Problems in nutritional science
 
Nutritional studies cannot be reliable
Nutritional recommendations formulated by national or international bodies are designed to provide information for planning a diet which will ensure people’s primary physiological needs for growth and function (the nutritional goal), and provide the basis for general good health (the public health goal). While knowledge of the minimum requirements to avoid deficiency symptoms may be considered good, there is significant uncertainty in terms of what the optimal nutritional intake for preventing widespread Western diseases. The key reason for this is that estimation methods that are available have serious limitations, and dietary studies have a relatively low priority within the field of medical research.
 
The main problem with epidemiological (observational) studies consists of confounding factors, i.e. factors that are associated with the dietary factor being examined and that are not being taken into considerations, or that cannot be fully adjusted for. A prudent lifestyle has many characteristics and some of them may not have any real health impact. Thus, confounding factors are not causal factors, they are only statistically related to the disease in question.
 
In molecular biology and physiology, the number of substances involved is unimaginably large and new substances are constantly being discovered. A vast number of yet undiscovered molecules are expected to have a large impact on human physiology. Mechanistic models in textbooks are strongly oversimplified and thereby potentially misleading.
 
The ideal situation would be to conduct a randomized controlled trial where the intervention consisted of a group of people who altered their diet in a predetermined way, but not in any other way. The problem is that the test subjects (the intervention group) often simultaneously improve their lifestyle in other respects. Unlike drugs, lifestyle changes cannot be studied in a double-blind manner.
 
Another potential source of errors can arise if studies with positive effects are published to a greater extent than those with absent or negative effects, making the intervention in question appear unduly positive (publication bias). More than 7,000 articles related to nutrition are published annually which makes it difficult for nutritional experts to get a complete overview.
 
An important source of inspiration for researchers themselves is the hope that their own hypotheses will be confirmed. It has long been known that this hope can affect the interpretation of study results, and obviously financial interest can amplify the problem. A recent study analysed financial sponsorship, or lack thereof, with regard to research articles on the health effects of soft drinks, juice, and milk in order to determine how sponsorship affects published conclusions [1]. A total of 206 articles were included in the study, of which 111 declared financial sponsorship. Of these, 22% had all industry funding, 47% had no industry funding, and 32% had mixed funding. Funding source was significantly related to conclusions when considering all article types (p=0.037). For interventional studies, the proportion with unfavorable conclusions was 0% for all industry funding versus 37% for no industry funding (p=0.009). Thus, 100% of industry-sponsored studies came to the conclusion that the beverage under investigation was beneficial to health.
 
One common effect of preconceived ideas is their tendency to direct the scientist’s focus towards those dietary aspect that are considered interesting in a trial, such as the proportion of fat/protein/carbohydrate or the amount of some other nutrient. An intervention diet based on meat, fish, vegetables and fruit may, depending on your particular belief system, be described as low glycemic-index, high-protein, low-starch, high-omega-3, low-fat, high-potassium, low-casein, high-folate, low-lectin, or numerous other options.
 
The uncertainty in nutritional science is considerable. The new approach of evidence-based medicine reveals these shortcomings of nutritional science. When nutritional recommendations are called ’evidence based’, the low grade of evidence behind dietary advice becomes obvious.
 
 
The basis of our opinion on diets
 
Science
epidemiology
physiology, molecular biology
randomised controlled trials (RCTs)
evolutionary biology
 
Other influences
belief systems
alliances, group dynamics, funding etc.
 
Ice cream? Placebo?