As mentioned in my opening post, I have planned to warble on about evidence based nutrition (EBN). I am trying to keep posts relatively brisk and to the point – as I can have a habit of heading down the rabbit hole for everything and this one is no exception.
But...
I am lucky with my place in life as a nutritionist. I have no requirement to provide public recommendations for health unless I really feel like doing it. I can have a complete independence form public health and having to provide something. I can sit on my high horse and ask for the best quality data, be critical of everything, and decry all public health campaigns as much as I feel like.
Evidence based medicine (EBM) fits me quite nicely. I had mentioned that EBM came out of Canada, and while many may be distrustful of Canadian exports thanks to Celine, but EMB is an especially good advancement in the area of medicine. The basic background and information of EBM is covered reasonable well in the bastion of all good referencing wikipedia.
I may be building my own strawman here, but EBM is a relatively simple idea, where the practice of medicine is guided from an evidence base built off a range of evidence. This evidence is then ranked based upon it’s ability to answer the question. A hierarchy of evidence. This evidence ranges from the anecdotes of patients, or clinicians (like a case report), right through case control, ecological, cohorts, uncontrolled trials, randomised clinical trials and more recently, systematic reviews and meta-analyses of clinical trials. Not accounted for in this is the basic art of interaction with the variable of the ‘sick’ human – which interplays the evidence base with the knowledge and creativity of the practitioner.
Now the nutrition world took this concept, but discovered that its not quite perfect mainly because of the differences between pharmacological/medical interventions and what we are capable of in nutritional interventions. Which in reality, isn’t much.
Blumberg (Blumberg et al, 2010) provided a nicer coverage on why EBM doesn't quite work in the context of nutrition. The same year, Jim Mann (Mann, JI, 2010) also presented a commentary , while unfortunately starting with bible quotes, highlights the World Cancer Research Fund (WCRF) approach, which is less randomised clinical trial (RCT) centric and provides more focus on the observational evidence. The WCRF report (World Cancer Research Fund, 2007) has a chapter on Evaluating the Evidence (Chapter 3) – in which they identify that they are so reliant on epidemiology - they spend quite some time explaining epidemiological style trials and a short time on RCT – mainly spending their time downplaying the value. That is because there is extremely few clinical trials on food and cancer outcomes, so they have nothing else to back it up on. So they highlight the value of the data they have available.
Following Blumbergs commentary, Fenton and Fenton (2010) wrote a letter to the editor, which in its title states the its entire intent “Nutrition Science mustn’t accept a lower level of evidence”. Basically, we cant just accept epidemiology, just because RCT’s are difficult.
To which Heaney and Blumberg (2010) replied
“Across its full range, from micronutrients to protein, nutrition does not have the kind of evidence the Fentons advocate. This is not an indictment of nutrition as a science; it is simply recognition of the following facts: 1) we cannot get RCT-level evidence for many, and perhaps most, nutritional questions, and 2) we have to make some recommendation.”
I understand why he is stating this, but I also say “bolderdash”.
Nutrition scientists have to stop making hard claims on soft epidemiology. Nutrition scientists, public health officials, academics, epidemiologists, non-governmental organisations, government health authorities, everyone – have to stop overselling their data. They have to stop saying the science is “settled” when it is largely based upon correlation. In the same breath, we need to stop taking pharma-risk factor/disease modification trials and whole heatedly assume that nutrition influencing these risk factors will have the same effect on disease.
If we do not have the data, we do not have to and should not make recommendations.
The public that funds an enormous amount of nutrition science deserves better.
References
Blumberg, J., Heaney, R. P., Huncharek, M., Scholl, T., Stampfer, M., Vieth, R., … Zeisel, S. H. (2010). Evidence-based criteria in the nutritional context. Nutrition Reviews, 68(8), 478–84.
Mann, J. I. (2010). Evidence-based nutrition: Does it differ from evidence-based medicine? Annals of Medicine, 42(7), 475–86.
World Cancer Research Fund, & American Institute for Cancer Research. (2007). Food, Nutrition, Physical Activity, and the Prevention of Cancer: a Global Perspective. (pp. 1–537). Washington DC: AICR.
Fenton, T. R., & Fenton, C. J. (2011). Nutrition science mustn’t accept a lower level of evidence. Nutrition Reviews, 69(7), 413–4; author reply 415–6.
Heaney, R. P., & Blumberg, J. (2011). Author reply: Nutrition science mustn’t accept a lower level of evidence. Nutrition Reviews, 69(7), 415–416.
