Heavy alcohol intake is a risk factor for esophageal squamous cell carcinoma among middle-aged men: A case-control and simulation study
- Authors:
- Published online on: July 4, 2013 https://doi.org/10.3892/mco.2013.142
- Pages: 811-816
Abstract
Introduction
The pathogenesis of esophageal cancer has not been fully elucidated and, despite the advances in surgical techniques and treatments, its prognosis remains poor, since the disease is usually diagnosed at an advanced stage (1–5). Thus, prevention plays an important role in reducing mortality. Numerous risk factors for esophageal cancer have been identified, including tobacco smoking, alcohol intake, hot beverage intake, family history of cancer and low fruit and vegetable intake (4–8), the first two are considered to be the most significant factors. These habits are correlated with the disease at the individual level and are potentially modifiable (9). Notably, the incidence of esophageal squamous cell carcinoma (ESCC) is on the decrease in the USA in association with a decline in smoking and alcohol intake (10).
Several studies investigating the effects of smoking and alcohol intake on the risk of esophageal cancer have demonstrated that long duration, high consumption and the interaction of these habits may increase the risk of ESCC (9,11–17). However, with the exception of two studies which demonstrated that the effects of smoking and alcohol intake on esophageal cancer may differ according to gender (9,18), to the best of our knowledge, only a limited number of large-scale studies have investigated the association of the amount and duration of smoking and alcohol intake with the risk of esophageal cancer, while accounting for gender-specific effects.
In this study, the incidence of ESCC among Chinese males was investigated by performing conditional logistic regression analysis of the combined effects of the amount and duration of smoking and alcohol intake on ESCC risk in a relatively large study population of 1,070 males from a hospital-based casecontrol study. The combined effects were estimated using a variable that combined amount and duration. The reliability of our results was subsequently confirmed by means of a simulation. The aim of investigating the combined effects of amount and duration of smoking and alcohol intake on the risk of ESCC was to provide sufficient data to guide improvements in prevention strategies.
Materials and methods
Design
In this case-control study, each patient was age-matched to a control subject (1:1 ratio) and a total of 535 pairs were enrolled.
Participants
A total of 571 Chinese male ESCC patients, aged <60 years, were enrolled from the Fourth Hospital of Hebei Medical University in Shijiazhuang, China. They were all residents of Shijiazhuang and between January, 2002 and December, 2006 had been diagnosed with histologically confirmed primary invasive ESCC, as defined by the International Classification of Diseases, 10th revision (ICD-10) diagnostic code (code, C15). In addition, 1,307 healthy male subjects, aged <60 years, were selected from the same area and during the same time to serve as controls. All controls had undergone medical checkups that verified the absence of any medical conditions. The participants were informed on the study objectives and content and provided informed consent prior to enrollment. The study protocol was approved by the local Ethics Committee and the Research Committee of Hebei Medical University.
Risk factors
Trained interviewers used a structured questionnaire to interview the participants. The questionnaire included questions on demographic factors such as age, gender, smoking behavior (age of initiation and cessation of smoking, duration of smoking and the number of cigarettes per day), alcohol intake (age of initiation and cessation of alcohol intake, duration of alcohol intake and type and the quantity of alcoholic beverages consumed) and the family history of esophageal cancer. A non-drinker was defined as someone who consumed <1 alcoholic drink/month. Daily alcohol intake (g ethanol/day) was calculated by considering the beverage type and daily consumption. Beverage type was classified as baijiu (the most popular Chinese alcoholic beverage), beer or other hard liquor. One unit of consumption was defined as 50 ml of baijiu, 400 ml of beer or 50 ml of hard liquor. The content of pure alcohol was calculated as follows: 40% for baijiu, 5% for beer and 40% for hard liquor. The resulting values were expressed in grams by converting 1 ml of pure ethanol to 0.789 g.
