FREE! Subscribe to News Fetch, THE daily wine industry briefing - Click Here

2019 Unified Wine and Grape Symposium

Why the new MACH15 moderate consumption study must correct its protocol in order to have any credibility at all


This editorial relates to the following article: 14 Reasons the new NIH/industry-funded moderate alcohol & health study could be a $100 Million CREDIBILITY flop



To the vast number of non-scientists — which is the vast majority of people — science and credibility are two very different things. This is especially true when you consider that for most Americans, science isn’t news.


This means that good science is often ignored, especially when it clashes with politics, regulatory authorities, and special interests with their own axes to grind.


The realm of alcohol consumption and health has its own emotional and political complications.


There are scores of solidly researched, peer-reviewed, respectably published scientific studies that show moderate consumers of alcohol live longer than abstainers or heavy drinkers due mostly to cardiovascular benefits which outweigh an increase in some cancers among some drinkers.


But none of those has convinced anti-alcohol advocates and their governmental proxies.


The problem is that those studies have been observational — epidemiological studies of population data. While epidemiological studies can tell you something is happening, links and associations cannot — by themselves — determine a cause.

Association is not causation.

While observational/epidemiological studies have pointed the way to solving numerous serious diseases — from cholera to AIDS — definitive causation can only come from a clinical trial.


This is why governmental bodies have a wide range of recommendations — from none to one or two standard drinks per day.


The MACH 15 study overview states:

“Observational studies document a lower risk of coronary heart disease and diabetes among moderate consumers relative to abstainers, but they also suggest a higher risk of breast and gastrointestinal cancers, and the possibility of residual confounding of these associations by other characteristics [underlining added] cannot be excluded.

“No clinical trial has been conducted to test the hypothesis that moderate alcohol consumption lowers risk of cardiovascular disease or diabetes compared to abstention, yet public policy continues to be made regarding safe limits of drinking.”

A clinical trial is obviously needed.


That is the key reason the MACH15 investigation is a good idea. But has credibility failure baked in because it fails to address numerous “residual confounding associations” which were pointed put in this artcle: “14 Reasons the new NIH/industry-funded moderate alcohol & health study could be a $100 Million CREDIBILITY flop.


In an email to Wine Industry Insight, the Principal Investigator of that study, Dr. Kenneth Mukamal said that

“Large randomized trials differ substantially from small, highly-controlled feeding studies. Randomization itself controls for all measured and unmeasured confounding factors, so long as the trial is large enough. It provides an unbiased, real-world answer irrespective of other cardiovascular risk factors, including diet, exercise, smoking, social status, etc.”

Large confounding datasets = large target

The questionable but often cited premise that valid conclusions can be derived from a large amount of data filled with uncontrolled and confounding factors may fly in purely scientific circles.


Even if that premise is valid, the overwhelming majority of the 14 reasons cited can be used by critics as blunt instruments batter its conclusions into public and regulatory futility.


If the MACH15 investigation has any chance to make a positive contribution to public debate and regulatory consistency, it must take seriously the lessons of “small, highly-controlled feeding studies” which must be precise, accurate, and non-confounding in their basic study protocol.

Protocol corrections should be made if conclusions are to be credible

Diet, smoking, exercise, cannabis use, life-saving CVD pharmaceuticals, and variations in medical treatment pose uncontrollable, and overwhelming confounding factors when taken individually but which expand into a complex and completely unknown and unknowable territory when all are combined.


At the very least. eliminating smokers and cannabis users from the study will remove those very large confounding factors.


Closer monitoring of diet and exercise and creating statistical categories that can be targeted will offer opportunities to tease meaningful distinctions out of the data.


Industry funding and the choice of an aging, unhealthy population for an investigation have practical organizational reasons. Those cannot be changed.


But eliminating other confounding factors will make it harder for critics to dismiss the finding as irrelevant to other alcohol-consuming demographics.


Course corrections need to be taken immediately to eliminate as many confounding factors as possible so as to create as small a target for the inevitable brickbats which will be thrown at it.