Annals of Oncology Advance Access originally published online on October 17, 2006
Annals of Oncology 2007 18(1):143-148; doi:10.1093/annonc/mdl352
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
© 2006 European Society for Medical Oncology
hematologic malignancies |
Alcohol consumption and risk of Hodgkin's lymphoma and multiple myeloma: a multicentre case–control study
1 Istituto Toscano Tumori, Environmental and Occupational Epidemiology Unit, Centre for Study and Prevention of Cancer, Florence
2 Epidemiology and Biostatistics Unit, National Cancer Research Institute, Genoa
3 National Cancer Institute Regina Elena, Rome
4 Oncology Department, Azienda USL Forlì, Forlì
5 Azienda Ospedaliera, Verona
6 Cancer registry & Pathology Unit, Azienda Ospedaliera "Civile M.P. Arezzo," Ragusa
7 Epidemiology Unit, National Cancer Institute, Milan
8 Pathology Institute, University of Siena, Siena
9 Local Health Unit, Novara
10 Cancer Epidemiology Unit, University of Turin, Italy
11 Imperial College—London, UK
* Correspondence to: Dr G. Gorini, Istituto Toscano Tumori, Environmental and Occupational Epidemiology Unit, Centre for Study and Prevention of Cancer, via di S. Salvi, 12-50135 Florence, Italy. Tel: +39-055-6268-347/345; Fax: +39-055-6268385/679954; E-mail: g.gorini{at}cspo.it
 |
Abstract |
|---|
 Top Abstract introductionmaterials and methodsresultsdiscussionAcknowledgementsReferences |
|---|
Background: Few studies have analysed the association between alcohol intake and Hodgkin's lymphoma (HL) or multiple myeloma (MM) risks.
Materials and methods: A multicentre population-based case–control study of 363 HL, 270 MM cases, and 1771 controls offered the opportunity to evaluate the relationship between alcohol and HL/MM risks. Unconditional logistic regression was carried out to estimate odds ratios (ORs) and 95% confidence intervals (CIs), associated with alcohol intake (servings per week, grams per day of ethanol intake) or duration of exposure (year).
Results: For HL, considering nonsmokers only, ever drinkers had a significantly decreased risk than never drinkers (OR = 0.46). Significantly lower risks in all levels of total alcohol intake were also detected, considering servings per week (OR for one to four servings per week = 0.51, 95% CI 0.32–0.82; OR for five to nine servings per week = 0.39, 95% CI 0.21–0.73; OR for 10–19 servings per week = 0.26, 95% CI 0.12–0.54; OR for 
20 servings per week = 0.34, 95% CI 0.15–0.79) and grams per day of ethanol intake (OR for 0.1–9.0 g/day = 0.45, 95% CI 0.27–0.74; OR for 9.1–17.9 g/day = 0.52, 95% CI 0.30–0.90; OR for 18.0–31.7 g/day = 0.27, 95% CI 0.13–0.57; OR for >31.7 g/day = 0.35, 95% CI 0.15–0.79). In the analysis for ever-smoking HL cases and controls, ever drinkers had the same risk as never drinkers. For MM, ever drinkers had a non-significantly decreased risk than non-drinkers (OR = 0.74), and ORs in almost all consumption levels were not significant (OR for 0.1–9.0 g/day = 0.93; OR for 9.1–17.9 g/day = 0.82; OR for 18.0–31.7 g/day = 0.47; 95% CI 0.28–0.81; OR for >31.7 g/day = 0.68). For HL and MM, the beverage type did not affect the risk significantly, and no consistent dose–response relationships were found, considering intensity or duration of alcohol consumption.
Conclusions: Our study indicates a protective effect of alcohol consumption for nonsmoking HL cases.
Key words: alcohol, case–control study, Hodgkin's lymphoma, multiple myeloma
|
introduction |
|---|
|
TopAbstractintroductionmaterials and methodsresultsdiscussionAcknowledgementsReferences |
|---|
Few studies have analysed the association between alcohol intake and Hodgkin's lymphoma (HL) [1–4] or multiple myeloma (MM) [1, 3–7].
The first study for HL [1] found a non-significantly negative association between alcohol intake and HL risk. The study by Bernard et al. [2] reported a significantly inverse association [odds ratio (OR) for wine drinker = 0.5, 95% confidence interval (CI) 0.3–0.8; OR for spirits drinker = 0.7, 95% CI 0.4–1.0], while the study by Tavani et al. [3] found no association. The most recent study [4] reported a strong inverse association in both gender (OR for men = 0.51, 95% CI 0.26–1.01; OR for women = 0.42, 95% CI 0.20–0.89).
