This year, the gloves have come off in the arena of disputes within lipid metabolism. No longer just slight hints of disdain and micro-correction among genteel academics at learned conferences, the battles now rage in shouty blogs, press headlines and anecdote-filled books.
The new message, the one that is very much counter to current nutrition policy and firm dietetic advice and now cast-in-concrete food labelling laws, is that saturated fat is not bad at all; cream is not ‘naughty but nice’ - it is just nice.
Should dietitians shift the long-steady rudder of dietary advice on saturates in relation to blood cholesterol levels and heart disease risk?
A year ago (October 2013), an opinion piece written by cardiologist Dr Aseem Malhotra, was published in the British Medical Journal. The article: ‘From the Heart: saturated fat is not the major issue - let’s bust the myth of its role in heart disease’, was very widely discussed in media. Then, in March 2014, a bombshell paper was published in the Annals of Internal Medicine. Lead author Rajiv Chowdhury is a cardiovascular epidemiologist at the University of Cambridge, and major funding for the study came from the British Heart Foundation and the Medical Research Council. Impeccable expertise went into the study; explosive results came out.
Chowdhury and colleagues did a systematic review and meta-analysis of published studies reporting dietary, circulating or supplement-source fatty acids and the risk of coronary disease. There were 32 studies with data on fatty acids from dietary intakes (from more than 510,000 participants), 17 studies with data of fatty acid biomarkers (from more than 25,000 participants) and 27 randomised controlled trials with data on fatty acid supplementation (from more than 105,000 participants).
Clarity on trans fats
Coronary outcomes in prospective cohort studies of dietary intakes showed a significant 16 percent increase in risk in people who consumed the top third of intakes of trans fatty acids compared to those with the bottom third of intakes. So there is clarity: higher intakes of dietary source trans fats increase the risk of coronary outcomes. However, the other observations were less consistent with expectations. The top versus bottom tertiles of saturated fatty acid intakes showed a small three percent increase in risk, and only very small reductions in risk of one percent and two percent could be observed for intakes of alpha linolenic and total n-6 fatty acids.
Long chain n-3 fatty acids were protective (a 13 percent risk reduction in top tertile intakes), so fish eating remains a good idea. In contrast, variations in intakes of monounsaturated fatty acids showed zero effects on coronary events, so should olive oil messaging be muted?
Looking at perhaps more accurate data from circulating fatty acids (in contrast to perhaps fuzzy food diary descriptions), another picture emerged. While total saturated fatty acids increased the risk of coronary outcomes by six percent, further breakdown by individual fatty acids showed the extremes of a 23 percent increase with stearic fatty acid (18:0), in contrast to a 33 percent reduction with margaric fatty acid (17:0).
Trans fats appear minimally bad (five percent increase), total monounsaturated fatty acids appear even worse (six percent increase), while total n-3 and n-6 appear mildly protective, with reductions of seven percent and six percent in the risk of a coronary outcome.
What about fatty acid supplements in relation to the risk of a coronary event? Perhaps obviously, there are no studies where participants are asked to take supplements with saturated fatty acids. Studies of alpha-linolenic acid supplements (n-3) show risk reductions of three percent long-chain n-3 supplements protect by six percent, and the surprise of more potent effects was that supplementary intakes with n-6 fatty acids protect by 14 percent.
In relation to diet and prospective risks of coronary disease, Chowdhury and colleagues conclude essentially ‘no effects’ with saturates or with n-6 polyunsaturates, or with monounsaturates, but some lower risk with dietary n-3 polyunsaturates.
Their final statement is, ‘current evidence does not clearly support cardiovascular guidelines that encourage high consumption of polyunsaturated fatty acids and low consumption of total saturated fats.’
Challenges and critiques
There were some academic responses to the Chowdhury review and some corrections were made post publication. The most immediate critique came the epidemiologists at the Harvard School of Public Health (although one of the HSPH staff, Dariush Mozaffarian, was in fact also a co-author of the Chowdhury review).
Professor Walter Willett and colleagues challenged the lack of effect described for n-6 fatty acids, and specifically corrected some of the data used in the review. Major studies that did show significant inverse associations between intakes of polyunsaturated fat intake (mainly as n-6) and the risk of coronary disease had not been included by Chowdhury. Further, Willett and colleagues also stated that most of the monounsaturated fat consumed in the studies were from red meat and dairy sources, and that the findings might not apply for analysis of plant source monounsaturates. The Harvard experts state, “the conclusions…regarding the type of fat being unimportant are seriously misleading and should be disregarded.” It is striking, however, that Willett and colleagues make no comment on the observations by Chowdhury of ‘no effect’ for intakes of saturates.
Other critiques, reported in Science magazine by K Kupferschmidt, are that diets replacing saturated fats with carbohydrates, and more dramatically diets with lower intakes of monounsaturates and polyunsaturates with higher energy intakes from carbohydrates, have been shown to increase the risk of coronary disease.
Professor Mozaffarian, the man in the uncomfortable position of straddling two stools, as both author and critic of his own paper, stated that he was not happy with the conclusions of the paper about polyunsaturated fats (but he supported the no effects for mono and saturated fats). Less impressive responses by the University of Cambridge researchers to the general excitement about the paper, was that the main problem had been that the paper had been, “wrongly interpreted by the media” (?), and that “more good trials were needed”.
Call for better nutritional studies
In fact, the paper is perhaps a general call to consider the limits of meta-analysis in relation to the population assessment of dietary data. Walter Willett was concerned that while drug trials are often a similar design, so it is possible to combine results, this was not true for nutritional studies, which vary widely in how they are set up. “Often strengths and weaknesses of individual studies get lost.”
A similar concern was expressed by Professor Bruce Griffin of the University of Surrey, to a London meeting of the Guild of Health Writers in September 2014. Many of the individual studies included in the review by Chowdhury and colleagues showed clear effects of fatty acids on coronary risk, but these were lost when data were merged. He illustrated the concept with the discrete traits of items within a fruit bowl, which become less distinct and identifiable when the same items are mashed into a blended salad or smoothie.
Fats, not carbs
About half of adults in the UK have elevated blood cholesterol levels, above 5.2 mmol/L. Current dietary intakes of saturated fats in the UK diet (as reported in the four year rolling programme of the National Diet and Nutrition Survey) are about 12.0-13.3% of energy, which is higher than the recommended amounts of no more than 10 percent of energy. Dietary fat modification appears to result in more successful outcomes in preventing cardiovascular disease than fat reductions (Hooper, 2011) and, while saturates may be less ‘bad’ than current dietary guidelines suggest, data supports the inclusion of monounsaturates and polyunsaturates in the diet (rather than some replacement of fats with carbohydrates).
Possible future reviews of dietary public health recommendations on saturated fats will certainly consider current evidence, and dietitians will be the authoritative channels to communicate the up-to-date guidance, whether constant or changed. However, a problem of today for all health professionals, is the occasional clash between the near-daily outcomes of latest published evidence and the long-developed guidelines and policies that guide medical advice.
This article originally appeared in Network Health Dietitians magazine in November 2014.