Butchers constitute an occupational group exposed to potentially harmful agents, including environmental stressors, such as cold, explosions, fires, and combustion products. They are also exposed suspected carcinogens, such as polycyclic aromatic hydrocarbons, nitrosamines, and potentially oncogenic animal viruses, which are formed during meat curing and slaughtering.[2–4] Despite its controversial health consequences, butchery is a prevalent occupation.
The number of reported cancers specific to butchers and related workers is relatively small. However, this finding may be biased because of the suspected excessive selective reporting of certain neoplasms.[5,6] Suspicions on the hazard in butchers originally arose from the routine investigations of occupational mortality and cancer incidence in Denmark and Sweden.[7,8] Subsequently, retrospective cohort studies of butchers in the USA also found increased risks of lung and colon cancer mortalities. Moreover, increased risks of leukemia and stomach cancer were reported by several papers[10,11] but have not been confirmed in recent investigations.[12,13]
Knowledge is inadequate on the risks of cancer mortality and incidence and the small in this occupational group, and the magnitudes of the expected increases in these risks are small. Thus, we performed a comprehensive meta-analysis of published studies to clarify the above-mentioned contradictory results and evaluate the relationships of the risks of cancer mortality and incidence to male and female butchers globally.
2.1 Search strategies
This study was conducted in accordance with the guidelines of the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) and reported in compliance with the guidelines developed by the Meta-Analysis of Observational Studies in Epidemiology group. We also registered the protocol on PROSPERO to document our methodological approach a priori. All analyses were based on previous published studies, thus no ethical approval and patient consent are required.
To identify eligible studies, 2 investigators (ZG and JW) performed a comprehensive literature search for eligible studies in the databases PubMed, Embase, Web of Science, Scopus, and Cochrane Library. Eligible studies included those on the relationships of the risks of cancer mortality and incidence to butchers from database inception to March 2017. Each database was searched without restrictions on language, publication type, and region using the following combinations of medical subject headings (MeSH) and non-MeSH search terms: “butcher”; “cancer, carcinoma, or tumor”; and “cancer risk, cancer mortality, or cancer incidence.” We also identified other potentially relevant studies by manual searches through reference lists of all the included studies and previous reviews on the topic of interest. We contacted the authors of unpublished studies (abstracts only) and the most recurrent studies. Any discrepancy was resolved by consulting an investigator not involved in the initial procedure.
2.2 Study selection criteria
Two independent investigators (ZG and YL) chose the studies that explored the potential relationships of butchery to risks of cancer mortality and incidence. These works were selected using the following inclusion criteria. The studies contained predefined diagnosis criteria for both butchers and cancer. Participants were selected without limitations on region, age, and social status. The studies presented sufficient original data (excluding reviews) on standardized mortality ratio (SMR), odds ratio (OR), risk ratio (RR), and hazard ratio (HR) estimates as well as associated 95% confidence intervals (CIs) for the correlation of butchery with cancer mortality and incidence. The works used either a case-control, cross-sectional, retrospective, or prospective design. Lastly, the population included butchers employed in industries and meat cutters working outside the meat industry (e.g., in department stores and other sectors of the food industry). Any disagreement was resolved through the adjudication of senior authors.
2.3 Data extraction
Data from the included studies were extracted and independently summarized by 2 investigators from our team (JW and LG) using a predefined data extraction form. Specifically, we read the reports and independently extracted and tabulated the valuable information into a standardized evidence table. The data included the study design, baseline population characteristics (i.e., mean age, sex, sample size, and country), cancer types, endpoints, adjusted factors, cancer incidence, mortality, and risk estimates from the most fully adjusted model with 95% CIs from all the included studies. We also checked the data for accuracy. Moreover, whenever possible, we contacted the primary authors of the studies with insufficient information to acquire and verify the data. Disagreements were resolved by discussion or consensus with a third reviewer.
2.4 Methodological quality assessment
The methodological quality of the included studies was assessed by 2 independent reviewers (SW and SG) using the modified Newcastle–Ottawa Scale (NOS). This scale consists of the following domains: patient selection, study group comparability, and outcome assessment. A score of 0 to 9 (denoted by stars) was allocated for observational studies. Disagreements were also settled through a discussion among the authors.