The first study on MM risk and alcohol consumption [1] reported a non-significantly inverse association, while the studies by Linet et al. [5] and Tavani et al. [3] reported no association. The first study by Brown et al. [6] found a slight positive association, while the second study of the same authors [7] found non-significantly decreased risks. The most recent study [4] found a non-significantly decreased risk for men with a non-significantly elevated risk for women (OR for men = 0.56, 95% CI 0.24–1.35; OR for women = 2.14, 95% CI 0.70–6.57).
Alcohol has immunomodulatory effects, which may be different according to the degree of intake. Although heavy alcohol consumption impairs immune function, light or moderate alcohol use might improve cellular and humoral immune responses [8]. Other mechanisms underlying a potential protective effect of moderate alcohol consumption may include antioxidants, such as resveratrol in wine or flavonoids in beer; improvement of insulin sensitivity by alcohol [9, 10]; or the interaction between retinol and ethanol metabolism [11, 12].
The aim of this study is to present an analysis of alcohol consumption and HL and MM risk in a subset of a large population-based case–control study conducted in Italy.
|
materials and methods |
|---|
|
TopAbstractintroductionmaterials and methodsresultsdiscussionAcknowledgementsReferences |
|---|
The Italian study has been described in detail elsewhere [13–16]. Briefly, the study covered 11 Italian areas (Varese, Forlì, Siena, Latina, Ragusa, Imperia, Florence, Novara, Vercelli, and Verona provinces plus the city of Turin), and included all incident HL [International Classification of Diseases (ICD-9) code: 201] and MM (ICD-9 code: 203) cases occurring from 1990 to 1993 in residents of both genders (age 20–74 years). In four areas (Novara, Varese, Vercelli, and Verona), only HL cases were collected. The overall study covered a population of
7 million residents. Cases were identified as follows: (i) systematic searches in departments of hematology, general medicine, and surgery, in all hospitals within the above-mentioned areas; (ii) specialized hospitals outside these areas, where such patients could be admitted; (iii) departments of pathology in the study areas; and (iv) cancer registries.
Diagnoses were based on histological analyses for HL cases and on morphological, cytochemical, and immunologic analyses for MM. All diagnoses of HL were classified by an experienced pathologist.
The control group was randomly selected from the population aged 20–74 years living in the same areas of cases. The sample was stratified according to gender and 5-year age groups.
Personal interviews, lasting 1 h each, were conducted in 1990–1994 by trained interviewers with subjects or with close relatives, if subjects were deceased or too ill to be interviewed. All subjects gave their informed consent to participate in the study.
A standardized questionnaire was used to obtain detailed information about sociodemographic characteristics, tobacco, and alcohol consumption. Cases and controls were asked if they had ever drunk alcohol at least once a month as an adult and, if so, the type of alcoholic beverage they drank. For each type of beverage reported, respondents were further asked how much they usually drank (per day, week, or month) up to 6 months before the date of the lymphoid neoplasm diagnosis and the age at start of alcohol consumption. Drinkers were defined as those individuals who consumed >0.01 g of ethanol a month [17].
Data for beer, wine (wine + fortified wine), and liquor (aperitif + spirit) consumption were analysed separately, and then summed to estimate total alcohol consumption. The intake of alcohol was analysed considering frequency in terms of servings per week or month and grams per day of ethanol intake, which was calculated by multiplying the frequency of consumption times its respective ethanol content (one beer serving = one can or bottle of beer = 330 ml, 1 ml beer = 0.046 g ethanol; one wine serving = one glass of wine = 125 ml, 1 ml wine= 0.104 g ethanol; one strong wine serving = one glass of strong wine = 90 ml, 1 ml strong wine = 0.167 g ethanol; one aperitif serving = one glass of aperitif = 40 ml, 1 ml aperitif = 0.25 g ethanol; one spirit serving = one shot of spirit = 30 ml, 1 ml spirit = 0.33 g ethanol) [17]. Duration of consumption was calculated from the age at start of alcohol intake to the age at inclusion in the study. Servings per week or month, grams of ethanol intake per day, and duration of exposure were categorized into quartiles, on the basis of the distribution in controls who drank alcohol.
