Variety articles
Back to overviewPrevious entry | Next entry
Evidence Based Nutrition?
13.08.2010 | Dr. Peter Prock
The concept of evidence based medicine (EBM) has long been firmly established. The core aspect of EBM is that all decisions concerning medical treatment are based solely on empirically proven efficacy. This means in practice that almost only findings from randomised, controlled trials (RCTs) should be implemented in practical medicine.
One of the questions addressed at the conference of the Council for Responsible Nutrition (CRN International) held in Geneva, Switzerland, on 3 July 2010 (1), was whether it is admissible to talk about "evidence based nutrition" (EBN) in parallel to EBM, what this would look like, and what consequences it would have for so-called health claims.
The conference was held against the background of the current situation within the EU where the assessment of health claims for foods is treated extremely restrictively. The aut
horities demand that the mandatory texts are based exclusively on "the highest scientific standards" and thus set these on a par with RCTs.
In his excellent lecture (2), Prof. Jeffrey Blumberg from Tufts University, Boston, USA, impressively proved not only that RCTs are of limited use in research into foods, but also that the evidence from the quantitative results of epidemiological studies is comparable to that from RCTs. He proved the latter point with high-ranking papers from the New England Journal of Medicine which compared the results from both designs (3, 4), and came to startlingly similar conclusions.
A detail from the Physicians’ Health Study II (PHS II) (5) deserves a mention with regard to the limitations of RCTs for intervention studies on micronutrients: there were dramatic differences in terms of overall mortality, mortality due to cardiovascular events and cancer between those study subjects who allowed themselves to be randomised into the study and those who did not. The considerably higher risk of mortality – up to 70% (!) – among the non-participants can be explained in part by concomitant diseases and lifestyle factors. However, for the PHS II this means that comparably healthy individuals were included in the study. These do not represent a normal cross-section of society, so it was even more difficult to identify any differences in mortality as an endpoint of the PHS II. Does this mean that micronutrients have no effect, as is often suggested? No, not by any means! The problem is, that no definitive conclusions can be drawn for this.
However, Blumberg was not satisfied with merely unveiling the fundamental limitations of this scientific viewpoint. Instead, he insisted that the totality of the scientific evidence must be applied in connection with any assessment of the effectiveness of foods for influencing health; and that such assessment must not be restricted to RCTs which were specifically developed for pharmaceutical research. He stated: “The popular belief that only randomized, controlled trials produce trustworthy results and that all observational studies are misleading does a disservice to patient care, clinical investigation, and education of health care professionals.“. Evidence based nutrition must have its own set of rules and must not be put on a par with, or confused with evidence based medicine.
This is an important message addressed to the authorities who ultimately have to make regulatory decisions and who, in protectionist undertones, often refer to the lack of RCTs and frequently take negative decisions in this respect.
Comments and references
(1) CRN-International, 1st Scientific Symposium: Scientif
ic Issues Related to Codex Goals, Geneva, Switzerland, July 3rd, 2010
(2) Blumberg J., Clinical Trials for Benefits: Evidence-Based Medicine or Nutrition?, CRN-I 2010
(3) Concato J, Shah N, Horwitz RI. Randomized, controlled trials, observational studies, and the hierarchy of research designs. N Engl J Med. 2000 Jun 22;342(25):1887-92
(4) Benson K, Hartz AJ. A comparison of observational studies and randomized, controlled trials. N Engl J Med. 2000 Jun 22;342(25):1878-86.
(5) Sesso HD, Gaziano JM, VanDenburgh M, Hennekens CH, Glynn RJ, Buring JE. Comparison of baseline characteristics and mortality experience of participants and nonparticipants in a randomized clinical trial: the Physicians' Health Study. Control Clin Trials. 2002 Dec;23(6):686-702.