Alcohol and cancers

 A recent massive data analysis assessed the risk of cancer associated with alcohol ingestion, finding a marked dose-response curve and several unexpected cancers not typically associated with alcohol (see alcohol cancer risk BJCancer2014 in dropbox or DOI:10.1038/bjc.2014.579)

Details:

-- 572 studies were included in this analysis, with 486,538 cancer cases identified

-- 409 case-control studies,163 cohort studies: 541 studies reported cancer incidence, 31 cancer mortality

-- site of study: North America in 236, Europe in 184, Asia in 101, mixed or other areas in 51

-- adjusted estimates reported in 297 studies, occasional drinkers in the reference group in 138

Results:

-- Relative risks (RRs) for heavy drinkers compared with nondrinkers and occasional drinkers, associated with a clear dose-risk relationship: 

    -- oral and pharyngeal cancer: RR 5.13

        -- pooled estimate for moderate and heavy drinkers vs nondrinkers and occasional drinkers: RR 1.13 (1.00–1.26) for light, RR 1.83 (1.62–2.07) for moderate and 5.13 (4.31–6.10) for heavy drinkers, in 52 studies

    -- esophageal squamous cell carcinoma: RR 4.95

        -- pooled estimate for moderate and heavy drinkers vs nondrinkers and occasional drinkers: RR 1.26 (1.06–1.50) for light, RR 2.23 (1.87–2.65) for moderate and 4.95 (3.86–6.34) for heavy drinkers, in 54 studies

    -- colorectal cancer: RR 1.44

        -- pooled estimate: moderate and heavy drinkers, but not light drinkers, was associated with an increased risk of cancer: RR 1.17 (1.11–1.24) for moderate and 1.44 (1.25–1.65) for heavy drinkers, in 66 studies

            -- though this study, as opposed to others, only found this association in men, with no significant effect in women

    -- laryngeal cancer: RR 2.65

        -- pooled estimate for moderate and heavy drinkers vs nondrinkers and occasional drinkers: RR 1.44 (1.25–1.66) for moderate and 2.65 (2.19–3.19) for heavy drinkers, 41 studies

    -- female breast cancer: RR1.61

        -- pooled estimate for moderate and heavy drinkers vs nondrinkers and occasional drinkers: RR 1.04 (1.01–1.07) for light, RR 1.23 (1.19–1.28) for moderate and 1.61 (1.33–1.94) for heavy drinkers, in 118 studies

-- other cancers with a higher risk in heavy drinkers:

    -- stomach cancer: RR 1.21 (1.07-1.36), 39 studies

    -- liver cancer : RR 2.07 (1.66 -2.58), 36 studies

    -- gallbladder: RR 2.64 (1.62 - 4.30), 8 studies

    -- pancreas cancer: RR 1.19 (1.11–1.28), 39 studies

    -- lung cancer: RR 1.15 (1.02–1.30), 34 studies

-- an indication of an association between consumption of alcohol and risk of:

    -- melanoma: RR 1.11 (0.97–1.27) for light and 1.20 (1.03–1.41) for moderate drinking; 14 studies (especially in studies in North America)

    -- prostate cancer: RR 1.04 (1.01–1.08) for light, RR 1.06 (1.01–1.11) for moderate and RR 1.09 (0.98–1.21) for heavy drinking; 43 studies

-- alcohol was not significantly associated with the risk of:

    -- adenocarcinoma of the esophagus and gastric cardia (25 studies)

    -- small intestine (5 studies)

    -- cervix (5 studies)

    -- endometrium (21 studies)

    -- ovary (20 studies)

    -- bladder (19 studies)

    -- brain (6 studies)

-- inverse association between alcohol and risk of:

    -- Hodgkin’s lymphoma: RR 0.73 (0.59–0.89) for light, RR 0.73 (0.60–0.87) for moderate and RR 0.63 (0.41–0.97) for heavy drinkers; 9 studies

    -- non-Hodgkin’s lymphoma: RR 0.88 (0.80–0.97) for light, RR 0.87 (0.81–0.95) for moderate and RR 0.75 (0.64–0.88) for heavy drinkers; 24 studies

    -- kidney: RR 0.92 (0.86–0.99) for light and RR 0.79 (0.72–0.86) for moderate drinkers; 17 studies

    -- thyroid: RR 0.81 (0.74–0.88) for light and RR 0.81 (0.71–0.94) for moderate; 9 studies

        -- the risk of thyroid and kidney cancers was significantly lower for light or moderate drinkers vs nondrinkers or occasional drinkers

-- the following table reviews the pooled relationships between alcohol and cancers, by type of study done:

-- and the following graphs reveal the dose-response curve of the relationship between the average number of grams of alcohol ingested per day (as a continuous variable) and cancer risk: 

Commentary:

-- this study is impressively large, allowing for more detailed statistical assessments of the relationship between alcohol consumption and many different cancers, across many different populations in many different countries, and scaled by amount consumed

    -- alcohol is also related to liver cirrhosis, epilepsy, poisoning, road traffic accidents, cardiovascular disease, anemia/leukopenia/thrombocytopenia, peptic ulcer disease,  birth defects, neuropsychiatric conditions, diabetes, and the array of personal/family/community adverse social effects

