CONSUMPTION OF ALCOHOLIC BEVERAGES - IARC Monographs on the

128 Pages·2012·2.59 MB·English

CONSUMPTION OF ALCOHOLIC BEVERAGES
Consumption of alcoholic beverages was considered by previous IARC Working Groups in
1987 and 2007 (IARC, 1988, 2010). Since that time, new data have become available, these
have been incorporated into the Monograph, and taken into consideration in the present
evaluation.
1. Exposure Data fruit, vegetables or parts of palm trees, by a fairly
simple production process.
1.1 Types and ethanol content of
1.1.2 Ethanol content of alcoholic beverages
alcoholic beverages
Percentage by volume (% vol) is used to indi-
1.1.1 Types of alcoholic beverages
cate the ethanol content of beverages, which is
also called the French or Gay-Lussac system.
The predominant types of commercially
Alcohol content differs according to the main
produced alcoholic beverages are beer, wine and
beverage type and may also vary by country.
spirits. Basic ingredients for beer are malted
Commonly, 4–5% vol are contained in beer,
barley, water, hops and yeast. Wheat may be
about 12% vol in wine and about 40% vol in
used. Nearly all wine is produced from grapes,
distilled spirits. However, lower or higher ethanol
although wine can be also made from other fruits
content in alcoholic beverages is also possible.
and berries. Spirits are frequently produced from
The ethanol content in beer can range from
cereals (e.g. corn, wheat), beet or molasses, grapes
2.3% vol to over 10% vol (lower alcohol content
or other fruits, cane sugar or potatoes. Main
in home- or locally produced alcoholic bever-
beverage types (i.e. beer, wine and spirits) may
ages such as sorghum beer), in wine from 8 to
be consumed in combination with each other to
15% vol, and in spirits from 20% vol (aperitifs) to
fortify the strength of an alcoholic beverage (e.g.
well over 40% vol (e.g. 80% vol in some kinds of
fortified wine, in which spirits are added to wine)
absinthe). There is a trend in recent years towards
(WHO, 2004).
higher (13.5–14.5%) alcohol volume in consumed
In addition to commercialized products, in
wines, associated with technology advances and
many developing countries different types of
increasing proportion in overall consumption
home- or locally produced alcoholic beverages
of wines produced outside the traditional wine-
such as sorghum beer, palm wine or sugarcane
growing regions of Europe (IARC, 2010).
spirits are consumed (WHO, 2004). Home-
To calculate the amount of ethanol contained
or locally produced alcoholic beverages are
in a specific drink, the amount (e.g. ml) of
produced through fermentation of seed, grains,
373
IARC MONOGRAPHS – 100E
alcoholic beverage consumed for each type of sugars, di- and tribasic carboxylic acids, colouring
beverage (e.g. a 330-mL bottle of beer) is multi- substances, tannic and polyphenolic substances
plied by the precentage of alcohol by volume, i.e. and inorganic salts (IARC, 2010).
the proportion of the total volume of the beverage Occasionally, toxic additives, that are not
that is alcohol (e.g. (330 mL) × (0.04) = 13.2 mL permitted for use in commercial production
of ethanol in a bottle of beer). Conversion factors have been identified in alcoholic beverages. These
may be used to convert the volume of alcoholic include methanol, diethylene glycol (used as
beverage into grams of ethanol, or volumes of sweetener) and chloroacetic acid or its bromine
alcohol may be recorded in ‘ounces’. Conversion analogue, sodium azide and salicylic acid, which
factors for these different measures (WHO, 2000) are used as fungicides or bactericides (Ough,
are as follows: 1987).
• 1 mL ethanol = 0.79 g Contaminants may also be present in alco-
• 1 UK oz = 2.84 cL = 28.4 mL = 22.3 g holic beverages. Contaminants are defined as
• 1 US fluid oz = 2.96 cL = 29.6 mL = 23.2 g substances that are not intentionally added but
are present in alcoholic beverages due to produc-
tion, manufacture, processing, preparation,
1.2 Chemical composition
treatment, packing, packaging, transport or
The main components of most alcoholic holding, or as a result of environmental contami-
beverages are ethanol and water. Some phys- nation. Contaminants and toxins found in alco-
ical and chemical characteristics of anhydrous holic beverages are nitrosamines, mycotoxins,
ethanol are as follows (O’Neil, 2001): ethyl carbamate, pesticides, thermal processing
Chem. Abstr. Services Reg. No.: 64–17.5 contaminants, benzene, and inorganic contami-
Formula: C H OH nants such as lead, cadmium, arsenic, copper,
2 5
Relative molecular mass: 46.07 chromium, inorganic anions, and organometals
Synonyms and trade name: Ethanol, ethyl (IARC, 2010).
