alcohol: even small amounts assoc with hypertension

 A recent article analyzed prior hypertension studies and found a linear relationship between the amount of alcohol consumed and hypertension, with no safe lower limit of alcohol (see htn alcohol intake HTN2023 in dropbox, or DOI: 10.1161/HYPERTENSIONAHA.123.21224)

 

Details:

-- this study used a new type of meta-analysis, a dose-response one-stage meta-analysis for aggregated data, combining longitudinal studies in healthy adults that reported on the association between alcohol intake and blood pressure

-- the included studies needed to be cohort or case-cohort investigations. Excluded studies included those with a cross-sectional design, those that had participants with previously diagnosed cardiovascular disease, diabetes, or cirrhosis, and those that only recruited participants with substantial alcohol use disorders or included binge drinkers

-- seven studies were found with 19,548 participants and a median follow-up 5.3 years (range 4-12 years)

    -- three of the studies were from the US: a cohort study of Black residents of Pitt County, Carolina, the multi-ethnic ARIC study from 4 US communities, and the CABL study using 24-ambulatory blood pressure monitoring in an elderly population in Manhattan

    -- four of the studies were from Asia: three from Japan and one from South Korea (the three Japanese studies included only men)

-- 12,710 men/6,650 women, four studies included smoking status as a potential confounder, five included BMI

 

Results:

-- overall, there was almost a linear positive association between baseline alcohol consumption and blood pressure change over time, with the slope being steeper and more linear for systolic than diastolic blood pressure [one standard drink is 10 g of alcohol]:

    -- 12 g/d greater daily alcohol consumption (versus no alcohol consumption): systolic blood pressure (SBP) 1.25 mmHg higher (0.49-2.01); diastolic (DBP) 1.14 mmHg higher (0.60-1.68)

    -- 24 g/d greater: systolic 2.48 mmHg higher (1.40-3.56), diastolic 2.03 mmHg higher (1.19-2.86)

    -- 48 g/d greater: systolic 4.90 mmHg (3.71-6.08), diastolic 3.10 mmHg (1.88-4.33)

 

-- sex-specific analysis (based on 5 studies with 12,196 men and 2 studies with 5,632 women):

    -- men:

        -- 12 g /d greater daily alcohol consumption (versus no alcohol consumption): systolic blood pressure 1.33 mmHg higher(0.22-2.44); diastolic 1.20 mmHg (0.47-1.93)

        -- 48 g/d greater: systolic 4.99 mmHg (3.70-6.20); diastolic 3.41 mmHg (2.06-4.77)

    -- women:

        -- 12 g /d greater daily alcohol consumption (versus no alcohol consumption): systolic blood pressure 0.82 mmHg more (-0.58 to 2.22); diastolic 1.45 mmHg more (0.70-12.20)

        -- 48 g/d greater: systolic 3.31 mmHg  more (-3.67 to 10.30); diastolic -1.27 mmHg less (-5.04 to 2.51)

 

-- Region-specific analysis:

    -- US: a positive almost linear association between alcohol intake and SBP, for DBP the curve flattened some after 24 g/d and then decreased

    -- Asia: a steeper positive almost linear curve for SBP and a less steep positive almost linear slope for DBP

 

            

 

-- there was a positive linear association between baseline BP and change in BP during follow-up as related to usual alcohol intake at baseline

-- duration of follow-up was positively associated with change in SBP during follow-up as related to usual alcohol intake at baseline, with the slope being steeper for SBP than DBP

-- the association between alcohol intake and BP was steeper in men, with a U-shaped diastolic blood pressure for women

 

-- sensitivity analysis excluding studies that did not include smoking or BMI: similar associations

 

Commentary:

-- this trial found that there was essentially a linear relationship between alcohol level and blood pressure, more profoundly with systolic blood pressure, in both men and women. This appeared to be true for any alcohol consumption

-- however, there were a few clear deviations from this result, specifically with women overall as well as with the US group overall, both having diastolic blood pressure tracking less clearly with alcohol consumption. Some potential explanations for these differences:

    -- for women:

        -- there were only two studies found where risk was stratified by sex. These were pretty different studies, with one having women drinking an average of 1.52 g per day, the other with women drinking from zero to >30g per day. Extrapolation by combinging these quite different studies (and with only two studies overall) limits the accuracy of their conclusions. It is unclear why diastolic blood pressure should be the only discrepancy. however, DBP in general tracked less well than SBP and DBP and may therefore be more susceptible to their evaluation in such heterogenous studies

        -- is there a role of sex hormones? Women with higher testosterone levels seem to have higher blood pressure (see https://www.nature.com/articles/s41371-021-00576-7 )

        -- is there an issue with the quality of the data itself? How accurate was the reporting of alcohol consumption? Was there any difference between men and women in this reporting? Were the assessments of alcohol intake done by validated questionnaires?