Unconditional logistic regression was carried out to estimate ORs and the corresponding 95% CIs, associated with daily intake of alcohol or with duration of intake, adjusting for age (10-year categories), gender, smoking status (never, current, former), area of residence (11 areas), educational level (three levels: illiterate + primary school, middle school + high school, university), and type of interview (direct, surrogate). Analyses of the linear trend for intensity were done by including variables for alcohol intake or duration of exposure as continuous variables in logistic regression models. All statistical tests were two-sided with an 
level of 0.05. Analyses were carried out with Stata software version 8.0 [18].
|
results |
|---|
|
TopAbstractintroductionmaterials and methodsresultsdiscussionAcknowledgementsReferences |
|---|
A total of 363 HL, 270 MM (88% of cases), and 1771 controls (81%) completed an interview [14]. Table 1 gives the distribution of the interviewed cases and controls by gender, age, residence area, educational level, type of interview, and smoking habits.
|
Seventy-one percent of HL, 74.8% of MM cases, and 78.3% of controls were ever drinkers; 63.5% of HL, 72.4% of MM cases, and 73.1% of controls ever drank wine; 44.6% of HL, 35.6% of MM cases, and 41.6% of controls ever drank beer; 29.3% of HL, 19.9% of MM cases, and 34.0% of controls ever drank liquor. Eighty-one percent of the daily intake of ethanol in controls derived from wine consumption, 72.2% in HL, 87.4% in MM cases.
Tables 2 and 3 show results for HL and MM, in association with total alcohol, wine, beer, and liquor intake, expressed as servings per week or month (Table 2), or as ethanol from alcohol intake, in grams per day (Table 3).
|
|
For HL, we presented separated analyses for ever smoker (cigarettes) and never smoker (cigarettes), since in the logistic regression model (Table 2) the interaction term between ‘ever/never drinker’ and ‘never/current/former smoker’ was significant (likelihood ratio test, chi square, 3 df = 19.27; P = 0.0002). In nonsmoker analysis, ever drinkers had a significantly decreased risk than never drinkers (OR = 0.46), whereas in the analysis for ever-smoking HL cases and controls, ever drinkers had the same risk as never drinkers. In nonsmoker analysis, significantly lower risks in all levels of total alcohol intake were detected, considering servings per week or ethanol intake in grams per day. In ever-smoker analysis, considering servings per week or grams of ethanol per day, risks were non-significantly different between ever and never drinkers.
Considering ever smokers and nonsmokers together, ever drinkers had a decreased risk than non-drinkers (OR = 0.65; 95% CI 0.48–0.88), and lower risks in all levels of alcohol intake were recorded [OR in the first quartile of consumption (0.1–9.0 g/day) = 0.63, 95% CI 0.44–0.91; OR for 9.1–17.9 g/day = 0.63, 95% CI 0.43–0.93; OR for 18.0–31.7 g/day = 0.62, 95% CI 0.41–0.95; OR for >31.7 g/day = 0.65, 95% CI 0.42–1.00).
For MM, ever drinkers had a non-significantly decreased risk than non-drinkers (OR = 0.74). Considering servings per week or grams of ethanol per day, risks were non-significantly different between ever and never drinkers.
Analyses for duration of total alcohol intake did not show any consistent dose–response curve in relation to the risk of HL or MM (data not shown).
|
discussion |
|---|
|
TopAbstractintroductionmaterials and methodsresultsdiscussionAcknowledgementsReferences |
|---|
For HL, we found a significantly protective effect of alcohol consumption, in particular for nonsmokers: our results are consistent with those recorded in the most recent case–control study on alcohol consumption and HL risk [2]. Alcohol has been previously described as a protective factor also for non-HL, in particular for Burkitt's, follicular, and diffuse lymphomas [19].
The role of smoking as risk factor for HL is controversial, since some studies did not find any direct association [20, 21], whereas other studies [22, 23] found significantly increased risks of HL in current smokers. Previous analyses of our study, considering cigarette smoking as a risk factor [13], found no association between smoking and risk of HL. In this analysis, smoking seems to modulate the protective effect of alcohol consumption, since never smokers recorded a strong negative association between alcohol intake and risk of HL, whereas ever smokers recorded no association. We did not find similar results in other studies on HL risk and alcohol consumption [1–4].
The protective effect of alcohol consumption on MM risk seems to be negligible in comparison with that recorded for HL, in particular for never smokers. Our results are similar to those observed in other studies for MM [1, 3–7].
In our study, both for MM and for HL, the type of alcoholic beverage consumed did not affect the risk significantly [4, 6], and no dose–response relation for intensity or duration of alcohol consumption was found [6].
For the overall interpretation of our results, some strengths and limitations of this study should be pointed out. To the strengths, we count the population-based design and the presentation of exposure information from standardized questionnaires obtained by trained interviewers [13, 14].