alcohol, ethyl hydroxide, ethyl hydrate, In view of the potential carcinogenicity of
absolute alcohol, anhydrous alcohol, dehy- acetaldehyde and its known toxic properties,
drated alcohol recent studies attempted to estimate exposure to
Description: Clear, colourless, very mobile, acetaldehyde from alcoholic beverages outside
flammable liquid, pleasant odour, burning ethanol metabolism at known levels of alcohol
taste exposure. The average exposure to acetaldehyde
Melting-point: –114.1 °C as a result of consumption of alcoholic beverages,
Boiling-point: 78.5 °C including “unrecorded alcohol,” was estimated
Density: d 20 0.789 at 0.112 mg/kg body weight/day (Lachenmeier
4
Refractive index: n 20 1.361 et al., 2009a). Levels of acetaldehyde in alcoholic
D
In addition to ethanol and water, wine, beverages vary from less than 1 g/hl of pure
beer and spirits contain volatile and non-vola- alcohol up to 600 g/hl, and high concentra-
tile compounds. Volatile compounds include tions of acetaldehyde were documented in alco-
aliphatic carbonyl compounds, alcohols, mono- holic beverages commonly consumed in many
carboxylic acids and their esters, nitrogen- and parts of the world, including distilled beverages
sulfur-containing compounds, hydrocarbons, from Brazil, the People’s Republic of China,
terpenic compounds, and heterocyclic and Guatemala, Mexico, and the Russian Federation,
aromatic compounds. Non-volatile extracts as well as calvados and fortified wines and fruit
of alcoholic beverages comprise unfermented and marc spirits from Europe (Lachenmeier &
374
Consumption of alcoholic beverages
Sohnius, 2008; Linderborg et al., 2008; Kanteres vol100E/100E-06-Table1.2.pdf), the countries
et al., 2009; Lachenmeier et al., 2009b). with the highest overall consumption of alcohol
per capita among the adult (15+ years) population
can be found in the the WHO Regional Office
1.3 Trends in consumption of
for European Region (12.2 L of pure alcohol per
alcoholic beverages
capita), and more specifically in eastern Europe.
The next highest alcohol consumption is in the
Volume, pattern and quality of consumed
WHO Region of the Americas (8.7 L of pure
alcohol are included in the description of differ-
alcohol per capita). Apart from some countries
ential exposure to alcohol.
in Africa and a few countries in other parts of
In a development of the WHO Global Alcohol
the world, alcohol consumption in the other
Database, WHO has developed the Global
regions is generally lower. The WHO Eastern
Information System on Alcohol and Health
Mediterranean Region ranks lowest with 0.7
(WHO, 2008). In 2008–09, WHO conducted the
litres of alcohol consumed per adult. Total adult
Global Survey on Alcohol and Health, collecting
(15+ years) per capita consumption in litres of
data on alcohol consumption, alcohol-related
pure alcohol by region and country is an indi-
harm and policy responses from its Member
cator of the alcohol consumption level of the
States.
adult population, irrespective of the number of
Total adult per capita consumption in litres
abstainers (i.e. people who do not drink alcohol)
of pure alcohol is defined as the total amount
in the country.
of alcohol consumed per person, taking into
Globally, men consume more alcohol than
account recorded consumption (i.e. alcoholic
women. This is reflected in the differences in
beverages consumed that are recorded in offi-
the number of lifetime abstainers, past year
cial statistics of production, trade or sales) and
abstainers and former drinkers (Table 1.2 on-line).
unrecorded consumption (i.e. alcoholic bever-
Lifetime abstainers are defined as the proportion
ages consumed that are not recorded in official
of people (15+ years) in a given population who
statistics and that can come from a variety of
have not consumed any alcohol during their life-
sources such as home- or informally produced
time, assessed at a given point in time. Past year
alcohol, illegal production and sale, smuggling
abstainers are people aged 15+ years who did
and cross-border shopping), and subtracting
not consume any alcohol during the past year. A
consumption by tourists, if possible. Recorded
former drinker is a person who did not consume
adult per capita consumption is calculated from
any alcohol during the past year. Generally, the
production, export and import data, or sales data.
percentage of lifetime and past year abstainers is
Unrecorded consumption is computed from
higher in women than in men. The prevalence
representative surveys, specific empirical inves-
of lifetime, past-year abstainers, and former
tigations or expert opinion. The percentage of
drinkers are calculated from large representative
lifetime and past-year abstainers provide impor-
surveys.