2.5 Statistical analyses
For the meta-analysis, the total effectiveness rates of the extracted data were pooled using SMR and OR with associated 95% CIs to determine the relationships of the risks of cancer mortality and incidence to male and female butchers worldwide. Since the absolute risk of caner is low, the measures of association are expected to yield similar estimates of ORs. Consequently, we therefore reported all results as the OR simplicity, as appropriate, so that comprehensiveness of the analysis and maximization of the statistical power are ensured.[17–19] The aggregated results and 95% CIs for the effect size were calculated using inverse-variance weighted meta-analysis. An I-square (I2) test was performed to assess the effect of study heterogeneity on the meta-analysis results. In this test, I2 values of 0%, 25%, 50%, and 75% represented no, low, moderate, and high heterogeneities, respectively. According to the Cochrane review guidelines, a severe heterogeneity of I2 ≥ 50% warrants the use of the random-effects model. Otherwise, a fixed-effects model is appropriate. Statistical significance was set at P < .05. Sensitivity analysis was conducted to eliminate 1 study at a time by evaluating the quality and consistency of the results. Funnel-plot visual inspection and Egger linear regression test were conducted to assess for publication bias. Subgroup analyses by country, sex, and study design were performed.
3.1 Study selection process
Figure 1 presents a flowchart describing the selection process of our literature screening. Our search yielded 241 unique reports, from which only 189 studies were retrieved after the removal of duplicates. After screening the titles and abstracts, only 36 studies were retained and the reasons are as follows: 36 studies did not match the butcher definition, 30 studies were not human studies, 24 studies were not related to cancer, 28 studies were mechanistic/genetic studies, 22 studies did not provide eligible endpoints, 8 studies were reviews, and 5 studies were published as abstracts. Finally, 19 full-text articles were discarded for the following reasons: 4 studies did not provide eligible outcomes, 4 studies did not match the butcher definition, and 11 studies did not provide sufficient data for extraction. Therefore, 17 observational studies[9–13,20–31] comprising 397,726 participants were included in our meta-analysis on the basis of the inclusion criteria.
3.2 Study characteristics and methodological quality
The characteristics of the 17 included studies[9–13,20–31] are shown in Table 1. Among the included works, 11 were case-control studies,[11,21–25,27–31] 5 were retrospective cohort studies,[9,10,12,13,20] and 1 was a prospective cohort study. Their publication years ranged from 1989 to 2011. Among the studies, 6 were conducted in the USA,[9,20,25,27,28,31] 3 in Sweden,[10,12,24] 2 in New Zealand,[11,22] 2 in Uruguay,[29,30] and 1 in the Netherlands. Moreover, 1 was conducted in 10 participating centers in the following locations: Denmark, Italy, the Netherlands, Norway, Spain, Sweden, UK, Germany, Greece, and France. Finally, 1 was performed in 7 participating centers located in Romania, Hungary, Poland, UK, Russia, Slovakia, and the Czech Republic. Meanwhile, 14 studies were performed among populations older than 18 years,[9–11,13,20,22–24,26–31] but 3 studies did not report the exact ages.[12,21,25] The cancer types varied across the included studies, and only 2 studies reported estimates without adjustments.[9,21] The follow-up length ranged from 2 years to 99 years. In the included clinical trials, the sample sizes varied between 159 and 348,555 participants.
In addition, 11 studies were of high methodological quality,[9,10,13,20,22,23,26–28,30,31] 3 studies[11,24,29] were of moderate quality, and 3 studies[12,21,25] were of poor quality based on the modified NOS. The main deficiency was the selection bias related to the insufficient adjustment of confounding factors among the included studies.
3.3 Association between butchers and the risk of cancer mortality
The data on cancer mortality were analyzed from 3 studies (2 retrospective cohort studies[9,20] and 1 case-control study; Fig. 2) with the random-effects model. We found that the butchers did not show a significantly different all-cancer mortality risk from that in the general population. The pooled overall SMRs were 1.07 (95% CI 0.96–1.20), and 1.07 (95% CI 0.93–1.21) for lung cancer, 1.54 (95% CI 0.59–4.01) for colon cancer, respectively. Nevertheless, moderate heterogeneity was found (I2 = 69.3%, P = .011), and the subgroup analyses were restricted by the small number of studies evaluated.
