It is time to revisit current dietary recommendations for saturated fat

Although there is no absolute consensus on the recommendation for total fat and SFA intake between governing bodies and health organizations, there is a general sense of convergence (see Supplementary Table S1¹ ). All guidelines currently suggest that total fat should not exceed 35% of daily calories. Although most guidelines propose a target for dietary SFA, there is no consensus on the value to aim for. European agencies propose that SFA intake (as well as trans-fat) be “as low as possible” (EFSA Panel on Dietetic Products, Nutrition and Allergies (NDA) 2010), while there is no specific target for SFA in Canada’s Food guide (Health Canada 2010). Two American advisory committees/agencies propose that dietary SFA should be lower than 7% of daily calories (Unites States Department of Agriculture (USDA) and United States Department of Health and Human Services 2010; Eckel et al. 2014.). Finally, the World Health Organization suggests that SFA should account for less than 10% of daily calories (Elmadfa and Kornsteiner 2009).

The first American food guide was defined in 1916. It included 5 “food groups” ((i) meat, milk, other; (ii) cereals, other; (iii) vegetable, fruits; (iv) fatty foods; (v) sugars) and suggested that dietary fats should constitute about 30% of daily calories (Hunt 1916). These recommendations were of course based on rather primitive scientific evidence, as nutrition research in those days had not yet bloomed to its current status.

The 1950s and the 1960s have produced the first of several determinant moments in nutrition research. The Seven Countries Study has provided the first evidence suggesting that dietary SFA intake increases the risk of coronary death (Keys 1957). It revealed that areas such as Crete and other Mediterranean countries with dietary SFA intake corresponding to less than 7% of calories had a very low rate of coronary death (Aravanis et al. 1970). Although criticized for not having used all data available to them at the time of analysis (Yerushalmy and Hilleboe 1957), the pioneering studies by Keys et al. had an immense impact on identifying SFA as the villain in subsequent research efforts.

In 1977, the US Senate Select Committee on Health and Human Needs published 2 editions of a report entitled Dietary Goals for the United States (US Senate Committee 1977b, 1977a). It was argued that the research and health professional communities were relatively unanimous in identifying intake of fat and SFA as 2 of the most important nutritional risk factors for chronic diseases, including cardiovascular disease (CVD). The US Senate Select Committee proposed specific dietary goals, including “… reducing overall fat consumption from 40% of calories to about 30%, and reducing saturated fat consumption from 16% of calories to about 10%.” The uncertainties surrounding the proposed targets for total fat and SFA are obvious, as emphasized by the wording of the recommendations.

The USDA in 1977 did not agree with the US Senate Committee’s position and argued that there was no absolute scientific proof of the danger and risk posed by dietary fat and SFA (Gifford 2002). USDA through the Dietary Guidelines for American Committee (DGAC) released their first dietary guidelines for Americans in 1980, with revisions every 5 years thereafter. Before 1985, recommendations regarding total dietary fat and SFA were written as “Avoid too much”, with no specific targets (Gifford 2002). But the legacy of the senatorial report was so strong that even DGAC in its 1990 revision of their dietary recommendations “gave in” and identified for the first time the ≤30% of energy target for total fat and ≤10% of energy for SFA. In 2005, DGAC changed their recommendations for total fat to 20%–35% for SFA to less than 7%.

In their 2010 revision of its dietary guidelines, DGAC proposed for the first time that the various SFAs should be considered differently according to their impact on blood lipids. Their recommendation is that stearic acid (C18:0) should not be categorized as a cholesterol-raising fatty acid, unlike lauric (C12:0), myristic (C14:0), and palmitic (C16:0) acids and industrially produced trans-fatty acids. DGAC also indicated in its 2010 report that setting the recommended percent of energy from these cholesterol-raising fats to less than 5% to 7% will help maintain blood cholesterol at desirable concentrations (USDA and US Department of Health and Human Services 2010). Modelling of current food patterns indicated that if all solid fats were removed and isocalorically replaced with oils, total SFA would be decreased to 7.0%–7.5% of calories and the cholesterol-raising fatty acids would be decreased to 5.0%–5.5% of calories. This was the rationale for proposing the “7% or less” cut-off point for SFA intake. Of course, this is a highly theoretical scenario since SFA cannot be fully substituted by PUFAs because these nutrients come as part of very different foods, with each having different nutrient profiles.

The recent 2010 DGAC report has been severely criticized for having included an incomplete body of relevant science, for inaccurately representing, interpreting, or summarizing the literature, and for drawing conclusions and (or) making recommendations that do not reflect the limitations or controversies in the science (Hite et al. 2010; Hoenselaar 2012). Thus, it appears that even the interpretation of the current body of knowledge on SFA and CVD risk poses a challenge within the scientific community. The next section provides a brief overview of some of the available evidence linking dietary SFA to CVD.

Most arguments advocating a lowering of SFA in the diet are based on the undisputable low-density lipoprotein (LDL)-raising effects of SFA. Specifically, meta-analysis of available clinical trials published prior to 2003 has shown that for each 5% of energy substitution of carbohydrate for SFA, plasma LDL-cholesterol increases in average by approximately 5% (Mensink et al. 2003), which in turn predicts a 5% increase in the risk of coronary heart disease (CHD). The Institute of Medicine (IOM) indicates that such extrapolations must be interpreted with caution (Committee on Qualifications of Biomarkers and Surrogate Endpoints in Chronic Disease (IOM) 2010), and that the effects of food on LDL-cholesterol should not be used alone to predict a change in CHD/CVD risk. Whether a diet-induced change in plasma LDL-cholesterol has the same cardiovascular benefit as a statin-induced change in LDL-cholesterol is unknown (Committee on Qualifications of Biomarkers and Surrogate Endpoints in Chronic Disease (IOM) 2010).