tant information about drinking in a population
Table 1.2 (on-line) provides information
and complement the indicator on total adult (15+
about the trend (i.e. robust estimate of five-year
years) per capita consumption.
change) in per capita consumption (2001–05) of
Overall, there is a wide variation in the
recorded alcohol, that is, indicates if consump-
volume of alcohol consumed across countries.
tion remained stable, increased, decreased, or if
As presented in Table 1.1 and Table 1.2 (available
no conclusion could be drawn. To estimate five-
at http://monographs.iarc.fr/ENG/Monographs/
year change in recorded adult (15+ years) per
375
IARC MONOGRAPHS – 100E
Table 1.1 Estimate for total adult (15+ years) per capita consumption, by WHO region, average
2003–05
WHO Region Adult (15+) per capita consumption*, total (recorded and unrecorded),
average 2003–05
African Region (AFRO) 6.2
Region of the Americas (AMRO) 8.7
Eastern Mediterranean Region (EMRO) 0.7
European Region (EURO) 12.2
South-East Asian Region (SEARO) 2.2
Western Pacific Region (WPRO) 6.3
* in litres of pure alcohol
Source: WHO (2008)
AFRO: Algeria, Angola, Benin, Botswana, Burkina Faso, Burundi, Cameroon, Cape Verde, Central African Republic, Chad, Comoros, Congo,
Côte d’Ivoire, Democratic Republic of the Congo, Equatorial Guinea, Eritrea, Ethiopia, Gabon, Gambia, Ghana, Guinea, Guinea-Bissau, Kenya,
Lesotho, Liberia, Madagascar, Malawi, Mali, Mauritania, Mauritius, Mozambique, Namibia, Niger, Nigeria, Rwanda, Sao Tome and Principe,
Senegal, Seychelles, Sierra Leone, South Africa, Swaziland, Togo, Uganda, United Republic of Tanzania, Zambia, Zimbabwe
AMRO: Antigua and Barbuda, Argentina, Bahamas, Barbados, Belize, Bolivia, Brazil, Canada, Chile, Colombia, Costa Rica, Cuba, Dominica,
Dominican Republic, Ecuador, El Salvador, Grenada, Guatemala, Guyana, Haiti, Honduras, Jamaica, Mexico, Nicaragua, Panama, Paraguay,
Peru, Saint Kitts and Nevis, Saint Lucia, Saint Vincent and the Grenadines, Suriname, Trinidad and Tobago, United States of America, Uruguay,
Venezuela (Bolivarian Republic of)
EMRO: Afghanistan, Bahrain, Djibouti, Egypt, Islamic Republic of Iran, Iraq, Jordan, Kuwait, Lebanon, Libyan Arab Jamahiriya, Morocco,
Oman, Pakistan, Qatar, Saudi Arabia, Somalia, Sudan, Syrian Arab Republic, Tunisia, United Arab Emirates, Yemen
EURO: Albania, Andorra, Armenia, Austria, Azerbaijan, Belarus, Belgium, Bosnia and Herzegovina, Bulgaria, Croatia, Cyprus, Czech Republic,
Denmark, Estonia, Finland, France, Georgia, Germany, Greece, Hungary, Iceland, Ireland, Israel, Italy, Kazakhstan, Kyrgyzstan, Latvia,
Lithuania, Luxembourg, Malta, Monaco, Montenegro, Netherlands, Norway, Poland, Portugal, Republic of the Republic of Moldova, Romania,
Russian Federation, San Marino, Serbia, Slovakia, Slovenia, Spain, Sweden, Switzerland, Tajikistan, The former Yugoslav Republic of Macedonia,
Turkey, Turkmenistan, Ukraine, United Kingdom, Uzbekistan
SEARO: Bangladesh, Bhutan, Democratic People’s Republic of Korea, India, Indonesia, Maldives, Myanmar, Nepal, Sri Lanka, Thailand, Timor-
Leste
WPRO: Australia, Brunei Darussalam, Cambodia, China, Cook Islands, Fiji, Japan, Kiribati, Lao People’s Democratic Republic, Malaysia,
Marshall Islands, Federated States of Micronesia, Mongolia, Nauru, New Zealand, Niue, Palau, Papua New Guinea, Philippines, Republic of
Korea, Samoa, Singapore, Solomon Islands, Tonga, Tuvalu, Vanuatu, Viet Nam
capita consumption, three-year moving averages Asia and Western Pacific Regions, which probably
were calculated for per capita consumption of reflects economic development and increases in
recorded alcohol for each year in the five-year consumers’ purchasing power as well as increases
period from 2001 to 2005. in the marketing of branded alcoholic beverages
Recent data on trends in consumption of (WHO, 2007).