3.4 Association between butchers and the risk of cancer incidence
The pooled ORs revealed that butchers were at an elevated risk of total cancer incidence (OR, 1.51; 95% CI, 1.33–1.73) (Table 2, Figs. 3–6). Significant associations were demonstrated for glioma (OR, 1.95; 95% CI, 1.19–3.17, P = .584) as well as cancers of the stomach (OR, 1.41; 95% CI, 1.14–1.76, P = .726), oral cavity and pharynx (OR, 1.60; 95% CI, 1.07–2.40, P = .989), lung (OR, 1.47; 95% CI, 1.23–1.74, P = .356), and liver (OR, 2.56; 95% CI, 1.52–4.32, P = .682). However, no significant association was found for leukemia (OR, 1.58; 95% CI, 0.97–2.57, P = .312) and lymphoma (OR, 0.79; 95% CI, 0.13–4.95, P = .097) as well as cancers of the larynx (OR, 1.86; 95% CI, 0.92–3.76, P = .321), colon (OR, 1.35; 95% CI, 0.38–4.81, P = .356), kidney (OR, 1.14; 95% CI, 0.68–1.89, P = .357), and prostate (OR, 1.44; 95% CI, 0.97–2.14, P = .070). The between-study heterogeneity was insignificant among the individual studies (I2 = 40.4%, P = .013). Thus, the fixed-effects model was used to estimate the cancer incidence among women. Moreover, the combined results were further confirmed by sensitivity analysis, and subgroup analyses were conducted to investigate the potential factors that may substantially affect the between-study heterogeneity.
3.5 Subgroup analyses
In the subgroup analyses by study design and country (Table 3), the findings on the associations of butchers with the risks of cancer mortality and incidence were consistent. Accordingly, under stratification analysis by country, the pooled ORs showed that working as butchers was significantly related to an elevated risk of cancer incidence in the USA, Sweden, New Zealand, and the Netherlands but not in Uruguay. Under stratification analysis by study design, the pooled ORs revealed that butchers were significantly related to an elevated risk of cancer incidence in all study designs. However, subgroup analyses were not performed for age and sex considering the limited number of independent datasets.
3.6 Sensitivity analysis
Sensitivity analysis was conducted to determine whether a certain study strongly influences the estimates between butchers and the risks of cancer mortality and incidence or affected the final heterogeneity. We evaluated the effect of each study on the methodological quality by the sequential exclusion of each study. The stability of the results did not significantly change (Fig. 7); thus, the rationality and reliability of our analysis were validated.
3.7 Evaluation of publication bias
Funnel-plot visual inspection and Egger linear regression test were conducted to check for publication bias (Fig. 8). The Egger test showed an insignificant result (P = .979), which indicates that our study possesses a low probability of publication bias.
In this study, we analyzed the associations of butchers with the risks of cancer-related mortality and incidence using a meta-analysis of 17 included studies[9–13,20–31] to obtain a powerful conclusion. To the best of our knowledge, this research is the first meta-analysis that provides comprehensive insights into the relationships between butcher occupation and risks of cancer mortality and incidence through a summary and review of previously published quantitative studies to answer various clinical questions related to this field. Overall, our results demonstrated that working as butchers significantly increased cancer incidence risk by 1.51 times, particularly for glioma as well as stomach, oral cavity, pharynx, lung, and liver cancers. By contrast, no positive association was found for leukemia; lymphoma; and laryngeal, colon, kidney, and prostate cancers. The current meta-analysis provided evidence that butchers were not associated with an elevated risk of total cancer mortality. The overall estimates were robust in the subgroup and sensitivity analyses. No evident publication bias was detected by visual inspection of funnel plot and Egger test.
Recent searches have documented that butchers hold a higher risk of developing cancers, such as lung and colon cancer.[20–22] Actually, zoonotic viruses and chemicals may contribute to the emergence of the lymphatic and hematopoietic system cancers among butchers and animals.[32,33] Moreover, some human retroviruses, such as sarcoma and avian leukosis virus, are known to cause cattle and chicken leukemia because of their frequent contact with the animals’ fluids and organs.[34–37] Several studies have confirmed that antibodies to avian reticuloendotheliosis viruses are present in the butcher's serum during meat curing and processing, respectively.[38,39] Therefore, future studies should further classify exposure not only by job category, but also by specific tasks and job titles. Our understanding of the effects of participant age and sex in the included studies on the overall results remains insufficient, although these factors have been investigated, but inadequately, in other studies. Thus, further research is needed to verify the findings of this meta-analysis with regard to different ethnic populations, low-bias risk, and adjusted confounding factors on extensive consequences. Nevertheless, the subgroup and sensitivity analyses did not modify the pooled results.
