArticlePubMedGoogle Scholar
- Jongeneelen FJ, Bos RP, Anzion RB, Theuws JL, Henderson PT: Biological monitoring of polycyclic aromatic hydrocarbons. Metabolites in urine. Scand J Work Environ Health. 1986, 12: 137-143.View ArticlePubMedGoogle Scholar
- Kang D, Rothman N, Cho SH, Lim HS, Kwon HJ, Kim SM, Schwartz B, Strickland PT: Association of exposure to polycyclic aromatic hydrocarbons (estimated from job category) with concentration of 1-hydroxypyrene glucuronide in urine from workers at a steel plant. Occup Environ Med. 1995, 52: 593-599.View ArticlePubMedPubMed CentralGoogle Scholar
- Merlo F, Andreassen A, Weston A, Pan CF, Haugen A, Valerio F, Reggiardo G, Fontana V, Garte S, Puntoni R, Abbondandolo A: Urinary excretion of 1-hydroxypyrene as a marker for exposure to urban air levels of polycyclic aromatic hydrocarbons. Cancer Epidemiol Biomarkers Prev. 1998, 7: 147-155.PubMedGoogle Scholar
- Zhao ZH, Quan WY, Tian DH: Urinary 1-hydroxypyrene as an indicator of human exposure to ambient polycyclic aromatic hydrocarbons in a coal-burning environment. Sci Total Environ. 1990, 92:145-54.: 145-154. 10.1016/0048-9697(90)90326-P.View ArticleGoogle Scholar
- Kamangar F, Strickland PT, Pourshams A, Malekzadeh R, Boffetta P, Roth MJ, Abnet CC, Saadatian-Elahi M, Rakhshani N, Brennan P, Etemadi A, Dawsey SM: High exposure to polycyclic aromatic hydrocarbons may contribute to high risk of esophageal cancer in northeastern Iran. Anticancer Res. 2005, 25: 425-428.PubMedGoogle Scholar
- Kang DH, Rothman N, Poirier MC, Greenberg A, Hsu CH, Schwartz BS, Baser ME, Groopman JD, Weston A, Strickland PT: Interindividual differences in the concentration of 1-hydroxypyrene-glucuronide in urine and polycyclic aromatic hydrocarbon-DNA adducts in peripheral white blood cells after charbroiled beef consumption. Carcinogenesis. 1995, 16: 1079-1085.View ArticlePubMedGoogle Scholar
- Sinha R: An epidemiologic approach to studying heterocyclic amines. Mutat Res. 2002, 506-507:197-204.: 197-204.View ArticleGoogle Scholar
- de Barros SG, Ghisolfi ES, Luz LP, Barlem GG, Vidal RM, Wolff FH, Magno VA, Breyer HP, Dietz J, Gruber AC, Kruel CD, Prolla JC: [High temperature "mate" infusion drinking in a population at risk for squamous cell carcinoma of the esophagus]. Arq Gastroenterol. 2000, 37: 25-30.View ArticlePubMedGoogle Scholar
- Victora CG, Munoz N, Horta BL, Ramos EO: Patterns of maté drinking in a Brazilian city. Cancer Res. 1990, 50: 7112-7115.PubMedGoogle Scholar
- Fonseca CA, Otto SS, Paumgartten FJ, Leitao AC: Nontoxic, mutagenic, and clastogenic activities of Mate-Chimarrao (Ilex paraguariensis). J Environ Pathol Toxicol Oncol. 2000, 19: 333-346.PubMedGoogle Scholar
- Ruschenburg U: Benzo[a]pyrene content of coffee and some other foodstuffs (Ger.). IIe Colloque Scientifique International sur le Café, Lomé 1985. 1985, Paris, Association Scientifique International du Café, 205-212.Google Scholar
- The pre-publication history for this paper can be accessed here:http://www.biomedcentral.com/1471-2407/6/139/prepub
This article is published under license to BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.