alcoholic beverages indicate that the European
Region and the Region of the Americas main-
tain a steady high consumption of alcoholic 2. Cancer in Humans
beverages, with 20% of all countries showing
an increase in consumption. Alcohol consump- 2.1 Description of cohort studies
tion remains low in the Eastern Mediterranean
Region. In the African Region, out of 50 coun- 2.1.1 Studies in the general population
tries, 20% show a decrease and 20% an increase
Cohort studies are classified by the country in
in consumption. There is a recent and continuing
which the study was conducted (Table 2.1 available
increase in alcohol consumption in several low
at http://monographs.iarc.fr/ENG/Monographs/
and middle-income countries in the South-East
vol100E/100E-06-Table2.1.pdf). The majority of
376
Consumption of alcoholic beverages
cohort studies have been conducted in the USA, availability of national registries of popula-
western Europe and Japan. Since the previous tions, inpatients and cancer, these studies were
IARC Monograph (IARC, 2010), data on the largely performed in Scandinavian countries.
association between alcohol consumption and The estimation of risk in these individuals is not
risk of cancer have been published from several based upon a comparison of exposed and unex-
cohorts, including updates of cohorts described posed subjects within the cohort, but with the
previously (Bongaerts et al., 2006, 2008; Li et al., expected rates of cancer in the general popula-
2006, 2009b; Weinstein et al., 2006; Ericson tion. Thygesen et al. (2009) is the only report
et al., 2007; Ferrari et al., 2007; Ishihara et al., from cohorts of special populations that has been
2007; Ozasa, 2007; Sutcliffe et al., 2007; Thygesen published since IARC (2010).
et al., 2007, 2008a, b; Fan et al., 2008; Ide et al.,
2008; Kabat et al., 2008; Nielsen & Grønbaek,
2.2. Cancers of the upper
2008; Rohrmann et al., 2008, 2009; Shimazu
aerodigestive tract
et al., 2008; Friberg & Wolk, 2009; Ishiguro
et al., 2009; Heinen et al., 2009; Klatsky et al.,
2.2.1. Cancer of the oral cavity and pharynx
2009; Rod et al., 2009; Thun et al., 2009; Weikert
et al., 2009) and reports from recently established It was concluded in the previous IARC
cohorts and some older cohorts from which data Monograph (IARC, 2010) that consumption of
on alcohol consumption and risk of cancer were alcoholic beverages is causally related to cancer
not available (Nakaya et al., 2005; Velicer et al., of the oral cavity and pharynx, and that the risk
2006; Akhter et al., 2007; Chlebowski et al., increases in a dose-dependent manner.
2007; Freedman et al., 2007a, b; Friborg et al.,
(a) Overview of cohort and case–control studies
2007; Gwack et al., 2007; Khurana et al., 2007;
Lim et al., 2007; Mørch et al., 2007; Sung et al., The association of consumption of alcoholic
2007; Tsong et al., 2007; Visvanathan et al., 2007; beverages and risk of cancer of the oral cavity
Zhang et al., 2007a; Ansems et al., 2008; Brinton and/or pharynx has been assessed in six cohort
et al., 2008, 2009; Chao et al., 2008; Lim & Park, studies (Freedman et al., 2007a; Friborg et al.,
2008; Muwonge et al., 2008; Ohishi et al., 2008; 2007; Ide et al., 2008; Muwonge et al., 2008;
Toriola et al., 2008, 2009; Allen et al., 2009; Duffy Allen et al., 2009; Weikert et al., 2009; Table 2.3
et al., 2009; Engeset et al., 2009; Gibson et al., available at http://monographs.iarc.fr/ENG/
2009; Gong et al., 2009; Jiao et al., 2009; Johansen Monographs/vol100E/100E-06-Table2.3.pdf).
et al., 2009; Lew et al., 2009; Setiawan et al., 2009; Significant increases in risk were found with
Chao et al., 2010). increasing amount of alcohol consumption in
all studies (Freedman et al., 2007a; Friborg et al.,
2.1.2 Studies in special populations 2007; Ide et al., 2008; Allen et al., 2009; Weikert
et al., 2009), increasing frequency of consumption
This group of studies is characterized by the
(Friborg et al., 2007; Muwonge et al., 2008), and
assumption that the study subjects have a pattern
duration of consumption (Muwonge et al., 2008).
