soft drinks increase mortality

Another study found significant increases in mortality associated with drinking either sugar-sweetened or artificially-sweetened soft drinks (see soft drinks in mortality jamaintmed2019 in dropbox or doi:10.1001/jamainternmed.2019.2478). Though I have blogged previously on the likely harms of soft drinks, I bring up this study because of its huge size and great database.

Details:
--EPIC study (European Prospective Investigation into Cancer and Nutrition), a population-based cohort study from 10 European countries (Denmark, France, Germany, Greece, Italy, the Netherlands, Norway, Spain, Sweden, the UK)
--451,743 people recruited from 1992-2001, excluding those who reported cancer, heart disease, stroke, or diabetes at baseline; data analyzed in 2018 with mean followup of 16.4 years
    --87% women, mean age 51, BMI 25, current smoker 23%, physically active 25%, total energy 2000 kcal/d (red meat 70 g/d, fruit and veges 400 g/d, alcohol 6 g/d, coffee 400 g/d, fruit/vege juices 20 g/d), menopausal 65%/menop hormone therapy 27%
    --there were some baseline differences in the groups: those who consumed more soft drinks tended to be younger, more likely to smoke, and more likely to be physically active
--main outcomes: total and cause-specific mortality, adjusting for other mortality risk factors (smoking status/intensity/duration, alcohol, BMI, physical activity, educational status, menopausal status/use of hormone replacement therapy, dietary intakes of total energy/red and processed meats/coffee/fruit and vegetable juices/fruits and vegetables/dietary fiber)
--41,693 deaths: 18,003 (43%) from cancer, 9106 (22%) from circulatory diseases, 1213 (3%) from digestive diseases

Results:
in those consuming 2 or more glasses/d, vs <1 glass/month:
    --17% higher all-cause mortality, HR 1.17 (1.11-1.22), p<0.001
        --sugar-sweetened soft drinks: 8% increase, HR 1.08 (1.01-1.16), p=0.004
        --artificially-sweetened soft drinks: 26% increase, HR 1.26 (1.16-1.35), p<0.001
    --circulatory disorder deaths:
            --sugar-sweetened soft drinks: not statistically significant (slight 11% trend to increase)
            --artificially-sweetened soft drinks: 52% increase, HR 1.52 (1.30-1.78), p<0.001
    --deaths from digestive diseases:
            --sugar-sweetened soft drinks: 59% increase, HR 1.59 (1.24-2.05), p<0.001
            --artificially-sweetened soft drinks: not statistically significant and no trend to difference
    --cancer deaths:
            --no statistically significant relationship with cancers overall, or breast or prostate cancer specifically, but there was a 25% increase in colorectal cancer deaths, HR 1.25 (1.07-1.47), p=0.004; and there were trends to increases in the artificially-sweetened group [HR 1.22 (0.91-1.64)], moreso than in the sugar-sweetened group [HR 1.10 (0.86-1.40)]
    --Parkinson’s disease mortality:
        --59% higher mortality, HR 1.59 (1.07-2.36), p=0.02
        --similar trend to increase with both sugar- and artificially-sweetened soft drinks
    --Alzheimer’s disease mortality: no association
    --deaths from external causes (their negative control): no differences with consumption of these soft drinks
--the above results were nonlinear: eg for all-cause mortality, higher risks were found for artificially-sweetened soft drinks at 125 ml/d (about ½ cup) and at 225 ml/d (about 1 cup) for sugar-sweetened ones
--sensitivity analyses produced similar results by adjusting for BMI, by comparing those who had low vs high BMI, by excluding deaths in the first 8 years of followup, or by assessing consumers of only sugar-sweetened vs only artificially-sweetened beverages

Commentary:
--this study is pretty consistent with other large studies, including the Health Professionals Follow-up Study and the Nurses’ Health Study, finding higher mortality with consumption of artificially-sweetened beverages, as well as with the Women’s Health Initiative Observational Study (see http://gmodestmedblogs.blogspot.com/2019/02/artificial-sweeteners-and-stroke.html , which reviews the WHI study finding an increased incidence of artificial sweeteners and stroke, CAD, and all-cause mortality)
--striking that the increases in overall mortality, as well as circulatory deaths, were actually higher in those drinking artificially-sweetened soft drinks over sugar-sweetened ones.
    --One concern in a study such as this one is reverse causality: are those drinking the artificially sweetened ones doing so because of underlying comorbidities (such as obesity) that might independently predispose them to higher mortality?? But:
        --there was no apparent effect associated with obesity: no effect by mathematically controlling for obesity (though, it is important to recognize that in controlling for something like obesity, there is a complex interaction between obesity and several other potential risk factors, perhaps some unanticipated confounders; so, controlling for a potential confounder as BMI is much more complex than a totally independent potential confounder). But they also found no difference in association just looking at those people with normal vs elevated BMIs at study entry.
        --they did not find any significant difference in the analysis after 8 years, decreasing the likelihood of reverse causality, since at least some of those with unanticipated comorbidities that could affect mortality might not have been alive before this 8 year cutpoint.
        --and, adding validity to the study, they did not find any association with deaths from their negative control of external causes
--interesting finding that only artificially-sweetened soft drinks led to increased circulatory system deaths and only sugar-sweetened ones with digestive system deaths.  I’m not sure how to explain this:

