Thursday, June 10, 2010

Dairy prevents heart attacks?

A sarcastic thank you to Scott Gavura from Science Based Medicine and Science Based Pharmacy whose tweet lead me to suffer through reading the latest in a never ending stream of Big Milk sponsored published info-studies.

This one?

Funded by Big Milk with a first author who has had speaking gigs with the Swedish Dairy Association and the International Dairy Federation this study has been reported as being proof that dairy consumption reduces the risk of heart attacks.

The study?

The authors, via a prospective case-control study, looked at blood levels of dairy biomarkers vs. heart attacks in 444 men and women and 556 controls over a 12 year period.

The findings?

Well according to Reuters,

"The researchers found that people with the highest levels of milk fat biomarkers, suggesting they consumed the most dairy fat, were actually at lower risk of heart attack; for women, the risk was reduced by 26 percent, while for men risk was 9 percent lower."
But of course that's before the authors controlled for confounders.

What confounders might there be?

Well surely diet's a huge one, right? Sure they're measuring dairy biomarkers but there are umpteen other dietary confounders that impact on heart attack risk - processed meats, whole grains, sodium, etc.

Well guess what?

Not only was the statistical control for diet ridiculously weak (they only controlled for reported intake of fruits and vegetables), it was incomplete as the authors admitted that dietary information was absent for at least 19% of the participants where they didn't fill out dietary surveys at all and that there was also a percentage of participants where their dietary surveys were excluded because they weren't scanned in and another group where their surveys were excluded because they were incomplete.

But rather than pick apart the study myself I'll simply let the authors shout out the value of their own research with direct quotes from their paper,
"It was not possible to delineate the exact mechanism behind the relations or exclude a more beneficial lifestyle pattern in milk fat consumers."

"The significant trends for both men and women were, however, lost after multivariable adjustment"

"Intakes of total milk products and 15:0+17:0 were not related to a first MI in this study"
So to summarize?

After doing a terrible job of controlling for confounders utilizing an underwhelming design with likely at least 25% of subjects not having dietary records, there was no statistically significant trend to report between dairy biomarkers and risk of heart attacks and yet Reuters and hundreds of other media outlets reported the findings as proof of magic dairy goodness and the American Journal of Clinical Nutrition happily published the non-findings.

Is there any wonder why Big Milk funds these studies?

(and a side question, how do these studies make it through peer review?)

Warensjo, E., Jansson, J., Cederholm, T., Boman, K., Eliasson, M., Hallmans, G., Johansson, I., & Sjogren, P. (2010). Biomarkers of milk fat and the risk of myocardial infarction in men and women: a prospective, matched case-control study American Journal of Clinical Nutrition DOI: 10.3945/ajcn.2009.29054