of consumption of alcoholic beverages that is
In one case–control study conducted in Taiwan,
different from that of the general population,
China among patients attending a hospital clinic
e.g. alcoholics, brewery workers, members of a
(Yen et al., 2008) no association was found
temperance organization (Table 2.2 available at
among non-smokers and a positive association
http://monographs.iarc.fr/ENG/Monographs/
among drinkers and smokers (Table 2.4 available
vol100E/100E-06-Table2.2.pdf). Because of the
at http://monographs.iarc.fr/ENG/Monographs/
377
IARC MONOGRAPHS – 100E
vol100E/100E-06-Table2.4.pdf). [No information There is increasing evidence from recent cohort
on the quantity of alcohol or tobacco consump- studies that risk may already be increased at
tion was available.] In a case–control study more moderate intake, particularly in women
in Uruguay (De Stefani et al., 2009) a signifi- (Freedman et al., 2007a; Allen et al., 2009;
cant positive association with a predominantly Weikert et al., 2009).
alcohol-based dietary pattern and cancer of the A pooled analysis of the International Head
oral cavity and pharynx was found (data not and Neck Cancer Epidemiology Consortium,
shown). [No specific assessment of alcohol intake which specifically examined the association of
was presented and the contribution of other foods alcohol consumption and duration, found that
to this dietary pattern was not known.] among drinkers of 10 drinks per day or less,
Undifferentiated nasopharyngeal carcinoma the association with total drink-years increased
(NPC), which is common in parts of Southern with increasing drinks/day, indicating that more
Asia, North Africa and the Arctic, is associated drinks/day for a shorter duration was more dele-
with Epstein-Barr virus and preserved foods terious than fewer drinks/day for a longer dura-
(see the Monograph on Chinese-style Salted Fish tion (Lubin et al., 2009).
in this volume). Friborg et al. (2007) confirmed
(c) Effect of cessation
earlier results suggesting limited or no associa-
tion between alcohol and undifferentiated NPC A meta-analysis of 13 case–control studies of
(Yu et al., 2002). However, in a Western popu- cancer of the oral cavity and pharynx combined
lation where differentiated forms of NPC are found that compared with current drinkers, risk
more common, a significantly increased risk of did not decrease until 10 years or more after
NPC has been associated with heavy drinking cessation of drinking (odds ratio (OR), 0.67;
(> 21 drinks per week) (Vaughan et al., 1996), 95%CI: 0.63–0.73) (Rehm et al., 2007) (Table 2.6
indicating a difference in ethiology between available at http://monographs.iarc.fr/ENG/
differentiated and undifferentiated types of NPC Monographs/vol100E/100E-06-Table2.6.pdf).
(Table 2.4 on-line). Consistent with many earlier studies, risks
Thygesen et al. (2009) reported a significantly were found to be elevated among recent former
higher rate of cancer of the oral cavity and pharynx drinkers, most likely due to ill health directly
among Danish alcohol abusers compared with related to the cancer or its precursors.
national rates. [This cohort study provided no
(d) Types of alcoholic beverage
information on individual exposures or results
adjusted for potential confounders.] See Table 2.5 Some studies have assessed whether the
available at http://monographs.iarc.fr/ENG/ association of consumption of alcoholic bever-
Monographs/vol100E/100E-06-Table2.5.pdf. ages on risk varies by beverage type, and have
found broadly similar associations in wine, beer
(b) Intensity and duration
and spirit drinkers (Freedman et al., 2007a; Allen
Previous studies consistently showed that et al., 2009).
consumption of alcoholic beverages is associ-
(e) Population characteristics
ated with an increased risk of cancer of the oral
cavity and/or pharynx, although the nature The association of consumption of alcoholic
of the dose–response relationship is not fully beverages with risk of cancer of the oral cavity
understood (IARC, 2010). In most studies an and pharynx is increased in both men and women
approximate threefold increased risk was found (Freedman et al., 2007a; Weikert et al., 2009).
at relatively high levels of intake (i.e. > 60 g/day).
378
Consumption of alcoholic beverages
Studies have been hampered with low numbers and pharynx is described in Section 2.20 of the
of women at the highest levels of exposure. Monograph on Tobacco Smoking in this volume.