In general, our meta-analysis exhibited significant strengths. First, this meta-analysis is the first to assess the potential correlations of butcher occupation with the risks of cancer mortality and incidence in populations worldwide by a thorough systematic search and rigorous analytical approaches. Second, the rationality and reliability of our meta-analysis were remarkably improved by overall combined estimates based on a large sample size. Third, multivariable-adjusted risk estimates were used to minimize the confounding factors, such as smoking, which might have influenced the overall results. These estimates were also used to reflect the correlations of butchers with the risks of cancer mortality and incidence accurately and obtain well-founded conclusions. Sufficient subgroup analyses and sensitivity analyses were performed to ensure study reliability.
However, the current meta-analysis was restricted by several limitations. First, residual confounding and non-measurable factors were present in the included observational studies. The accuracy of the results may be increased by adjusting other confounding factors, such as age, sex, and ethnicity. Second, most of the included studies were performed in Europe, America, and Oceania. Therefore, the conclusions should be considered cautiously for other ethnic populations. We suggest that population-based cohort studies be conducted to explore the association in study under each ethnicity. Third, a potential publication bias likely exists despite the lack of evidence for this occurrence gained from our statistical tests. Lastly, only English language reports were included. Hence, we may have missed data from important studies published in other languages.
In summary, our meta-analysis suggests that the overall cancer incidence among butchers is higher than that among the general population, particularly for glioma as well as stomach, oral cavity, pharynx, lung, and liver cancers. However, butchers do not exhibit elevated cancer mortality risk, specifically, in colon and lung cancers. Additional studies should be conducted to confirm whether the mortality from other cancers can be attributed to chance. Further investigation must also be performed to clarify the potential biological mechanisms further and update the findings of this analysis.
We would like to thank KGSupport for editing the language of this study.
. Buehring GC, Philpott SM, Choi KY. Humans have antibodies reactive with bovine leukemia virus. AIDS Res Hum Retroviruses 2003;19:1105–13.
. McLean D, Cheng S, ’t Mannetje A, et al. Mortality and cancer incidence in New Zealand meat workers. Occup Environ Med 2004;61:541–7.
. Ortín A, Cousens C, Minguijón E, et al. Characterization of enzootic nasal tumour virus of goats: complete sequence and tissue distribution. J Gen Virol 2003;84:2245–52.
. Jakszyn P, Agudo A, Ibáñez R, et al. Development of a food database of nitrosamines, heterocyclic amines, and polycyclic aromatic hydrocarbons. J Nutr 2004;134:2011–4.
. Cousens C, Thonur L, Imlach S, et al. Jaagsiekte sheep retrovirus is present at high concentration in lung fluid produced by ovine pulmonary adenocarcinoma-affected sheep and can survive for several weeks at ambient temperatures. Res Vet Sci 2009;87:154–6.
. Scully C. Oral squamous cell carcinoma; from an hypothesis about a virus, to concern about possible sexual transmission. Oral Oncol 2002;38:227–34.
. Fox AJ, Lynge E, Malker H. Lung cancer in butchers. Lancet 1982;319:165–6.
. Deangelis TM, Shen L. Outbreak of progressive inflammatory neuropathy following exposure to aerosolized porcine neural tissue. Mt Sinai J Med 2009;76:442–7.
. Johnson ES, Ndetan H, Felini MJ, et al. Mortality in workers employed in pig abattoirs and processing plants. Environ Res 2011;111:871–6.
. Aragonés N, Pollán M, Gustavsson P. Stomach cancer and occupation in Sweden: 1971–89. Occup Environ Med 2002;59:329–37.
. Bethwaite P, McLean D, Kennedy J, et al. Adult-onset acute leukemia and employment in the meat industry: a New Zealand case-control study. Cancer Causes Control 2001;12:635–43.
. Boffetta P. Employment as butcher and cancer risk in a record-linkage study from Sweden. Cancer Causes Control 2000;11:627–33.
. Guberan E. Mortality and incidence of cancer among a cohort of self employed butchers from Geneva and their wives. Br J Ind Med 1993;50:1008–16.
. Moher D, Liberati A, Tetzlaff J, et al. Preferred reporting items for systematic reviews and meta-analyses: the PRISMA statement. BMJ 2009;339:b2535.