It is noteworthy that in our study few ever drinkers drank beer or liquor, and even the last quartiles of beer and liquor consumption (>3.0 and >3.5 g/day, respectively) showed a moderate consumption. Moreover, the most part (80%) of the daily intake of ethanol in cases and controls together derived from wine consumption, whereas only 12.1% from beer and 5.8% from liquor intake. These patterns of consumption were consistent with those observed in Italy in the same period [24]. Thus, our results on wine intake were more reliable, whereas findings on beer and liquor intake might be inconclusive, as few people consumed these beverage types and only a moderate consumption of beer and liquor were recorded among drinkers.
Unmeasured confounding by dietary patterns might have partially produced our results, if different levels of alcohol consumption had different dietary patterns which were also associated with lymphoma risk. Moderate consumption of wine correlates with specific behaviour and lifestyle characteristics, including a better diet [25–27]. Results of epidemiological studies on diet and lymphoma risk, however, are inconclusive, and diet cannot be considered an established risk factor for lymphomas [28].
Of concern is the possibility of inaccurate reporting of alcohol consumption. Drinkers tend to underreport the amount of alcohol drunk, and proxy responders may have been less likely than patients to recall details about drinking history, in particular in the distant past. The proportion of drinkers among directly and indirectly interviewed individuals, however, was the same.
Furthermore, in our study the small sample sizes of MM and HL cases could have widened the CI estimates.
In conclusion, our study indicates a significantly protective effect of alcohol consumption for nonsmoker HL cases. Non-significantly decreased risks were also detected for MM. For HL and MM, the beverage type did not affect the risk significantly, and no consistent dose–response relationship was found, considering intensity or duration of alcohol consumption.
Larger case–control studies focused on lifestyle habits, such as alcohol consumption, smoking, and dietary patterns, are needed to more fully explore a possible relationship between alcohol consumption and HL or MM risk.
|
Acknowledgements |
|---|
|
TopAbstractintroductionmaterials and methodsresultsdiscussionAcknowledgementsReferences |
|---|
This work has been carried out with the cooperation of S. Alberghini Maltoni, S. Barcellini, G. Barni, L. Bellesini, V. Cacciarini, R. Carlini, M. Casale, G. Castellino, G. Cremaschi, L. Davico, A. Fiorio, R. Gibilisco, L. Guzzo, R. Hirvas, S. Legrotti, L. Migliaretti, R. Monteleone, G. Osella, T. Palma, G. Panizza, C. Picoco, G. Piergiovanni, G. Righetti, R. Sguanci, M. Tedeschi, D. Tiberti, G. Tonini, P. Trada, T. Vescio, and M. Zanetta. We thank the following members of the panel of pathologists: S. Di Lollo, L. Fiore Donati, U. Magrini, F. Menestrina, D. Novero, G. Palestro, A. Paolucci, and M. Paulli. The authors thank the clinical department staff involved in patient recruitment. This work was funded by the USA National Cancer Institute (grant NCI CA51086), the European Community (Europe against Cancer Programme), and the Italian Alliance against Cancer.
Received for publication May 5, 2006. Revision received July 28, 2006. Accepted for publication August 8, 2006.
|
References |
|---|
|
TopAbstractintroductionmaterials and methodsresultsdiscussionAcknowledgementsReferences |
|---|
1. William RR and Horn JW. (1977) Association of cancer sites with tobacco and alcohol consumption and socio-economic status of patients: interview study from the Third National Cancer Survey. J Natl Cancer Inst 58:525–547.[Web of Science][Medline]
2. Bernard SM, Cartwright RA, Darwin CM, et al. (1987) Hodgkin's disease: case control epidemiological study in Yorkshire. Br J Cancer 55:85–90.[Web of Science][Medline]
3. Tavani A, Pregnolato A, Franceschi S, et al. (1997) Diet and risk of lymphoid neoplasms and soft tissue sarcomas. Nutr Cancer 27:256–260.[Web of Science][Medline]
4. Nieters A, Deeg E, Becker N. (2006) Tobacco and alcohol consumption and risk of lymphoma: results of a population-based case-control study in Germany. Int J Cancer 118:422–430.[CrossRef][Web of Science][Medline]
5. Linet MS, Harlow SD, McLaughlin JK. (1987) A case-control study of multiple myeloma in whites: chronic antigenic stimulation, occupation, and drug use. Cancer Res 47:2978–2981.
6. Brown LM, Gibson R, Burmeister LF, et al. (1992) Alcohol consumption and risk of leukemia, non-Hodgkin's lymphoma, and multiple myeloma. Leuk Res 16:979–984.[CrossRef][Web of Science][Medline]
7. Brown LM, Pottern LM, Silverman DT, et al. (1997) Multiple myeloma among blacks and whites in the United States: role of cigarettes and alcoholic beverages. Cancer Causes Control 8:610–614.[CrossRef][Web of Science][Medline]
8. Diaz LE, Montero A, Gonzalez-Gross M, et al. (2002) Influence of alcohol consumption on immunological status: a review. Eur J Clin Nutr 56:Suppl 3, S50–S53.