(f) Histological subtype
2.2.2 Cancer of the larynx
Very few studies have examined the asso-
It was concluded in the previous IARC
ciation of consumption of alcoholic beverages by
Monograph (IARC, 2010) that consumption of
histological subtype for cancers of the oral cavity
alcoholic beverages is causally related to cancer
and pharynx. From a large-scale cohort study,
of the larynx, and that the risk increases in a
Weikert et al. (2009) reported that both baseline
dose-dependent manner.
and lifetime alcohol intake were associated with
an increased risk of squamous cell carcinoma (a) Overview of cohort and case–control studies
of the oral cavity and the pharynx, with an
Since IARC (2010), the association of
increased risk of 10% (95%CI: 8–13%) per 10 g
consumption of alcoholic beverages and risk of
per day increase in lifetime alcohol intake.
cancer of the larynx has been assessed in three
(g) Association among non-smokers general-population cohort studies (Freedman
et al., 2007a; Allen et al., 2009, Weikert et al.,
There is evidence from a pooled analysis
2009; Table 2.7 available at http://monographs.
of 15 case–control studies that increasing
iarc.fr/ENG/Monographs/vol100E/100E-06-
consumption of alcoholic beverages increases
Table2.7.pdf) and one case–control study
risk of cancer of the oral cavity and oropharynx/
(Garavello et al., 2006; Table  2.8 available at
hypopharynx cancer among never smokers, with
http://monographs.iarc.fr/ENG/Monographs/
risk estimates of 1.23 (95%CI: 0.59–2.57) and
vol100E/100E-06-Table2.8.pdf), all of which
5.50 (95%CI: 2.26–13.36), for 5 or more drinks/
found significant increases in risk associated
day versus never drinkers, for the two cancer
with alcohol consumption. In one further case–
sites, respectively (Hashibe et al., 2007; Table 2.6
control study in Uruguay (De Stefani et al., 2009)
on-line). One study in Taiwan, China found no
a significant positive association with a predomi-
association with alcohol consumption among
nantly alcohol-based dietary pattern and cancer
non-smoking ever-drinkers for cancer of the oral
of the larynx was found (data not shown).
cavity (Yen et al., 2008). [No information was
[No specific assessment of alcohol intake was
provided on the quantity of alcohol consumed.]
presented and the contribution of other foods to
(h) Joint effect of alcoholic beverages and this dietary pattern was not known.]
tobacco smoking Thygesen et al. (2009) reported a signifi-
cantly higher rate of cancer of the larynx among
It is well established that there is a joint effect
Danish alcohol abusers compared with national
of tobacco smoking and consumption of alcoholic
rates (Table 2.5 on-line). This study provided no
beverages on the risk of cancer of the oral cavity
information on individual exposures or results
and pharynx, with very high risks observed in
adjusted for potential confounders
individuals who are both heavy drinkers and
heavy smokers, corresponding to a greater than (b) Intensity and duration
a multiplicative interaction (IARC, 2010). The
Previous studies consistently showed that
joint effect of alcohol consumption and tobacco
increasing alcohol consumption is associated
smoking on the risk of cancers of the oral cavity
with an increased risk of cancer of the larynx
(IARC, 2010). Bagnardi et al. (2001) reported risk
379
IARC MONOGRAPHS – 100E
estimates of 1.38 (95%CI: 1.32–1.45) for intake of current drinkers; the risk for never-drinkers was
25 g alcohol per day, 1.94 (95%CI: 1.78–2.11) for 0.56 (95%CI: 0.31–0.99).
50 g per day, and 3.95 (95%CI: 3.43–4.57) for 100
(d) Types of alcoholic beverage
g per day from a meta-analysis of 20 case–control
studies, including over 3500 cases. An increased Evidence suggests that the most frequently
risk for cancer of the larynx was found for consumed beverage in a population tends to be
women drinking above one drink/day in a large associated with the highest risk of cancer of the
cohort of United Kingdom women [equivalent to larynx. Data published recently largely supports
an increased risk of 1.44 (95%CI: 1.10–1.88) per this view (Garavello et al., 2006). The NIH-AARP
10 g alcohol per day] (Allen et al., 2009). This is Diet and Health Study found a stronger associa-
consistent with the 1.38 (95%CI: 1.10–1.73) esti- tion for spirits than for beer or wine consump-
mate per 10 g per day reported among women tion among men (Freedman et al., 2007a). [The
in the European Prospective Investigation into Working Group noted the small number of
Cancer and Nutrition (Weikert et al., 2009). cases.]
Compatible with this, Freedman et al. (2007a)
(e) Risk among non-smokers
reported a risk estimate of 2.15 (95%CI: 0.82–
5.65) among women associated with 3 or more There is evidence from a pooled analysis of 11
drinks/day from the NIH-AARP Diet and Health case–control studies, based on 121 cases of laryn-
Study. Among men, the dose–response relation- geal cancer, that increasing alcohol consump-
ship is slightly weaker (Freedman et al., 2007a; tion increases the risk for cancer of the larynx
Weikert et al., 2009), although it is difficult to among never smokers, with a risk estimate of
determine whether these differences are due to 2.98 (95%CI: 1.72–5.17) for 5 or more drinks/
chance because of the relatively low number of day versus never drinkers (Hashibe et al., 2007;
cases in women. In a large case–control study in Table 2.6 on-line).