-- and, of course, this is particularly important because of the associated huge global mortality (about 2.5 million deaths per year and 4.5% of the global burden of disease and injury; also, about 500,000 cancer deaths worldwide in 2004)

-- this study also extends our understanding of the breadth of alcohol-related cancers, beyond the standard-bearers of oral, pharyngeal, laryngeal, esophageal (squamous), liver (hepatocellular), colorectal, and breast cancers

-- and the existence of the dose-response of finding increased risk of many of the cancers with increasing alcohol consumption does support the relationship between alcohol with cancer (though, of course, increasing alcohol consumption may also lead to adverse changes in diet/exercise/stress levels/etc that affect the risk of cancer)

-- of note, there were a couple of studies finding a relationship between alcohol consumption and prostate cancer, including a large Canadian study finding that the lifetime intake of alcohol increased the risk of both non-aggressive and aggressive prostate cancer cases by 78%, especially by the lifetime beer intake: https://www.jstor.org/stable/24717870

-- a prior study including 222 articles found that light drinking (up to 1 drink/day), comparing 92,000 light drinkers to 60,000 non-drinkers with cancer, found a 17% increase in oropharyngeal cancers, a 30% increase of esophageal squamous cell cancers, and 5% increase in female breast cancer, thereby estimating that in 2004, approximately 5000 deaths from oral and pharyngeal cancer, 24000 for esophageal squamous cell cancer and 5000 from breast cancer  were attributable to light drinking worldwide: alcohol light drinking cancer AnnOnc2012 in dropbox, or doi:10.1093/annonc/mds337

-- a more recent Australian study of 226,162 participants aged 45 and older followed 5.4 years found 17,332 cancers. and increasing levels of alcohol were associated with a 19% increase in aerodigestive tract cancer, 18% increase in mouth and pharynx cancer, 22% increase in esophagus cancer, 9% increase in colorectal cancer, 13% increase in colon cancer, 22% increase in liver cancer, and 11% increase in breast cancer . And this was a study with a short-term followup: alcohol cancer Australia BritJCancer2021 in dropbox, or doi.org/10.1038/s41416-020-01101-2

-- in terms of mechanisms of the alcohol-cancer relationship:

    --some are clearly related to the effect of the degradation product of alcohol to the carcinogen acetaldehyde (see next blog for more detailed studies on this); also the potential  effect of alcohol in the presence of UV radiation may enhance cellular damage leading to skin cancers  (the studies in the above analysis were mostly done in the US), or in the immunocompromise form alcohol perhaps paralleling other immunosuppressed conditions associated with melanoma

    -- hard to explain alcohol's benefit found in lymphomas.?? people stopping drinking and being included in the non-drinking or light-drinking categories?? of note, a recent European study of 120,852 people did not reveal an inverse association

Limitations:

-- though this is a very impressive analysis of a huge database including lots of diverse studies, as observational studies they cannot determine causality, only associations

-- and since they are pooling many studies together, the authors are combining different populations, of different ages and demographics, different underlying medical conditions, different medications taken, and different psychosocial situations and conditions, including different diets, exercise patterns and other culturally-specific differences. And this aggregation limits the generalizability of the results to the individual patient we are assessing and treating

-- there may also be residual confounding for those cancers with very strong relationships with the combination of alcohol and tobacco, where statistical manipulations to "control" for smoking may obscure some confounding factors. And, as mentioned above, increased drinking is associated with other adverse health behaviors including diet/exercise/perhaps accessing health care as potential confounders that are hard to disaggregate.

-- many of the determinations of alcohol consumption in the studies are by self-reporting, with no hard evidence other than the known tendency for self reports to under-report actual consumption

-- there were some differences in the statistical associations with alcohol consumption by the type of study assessed: case-controlled vs cohort studies.

so, 

-- eye-opening study of the array of potentially serious cancer associations with alcohol ingestion, with the increases found for many of the common cancers being with light to moderate alcohol ingestion

-- and this means that for some very common cancers, even small increases in the percentages of these cancers conferred by alcohol may well translate into lots of cancer-related morbidity and mortality

   -- there are also concerns about reverse causation: did people with early cancers or otherwise be at high risk of cancer decrease their drinking and thereby be included as non- or light-drinkers??  we know from the newer cardiovascular studies that the older conclusion that moderate alcohol intake was cardioprotective has been disproved, and that zero alcohol intake is the appropriate advice to the population: https://gmodestmedblogs.blogspot.com/2023/04/mendelian-randomization-alcohol-does.html 

-- though the current study extends the potential carcinogenic effect of alcohol to cancers not considered to be alcohol-related, it would be important to confirm these associations in other studies, preferably in persons with more extensive background information (eg underlying medical comorbidities, social conditions, diet, exercise...)

    -- but this study also reinforces the risk for cancer for the usual suspects: the oral cavity and pharynx, esophagus, colorectum, liver, larynx and female breast.

The next blog will be on the benefits of decreasing alcohol intake on cancer risk, as well as studies on the potential mechanisms of action leading to the alcohol/cancer relationship

geoff

-----------------------------------

If you would like to be on the regular email list for upcoming blogs, please contact me at gmodest@bidmc.harvard.edu

to get access to all of the blogs:  go to http://gmodestmedblogs.blogspot.com/ to see the blogs in reverse chronological order

or you can just click on the magnifying glass on top right, then type in a name in the search box and get all the blogs with that name in them