alcohol consumption: small amounts not assoc with decreased mortality

 A recent systematic review and meta-analysis assessed daily alcohol intake and the associated risk of all-cause mortality, finding that daily low-to-moderate alcohol intake was not associated with lower all-cause mortality risk as found in many prior observational studies (see alcohol low levels and inc mortality JAMA2023 in dropbox, or doi:10.1001/jamanetworkopen.2023.6185)

 

Details:

-- 107 cohort studies involving alcohol use and all-cause mortality were evaluated, from 1980 to 2021

    -- 4,838,825 participants and 425,564 deaths were analyzed

-- this review assessed these cohort studies specifically to determine their potential biases in controlling for known mortality risk factors as well as theie definition of people who were identified as abstainers and occasional drinkers (see below)

-- studies were stratified by age (<56yo vs >56yo), and sex

-- location of studies: 38% North and Central America, 49% Europe and Australia, 13% Asian

-- abstainer biases: only 21 studies (20%) did not include former drinkers and/or occasional drinkers in their abstainer reference group

    -- the issue here is that many "abstainers" in fact had consumed alcohol in the past (and perhaps lots of it): they may have transitioned to being "abstainers" because of medical problems (perhaps related to alcohol) and were at higher risk of death because of these medical problems. And analysis by current alcohol abstention might make it seem that continued low-intensity alcohol use is protective as compared to these "abstainers"

-- other risk factors controlled for: smoking 83%, socioeconomic status 64%, race 26%, diet 16%, exercise 41%, BMI 56%

-- participants were divided according to the grams of alcohol consumed (a standard drink has 14 g of alcohol)

    -- former drinkers (but completely abstaining at the time of the study)

    -- current occasional drinkers: <9.1 grams of alcohol per week

    -- low-volume drinkers: 1.3-24.0 g per day

    -- medium-volume: 25.0-44.0 g per day

    -- high-volume: 45.0-64.0 g per day

    -- highest-volume: more than 65 g per day

-- Main outcome: relative risk estimates for an association between mean daily alcohol intake and all-cause mortality

 

Results:

-- from the 107 studies, they analyzed 724 risk estimates of all-cause mortality due to alcohol intake

-- as compared to lifetime nondrinkers, adjusting for potential confounding effects of sampling variation, former drinker bias, and other prespecified study-level quality criteria:

    -- occasional drinkers: no statistically reduced risk of all-cause mortality, RR 0.96 (0.86-1.06), p=0.41

    -- low-volume drinkers: no statistically reduced risk of all-cause mortality: RR 0.93 (0.85-1.01), p=0.08

    -- medium-volume drinkers: no statistically reduced risk, RR 1.05 (0.96-1.14), p=0.28

    -- high-volume drinkers: 19% increased risk, RR 1.19 (1.07-1.32), p<0.001

    -- highest volume drinkers: 35% increased risk, RR 1.35 (1.23-1.47),  p<0.0001

    -- former female drinkers: larger risk of mortality versus lifetime nondrinkers, 22% increased, RR 1.22 (1.02-1.46), p=0.03; however there was a 15% lower risk for low-volume drinkers, RR 0.85 (0.81-0.88), p=0.001

-- BUT, in the unadjusted model, as compared to lifetime nondrinkers, occasional drinkers had a nearly significant 8% decreased mortality (p=0.08), and low-volume drinkers a highly statistically significant 15% decrease (p<0.001)

-- overall, analysis for low-volume alcohol consumption (1.3-24 g/d), found that the fully adjusted model (which controlled for smoking, drinking pattern, former drinker bias/“unhealthy abstainers” as well as diet, exercise, BMI) was associated with pretty significantly higher all-cause mortality than found in the unadjusted models (may be hard to see clearly, but the top of the Forrest plot below is the fully adjusted model and the bottom is the unadjusted one):

 

 

Commentary:

-- many observational studies have found that people who were low-to-moderate drinkers had longer life expectancy and less cardiovascular mortality than abstainers, with a J-shaped curve of increased cardiovascular disease and mortality in nondrinkers, then decreasing in moderate drinkers, and then increasing in a dose-response manner with higher levels of alcohol intake (ie, a cardioprotective and mortality benefit of lower levels of alcohol consumption vs non-drinkers and heavier drinkers)

-- there were, however, several concerns articulated over the years:

    -- there were systematic biases, especially a few studies that found that light-to-moderate alcohol consumers were healthier than nondrinkers (e.g. studies had found that light drinkers had better dental hygiene, exercised more, ate better, had better weight control, and higher income than nondrinkers; this finding by itself might explain the "J-shaped" curve in light-to-moderate drinkers)