All continuous variables (number of cigarettes/day, duration of smoking, daily alcohol intake and duration of alcohol consumption) were assigned to categories of zero, low and heavy. Absence of smoking or alcohol intake was categorized as zero (reference group). To ensure equal sample sizes, median values were used to divide tobacco smokers and alcohol drinkers into low and heavy groups. Values ranging from >0 to the median were classified as low and values greater than the median were classified as heavy. The duration and amount categories were subsequently combined to investigate the association of duration and amount of smoking or alcohol intake with ESCC risk. The combined smoking variable comprised 5 categories: no smoking, ≤20 cigarettes/day for ≤20 years, ≤20 cigarettes/day for >20 years, >20 cigarettes/day for ≤20 years and >20 cigarettes/day for >20 years. Five combined categories were created for alcohol intake: no alcohol intake, ≤53.3 g ethanol/day for ≤20 years, ≤53.3 g ethanol/day for >20 years, >53.3 g ethanol/day for ≤20 years and >53.3 g ethanol/day for >20 years.
Statistical analysis
The individual risk factors for ESCC (i.e., family history of esophageal cancer, combined smoking variable and combined alcohol variable) were assessed by univariate conditional logistic regression analysis, and unadjusted ORs and 95% CIs were estimated. Multivariate conditional logistic regression analysis was used to assess the association between the amount and duration of smoking and alcohol intake and the risk of ESCC. Fisher’s exact test was used to compare the proportion of ESCC cases between males aged 30–49 and 50–59 years for each category of the combined alcohol variable. Each proportion was shown in a bar graph to demonstrate the likelihood of developing ESCC at a younger or older age. P<0.05 was considered to indicate a statistically significant difference. P values were derived from two-sided statistical tests. Statistical analysis was performed using SAS statistical software version 9.2 (SAS Institute Inc., Cary, NC, USA).
Although the control for each case was randomly selected from the candidate pool comprising several controls of matching age, the OR estimates may have been inconsistant, depending on the control selected. Therefore, a simulation study was performed to evaluate the reliability of the OR estimates obtained from the multivariate conditional logistic regression analysis. Each simulation dataset included the 535 age-matched pairs. The controls were randomly extracted from the original database of 1,307 male controls, without replacement. For each dataset, conditional logistic regression analysis was performed with the previous three variables. This process was repeated 10,000 times, thus generating 10,000 estimates of coefficients and standard errors for the variables. The simulation generated means for coefficients and standard errors, which were used to produce a summary of ORs and 95% CIs.
Results
Participant characteristics
Selected characteristics of patients and controls are presented in Table I. A total of 772 subjects (72.2%) were aged 50–59 years and 270 (25.2%) were aged 40–49 years. These age groups represented 97.4% of the participants. Overall, 82.8% of the alcohol drinkers also had a smoking habit.
Univariate conditional logistic regression
The univariate conditional logistic regression analysis revealed that family history of esophageal cancer, the combined smoking variable and the combined alcohol variable were significantly associated with ESCC risk (P<0.001) (Table I). With regard to smoking, a duration of ≤20 years was not a significant risk factor, irrespective of the quantity of cigarettes smoked. With regard to alcohol intake, all categories of the combined alcohol variable, apart from an intake of ≤53.3 g ethanol for ≤20 years, were significantly associated with ESCC risk.
Multivariate conditional logistic regression
Multivariate analysis revealed that a family history of esophageal cancer and the combined alcohol variable were significantly associated with ESCC risk (P<0.001), although the combined smoking variable was not (Table II). The OR (95% CI) associated with a family history of esophageal cancer was 2.06 (1.48–2.85). With regard to the combined alcohol variable, the OR (95% CI) associated with an intake of ≤53.3 g ethanol/day for ≤20 years was 1.20 (0.83–1.74); with an intake of ≤53.3 g ethanol/day for >20 years, 2.28 (1.32–3.94); with an intake of >53.3 g ethanol/day for ≤20 years, 1.91 (1.25–2.92); and with an intake of >53.3 g ethanol/day for >20 years, 7.25 (3.12–16.83). All the categories of the combined alcohol variable, apart from an intake of ≤53.3 g ethanol/day for ≤20 years, were significantly associated with risk. Alcohol intake for >20 years was a risk factor, as was a heavy alcohol intake of >53.3 g ethanol/day. Heavy alcohol intake was significantly associated with ESCC risk, even if the duration of alcohol intake was ≤20 years.