    --both sugar-sweetened and artificially sweetened soft drinks have been associated with higher risks of obesity, diabetes, and cardiovascular disease (artificial sweeteners are associated with significant gut microbiome changes and a predisposition to diabetes and obesity).

Not sure why the increased cardiovasc events was so different in the 2 groups???

    --and there are reported microbiome changes associated with consumption of both of these types of sweeteners, with the sugar-sweetened ones often using fructose, which promotes lipogenesis, NAFLD and decreased insulin sensitivity (eg, see http://gmodestmedblogs.blogspot.com/2016/08/non-alcoholic-fatty-liver-disease-2.html). see http://gmodestmedblogs.blogspot.com/2017/01/artificial-sweeteners-microbiome1.html, which reviews some studies on the harmful effects of artificial sweeteners on the microbiome

--A recent study in BMJ linked sugary drinks to cancer (see cancer sugary drinks bmj2019 in dropbox, doi: https://doi.org/10.1136/bmj.l2408 :

    --100K French adults followed 5 years found that of 2200 incident cancers diagnosed, about one third of which were breast cancers, mostly premenopauseal

    --total cancer risk increased with each 100-mL/day increase in consumption of sugar-sweetened beverages, including soft drinks and 100% fruit juices.

--there were some unexpected findings to me in this study:
    --for total mortality: current smokers had a lower risk from soft drinks than never smokers (10% increase vs 13%, and 17% vs 31% for those drinking artificially-sweetened soft drinks), BMI tracked at 18% increase if <25, decreasing to 5% if 25-30, and increasing back to 23% if >30 [part of this could be selection bias, if those with lower BMI actually were sicker leading to their lower BMI, though there was no significant difference by excluding the first 8 years of followup], those less physically active (16% vs 8%, as well as those specifically drinking artificially-sweetened soft drinks as 26% vs 9%), and in those drinking less alcohol
    -- and, generally, mortality outcomes tend to be worse in those with multiple concurrent risk factors, which was not found in this study
--limitations of the study:
    --this was an observational study, so unable to determine causality (eg, there may have been some unanticipated confounders).
    --The study relied on baseline data, and that most likely changed significantly over the long course of this study: ie, they did not update their evaluation of the patients over time, and many parameters likely changed a lot over the course of the study: eg, did many smokers stop in the first year?? Did others continue throughout?? Did some start after the study enrollment?? Did people change their soft drink patterns over time?? What about changes in weight/BMI? Or diet and exercise????....
        --The fact that the relative risks associated with drinking these soft drinks was pretty low, or that there was no additive effect of other mortality risk factors suggests (to me) that there were likely cross-overs from smoking to nonsmoking, changes in soft drink consumption patterns, etc. Also, Italy, Sweden and Spain did not differentiate the types of soft drinks consumed, which limited the statistical analysis.

So, to me this study just reinforces the likely harmful effects of sweetened soft drinks, either by sugars or artificial sweeteners. There is no clear randomized controlled trial, nor is there ever likely to be one (hard to assign a large group of people to drink varying amounts of these different soft drinks and look at outcomes many years later… not likely to get lots of financial support by Coca Cola). So we are left with large prospective epidemiologic studies like this one, and some pretty supportive mechanistic studies in animals and humans (eg on the effects of these sweeteners on the microbiomes, including glucose intolerance, or the effects of high glycemic index foods on metabolism) to reach probably reasonable public health conclusions. And, since there is no clear necessity for these soft drinks (water works well…), it does seem to me that this is the low-hanging fruit on dietary change (though there are many fruits such as lemons or limes may even make the water taste better…).

Bottom line: I think that we clinicians should be clear to patients that soft drinks are not healthy. And many patients don't know that artificially sweetened beverages are particularly bad, and may even predispose people to diabetes and obesity...

An interesting prior blog:

http://gmodestmedblogs.blogspot.com/2016/09/sugar-industrys-role-in-creating-policy.html has an interesting history of the sugar industry, including the chair of Harvard School of Public Health being a major backer (and board member) of the sugar industry's push to publish medical articles to undercut the evidence of the link between sugar and coronary artery diesease

geoff​

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

to get access to all of the blogs:

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

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

3. 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

please feel free to circulate this to others. also, if you send me their emails, i can add them to the list