‘State-of-the-Art’ Review Sows Confusion About Saturated Fats

Background

Recently, The Journal of the American College of Cardiology (JACC) published a state-of-the-art review entitled, “Saturated Fats and Health: A Reassessment and Proposal for Food-Based Recommendations.” The authors presented the argument that current evidence does not support the dietary guidelines recommendations to limit saturated fatty acid (SFA) intake to less than 10% of daily calories to prevent common chronic diseases, most notably cardiovascular disease (CVD). They remark that although SFAs raise low-density lipoprotein cholesterol (LDL), this is due to increases in large (and not smaller, dense) particles, which are not as strongly related to CVD risk. Further, they exclaim that not all fatty acid sources impart similar health effects due to differences in their overall structure and the complex matrices of the given food. They explicitly emphasize that dark chocolate, unprocessed meat, and whole-fat dairy are not associated with CVD risk and recommendations to limit their intake solely due to their SFA content are unsubstantiated1. Although some sentiments they offer are agreeable, and despite their claims otherwise, the totality of available evidence does not support many of their arguments. Accordingly, their publication will likely cause more confusion to the American population already struggling to decide whom they can trust among the disordered field of nutritional science, resulting in more harm than benefit.

Key Takeaways

1. Authors of this JACC State of the Art review state that the United States’ Dietary Guideline recommendations to reduce saturated fat are not justified due to the existence of evidence suggesting it may not increase the risk of CVD.

• The trials and meta-analyses they discuss are laden with critical flaws that go overlooked, including adjustment for a moderator variable (LDL cholesterol), a small overall variance in saturated fat intake, inclusion of multiple cohorts with intakes predominately above or below the threshold where the majority of the increase in CVD risk is observed, failure to disclose review protocols, and inclusion of only CVD mortality metrics.
• Consistent evidence from dozens of RCTs and meta-analyses of prospective cohorts including millions of participants has confirmed the role of saturated fat in increasing CVD risk, and the benefit of replacing it with PUFAs and high quality carbohydrates.

2. It is repeatedly mentioned that replacing dietary saturated fat with carbohydrates does not decrease the risk of cardiovascular disease, and may even increase it, along with the risk of metabolic syndrome. Subsequently, it is implied that carbohydrates are of greater concern with respect to CVD and type 2 diabetes.

• Despite continually admonishing the creators of the dietary guidelines for failing to distinguish between nutrients, the foods that provide them, and the corresponding impacts they have on health, the authors do exactly that with these remarks by choosing not to mention these associations are restrained to refined carbohydrates, and mostly in the context of diets also high in saturated fats, trans fats, and cholesterol.
• This rhetoric inappropriately instills a fear of carbohydrates in the reader due to their perceived negative impact on health. Such claims are highly misleading given that many foods rich in carbohydrates, such as legumes, whole grains, fruits, and vegetables, offer a myriad of health benefits and reduce the risk of CVD, metabolic syndrome, diabetes, hypertension, and many other common diseases.
• Current recommendations clearly emphasize replacing saturated fat with PUFAs, MUFAs, and/or unrefined carbohydrates, so this discussion is tangential and unproductive.

3. Authors postulate that reduction of LDL through diet cannot be inferred to result in CVD benefit with having the means to assess other biological effects that accompany it. They then bring up cases where increases or decreases in LDL aren’t reflected in the corresponding changes in risks of CVD in an attempt to strengthen their case. Finally, they suggest that despite these points, saturated fat reduction doesn’t decrease the concentration of plasma sdLDL, which has a stronger association with CVD, and therefore can’t be assumed to reduce risk.

• Irrespective of having a means to assess the biological effects coinciding with diet driven reductions in LDL, or that saturated fat reduction doesn’t decrease sdLDL concentrations, the fact remains that it has clearly been shown to elicit strong and significant reductions in the risk of CVD.
• Most of the exceptions the authors present to strengthen their argument (CETP inhibitors, progestin and estrogen therapy, and Mediterranean diet interventions) are actually characterized by changes (or lack thereof) in CVD risk consistent with those seen in LDL, and others are explained by observed changes in other metrics known to influence risk.
• While small, dense LDL can correlate strongly with increased CVD in univariate analysis, it typically fails to maintain predictive power after multivariate adjustment for triglycerides and HDL. This is likely because it is part of a broader pathophysiology (e.g. high triglycerides, low HDL cholesterol, increased LDL particle number, obesity, insulin resistance, diabetes, metabolic syndrome) that accelerates atherosclerosis, not due to it have a greater intrinsic ability to increase the risk of cardiovascular disease.