Comparison of the proportion of ESCC among males aged 30–49 and 50–59 years by category of combined alcohol variable
Among males who have consumed alcohol for ≤20 years, the proportion of ESCC cases was higher among older males compared with younger males with an intake of ≤53.3 g ethanol/day (P=0.035) (Fig. 1). By contrast, the proportion was higher among younger males compared with older males with an intake of >53.3 g ethanol/day, although the difference was not statistically significant (P=0.250). Thus, heavy alcohol intake may lead to the development of ESCC at a younger age.
Simulation
The multivariate conditional logistic regression and corresponding simulation study demonstrated that the same factors were significantly associated with ESCC risk (Table II).
Discussion
Our analysis indicated that heavy alcohol intake (>53.3 g ethanol/day) is an important risk factor for ESCC in middle-aged Chinese males. In addition, heavy alcohol intake, even of short duration, may lead to the development of ESCC at a younger age. Smoking was not identified as a significant risk factor for ESCC.
We investigated ESCC in a cohort of middle-aged Chinese males, aged <60 years, for the following reasons: although alcohol intake may be a dominant risk factor for both sexes, a previous study by Wu et al(18) revealed that alcohol intake increased the risk of ESCC risk in males, with an OR of 1.76 (95% CI: 1.48–2.09). However, the association was not significant in females, a finding which was supported by those of Castellsagué et al(9), who observed that alcohol intake was associated with esophageal cancer in males, with an OR of 4.4 (95% CI: 3.1–6.2) and to a lesser extent in females, with an OR of 2.2 (95% CI: 1.3–3.9). Against this background, Chinese males were focused on in this study, as it would appear that they may be at higher risk of developing ESCC compared to females. Moreover, with regard to age, patients and controls aged <60 years were recruited in this study in order to avoid the numerous factors associated with cancer differentiation and development in elderly adults.
Our study demonstrated a stronger association of ESCC development with heavy alcohol intake, compared to long duration of alcohol intake, confirming the findings of previous studies (14,16,17). When daily ethanol consumption exceeded 53.3 g, the risk of developing ESCC within the next 20 years was increased. The exact mechanism by which ethanol causes esophageal cancer has not been elucidated, although several possible pathways have been proposed. Although ethanol itself is not carcinogenic, its major intermediary metabolite, acetaldehyde, is a known carcinogen in animals; alcohol may act as a solvent that enhances the penetration of carcinogens from other environmental sources; the regular intake of alcohol consumption may reduce the intake and bioavailability of certain nutrients that have chemopreventive properties; and alcohol may directly irritate the esophageal epithelium, creating the potential for ESCC pathogenesis (20). Whereas heavy alcohol intake may yield more acetaldehyde and reduce the intake and bioavailability of certain nutrients, the solvent properties of alcohol and its ability to directly irritate the esophageal epithelium are most likely causes for our observations in middle-aged Chinese males. Firstly, the majority (82.8%) of the alcohol drinkers in our study population were also smokers. Tobacco contains the chemical carcinogen nitrosamine, as well as other cancer-promoting agents (5). These agents may act synergistically with ethanol to increase ESCC risk (5) and our results may be explained by the concurrence of the two habits. Secondly, it has previously been demonstrated that exposure to a high concentration of ethanol (40%) severely damaged the esophageal mucosa of rabbits, whereas a lower concentration (20%) exerted a notably lower adverse effect (21). Furthermore, consumption of undiluted hard liquor (>40% ethanol) was found to be significantly harmful, whereas wine, beer and hard liquor with a non-alcoholic mixer were not (11,20). With an ethanol content of 40–60%, baijiu is consumed undiluted (22) and is the beverage of choice among Chinese males, accounting for >1/3 of all alcoholic beverages consumed in China (22). The high concentration of alcohol may directly irritate the esophageal epithelium and lead to the development of ESCC within a short period of time, i.e., at a younger age.