Italy there was clear evidence of a dose–response
(f) Joint effect of alcoholic beverages and
relationship for men and women combined
tobacco smoking
(Garavello et al., 2006).
A pooled analysis of the International Head Evidence suggests that there exists a joint
and Neck Cancer Epidemiology Consortium, effect of tobacco smoking and consumption of
which specifically examined the association of alcoholic beverages on the risk of cancer of the
quantity and duration of alcohol consumption, larynx, with very high risks observed in indi-
found that among drinkers of 10 drinks per day viduals who are both heavy drinkers and heavy
or less, the association with total drink-years smokers. More recent studies that have exam-
increased with increasing drinks/day, indicating ined the joint effect of alcohol consumption
that more drinks/day for a shorter duration was and tobacco smoking on the risk of cancer of
more deleterious than fewer drinks/day for a the larynx are described in Section 2.20 of the
longer duration (Lubin et al., 2009). Monograph on Tobacco Smoking in this volume.
(c) Effect of cessation
2.2.3 Cancer of the oesophagus
Few studies have assessed whether the risk
It was concluded in the previous IARC
for cancer of the larynx declines since stopping
Monograph (IARC, 2010) that consumption of
drinking. Altieri et al. (2002) reported a risk
alcoholic beverages is causally associated with
estimate of 0.53 (95%CI: 0.15–1.94) for stopping
cancer of the oesophagus. The increased risk is
drinking for 20 years or more ago compared with
380
Consumption of alcoholic beverages
largely restricted to squamous cell carcinoma, of alcohol intake have been found by analysis
with little or no association for adenocarcinoma of large cohorts in Europe, with a significant
of the oesophagus. increased risk of approximately 20% per 10 g
alcohol per day (Allen et al., 2009; Weikert et al.,
(a) Overview of cohort and case–control studies
2009).
Since IARC (2010), the association of consump- In several studies an increased risk has been
tion of alcoholic beverages and risk of cancer of found with duration of drinking (Lee et al., 2005;
the oesophagus has been assessed in six cohort Ozasa et al., 2007; Fan et al., 2008, Vioque et al.,
studies (Freedman et al., 2007b; Ozasa et al., 2007; 2008), frequency of drinking (Ozasa et al., 2007),
Fan et al., 2008; Allen et al., 2009; Ishiguro et al., a lower age at starting drinking (Ozasa et al.,
2009; Weikert et al., 2009; Table 2.9 available at 2007; Fan et al., 2008), or cumulative intake
http://monographs.iarc.fr/ENG/Monographs/ (Benedetti et al., 2009). Risk is similar when
vol100E/100E-06-Table2.9 pdf) and four case– alcohol consumption is based on measures of
control studies (Lee et al., 2005; Vioque et al., either baseline or lifetime alcohol consumption
2008; Benedetti et al., 2009; Pandeya et al., 2009; (Fan et al., 2008; Weikert et al., 2009).
Table 2.10 available at http://monographs.iarc.fr/
(c) Effect of cessation
ENG/Monographs/vol100E/100E-06-Table2.10.
pdf), all of which found significant increases in In several studies the risk for cancer of the
risk with alcohol consumption. A case–control oesophagus was reduced with increasing time
study in Uruguay (De Stefani et al., 2009) found since cessation of drinking. In a meta-analysis of
a significant positive association with a predomi- 5 case–control studies Rehm et al. (2007) reported
nantly alcohol-based dietary pattern and cancer that compared with current drinkers, risk was
of the oesophagus (data not shown). [No specific not reduced until 5 years or more after cessation
assessment of alcohol intake was presented and of drinking (OR, 0.85; 95%CI: 0.78–0.92) and
the contribution of other foods to this dietary approached that of nondrinkers after 15 years or
pattern was not known.] more since quitting drinking (Table 2.6 on-line).
Thygesen et al. (2009) reported a significantly Similar results were obtained from a cohort
higher rate of cancer of the oesophagus among study (Ozasa et al., 2007). Risks are elevated
Danish alcohol abusers compared with national among more former drinkers, who most likely
rates (Table 2.5 on-line). [This study provided no cease drinking due to ill health directly related
information on individual exposures or results to the cancer or its precursors (Rehm et al., 2007;
adjusted for potential confounders.] Vioque et al., 2008; Ozasa et al., 2007).