4. The JACC review briefly discusses the role of dairy and meat from domesticated animals in human diets of the past, how many plant sources oils are relatively new with respect to these, and that process contaminants in these oils may elicit detrimental effects on health. It also mentions that the scientific community has failed to discover the “one true diet” to optimize metabolic health for all, that results of dietary intervention trials are inconsistent, and that genetics may modulate the relationship between saturated fat and certain health outcomes.

• Aside from not providing any, or weak animal and mechanistic evidence for these points, many are fallacious and incredibly inappropriate in the context of determining population-level nutritional recommendations for supporting a healthy life.
• The evidence on what constitutes a healthy dietary pattern is fairly strong and consistent, with most heterogeneity in outcomes from intervention trials being explained by differences in macronutrient quality, adherence, age, baseline, disease, and other well-acknowledged factors.
• Genetic differences in responses to saturated fat discussed by the authors here don’t provide any rationale for dismissing its role in disease risks. If anything, it emphasises that some groups may benefit to an even greater extent than others.

5. They exclaim that the nutrients, food matrix, and chemical structure of fatty acids found in particular foods, such as full-fat dairy, unprocessed red meat, and dark chocolate, outweigh their saturated fat content and explain their supposed neutral or beneficial associations with CVD.

• Except for dark chocolate (in moderate quantities), they fail to provide any evidence suggesting these characteristics meaningfully change their relationship with CVD/type 2 diabetes.
• The evidence given for full-fat cheese and yogurt is riddled with problematic methodology, including adjustments for serum cholesterol, lack of consideration for comparators, and populations with low mean intakes that prevent meaningful observations of the resulting CVD risk from being made.
• Isocaloric substitution of dairy fats with omega 6 fatty acids, alpha-linoleic acid, marine omega 3 fatty acids, and carbohydrates from whole grains significantly reduce CVD and stroke risk.
• In accordance with these findings, controlled trials on cheese indicate it may increase LDL slightly less than butter, but only in those with baseline high values. However, it significantly increases LDL when compared to tofu, reduced-fat cheese, and egg whites. Trials specifically on yogurt are lacking, but those involving an increase in full-fat yogurt, cheese, and milk consumption reveal an LDL cholesterol-raising effect.
• RCTs on red meat demonstrate a similar cholesterol-raising effect compared to lower saturated fat alternatives. Numerous meta-analyses and large pooled cohorts repeatedly demonstrate it significantly elevates CVD, type 2 diabetes, colorectal cancer, and stroke risk.

6. Furthermore, they posit that avoidance of these foods resulting from dietary recommendations to limit saturated fat intake will result in reduced intakes of important nutrients, other potentially beneficial components, and the ostensible reductions in disease risk, as well as elimination of important low cost, nutrient-dense foods to malnourished and elderly populations.

• Lower fat alternatives to yogurt/cheese and unprocessed meat contain similar (and in some cases even greater) amounts of the nutrients the authors suggest individuals will be missing out on in the full-fat counterparts, so the importance of this point is questionable.
• Additionally, there are numerous other foods, such as legumes, whole grains, tubers, nuts/seeds, and fish, that would offer a broad variety of nutrients and confer substantial protection against many of the most common chronic diseases rampant in current day societies.
• The dietary guidelines aren’t directed towards subgroups such as the elderly and malnourished, again leaving one questioning why such remarks are being made. Regardless, the aforementioned foods could serve a similar role as cheap, nutrient-dense staples, and evidence suggests such patterns are associated with beneficial impacts on the health of the elderly.