   -- the abstainer group, used as the reference group for clinical outcomes, usually included “sick quitters” and former drinkers: many of these individuals may have cut down or stopped drinking because of health reasons, perhaps some related to their former drinking. In fact, of the 107 studies in this current meta-analysis, 86 included former drinkers and/or occasional drinkers in the abstainer group

    -- many of the studies had overrepresentation of old white men

   -- Mendelian randomization studies in the past, which decrease the confounding effects of sociodemographic and environmental factors, did not find evidence of cardioprotection (see http://gmodestmedblogs.blogspot.com/2019/05/stroke-risk-lowest-if-zero-alcohol.html ). this Chinese study found that a genetic variant, when present, predicted the degree of alcohol consumption in the population, and that this variant also predicted the occurrence of strokes in the population in a linear fashion, though there was no benefit in the occurrence of MIs (ie no “J-curve”). The next blog will review a new Mendelian randomization study that was published last year

 

-- overall conclusions of this new meta-analysis:

    -- no significant mortality reduction associated with occasional or light drinkers (moderate alcohol consumption)

    -- increased mortality risk in those consuming 25 or more grams of alcohol/day, and this mortality risk increased progressively with higher alcohol consumption

    -- analysis that controlled for important mortality risk factors (eg diet, exercise, BMI, drinking pattern) led to decreasing association of mortality benefit with levels of alcohol consumption

    -- and controlling for individuals stating they were non-drinkers or occasional drinkers in the fully adjusted model above found no mortality benefit. These non- and occasional-drinkers included “sick quitters” who had consumed alcohol in the past and stopped drinking because of severe medical problems, as noted above

    -- this study reinforced some of the older studies finding that younger people (eg <50yo) who were followed at least 10 years had significantly higher mortality at lower levels alcohol use. the younger cohort is likely to have less selection bias, since they are less likely to have as many  accumulated cardiovascular and other mortality risk factors. this age-based finding reinforces the importance of controlling for known mortality risk factors and that alcohol is the culprit

 

-- Also, it is important to remember that alcohol is associated with an array of cancers, including with low alcohol consumption (see http://gmodestmedblogs.blogspot.com/2019/04/a-bottle-of-wine-week-and-cancer-risk.html, reviewing a study finding that drinking one bottle of wine/week was equivalent to smoking 5 cigarettes/week for men and 10 cigarettes/week for women. And noting that only 13% of people in a national population survey considered cancer as a health risk of alcohol (?a public health system failure??)

-- and, several studies have found increases in cardiovascular disease in alcohol "abstainers" (ie, no significant protective effect of alcohol-induced increases in HDL or decreases in platelet function), presumably for all of the above reasons: increased risk factors and sick quitters in the "abstainers"

 

Limitations:

-- only 21 of the 107 studies actually assessed "never drinkers", and therefore limited the rigor of this overall analysis, so the actual sample size was much smaller than the 4.8 million people in the full analysis

-- questionnaires about alcohol consumption are pretty notoriously inaccurate, with people minimizing their self-reported alcohol intake, thereby limiting the accuracy of this report

-- they did analyze lots of potential mortality risk factors, but still may have missed other important ones, specifically those associated with psychosocial issues (eg, how extensive/granular was perceived situational stress evaluated: stress in many studies has been associated with cardiovascular disease, for example. see http://gmodestmedblogs.blogspot.com/2021/11/mental-stress-and-heart-disease.html and http://gmodestmedblogs.blogspot.com/2022/01/stress-induced-cardiovascular-disease.html )

 

So, it is finally time to put to bed the concept that small amounts of alcohol are beneficial. Not only is alcohol a carcinogen even at low levels, but it is also associated with increased risk of heart disease and all-cause mortality in a dose-dependent manner. The next blog will be on a new mendelian randomization study, which reinforces this conclusion through genetic analysis

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 (2 options):

1. go to http://gmodestmedblogs.blogspot.com/ to see the blogs in reverse chronological order

  -- click on 3 parallel lines top left, if you want to see blogs by category, then click on "labels" and choose a category​

  -- 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

2. go to https://bucommunitymed.wpengine.com/, a website from the Community Medicine section at Boston Medical Center.  This site does have a very searchable and accessible list of my blogs and is the easiest to view blogs and displays more at a time.

if you would like to see the article, please email me. 

please feel free to circulate this to others. also, if you send me their emails (gmodest@bidmc.harvard.edu), i can add them to the list