Connecticut Tumor Registry, Connecticut Department of Public Health, 410 Capital Avenue, Hartford, CT 06134-0308, USA
Correspondence: Tel.: +1860 509 7163; Fax: +1860 509 7161; E-mail: Anthony.polednak{at}po.state.ct.us
(Received 15 February 2005; first review notified 22 February 2005; in revised form 1 March 2005; accepted 2 March 2005; Advance Access publication 29 March 2005)
![]() |
ABSTRACT |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
![]() |
INTRODUCTION |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
According to analyses of data from the US National Alcohol Survey, the prevalence of current alcohol drinking (including weekly heavy drinking) declined considerably from 1984 to 1990, but then increased slightly from 1990 to 1994 (Greenfield et al., 2000). Data from the US Centers for Disease Control and Prevention (CDC) Behavioral Risk Factor Surveillance System (BRFSS) of telephone surveys of adults (age
18 years) indicated that trends in the prevalence of binge drinking from 1991 to 1999 (i.e.
5 drinks on any occasion during the previous month) differed by state but prevalence increased in 19 states (Nelson et al., 2002
). Therefore, surveillance of trends in incidence rates for alcohol-related cancers in recent years should be of interest.
High-quality cancer registries in some states allow the monitoring of trends in cancer incidence rates, rather than relying on mortality data that are less likely to reflect any effects of recent changes in prevalence of risk factors (such as alcohol use patterns). This report examines the recent trends in age-adjusted incidence rates for selected alcohol-related cancers using data from the latest 10-year period of diagnosis (19922001) for high-quality population-based cancer registries. Consistency in the temporal trends in incidence rates across geographic areas, as well as recent trends in alcohol consumption and adult prevalence of binge drinking by geographic area was also examined.
![]() |
SUBJECTS AND METHODS |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
ASIRs 100 000 population/year for 19922001 were calculated by gender for all of the SEER geographic areas combined. For the latest 10-year period (years of diagnosis 19922001), the 12 SEER areas or populations are metropolitan Atlanta GA, Connecticut, metropolitan Detroit MI, Hawaii, Iowa, New Mexico, San Francisco-Oakland CA, Los Angeles County CA, San Jose-Monterey CA, Seattle-Puget Sound WA, Utah, and Alaska Natives (National Cancer Institute, 2004; Ries et al., 2004
). SEER data files for several of these registries started only with cancers diagnosed in 1992.
The three selected alcohol-related cancer site groups were: oral cavity and pharynx (International Classification of Diseases for Oncology, site codes C000148), esophagus (codes C150159) and larynx (C320329). Routinely published SEER statistics on incidence do not include data on histologic types. Squamous cell carcinomas (ICD-O morphology codes 80518084) (Percy et al., 1995) were selected, and have accounted for >95% of almost all cancers in the upper aerodigestive tract in SEER statistics (Muir et al., 1995
). For the esophagus, adenocarcinomas are much less strongly associated with alcohol (and tobacco) than squamous cell carcinomas (Adami et al., 2002
). Certain other cancer sites (liver, breast and colorectal) have been weakly associated with alcohol use (Bagnardi et al., 2001
; Rehm et al., 2003
; Poschl and Seitz, 2004
) but were excluded because temporal trends in ASIRs have been strongly influenced by hepatitis (for liver cancer) and by cancer screening rates (for breast and colorectal).
ASIRs were calculated by the direct method using age-specific incidence rates in five-year age intervals (through age 85+ years) and weights based on the age distribution of the US population in 2000. Version 5.2 of the SEER*Stat software was used to obtain the annual percentage change (APC) in ASIRs for years of diagnosis 1992 through 2001, using weighted least squares; 95% confidence intervals (CIs) were calculated based on both ASIRs and APCs (National Cancer Institute, 2004). ASIRs within younger adults, such as age 2044 years at diagnosis, may be more strongly affected by recent changes in risk-factor prevalence and may predict future cancer trends (Doll, 1991
). For this study, the age group 2054 years was used because of the low incidence rates before the age of 45 years; for ASIRs within age 2054 years, standardization involved the use of weights based on the distribution of the US population within this age group (National Cancer Institute, 2004
).
