Monday, July 11, 2011

The American Journal of Clinical Nutrition's recent peer review failure.

There's got to be a better way than peer review, because if papers like this one can make it through, the system is broken.

The paper's entitled, Dairy attenuates oxidation and inflammatory stress in metabolic syndrome, and it's written by Renee Stancliffe, Teresa Thorpe and Michael Zemel.

The paper's aim was to study the impact of dairy on oxidative and inflammatory biomarkers in individuals with metabolic syndrome, following 12 weeks of being randomized to different levels of dietary dairy. They also measured waist circumference.

So what's got my feathers ruffled?

The whole point of this study was to determine the impact dairy intake had on oxidative stress and inflammation, yet the design of the study had the folks in the "low" dairy group substituting dairy with 3 daily servings of prepackaged non-dairy foods so as to maintain comparable macronutrient intakes between groups.

So? Wouldn't ensuring a uniform macronutrient intake be a good idea?

Absolutely, but not if the foods you use to ensure uniformity have a known impact on oxidative stress and inflammation! And so what was on the list of foods provided to the low-dairy folks? Among other things processed luncheon meats and trans-fat containing peanut butter crackers.

The authors do in fact mention this as a potential confounder in their conclusion, but certainly for me, given that there was no need whatsoever to provide the subjects trans-fat containing foods (there are plenty of trans-fat free cracker choices) or processed meats, it'd have been a peer review deal breaker.

Unfortunately, there's more.

Waist circumference was also a variable they measured, and they concluded that the higher dairy group significantly reduced their waist circumference and trunk fat. Yet they chose to measure waist circumference, not from an anatomical landmark, but rather from midway between the lateral lower rib margin and the ileac crest at mid-exhalation. While the NHLBI and NIH recommend using an anatomical landmark (the ileac crest), the WHO recommends the midway point described in this paper. However no one recommends mid-exhalation because that's a fairly impossible thing to measure. What this translates to mean is that there was ample room for subjective error, all the more so given that there's no mention of blinding at all in the study methodology.

Next there's their description of their diet modalities. They describe their dairy arms as, "low" and "adequate". Adequate is a subjective term. To me, adequate means sufficient, but certainly not ideal. While they might argue they used the term "adequate" because that arm consumed the USDA recommended number of dairy servings, then perhaps they could have used, "USDA recommended" in place of "adequate". Subjective adjectives have no place in a research paper, and it clearly suggests author bias as to the benefits of more dairy.

So do the authors have a conflict of interest?

Not according to them. The last line of the published paper clearly states,

"The authors reported no conflicts of interest"
Yet conflicts of interest are almost unavoidable in research. If you're an expert at something, chances are you've benefited from your expertise, and even if you feel that your benefits don't put you into conflict, conflicts of interest must be disclosed if there's a possibility that a person's interests could be even just be fairly perceived as a conflict.

So let me ask you, do you think it could fairly be perceived as a conflict of interest to be an author of a study aiming to look at the outcomes of differing degrees of dairy on inflammatory biomarkers and waist circumferences, that's designed with an entirely unnecessarily built in trans-fat containing inflammatory confounder, a lack of blinding, and a subjective means of measuring waist circumference, when you yourself have a patent on the use of dairy to affect fat loss? Or how about if you authored a diet book called, The Calcium Key, where one of your claims was that dairy would,
"Triple the rate of fat lost from your abdominal area–the "belly fat" that puts you at greatest risk for high blood pressure, high cholesterol, heart disease, and Type 2 diabetes"
If you answer, "yes", then the next question you've got to ask yourself is why Michael Zemel reported no conflicts of interest to the American Journal of Clinical Nutrition, and also just who the heck peer reviewed this paper, as clearly, if they didn't see a conflict with Zemel, then they don't know dairy literature well enough to have been chosen as a peer reviewer.

Of course you'd think the editors of the journal ought to have known about the conflict of interest too.

(BTW - This isn't the first time Zemel's withheld his conflicts of interest from a journal - see this note published by the editor of the International Journal of Obesity)

Stancliffe RA, Thorpe T, & Zemel MB (2011). Dairy attentuates oxidative and inflammatory stress in metabolic syndrome. The American journal of clinical nutrition PMID: 21715516