Articles about nutrition myths always make us think that there are hundreds of studies with thousands of participants as proof – they can not all be wrong, can they? As I went through some websites to read about recent myths I noticed something. The references (where mentioned at all) became more and more familiar to me – because the same studies were used over and over again.
Identifying the myth
Actually it isn’t that hard to identify myths – you just have to go to different websites and check out the references instead of getting your information only from one source. For example, I was looking for information for my article Milk & Bones, which discusses the connection between milk intake and risk of fracture. I inevitabely encountered some websites and articles that where criticizing milk quite harshly. Of course I checked their references and noticed that most of them were actually the same – which led me to the list of websites you can find below.
So all together I found about six studies that were cited noticeably frequent in these articles. Actually, most of them are quite “harmless” and only state that there is “no overall association” between milk intake and risk of fracture or at least “no increased risk” for it.
But the one from Michelsson (2014) was the one used most. Interestingly (but not surprising), it is also the one with the most provoking finding. It is stating that
High milk intake was associated with […] higher fracture incidence in women.
which is propably the most cited passage from the study. But in all of them they skip the second part of the conclusion which says
Given the observational study designs with the inherent possibility of residual confounding and reverse causation phenomena, a cautious interpretation of the results is recommended.
Well, that didn’t quite work out Michaeelsson, did it?
Despite the provoking statements, Michaelsson himself already mentions the weaknesses of this study in the conclusion.
- Observational study design: regarding the food frequency questionaires, which are used for determining the milk intake, but which are not the most precise assessment method. The participants “reported their average frequency of consumption […] during the past year” so they actually had to remember or estimate what they ate and drink during the past 12 months. Do you remember exactly what you had for lunch last week?
- Confounding and reverse causation: so basically a problem of correlation and causation. Because actually the study did not prove any causality between milk and fractures, rather showing that a positive correlation exists. You can not exclude that another factor beneath milk intake could cause the higher incidence of fractures (confounding). Furthermore, reverse causation says that the correlation could also be the other way around, meaning people with high risk of fracture tend to consume more milk (due to common sense that it’s beneficial). You should always take biases like that into account when evaluating study results.
I don’t want to talk down this study, On the contrary, I want to make clear that no study is ever flawless. Michaelsson made it clear that a cautious interpretation of the study results is mandatory. But for me, the mentioned articles do not take that into account, but instead exaggerate and distort the conclusion and therefore manipulate the intended message.
One in a million
On the other hand there are (maybe not a million but a lot) other studies, which do not support these myths. When I started researching the connection between milk intake and fracture risk there were in fact a lot high-quality and up-to-date studies (and meta-analyses) from independent researchers. It’s all a matter of where you look (quality) and how diversified your research is (quantity). But you must not base your opinion (or article) just on a single study and ignore all others that are out there.
Well, in conclusion there is no conclusion. There are studies that show one thing and studies that refute it. It’s the endless wheel of research. What I want to point out is the cautious interpretation and the careful handling of results, which the websites and articles mentioned before are sadly failing to do. So drink milk or leave it be, but don’t fuel (unfounded) fears and myths through misinterpretation.
Freuquently used publications