Geographic variation in ASIRs was examined only in males with the SEER race item (usually based on reports from hospitals) coded as white in the SEER database; populations of other racial groups in many SEER areas were small, and ASIRs for females were low (with statistically unreliable rates). All the selected alcohol-related cancer types were combined, to increase the statistical reliability of the ASIRs in each SEER area. In interpreting geographic variation in these ASIRs, data were tabulated on trends in the estimated per capita (PC) consumption (age 14 years) of alcohol (ethanol in gallons) by state for 1980, 1990 and 2000 (Nephew et al., 2003
). PC consumption is a standard measure for temporal and geographic analyses of alcohol use (Greenfield and Kerr, 2003
). Data for 1980 were included because of the unknown average lag period between changes in consumption and changes in ASIRs. Declines in consumption have been due almost entirely to spirits vs wine and beer (Greenfield et al., 2000
), but total alcohol ingestion (all types of beverages) has been regarded as relevant to cancer risk (Poschl and Seitz, 2004
).
Data on estimated prevalence of heavy drinking (defined as an average of >2 drinks/day for men, and >1 drink/day for women, during the preceding 30 days) were available by state from the 1991 and 2001 BRFSS surveys (Ahluwalia et al., 2003), and for both states and counties or selected metropolitan areas from the BRFSS 2002 survey only (Centers for Disease Control, 2004
). Published data on trends in the prevalence of binge drinking among adults from 1991 to 1999 (Nelson et al., 2002
) were also tabulated. Patterns of drinking, as well as total consumption measures, are important in assessing alcohol-related diseases (Midanik and Room, 2005
). In epidemiologic studies, heavy alcohol use accounts for much of the estimated attributable fraction of cancers related to alcohol (Zeka et al., 2003
), including squamous cell carcinoma of the esophagus (Brown et al., 2001
).
![]() |
RESULTS |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
|
|
![]() |
DISCUSSION |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
Study limitations include use of data from the SEER program, in lieu of unknown ASIRs for the entire country. In addition, delayed reporting of cancers from hospitals to central registries may lead to slight underestimation of incidence rates for recent years of diagnosis, particularly for lung cancer (Ries et al., 2004). This study was limited to the upper aerodigestive system cancers, and was not intended to assess trends in all alcohol-related cancers. In interpreting trends and geographic differences in ASIRs, data on alcohol use were limited. Binge drinking is under surveillance as a leading US health indicator in Healthy People 2010 (Miller et al., 2004
) and is a standard measure of heavy drinking episodes (Midanik and Room, 2005
), but neither binge drinking nor heavy drinking (as defined in BRFSS surveys) may capture the drinking patterns that explain most of the cancers attributable to alcohol.
In interpreting trends in ASIRs in relation to alcohol use, confounding with trends in other cancer risk factors is a potential problem. Fruit and vegetable consumption, while associated with reduced risk of upper aerodigestive cancers (Adami et al., 2002), has changed little in the US in recent decades (Serdula et al., 2004
). For laryngeal cancer, relative risks are higher for tobacco than alcohol (Bosetti et al., 2002
), but the reverse is true for squamous cell carcinoma of the esophagus (Engel et al., 2003
) and declines in ASIRs were greater for esophageal cancers than for oral cavitypharynx and larynx (Table 1). The estimated alcohol-attributable fraction owing to any alcohol consumption is high for esophageal squamous cell carcinoma (i.e. 72% for both sexes combined) in the US (Engel et al., 2003
). Declines in ASIRs for the alcohol-related sites (Table 1) were larger than those reported for lung cancer, which is strongly associated with tobacco but not alcohol (Rehm et al., 2003
) in all SEER areas combined (particularly in females, with an APC of only 0.2%) in 19922001 (Jemal et al., 2004
). This also supports the conclusion that trends in alcohol consumption had an impact on trends in ASIRs for the alcohol-related cancers, independent of the impact of trends in tobacco use.
A lag of several years duration has been described between changes in the per capita total alcohol consumption and changes in the mortality rates of cirrhosis, although trends in the type of alcohol (i.e. spirits) and hepatitis may also be relevant (Mann et al., 2003). Reductions in the level of alcohol consumption would affect liver disease deaths in the short term, while changes in abstention rates would require many decades to affect mortality (Mann et al., 2003
). Lag times for impacts of changes in alcohol use on cancer rates are uncertain but alcohol is believed to act mainly as a promoter and cocarcinogen (through its metabolite, acetaldehyde, interacting with environmental carcinogens) (Seitz et al., 2004
). The effect of quitting alcohol (along with tobacco) showed a 70% reduction in esophageal cancer risk within 59 years (Castellsague et al., 2000
), while longer time periods may be required for risks of laryngeal cancers to decline after stopping drinking (Altieri et al., 2002
). Therefore, any impact of increases in alcohol consumption and prevalence of heavy or binge drinking (Table 2) may not yet be apparent in SEER data, and a continued surveillance of ASIRs for alcohol-related cancers is needed.
