Excessive alcohol consumption is a major cause of ill health worldwide.
1
However, results from various studies have suggested that modest alcohol consumption
can lower risk of type 2 diabetes and coronary artery disease.
1
,
2
The relationship between alcohol consumption and chronic disease risk appears to be
complex, not only depending on the specific disease outcome, but also on the pattern
of alcohol consumption. The same amount of alcohol consumed over a certain period
is likely to have a different effect for binge drinking as compared with lower but
regular consumption. Furthermore, the effect of alcohol intake is likely to be modified
by sex, with possible beneficial effects of moderate alcohol consumption on risk of
type 2 diabetes restricted to women.
2
In addition, common variants in the alcohol dehydrogenase (ADH) and aldehyde dehydrogenase
(ALDH) gene affect the activity of enzymes that metabolize alcohol.
3
In persons of East Asian ancestry, genetic variants that lead to slower metabolism
of alcohol and its metabolite acetaldehyde are much more common than in persons of
European ancestry.
4
Thus, the impact of alcohol consumption on the development of type 2 diabetes is likely
to depend on patterns of alcohol consumption, sex, and ethnicity.
5
In this issue of the Journal of Epidemiology, Jisun Lim et al report an association
between alcohol consumption and the prevalence of impaired fasting glucose (IFG) and
undiagnosed diabetes in the Korea National Health and Nutrition Examination Survey
(KNHANES).
6
This study had a cross-sectional design, but reverse causation seems unlikely, as
participants with known diabetes were excluded. The study had several strengths. First,
the authors considered patterns of alcohol consumption by examining the frequency
of alcohol consumption in relation to diabetes separately for persons with a high-risk
level of consumption and participants with lower levels of consumption. Second, instead
of combining all participants with no current alcohol consumption as the reference
group, they distinguished between never consumers, past consumers, and participants
with a very low frequency of consumption (<1/month) and used the latter as the reference
group. This is relevant, because past alcohol consumers are likely to include individuals
who quit consumption because of ill health, thus making them inappropriate as a reference
group. In fact, a recent meta-analysis only showed an association between moderate
alcohol consumption and a lower diabetes risk in studies using all non-current consumers
as a reference group and not in studies using never consumers as the reference group.
2
The study by Lim et al is a valuable addition to the literature because only a limited
number of studies attempted to distinguish frequency and level of consumption or used
a reference group not including past alcohol consumers.
Lim et al report that a high-risk level of consumption was significantly associated
with a higher prevalence of IFG in both men and women and a higher prevalence of diabetes
in men. This association was observed even for high-risk drinking with a frequency
of only 2 to 4 times per month. High-risk drinking was defined as at least 7 drinks
per sitting in men and at least 5 drinks per sitting in women. Thus, the amount of
alcohol consumed at a single sitting was very high, but the average daily alcohol
intake for a frequency of 2 to 4 times per month was only 8 g in women and 12 g in
men. These results suggest that binge drinking may be associated with a higher risk
of type 2 diabetes, even in the absence of high average daily consumption. Furthermore,
high-frequency binge drinking was more strongly associated with IFG and diabetes risk
than low-frequency binge drinking.
For modest levels of alcohol intake, regular drinking (on most or all days of the
week) was associated with a lower diabetes prevalence in women as compared with rare
consumption. The average alcohol consumption in this group was 17 grams per day corresponding
with about 1.5 standard alcoholic drinks. These results should be interpreted with
caution, as this consumption pattern was not associated with prevalence of IFG and
no trend for increasing frequency of consumption was observed suggesting that this
association might be due to chance or residual confounding. However, these results
do fit into the overall evidence from meta-analyses of prospective cohort studies
and short term randomized trials. In prospective cohort studies, moderate alcohol
consumption was associated with a lower risk of type 2 diabetes in women but not in
men.
2
Similarly, moderate alcohol consumption as compared with non-consumption in trials
reduced fasting plasma insulin concentrations in women only, suggesting a sex-specific
beneficial effect on insulin sensitivity.
7
It has been postulated that the sex difference in the association between alcohol
consumption and diabetes risk may be due to the impact of alcohol on sex hormones.
2
Alcohol lowers testosterone levels, which are associated with a higher diabetes risk
in women but with a lower diabetes risk in men.
In men participating in KNAHES, even frequent low-risk drinking, with an average intake
of 27 grams per day, was associated with higher risk of diabetes. In previous studies,
very high total alcohol consumption (>60 g in men and >50 g in women per day) was
associated with a higher diabetes risk.
2
The higher prevalence of variants in alcohol metabolizing genes that can lead to higher
circulating concentrations of alcohol and acetaldehyde for the same level of alcohol
intake may increase diabetes risk at a lower level of consumption in East Asian populations.
4
Interestingly, a Mendelian randomization study in a Korean population also suggested
that moderate alcohol consumption increases fasting blood glucose concentrations in
men.
4
This study used ALDH gene variants as an instrumental variable reflecting alcohol
consumption. Mendelian randomization has been suggested to reflect causal relationships,
potentially avoiding confounding and reverse causation that can affect traditional
epidemiological approaches. However, in the case of alcohol consumption, this approach
may have limitations, as the genetic instrument cannot clearly distinguish patterns
of alcohol consumption and reflects not only levels of consumption but also differences
in alcohol metabolism that may affect diabetes risk.
In sum, the study by Lim et al suggests that binge drinking has adverse effects on
the development of type 2 diabetes. The high prevalence of binge drinking reported
in this population of adult Koreans over 30 years of age is alarming. Regular consumption
of moderate amounts of alcohol consumption may lower risk of type 2 diabetes in women.
However, this level of consumption can still increase risk of traffic accidents, future
transition to alcoholism, and several common forms of cancer, including common cancers
such as breast and colorectal cancer.
1
In light of these health risks associated with alcohol consumption, recent national
guidelines recommend consumption of a maximum of 14 standard alcoholic drinks per
week in the United Kingdom
8
and a maximum of one drink per day in the Netherlands.
9
In East Asian populations, high alcohol consumption may be of even greater public
health concern because a large proportion of the population is genetically predisposed
to high levels of the carcinogenic alcohol metabolite acetaldehyde.
3
,
4
Thus, light-to-moderate regular consumption of alcohol may be a beneficial consumption
pattern for women for the prevention of diabetes, but even for this pattern of consumption,
the overall health impact is likely to be negative for many individuals.