7. Finally, authors conclude previous guidelines have a saturated fat bias, and that new guidelines should inform the public that foods rich in saturated fat may play a role in reaching nutritional recommendations, acknowledge that low-carbohydrate diets rich in saturated fat may improve metabolic disease endpoints, and should shift from the current paradigm that emphasizes the saturated fat content of foods as key for health towards one that focuses on healthy dietary patterns appropriate for different cultures and populations.

• None of the evidence put forth by the review substantiates the existence of a “bias” against saturated fat, or that most of the foods they claim are neutral with respect to their impact on the risk of CVD and T2D.
• No nutrients found in full-fat dairy, unprocessed red meat, or chocolate that can’t be sourced from lower fat alternatives or other health promoting foods are even discussed.
• In addition to the reality that a high saturated fat intake is not a necessity for a lower carbohydrate diet, the authors only cite a single short term trial that suggests they possess any sort of a beneficial effect on metabolic disease endpoints, hardly enough to even spark interest, and nowhere near that required to justify incorporating into population-wide guidelines.
• Even though the review incessantly attacks the dietary guidelines for giving advice centered only on nutrients, the reality is that they primarily emphasize exactly what the authors are demanding; specific foods and food groups that constitute a healthy dietary pattern. Only in addition to these do they offer specific nutrient recommendations that are based on the totality of the high-quality evidence, not an inherent bias.

The Origin of Recommendations to Reduce Saturated Fat Intake

The review begins by briefly discussing the history behind the initiation of dietary goals and recommendations to lower saturated fat intake dating back to the 1970s. It details that since the 80s, there have been specific goals of reducing saturated fat intake to less than 10% of total calories to reduce CVD risk. The authors declare that their main intention is to answer the question posed by the United States Department of Agriculture and Health and Human Services’ in 2018; “What is the relationship between saturated fat consumption (types and amounts) and the risk of CVD in adults?” by reviewing the effects of saturated fats on health outcomes, risk factors, and mechanisms underlying CVD and metabolic outcomes. The question is whether the review provides a sufficient answer that is backed by a substantial body of evidence, which will become quickly apparent is far from the case.

The following paragraph states: “The relationship between dietary SFAs and heart disease has been studied in about 400,000 people and summarized in several systematic reviews of observational studies and randomized controlled trials (RCTs). Some meta-analyses find no evidence that reduction in saturated fat consumption may reduce CVD incidence or mortality (3–6), whereas others report a significant — albeit mild — beneficial effect (7,8). Based on this collection of research, they conclude that the basis for recommending a low saturated fat diet is unclear and that they intend to propose an evidence-based recommendation for intake of SFA food sources. Right from the start, the authors have demonstrated that they are not genuinely reaching their conclusions by considering the totality of the evidence. They mention five publications designed to observe the relationship between saturated fat and CVD that contain about 400,000 subjects. In reality, dozens of systematic reviews and meta-analyses of observational studies and RCTs encompassing millions of people indicate that reducing or replacing saturated fat with other nutrients results in significantly decreased CVD morbidity or mortality (8–19). Recent publications on this relationship include over double the number of participants they imply it has been studied in (16,18). While there are also a few additional reviews suggesting that it has no observable effect (20,21), along with those that the authors of the JACC review mention (2–5), they involve critical flaws that impair their ability to effectively assess the relationship between SFAs and CVD morbidity/mortality, and are heavily criticized. Following is a discussion of each of these trials, their respective pitfalls, and additional comments from other parties.

Inconsistent Results: A Cause for Confusion?

Harcombe et al., de Souza et al., Siri-Tarino et al, and Zhu et al.