While geographic variation in ASIRs for the alcohol-related cancer was limited among white males, the low ASIRs for Utah were consistent with data on apparent alcohol consumption and prevalence of binge or heavy drinking in all adults (Table 2). The high ASIRs for white males in Hawaii are unexplained, but the 95% CI were wide and in 2001, overlapped with those for several other SEER areas. Previous studies have shown that Hawaii had the highest ASIRs among SEER areas for cancers of oral cavitypharynx and larynx but not esophagus (Ries et al., 2004), and also high rates for hepatocellular carcinoma (Davila et al., 2003
).
The consistency across all SEER areas in decline in ASIRs for the alcohol-related cancers combined (for white males) (Table 1) support the conclusion that declines are real, and all states involved in the SEER Program showed declines in apparent alcohol consumption (Table 2). However, BRFSS data (from telephone surveys) may underestimate the prevalence of binge drinking, in comparison to the National Survey on Drug Use and Health (an in-person survey) for the 1999 and 2001 surveys combined (Miller et al., 2004), and the prevalence of binge drinking increased from 1991 to 1999 in the entire US and in some states with a SEER registry (Table 2) (Nelson et al., 2002
). The prevalence of heavy drinking also increased from 1991 to 2001 in the US and in almost all states with a SEER registry (Ahluwalia et al., 2003
), while apparent total alcohol consumption (Nephew et al., 2003
) continued to decline (albeit very little in some states) (Table 2). These data support the need for continued surveillance of ASIRs for alcohol-related cancers, to determine if plateaux or increases occur in any geographical areas. Surveillance of ASIRs in the US will be enhanced as the number of high-quality cancer registries continues to grow (Wingo et al., 2003
).
Reducing the prevalence of heavy and binge drinking in the US population would be expected to lead to further reductions in incidence rates for alcohol-related cancers, particularly for squamous cell carcinoma of the esophagus (Brown et al., 2001), and the apparent cardiovascular benefits of moderate alcohol consumption are reversed with heavy or binge drinking (Buemann et al., 2002
).
![]() |
ACKNOWLEDGEMENTS |
---|
![]() |
REFERENCES |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
Ahluwalia, I. B., Mack, K. A., Murphy, W. et al. (2003) State-specific prevalence of selected chronic-disease-related characteristicsBehavioral Risk Factor Surveillance System, 2001. Morbidity and Mortality Weekly Reports Surveillance Summaries 52, 180.
Altieri, A., Bosetti, C., Talamini, R. et al. (2002) Cessation of smoking and drinking and the risk of laryngeal cancer. British Journal of Cancer 87, 12271229.[CrossRef][ISI][Medline]
Bagnardi, V., Blangiardo, M., La Vecchia, A. et al. (2001). A meta-analysis of drinking and cancer risk. British Journal of Cancer 85, 17001705.[ISI][Medline]
Bosetti, C., Gallus, S., Franceschi, S. et al. (2002) Cancer of the larynx in non-smoking alcohol drinkers and in non-drinking tobacco smokers. British Journal of Cancer 87, 516518.[CrossRef][ISI][Medline]
Brown, L. M., Hoover, R., Silverman, D. et al. (2001) Excess incidence of squamous cell esophageal cancer among US black men: role of social class and other risk factors. American Journal of Epidemiology 153, 114122.
Buemann, B., Dyerberg, J. and Astrup, A. (2002) Alcohol drinking and cardiac risk. Nutrition Research Reviews 15, 91121.[ISI]
Castellsague, X., Munoz, N., De Stefani, E. et al. (2000) Smoking and drinking cessation and risk of esophageal cancer (Spain). Cancer Causes and Control 11, 813818.[CrossRef][ISI][Medline]
Centers for Disease Control and Prevention (2004) Surveillance for certain health behaviors among selected local areasUnited States, Behavioral Risk Factor Surveillance System, 2002. Morbidity and Mortality Weekly Reports 53 (No.SS-5).
Corrao, G., Bagnardi, V., Zambon, A. et al. (2004) A meta-analysis of alcohol consumption and the risk of 15 diseases. Preventive Medicine 38, 613619.[CrossRef][ISI][Medline]
Davila, J. A., Petersena, N. J., Nelson, H. A. et al. (2003) Geographic variation within the United States in the incidence of hepatocellular carcinoma. Journal of Clinical Epidemiology 56, 487491.[CrossRef][ISI][Medline]
Doll, R. (1991) Progress against cancer: an epidemiologic assessment. American Journal of Epidemiology 134, 675688.[ISI][Medline]
Engel, L. S., Chow, W. H., Vaughan, T. L. et al. (2003) Population attributable risks for esophageal and gastric cancers. Journal of the National Cancer Institute 95, 14041413.