This is an important consideration in light of the fact that SFA may affect other plasma lipid risk factors in a neutral and perhaps even a beneficial way. The meta-analysis by Mensink et al. (2003) has shown that most dietary SFA increase plasma high-density lipoprotein (HDL)-cholesterol concentrations compared with dietary carbohydrate, with therefore little impact on the total cholesterol/HDL-cholesterol ratio, a powerful lipid risk factor for CVD. The extent to which the increase in HDL-cholesterol counterbalances the LDL-raising effects of SFA compared with dietary carbohydrates is unclear and certainly deserves more scrutiny. Accumulating evidence suggest that apoB, which reflects the number of atherogenic particles in the blood (not just LDL but also very low-density lipoprotein), may be superior to LDL-cholesterol in predicting the risk of CVD (Barter et al. 2006). We have shown that individuals with increased number of small dense LDL, as denoted by high plasma apoB concentrations and LDL of reduced size, are at high risk of CHD irrespective of their plasma LDL-cholesterol concentrations (Lamarche et al. 1997). Data suggest that replacing dietary carbohydrates by SFA has no significant impact on apoB concentrations (Mensink et al. 2003). Dietary SFA has also been shown to increase the size of LDL particles compared with other nutrients (Desroches and Lamarche 2004). SFA may reduce plasma triglyceride concentrations at least compared with dietary carbohydrates (Dreon et al. 1994). Data from a large multi-center intervention study suggest that isoenergetic replacement of SFAs with monounsaturated fatty acids (MUFAs) or carbohydrates has no deleterious effect on insulin sensitivity (Jebb et al. 2010), a central component of metabolic syndrome and CVD risk. Replacement of SFAs with MUFAs or carbohydrates in healthy subjects has also been shown to have no effect on vascular function (Sanders et al. 2013). SFAs are not susceptible to lipid peroxidation, and are therefore not considered to be involved in this process (Lawrence 2013). Finally, the impact of SFA on inflammation remains unclear and further research on this topic is required (Calder et al. 2011).

We stress again that this whole body of research has essentially been ignored in the analysis of existing data on dietary SFA to support most current dietary guidelines for the prevention of CVD.

Systematic reviews of the available epidemiological data have also fuelled the controversy on the association between dietary SFA and CVD risk. The meta-analysis by Siri-Tarino et al. concluded that higher intakes of SFA were not associated with an increased risk of CHD, stroke, or CVD (Siri-Tarino et al. 2010). As is often the case, the paper has been criticized for over-interpreting the data and for having excluded important studies on the topic (Scarborough et al. 2010). Another meta-analysis of prospective observational cohort studies has shown, based on mathematical modelling of the dietary data, that for each 5% lower energy intake from SFA being replaced by PUFA there was a significant 13% reduction in CHD risk and 26% reduction in coronary deaths (Jakobsen et al. 2009). Mozaffarian et al. have meta-analyzed 8 randomized clinical trials (RCTs) documenting the impact of increasing PUFA in place of SFA on CHD risk (Mozaffarian et al. 2010). Data from this analysis suggested that each 5% energy increase from PUFA reduced the occurrence of CHD events by 10% (Mozaffarian et al. 2010). However, PUFA in these RCTs replaced not only SFA but also trans-fatty acids from common “hard” margarines and industrial shortenings. This important confounding role of trans-fatty acid in the analysis has not been fully appreciated (Ramsden et al. 2010). Finally, combining recovered data from the Sydney Diet Heart Study to another meta-analysis of existing data on this topic (Ramsden et al. 2010) showed that selectively increasing n-6 LA in place of SFA, without a concurrent increase in n-3 PUFA, elevates the risk of coronary death, particularly in a secondary prevention setting (Ramsden et al. 2013).

In sum, while the impact of dietary SFA on blood cholesterol is undisputable, the resulting impact on CVD risk is not as straightforward with data giving place to very discordant interpretations.

Dietary recommendations have changed little over the years and the early focus on SFA remains omnipresent in most guidelines. Based on recent controversial and inconsistent evidence from epidemiological and intervention studies, it seems reasonable to revisit this SFA–CVD scheme (Astrup et al. 2011). We believe that evidence regarding the impact of dietary SFA on CVD risk factors other than LDL-cholesterol and evidence regarding whole foods rather than just SFA, both from clinical as well as epidemiological perspectives, need to be considered in the future. Because we believe that focussing on SFA in dietary guidelines may not have yielded full benefits in terms of cardiovascular prevention, considering the impact of individual SFA on CVD risk may also be an inappropriate route to undertake in the future. There are currently no harmonized dietary guidelines in Canada for the prevention of CVD. We therefore urge the Canadian nutrition community to be proactive in this process and to hopefully arrive at a consensus that will reflect both our own food reality and the latest science. Perhaps in the end the conclusions will remain that reducing SFA intake per se is indeed a justified dietary target. But at least we will have a sense of having gone through the process with an open mind, using the totality of the evidence. And perhaps the conclusion will be that reducing SFA intake should no longer be a target to advocate in future dietary guidelines.

Drs. Lamarche and Couture have received research grants from the Agriculture and Agri-Food Canada Dairy Cluster. Dr. Lamarche has received research funding from the Danone Institute, Atrium Innovations, and Merck Canada, and speaking honoraria from Unilever, Danone, and the Dairy Farmers of Canada. Dr. Lamarche is Laval University’s Chair in Nutrition, supported in part by Provigo/Loblaws, Pfizer, and la Banque Royale du Canada.