Although a previous meta-analysis of worldwide data demonstrated that smoking is an established risk factor for esophageal cancer (23), according to our data it was not a significant risk factor among our cohort of middle-aged Chinese males. Evidence on the ESCC risk associated with smoking in China is inconsistent. Weak associations with tobacco have been reported in Chinese regions exhibiting a significantly high risk for esophageal cancer, such as Linxian, whereas strong associations have been reported in low-risk urban areas (5,6,20,24). The reason for the weaker association in these areas compared to the rest of the world has not yet been elucidated. One hypothesis is that there are other important risk factors in these high-risk areas and that these factors account for the majority of ESCC cases, thus reducing the effect of smoking (6). The lack of association in the present study may also be due to the effect of other risk factors.
This hospital-based case-control study has several limitations. Although the age-matched case-control design was used to decrease the effects of potential confounding factors, ESCC risk was estimated by only using information on alcohol intake, smoking and family history of esophageal cancer. Other categories of smoking and drinking, such as currently or formerly practiced, were not investigated. Moreover, heavy alcohol intake may be correlated with nutritional, socioeconomic or educational status, although we were unable to collect data on other possible risk factors, such as diet and social status. However, we believe that the limitations mentioned above did not significantly affect the study findings, as the differences among groups were too clear to have resulted from such bias.
In conclusion, heavy alcohol intake, even of short duration, is an important risk factor for ESCC in middle-aged Chinese males. Specifically, middle-aged Chinese males who consume >53.3 g ethanol/day (~170 ml of baijiu) are at high risk of developing ESCC, even with a duration of heavy alcohol intake of <20 years. This finding indicates that such individuals are at high risk of developing ESCC at a younger age. The present findings strongly suggest that limiting the consumption of undiluted hard liquor may decrease the incidence of ESCC in the investigated population.
Acknowledgements
The authors would like to thank the study participants and the staff at the Fourth Hospital of Hebei Medical University for their assistance.
Abbreviations:
|
CI |
confidence interval |
|
ESCC |
esophageal squamous cell carcinoma |
|
OR |
odds ratio |
References
|
Tew WP, Kelsen DP and Ilson DH: Targeted therapies for esophageal cancer. Oncologist. 10:590–601. 2005. View Article : Google Scholar : PubMed/NCBI | |
|
Chang AC, Ji H, Birkmeyer NJ, Orringer MB and Birkmeyer JD: Outcomes after transhiatal and transthoracic esophagectomy for cancer. Ann Thorac Surg. 85:424–429. 2008. View Article : Google Scholar : PubMed/NCBI | |
|
Enzinger PC and Mayer RJ: Esophageal cancer. N Engl J Med. 349:2241–2252. 2003. View Article : Google Scholar : PubMed/NCBI | |
|
Kollarova H, Machova L, Horakova D, Janoutova G and Janout V: Epidemiology of esophageal cancer - an overview. Biomed Pap Med Fac Univ Palacky Olomouc Czech Repub. 151:17–20. 2007. View Article : Google Scholar : PubMed/NCBI | |
|
Mao WM, Zheng WH and Ling ZQ: Epidemiologic risk factors for esophageal cancer development. Asian Pac J Cancer Prev. 12:2461–2466. 2011.PubMed/NCBI | |
|
Kamangar F, Chow WH, Abnet CC and Dawsey SM: Environmental causes of esophageal cancer. Gastroenterol Clin North Am. 38:27–57. 2009. View Article : Google Scholar | |
|
Melhado R, Alderson D and Tucker O: The changing face of esophageal cancer. Cancers. 2:1379–1404. 2010. View Article : Google Scholar : PubMed/NCBI | |
|