Reviews on prospective cohort studies and case-control assessing the relationship between saturated fat intake and CVD by Harcombe et al., de Souza et al., Siri-Tarino et al., and Zhu et al. fall victim to the same major methodological problem. This is the inclusion of an appreciable amount of cohorts that adjusted for serum cholesterol or baseline hypercholesterolemia (5/10, 1/3, and 5/11 cohorts in de Souza for total CHD incidence, CVD mortality, and CHD mortality, 7/16 and 4/8 in Siri-Tarino for CHD and stroke events, 3/6 in Harcombe et al., and 14/40 and 7/22 in Zhu et al. for CHD in the highest vs lowest comparison and dose-response analysis, respectively). The action of adjusting for a moderator variable (in this case LDL-c, which is increased by saturated fat and increases the risk of CVD) that lies on the causal chain of the outcome of interest pulls the results of the analyses towards null, creating a biased and inaccurate observation of the real relationship. Because nearly half of the cohorts in each of these reviews make such adjustments, any effect would be significantly muted. Scarborough et al. discuss this very issue in their comment responding to Siri-Tarino et al.’s authors closely following its publication (22). While there are additional issues that may also contribute to the lack of effect, such as little to no variance in saturated fat intake among the cohort’s sample population, the inclusion of multiple cohorts with intakes all above or below the threshold where the majority of the observed increase in CVD risk is detected, failure to disclose review protocols, and inclusion of only CVD mortality metric; just this alone is enough to invalidate the results of these reviews. Lastly, despite these over-adjustments, trans fat was still associated with significant increases with CHD mortality and CVD in Zhu et al. and de Souza et al., which will prove to be an important consideration concerning some of the other reviews.

Ramsden and Hamley

Apart from the four reviews just discussed, two additional reviews have suggested there may not be an association between saturated fat intake and CVD, one of which the authors of the JACC mentioned in their brief comments on the subject. Ramsden et al. and Steven Hamley carried out these reviews, both including randomized controlled trials focused on determining the potential benefit of replacing SFAs with mostly polyunsaturated fatty acids (PUFA). Ramsden et al’s review included discussion of recovered data from the Minnesota Coronary Experiment (MCE) and also carried out a meta-analysis of an additional 4 RCTs, the Sydney Diet Heart Study (SDHS), the Rose Corn Oil Trial (RCOT), the Los Angeles Veterans Trial (LA Vet), and the Medical Research Council Soy Oil Trial (MRC-Soy). They also performed a sensitivity analysis on the previous 5 in addition to 3 more, the Diet and Re-Infarction Trial (DART), the Oslo Diet-Heart Study (ODHS), and the St. Thomas Atherosclerosis Regression Study (STARS). Not only was the MCE flawed in numerous ways that prevented meaningful conclusions from being drawn, but there were also multiple issues with the other smaller trials included in their analysis (for which MCE ended up weighed the most). As for the MCE, the main problems were the insufficient power to detect effects on mortality due to the high dropout rate (>75%), utilization of likely trans-fat-containing margarine in the intervention group, the main difference in mortality being observed only in subjects over 65 years of age, and the lack of essential metrics such as smoking status, LDL cholesterol, detailed dietary intake data, weight loss, and coronary disease status. The smaller size (and weaker statistical power), exclusion of mortality as the primary endpoint, and inclusion of trans-fat-based margarine in the experimental group of another trial (SDHS) were further issues that rendered the findings of the analyses unuseful. As for Hamley’s review of RCTs, he chose to exclude individual trials based on “inadequate control,” and other factors that he posits would impact the results, including suspicions that the control group had a higher trans fat intake, were exposed to cardiotoxic medications and had lower vitamin E intake. His meta-analysis ended up incorporating the same trials as Ramsden et al. ‘s, with the one exception being LA VET, which he replaced with DART. Ironically, this means he included both MCE and SDHS, which exposed the intervention groups to higher trans fat intake, alongside other small studies underpowered to detect meaningful effects on the chosen endpoints.

2020 Cochrane Review

Directly contradicting both of these trials is the recent Cochrane Review on reducing saturated fat for cardiovascular disease. This publication was subject to far more rigorous pre-review protocols and analyzed the results of 15 RCTs (even including SDHS) to find that long-term reductions in saturated fat intake resulted in significant reductions in combined cardiovascular events (17). Furthermore, they conducted a meta-regression that demonstrated more significant reductions in saturated fat (and consequently, more significant reductions in cholesterol) led to greater events reductions. Unfortunately, no such regression was carried out for CVD mortality. However, in the subgroup analysis stratifying by absolute saturated fat reduction, a clear trend towards significant decreases in mortality can be observed, with the trial in which the most notable reduction was achieved reaching significance (Veterans Admin).