9. van de Wiel A. (2004) Diabetes mellitus and alcohol. Diabetes Metab Res Rev 20:263–267.[CrossRef][Web of Science][Medline]
10. Sierksma A, Patel H, Ouchi N, et al. (2004) Effect of moderate alcohol consumption on adiponectin, tumor necrosis factor-alpha, and insulin sensitivity. Diabetes Care 27:184–189.
11. Wang XD. (1999) Chronic alcohol intake interferes with retinoid metabolism and signaling. Nutr Rev 57:51–59.[Web of Science][Medline]
12. Bi J, Hu X, Zhou FC, et al. (2001) Up-regulation of cellular retinoic acid-binding protein I expression by ethanol. Dev Growth Differ 43:553–561.[CrossRef][Web of Science][Medline]
13. Stagnaro E, Ramazzotti V, Crosignani P, et al. (2001) Smoking and hematolymphopoietic malignancies. Cancer Causes Control 12:325–334.[CrossRef][Web of Science][Medline]
14. Seniori Costantini A, Miligi L, Kriebel D, et al. (2001) A multicenter case-control study in Italy on hematolymphopoietic neoplasms and occupation. Epidemiology 12:78–87.[CrossRef][Web of Science][Medline]
15. Vineis P, Crosignani P, Sacerdote C, et al. (2000) Haematopoietic cancer and medical history: a multicentre case control study. J Epidemiol Community Health 54:431–436.
16. Vineis P, Miligi L, Crosignani P, et al. (2000) Delayed infection, family size and malignant lymphomas. J Epidemiol Community Health 54:907–911.
17. Sancho-Garnier H, Anderson A, Biedermann A, et al. (2004) Evidence-Based Cancer Prevention. A UICC Handbook for Europe. (UICC, Geneva, Switzerland).
18. Stata Corporation. Stata Statistical Software: Release 8.0. College Station, TX 2003.
19. Morton LM, Zheng T, Holford TR, et al. (2005) Alcohol consumption and risk of non-Hodgkin lymphoma: a pooled analysis. Lancet Oncol 6:469–476.[CrossRef][Web of Science][Medline]
20. Fernberg P, Odenbro A, Bellocco R, et al. (2006) Tobacco use, body mass index and the risk of malignant lymphomas—a nationwide cohort study in Sweden. Int J Cancer 118:2298–2302.[CrossRef][Web of Science][Medline]
21. Gallus S, Giordano L, Altieri A, et al. (2004) Cigarette smoking and risk of Hodgkin's disease. Eur J Cancer Prev 13:143–144.[CrossRef][Web of Science][Medline]
22. Glaser SL, Keegan TH, Clarke CA, et al. (2004) Smoking and Hodgkin lymphoma risk in women United States. Cancer Causes Control 15:387–397.[CrossRef][Web of Science][Medline]
23. Briggs NC, Hall HI, Brann EA, et al. (2002) Cigarette smoking and risk of Hodgkin's disease: a population-based case-control study. Am J Epidemiol 156:1011–1020.
24. Italian Statistics Institute (ISTAT) Lifestyle Habits and Health. Annual Multi-Purpose Survey ‘Quotidian Life Characteristics’. (1993) (ISTAT, Rome, Italy).
25. McCann SE, Sempos C, Freudenheim JL, et al. (2003) Alcoholic beverage preference and characteristics of drinkers and non-drinkers in western New York (United States). Nutr Metab Cardiovasc Dis 13:2–11.[CrossRef][Web of Science][Medline]
26. Ruidavets JB, Bataille V, Dallongeville J, et al. (2004) Alcohol intake and diet in France, the prominent role of lifestyle. Eur Heart J 25:1153–1162.
27. Rosell M, De Faire U, Hellenius ML. (2003) Low prevalence of the metabolic syndrome in wine drinkers: is it the alcohol beverage or the lifestyle? Eur J Clin Nutr 57:227–234.[CrossRef][Web of Science][Medline]
28. Chiu BC and Weisenburger DD. (2003) An update of the epidemiology of non-Hodgkin's lymphoma. Clin Lymphoma 4:161–168.[Web of Science][Medline]
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  S Deandrea, P Bertuccio, L Chatenoud, S Franceschi, D Serraino, and C La Vecchia Reply to 'Alcohol consumption and risk of Hodgkin's lymphoma and multiple myeloma: a multicentre case-control study' by Gorini et al. Ann. Onc., June 1, 2007; 18(6): 1119 - 1121. [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||