Greenfield, T. K., Midanik, L. T. and Rogers, J. D. (2000) A 10-year national trend study of alcohol consumption, 19841995: is the period of declining drinking over? American Journal of Public Health 90, 4752.
Greenfield, T. K. and Kerr, W. C. (2003) Tracking alcohol consumption over time. Alcohol Research and Health 27, 3038.[ISI][Medline]
Hankey, B. F., Ries, L. A. and Edwards, B. K. (1999) The Surveillance, Epidemiology, and End Results program: a national resource. Cancer Epidemiology Biomarkers and Prevention 8, 11171121.
Jemal, A., Clegg, L. X., Ward, E. et al. (2004) Annual report to the nation on the status of cancer, 19752001, with a special feature regarding survival. Journal of the National Cancer Institute 101, 327.
Mann, R. E., Smart, R. G. and Govini, R. (2003) The epidemiology of alcohol liver disease. Alcohol Research and Health 27, 209219.[ISI][Medline]
Merrill, R. M. and Dearden, K. A. (2004) How representative are the Surveillance, Epidemiology, and End Results (SEER) Program cancer data of the United States? Cancer Causes and Control 15, 10271034.[CrossRef][ISI][Medline]
Midanik, L. and Room, R. (2005) Contributions of social science to the alcohol field in an era of biomedicalization. Social Science and Medicine 60, 11071116.[CrossRef][ISI][Medline]
Miller, J. W., Gfroerer, J. C., Brewer, R. D. et al. (2004) Prevalence of adult binge drinking: a comparison of two national Surveys. American Journal of Preventive Medicine 27, 197204.[CrossRef][ISI][Medline]
Muir, C. and Weiland, L. (1995) Upper aerodigestive tract cancers. Cancer 75, 147153.[ISI][Medline]
National Cancer Institute (2004) Surveillance, Epidemiology, and End Results (SEER) Program Public-Use CD-ROM (19732001), National Cancer Institute, DCPC, Surveillance Program, Cancer Statistics Branch, released April 2004.
Nelson, D. E., Bland, S., Powell-Griner, E. et al. (2002) State trends in health risk factors and receipt of clinical preventive services among US adults during the 1990s. Journal of the American Medical Association 287, 26592667.
Nephew, T. M., Williams, G. D., Yi, H. et al. (2003) Apparent per capita alcohol consumption: national, state, and regional trends, 19772000. Surveillance Report #62. National Institute on Alcohol Abuse and Alcoholism. Bethesda, MD, August 2003. Available at www.niaaa.nih.gov/publications/surveillance62/CONS00.htm
Percy, C., Young, J. L., Muir, C. et al. (1995) Introduction. Cancer 75, 140146.[ISI][Medline]
Poschl, G. and Seitz, H. K. (2004) Alcohol and cancer. Alcohol and Alcoholism 39, 155165.
Rehm, J., Room, R., Graham, K. et al. (2003) The relationship of average volume of alcohol consumption and patterns of drinking to burden of disease: an overview. Addiction 98, 12091228.[CrossRef][ISI][Medline]
Ries, L. A. G., Eisner, M. P., Kosary, C. L. et al. (eds) (2004) SEER Cancer Statistics Review, 19752001. National Cancer Institute. Bethesda MD. Available at http://seer.cancer.gov/csr/1975_2001
Seitz, H. K., Stickel, F. and Homann, N. (2004) Pathogenic mechanisms of upper aerodigestive tract cancer in alcoholics. International Journal of Cancer 108, 483487.[CrossRef][ISI]
Serdula, M. K., Gillespie, C., Kettel-Khan, L. et al. (2004) Trends in fruit and vegetable consumption among adults in the United States: behavioral risk factor surveillance system, 19942000. American Journal of Public Health 94, 10141018.
Wingo, P. A., Jamison, P. M., Hiatt, R. A. et al. (2003) Building the infrastructure for nationwide cancer surveillance and control. Cancer Causes and Control 14, 175194.[CrossRef][ISI][Medline]
Zeka, A., Gore, R. and Kriebel, D. (2003) Effects of alcohol and tobacco on aerodigestive cancer risks: a meta-regression analysis. Cancer Causes and Control 14, 897906.[CrossRef][ISI][Medline]
|