Taking into account the extensive body of high-quality evidence from RCTs and observational trials that demonstrated a beneficial effect of reducing saturated fat intake on CVD morbidity and mortality, and the methodological flaws in conflicting studies, it should be incredibly clear the author’s statement about a lack of clarity regarding the basis for reducing saturated fat intake is preposterous.

Chemical Structure of Saturated Fatty Acids

Next, the review moves on to discuss the variation in chemical structures of SFAs found in a wide variety of foods, explaining that SFAs vary based on their carbon chain length, their melting point, and state at room temperature (solid vs liquid), the presence or absence of methyl branches (branched vs straight-chain fatty acids), and their origin (exogenous vs endogenous). Furthermore, they exclaim that food sources of such SFAs contain different proportions of specific fatty acids and other nutrients that impact their physiological and biological effects. They continue, saying “Branched-chain SFAs are found primarily in dairy, beef, and other ruminant-derived foods (13), and have similar physicochemical properties as unsaturated fatty acids, in particular lower melting point (or more accurately, phase transition temperature). In experimental animal studies, branched-chain fatty acids alter the microbiota composition in the direction of microorganisms that use these fatty acids in cellular membranes (14), and because they are normal constituents of the healthy human infant gut (15), these fatty acids could play a role in normal colonization.” It seems odd for them to be focusing on mechanistic studies in animal models (given the variability in their relevance to humans) in order to insinuate some innate requirement for these fatty acids, especially given that they also bring up the ability of intestinal flora and the liver to synthesize them shortly after that. That being said, the comment is followed by a remark that fatty acids are synthesized de novo from “excess carbohydrate and protein” and reference to a study linking specific plasma phospholipid concentrations of fatty acids and adverse cardiovascular outcomes, seemingly hinting at the notion the process is inherently harmful. However, that may not be the case. They then re-emphasize the earlier point that different circulating saturated fatty acids exert differing effects on blood lipids, glucose-insulin homeostasis, insulin resistance, and diabetes. While this is certainly a valid point, it does not purely suggest such effects follow changes directly resulting from consumption of the respective fatty acids given the potential for de novo synthesis and various pathophysiological processes to exert effects on plasma concentrations. This section’s final paragraph raises concerns regarding the failure to distinguish between fat and fatty acids, saying that the former comprises fatty acids in differing proportions and other components such as glycerol. Another valid point, but not necessarily pertinent to any of the claims they have put forth so far.

The Effects of Saturated Fat on Health

Old Research, Shifting Dietary Patterns, and Lessons from Finland

The next section of the review directs its focus towards the evidence on saturated fat’s health effects. Like the earlier paragraph briefly discussing the matter, the authors make numerous claims that are extremely misleading and not backed by the majority of the scientific literature. They begin by offering up typical rhetoric about current guidelines being based upon observations from ecologic research (including the Seven Countries Study) throughout the 1950s on diet and coronary heart disease. Next, they exclaim, “In recent decades, however, diets have changed substantially in several regions of the world. For example, the very high intake of saturated fat in Finland has decreased considerably, with per capita butter consumption decreasing from ∼16 kg/year in 1955 to ∼3 kg/year in 2005, and the percent energy from saturated fat decreasing from ∼20% in 1982 to ∼12% in 2007 (28). Therefore, the dietary guidelines that were developed based on information from several decades ago may no longer be applicable.” First, the fact that diets have changed in recent decades is not reason to completely dismiss previously established guidelines. Even more astonishingly, the specific country they mention (Finland) is a prime example of evidence cutting directly against their position. The same article they reference explains the implementation of a widespread preventative health intervention originating in North Karelia (and later expanding to other regions in Finland) that aimed to modulate citizen’s dietary habits, mainly focusing on reducing saturated fat intake and replacing it with unsaturated fat, as well as increasing vegetable intake, and reducing sodium intake. This program’s results, which have been demonstrated to be attributed almost entirely to dietary changes, were that the working-age population experienced a considerable reduction in blood cholesterol levels and a remarkable 80% reduction in annual CVD mortality rates (23). However, the authors of the JACC review make no mention of this whatsoever.