    -- for Asian individuals:

        -- there is a high frequency of genetic variations leading to differences in alcohol metabolizing enzymes (alcohol dehydrogenase, aldehyde dehydrogenase). Those with these genetic variations tend to drink less because of significant flushing and discomfort with pretty much any alcohol consumption. But, we do not have individual-level data to disaggregate the Asian participants into those with these genetic differences: could these differences explain why the diastolic blood pressure increased more in Asian people than those from the US?

-- There have been a few studies finding that small increases in blood pressure, as observed in the above alcohol study, may well be associated with significant cardiovascular events when assessed on a community level:

    -- the Multiple Risk Factor Intervention Trial (MRFIT, one of the best acronyms of all time!!) found that those with SBP 2 mmHg lower had a lower annual mortality from stroke (6%), coronary heart disease (4%), and all-cause mortality (3%)

    -- the Atherosclerosis Risk in Communities study (ARIC, see https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4845128/ ), one of the studies used in the above analysis, found that a 1 mmHg lower SBP was associated with 13.5 fewer coronary heart disease events /100,000 persons per year in African-American participants, and 9.0 fewer events in white participants (though, as noted in many other blogs, these "racial" differences are really very likely related to different living situations, eg different quality of life issues, overwhelming issue of the effects of racism on a daily basis, etc). these differences which in the patst were attributed to race,  basically reflect a social construct and not a genetic one, and we know that stress is associated with hypertension (acute stress clearly raises blood pressure; chronic stress is pretty clearly associated with chronic hypertension, though a clear-cut causal relationship is really hard to prove); for example, see Social network factors and cardiovascular health among baltimore public housing residents - PMC (nih.gov)

 

Limitations:

-- The confidence intervals for women were large because of there being only 2 studies analyzed, which decreases the generalizability to broader groups of women (see above)

-- we do not have any information about the individual genotypes of Asian participants, specifically the role of individulas' genetic variations in alcohol dehydrogenase and aldehyde dehydrogenase enzymes in cardiovascular outcomes with alcohol consumption

-- though this study did average a median follow-up of 5.3 years, low levels of hypertension for many years (presumably the usual trend for those with mild hypertension) may lead to increasing cardiovascular events from progressive arterial changes in stiffness/endothelial function etc.

    -- and, the study had no information about the age of the participants: one might expect that younger people followed for a longer period of time might have more long-term complications

-- the basis of this type of meta-analysis done is the quality of evidence in the individual studies (as with all meta-analyses) and the ability to combine studies into a meta-analysis with very different inclusion/exclusion criteria, patient populations studied, important non-measured confounders (diet, exercise, stress levels, etc.), even blatant co-conspirators such as smoking (only included in four of the studies above!!!) and BMI (only included in 5 of the studies). Were there other unmeasured confounders?

-- In particular, the studies only had a few assessments of the amount of drinking at various time intervals, with mathematical adjustments to develop a continuous assessment (the baseline assessment of alcohol intake was derived from consumption at baseline):

    -- it is likely that people did not drink exactly the same amount consistently for the length of studies. How accurate was the mathematicial adjustment of the data in taking this into account? Was there a difference if participants drank more heavily for a period of time and less heavy other times versus drinking more consistently? What if patients stopped completely for one or two years: did they have less cardiovascular disease because they did stop for a period of time? What if patients decreased or increased their drinking more consistently: would that affect the results?

    -- eg, the study specifically assessed healthy individuals who did not have binge drinking. Do these results apply to those who drank significantly over the weekends but not consistently during the week (a rather common finding in my experience)?

-- We have no information about the type of alcohol consumed. Were there differences in people who drank red wine versus tequila? Is it just the alcohol content of the drink, or are there other factors that might be involved (e.g. antioxidants in the beverage...)?

-- Those with diabetes were also excluded, a large and increasing percentage of people in the US. Also those with cardiovascular disease were excluded, also a large population . All this limits generalizability of the results

 

So, this trial reiterates that any level of alcohol consumption is harmful. A few points:

-- it has long been held that low levels of alcohol consumption was actually better than none, that there was a J-shaped curve where low-to-moderate alcohol was associated with improved cardiovascular health

    -- this study contradicts this long-held belief, which had been attributed to people relaxing more when they were drinking.

        -- recent studies have disproved this J-shaped curve for alcohol consumption and cardiovascular disease: http://gmodestmedblogs.blogspot.com/2023/04/mendelian-randomization-alcohol-does.html

        -- this new study suggests that at least part of this cardiovascular association may be mediated through a continuous relationship between alcohol and hypertension

            -- ie, this study is in sharp contradistinction to the older studies finding that only heavy alcohol consumption was associated with increased blood pressure: https://www.nejm.org/doi/full/10.1056/nejm197705262962103 

            -- and, this "zero alcohol" goal is confirmed by the studies on stroke and other risks: stroke risk lowest if zero alcohol (gmodestmedblogs.blogspot.com) and alcohol consumption: small amounts not assoc with decreased mortality (gmodestmedblogs.blogspot.com) 

-- Which all means that the public health message should be zero alcohol…

 

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

  -- 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

 

if you would like to see the articles in this blog, 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