Total Pageviews

Tuesday, June 9, 2015

Real nutritional science re-affirms advice to lower dietary saturates

In the last few years there has been a flurry of scientific papers published showing no apparent association between deaths from cardiovascular disease and the consumption of dairy products leading to an amazing level of interest in the traditional and the social media in these new revolutionary findings. “Butter is back”, wrote the New York Times while Time Magazine featured a cover story headlining “Eat butter – the scientists labeled fat the enemy. Why they were wrong”. The Wall Street Journal ran a feature entitled “The questionable link between saturated fatty acids and heart disease. The meat, dairy and egg industries had a field day. However, little ol’ me pointed out in a blog entitled “Fats, facts and baloney[1]”, that these scientific assertions were based on a re-analysis of all existing data that examined the association between consumption of dairy foods, saturated fats and heart disease rates. Let me stress the word “association”. Epidemiological studies examine some aspect of lifestyle such as diet and some putative link to that lifestyle and seek to explore whether the link is positive, negative or otherwise on the basis of statistical analyses. No matter what statistically significant effect is seen and no matter how often it is repeated across time and space, it remains an association. A causal effect of the lifestyle factor and the putative end point of that lifestyle has not been proven. In certain areas of epidemiology, it is impossible to follow this association up with a human intervention study to prove the cause-effect association. Thus when epidemiologists studied the link between cancer and tobacco, experiments in humans were out of the question. To that end, a famous English epidemiologist, Sir Austin Bradford-Hill set out a list of criteria that should be considered in perusing epidemiological data to ensure that any conclusion drawn on such data, are, as far as is humanly possible, plausible and actionable. I have never seen or read of any nutritional epidemiologist citing the Bradford Hill criteria.

Now what makes nutritional issues different from the tobacco issue is that we can and do large human intervention studies in which one group has their food intake or nutrient intake manipulated, while another serves as a control. Examples include studies of diet and cancer and cardiovascular disease in women[2], studies of the Mediterranean diet[3] or studies of dietary lipids and the metabolic syndrome[4] to mention just a few.

In recent months, two papers have appeared from UK researchers, in which the level of dietary fat and the type of dietary fat were manipulated[5], [6]. The first in volved195 subjects aged 21 to 60 years with moderate risk of cardiovascular disease. They were randomly allocated to one of three intervention diets for 16 weeks in which the ratio of saturated to monounsaturated to polyunsaturated fats varied: 17, 11, 4 versus 9, 19, 4 versus 9, 13, 10 to provide diets high in saturates, monounsaturates or polyunsaturates. Replacing saturates with monounsaturates or polyunsaturates resulted in statistically significant falls in plasma cholesterol both total and LDL cholesterol (about 9% and 12% respectively). This was calculated to translate into a reduction on cardiovascular disease mortality by up to 20%. The second study involved 165 healthy adults assigned to a standard British diet or to a diet that attempted to meet the UK dietary guidelines.  The group assigned to the dietary guidelines arm of the treatment showed a reduction in the intakes of total, saturated and trans fat intakes with a partial substitution with increased intakes of polyunsaturates but especially, monounsaturated fats. Total and LDL cholesterol fell significantly over the 12 weeks in the group assigned to the dietary guidelines. The estimated overall reduction in heart disease mortality was between 15 and 30%.

Now it matters nothing as to how many epidemiologists dance on a decimal point, the facts remain that adhering to existing dietary guidelines on the composition of dietary fat, which have changed rather little in the last 50 years, will significantly reduce heart disease death rates. These data are experimental proofs of what was already proven many times many years ago. However, in an era where epidemiology and its statistical gymnastics rule, it was necessary once again, to show that the adherence to prevailing dietary guidelines is justified and will save lives. These papers won’t hit the Wall Street Journal, Time Magazine or the New York Times. The general public will never hear about them. I happen to care, hence another rant!



[2] http://whs.bwh.harvard.edu
[3] Estruch R et al (2013) N Eng J Med, 368, 1279-1290
[4] http://www.ucd.ie/lipgene/
[5] Vafeiadou K et al (2015) Am J Clin Nutr, e pub ahead of print
[6] Reidlinger DP et al (2015) Am J Clin Nutr, 101, 922-930