(b) Intensity and duration (d) Types of alcoholic beverage
Data reviewed previously (IARC, 2010) Most previous studies found no material
consistently showed that increasing consump- difference in the association of consumption
tion of alcoholic beverages is associated with of alcoholic beverages on risk of cancer of the
an increased risk of cancer of the oesophagus. oesophagus according to specific beverage types,
Consistently, a 3–8 fold increased risk with high with the most commonly consumed beverage
intakes of alcohol has been reported in more tending to be associated with the highest risk.
recent studies (Lee et al., 2005; Freedman et al., This is supported by data from more recent
2007b; Ozasa et al., 2007; Fan et al., 2008; Vioque studies (Lee et al., 2005; Freedman et al., 2007b;
et al., 2008; Ishiguro et al., 2009; Pandeya et al., Fan et al., 2008; Vioque et al., 2008; Allen et al.,
2009). Smaller increases in risk at lower amounts 2009; Pandeya et al., 2009).
381
IARC MONOGRAPHS – 100E
(e) Population characteristics increasing alcohol consumption among current
smokers compared to never smokers from a case–
All recent studies have found significant posi-
control study in Australia. [The low numbers
tive associations in both men (Fan et al., 2008;
of highly exposed cases among never smokers
Benedetti et al., 2009; Ishiguro et al., 2009) and
makes it difficult to draw any conclusions.]
women (Allen et al., 2009), and in studies that
have stratified by sex (Pandeya et al., 2009).
(i) Joint effect of alcoholic beverages and
tobacco smoking
(f) Risk associated with facial flushing response
Evidence suggests a joint effect of tobacco
One cohort study in Japan examined the
smoking and consumption of alcoholic bever-
association of consumption of alcoholic bever-
ages on the risk for cancer of the oesophagus,
ages with cancer of the oesophagus according
with very high risks observed in individuals
to whether the cohort members experienced a
who were both heavy drinkers and heavy
facial flushing response. Although the risk asso-
smokers. Recent studies that have examined the
ciated with a high alcohol intake among men
joint effect of alcohol consumption and tobacco
with a flushing response was higher than among
smoking on the risk of cancer of the oesophagus
those with no flushing response, the differences
are described in Section 2.20 of the Monograph
were not significant. Details of the association of
on Tobacco Smoking in this volume.
alcohol consumption according to genetic vari-
ants in alcohol-metabolizing genes related to the
2.2.4. Cancers of the upper aerodigestive tract
flushing response are presented in Section 2.19.
combined
(g) Histological subtypes
(a) Overview of cohort and case–control studies
In the previous IARC Monograph it was
In the previous IARC Monograph (IARC,
concluded that consumption of alcoholic bever-
2010) the association between consumption of
ages is causally related to squamous cell carci-
alcoholic beverages and risk of cancer of the
noma of the oesophagus (OSCC), with no or
upper aerodigestive tract combined was not eval-
little association with adenocarcinoma of the
uated. Since then, three cohort studies (Thygesen
oesophagus (IARC, 2010). Data published since a
et al., 2007; Allen et al., 2009; Weikert et al., 2009;
strong association with OSCC or non-adenocar-
Table 2.11 available at http://monographs.iarc.fr/
cinoma (Lee et al., 2005; Freedman et al., 2007b;
ENG/Monographs/vol100E/100E-06-Table2.11.
Fan et al., 2008; Vioque et al., 2008; Allen et al.,
pdf) and one case–control study (Zaridze et al.,
2009; Benedetti et al., 2009; Ishiguro et al., 2009;
2009; Table 2.12 available at http://monographs.
Pandeya et al., 2009; Weikert et al., 2009), and
iarc.fr/ENG/Monographs/vol100E/100E-06-
no association with alcohol consumption and
Table2.12.pdf) have examined the association
adenocarcinoma of the oesophagus (Freedman
of alcoholic beverage consumption and cancers
et al., 2007b; Allen et al., 2009; Benedetti et al.,
of the upper aerodigestive tract (i.e. oral cavity,
2009; Pandeya et al., 2009).
pharynx, larynx and oesophagus combined),
(h) Association among non-smokers and one reported on cancers of the oral cavity,
pharynx and larynx combined (Freedman et al.,
Data on the association of consumption of
2007a). All studies reported significant increases
alcoholic beverages with risk of cancer of the
in risk with alcoholic beverage consumption,
oesophagus among non-smokers are limited.
observed in both men and women (Freedman
Pandeya et al. (2009) reported higher risks with
382