. Stroup DF, Berlin JA, Morton SC, et al. Meta-analysis of observational studies in epidemiology: a proposal for reporting. JAMA 2000;283:2008e12.
. Hartling L, Milne A, Hamm MP, et al. Testing the Newcastle Ottawa Scale showed low reliability between individual reviewers. J Clin Epidemiol 2013;66:982–93.
. Greenland S. Quantitative methods in the review of epidemiologic literature. Epidemiol Rev 1987;9:1–30.
. Larsson SC, Mantzoros CS, Wolk A. Diabetes mellitus and risk of breast cancer: a meta-analysis. Int J Cancer 2007;121:856–62.
. Adami HO, Hunter D, Trichopoulos D. Textbook of Cancer Epidemiology. NY: Oxford University Press; 2008.
. Besson H, Banks R, Boffetta P. Cancer mortality among butchers: a 24-state death certificate study. J Occup Environ Med 2006;48:289–93.
. Coggon D, Wield G. Mortality of butchers and cooks identified from the 1961 census of England and Wales. Occup Environ Med 1995;52:157–9.
. Corbin M, McLean D, Mannetje A’, et al. Lung cancer and occupation: a New Zealand cancer registry-based case-control study. Am J Ind Med 2011;54:89–101.
. Durusoy R, Boffetta P, Mannetje A’, et al. Lung cancer risk and occupational exposure to meat and live animals. Int J Cancer 2006;118:2543–7.
. Gustavsson P, Fellenius E, Hogstedt C. Possible causes of increased lung cancer incidence among butchers and slaughterhouse workers. Scand J Work Environ Health 1987;13:518–23.
. Metayer C, Johnson ES, Rice JC. Nested case-control study of tumors of the hemopoietic and lymphatic systems among workers in the meat industry. Am J Epidemiol 1998;147:727–38.
. Neasham D, Sifi A, Nielsen KR, et al. Occupation and risk of lymphoma: a multicentre prospective cohort study (EPIC). Occup Environ Med 2011;68:77–81.
. De Roos AJ, Stewart PA, Linet MS, et al. Occupation and the risk of adult glioma in the United States. Cancer Causes Control 2003;14:139–50.
. Ruder AM, Waters MA, Carreón T, et al. The Upper Midwest Health Study: industry and occupation of glioma cases and controls. Am J Ind Med 2012;55:747–55.
. De Stefani E, Boffetta P, Oreggia F, et al. Occupation and the risk of laryngeal cancer in Uruguay. Am J Ind Med 1998;33:537–42.
. Stefani E. Occupational exposures and risk of adenocarcinoma of the lung in Uruguay. Cancer Causes Control 2005;16:851–6.
. Suarez L, Weiss NS, Martin J. Primary liver cancer death and occupation in Texas. Am J Ind Med 1989;15:167–75.
. Chen SL, Wang SC, Ho CJ, et al. Prostate cancer mortality-to-incidence ratios are associated with cancer care disparities in 35 countries. Sci Rep 2017;4:40003.
. Lloyd MM, Lyster WR, Lloyd OL. Birth sex ratios and prostatic cancer in butchers. Lancet 1987;7:561.
. Johnson ES. Poultry oncogenic retrovimses and humans. Cancer Detect Prev 1994;18:9–30.
. Gilsing AM, Schouten LJ, Goldbohm RA, et al. Vegetarianism, low meat consumption and the risk of colorectal cancer in a population based cohort study. Sci Rep 2015;28:13484.
. Brüske-Hohlfeld I, Möhner M, Pohlabeln H, et al. Occupational lung cancer risk for men in Germany: results from a pooled case–control study. Am J Epidemiol 2000;151:384–95.
. Johnson ES, Yau LC, Zhou Y, et al. Mortality in the Baltimore Union poultry cohort-non-malignant diseases. Int Arch Occup Environ Health 2010;83:543–52.
. Ortín A, Cousens C, Minguijón E, et al. Characterization of enzootic nasal tumour virus of goats: complete sequence and tissue distribution. Gen Virol 2003;84:2245–52.
. Tucker AW, Mellencamp MM, Donadeu M, et al. Retroviremia in commercial pigs and its preliminary association with poor health. J Clin Microbiol 2006;44:3846–7.
Keywords:Copyright © 2017 The Authors. Published by Wolters Kluwer Health, Inc. All rights reserved.
butcher; cancer; incidence; meta-analysis; mortality; systematic review