Castellsagué X, Muñoz N, De Stefani E, Victora CG, Castelletto R and Rolón PA: Influence of mate drinking, hot beverages and diet on esophageal cancer risk in South America. Int J Cancer. 88:658–664. 2000.PubMed/NCBI | |
|
Castellsagué X, Muñoz N, De Stefani E, Victora CG, Castelletto R, Rolón PA and Quintana MJ: Independent and joint effects of tobacco smoking and alcohol drinking on the risk of esophageal cancer in men and women. Int J Cancer. 82:657–664. 1999.PubMed/NCBI | |
|
Zhang HZ, Jin GF and Shen HB: Epidemiological differences in esophageal cancer between Asian and Western populations. Chin J Cancer. 31:281–286. 2012. View Article : Google Scholar : PubMed/NCBI | |
|
Castellsagué X, Quintana MJ, Martínez MC, et al: The role of type of tobacco and type of alcoholic beverage in oral carcinogenesis. Int J Cancer. 108:741–749. 2004.PubMed/NCBI | |
|
Franceschi S, Talamini R, Barra S, Barón AE, Negri E, Bidoli E, Serraino D and La Vecchia C: Smoking and drinking in relation to cancers of the oral cavity, pharynx, larynx, and esophagus in northern Italy. Cancer Res. 50:6502–6507. 1990.PubMed/NCBI | |
|
Muwonge R, Ramadas K, Sankila R, Thara S, Thomas G, Vinoda J and Sankaranarayanan R: Role of tobacco smoking, chewing and alcohol drinking in the risk of oral cancer in Trivandrum, India: a nested case-control design using incident cancer cases. Oral Oncol. 44:446–454. 2008. View Article : Google Scholar : PubMed/NCBI | |
|
Zambon P, Talamini R, La Vecchia C, Dal Maso L, Negri E, Tognazzo S, Simonato L and Franceschi S: Smoking, type of alcoholic beverage and squamous-cell oesophageal cancer in northern Italy. Int J Cancer. 86:144–149. 2000. View Article : Google Scholar : PubMed/NCBI | |
|
Lee CH, Lee JM, Wu DC, Hsu HK, Kao EL, Huang HL, Wang TN, Huang MC and Wu MT: Independent and combined effects of alcohol intake, tobacco smoking and betel quid chewing on the risk of esophageal cancer in Taiwan. Int J Cancer. 113:475–482. 2005. View Article : Google Scholar : PubMed/NCBI | |
|
Launoy G, Milan CH, Faivre J, Pienkowski P, Milan CI and Gignoux M: Alcohol, tobacco and oesophageal cancer: effects of the duration of consumption, mean intake and current and former consumption. Br J Cancer. 75:1389–1396. 1997. View Article : Google Scholar : PubMed/NCBI | |
|
Chen J, Zhang N, Wakai T, Wei L, He Y, Kumagai N, Kitsu K, Wang S and Akazawa K: Effect of the interaction between the amount and duration of alcohol consumption and tobacco smoking on the risk of esophageal cancer: A case-control study. Exp Ther Med. 1:991–997. 2010.PubMed/NCBI | |
|
Wu M, Zhao JK, Zhang ZF, et al: Smoking and alcohol drinking increased the risk of esophageal cancer among Chinese men but not women in a high-risk population. Cancer Causes Control. 22:649–657. 2011. View Article : Google Scholar : PubMed/NCBI | |
|
Lundell LR: Etiology and risk factors for esophageal carcinoma. Dig Dis. 28:641–644. 2010. View Article : Google Scholar : PubMed/NCBI | |
|
Fan Y, Yuan JM, Wang R, Gao YT and Yu MC: Alcohol, tobacco, and diet in relation to esophageal cancer: the Shanghai Cohort Study. Nutr Cancer. 60:354–363. 2008. View Article : Google Scholar : PubMed/NCBI | |
|
Salo JA: Ethanol-induced mucosal injury in rabbit oesophagus. Scand J Gastroenterol. 18:713–721. 1983. View Article : Google Scholar : PubMed/NCBI | |
|
Cochrane J, Chen H, Conigrave KM and Hao W: Alcohol use in China. Alcohol Alcohol. 38:537–542. 2003. View Article : Google Scholar : PubMed/NCBI | |
|
Gandini S, Botteri E, Iodice S, Boniol M, Lowenfels AB, Maisonneuve P and Boyle P: Tobacco smoking and cancer: a meta-analysis. Int J Cancer. 122:155–164. 2008. View Article : Google Scholar : PubMed/NCBI | |
|
Tran GD, Sun XD, Abnet CC, Fan JH, Dawsey SM, Dong ZW, Mark SD, Qiao YL and Taylor PR: Prospective study of risk factors for esophageal and gastric cancers in the Linxian general population trial cohort in China. Int J Cancer. 113:456–463. 2005. View Article : Google Scholar : PubMed/NCBI |