Queue the War on Carbohydrates

After this, they bring up what they claim to be a few large and well-designed prospective cohorts carried out recently that supposedly demonstrate replacing saturated fat with carbohydrate does not result in a lower risk of coronary heart disease, and may increase mortality risk (8, 24,25). However, one of these is an editorial (Hu 2010) and echoes what the results from the other two (Jakobsen et al. 2009 and Liu et al. 2000) demonstrate, which is that replacing saturated fat with refined carbohydrate does not reduce, and may increase, the risk of CVD. Additionally, Jakobsen et al. showed that the replacement of SFAs with PUFAs elicited significant reductions in both CVD events and mortality. Hu’s editorial declared that increased intake of refined carbohydrates and SFA were independent risk factors for CVD. Moreover, although he believes that reducing refined carbohydrate intake may be most important, he states that low fat, high complex carbohydrate or moderately restricted carbohydrate diets rich in fat and protein from vegetable sources may confer protection against CVD. This is surprising given that the authors spent quite a bit of time discussing the importance of not looking at saturated fat as a single uniform nutrient, yet they are willing to do it for carbohydrates. Once again, this shows they are making misleading statements and providing studies that either do not relate to their position or invoke evidence contradicting it. They continue by saying, “Furthermore, a number of systematic reviews of cohort studies have shown no significant association between saturated fat intake and coronary artery disease or mortality, and some even suggested a lower risk of stroke with higher consumption of saturated fat (3,6,32,33).” The significant issues with three of these have been addressed previously (Zhu et al., de Souza et al., and Siri-Tarino et al.), and the fourth is on saturated fat intake and its potential protective effect against stroke (26), which is a separate consideration entirely. The authors end this paragraph by commenting on a meta-analysis of prospective cohorts demonstrating that biomarkers of long-chain SFAs in plasma or serum were not associated with CHD (27). It is entirely unclear why this was included, given that previous similar publications have noted that these SFAs in plasma are unlikely to result from diet (28). Not only that, only shortly after the authors claim that “it is important to distinguish between dietary saturated fat and serum SFAs”.

PURE and UK Biobank

Next, they refer to the PURE study, and although this study was an incredibly ambitious undertaking, it is unfortunately ridden with numerous critical flaws that dozens of researchers have discussed via comments responding to the original publication. As the JACC review’s authors discuss, 80% of the sample population was from low- and middle-income countries, which explains the origin of one of its most massive blunders; bias from malnutrition. In addition to focusing on lower-income countries, authors included energy intakes down to as low as 500 kcal/day. Higher carbohydrate consumption as a percentage of energy and the consequential lower percentage from fat was strongly correlated with low income, food availability, pollution, and healthcare access (29,30). Therefore, comparing increasing quintiles of fat consumption to the lowest quintile is essentially comparing to a likely malnourished population. Even just minor decreases in the incidence of death and disease would deem fat intake protective when, in reality, an unfair comparison is just skewing the results. Further issues are failure to report the type of food frequency questionnaire used, lack of baseline health status and adjustment, failure to distinguish between types of carbohydrates, and large differences in the data from PURE and China Health and Nutrition Survey (CHNS) on China’s fat intake (31–33). Finally, it seems the JACC review’s authors again selectively choose to showcase the observations they feel support their assertions and ignore those that do not, remarking that stroke incidence is reduced with increased saturated fat consumption. They fail to mention that rates of myocardial infarction are highest in the top quintile (the fact this does not factor in adjustments for confounders prior to calculation of an HR is unfortunate, and it would undoubtedly be of interest to see the resulting values). Further, mortality and incidence of all other observed outcomes were lowest in the third and fourth quintiles of saturated fat intake, which are unsurprisingly those where subject’s intakes were below 10% of their total caloric intake. In a demonstration of consistency, they then cite a recent prospective cohort of UK Biobank participants, exclaiming that it does not show an association between saturated fat and incident CVD, and that “Although there was also a positive relation of saturated fat intake with all-cause mortality, this became significant only with intakes well above average consumption.” While the former was shockingly correct, the latter was not, as the relation between saturated fat intake and all-cause mortality became significant at above around 12% of calories (Figure 1), with the average intake of the study’s sample being over 13% (seen in Table S2) (34). To follow that up, in a sentence mentioning the macronutrient distributions associated with the lowest hazard ratio for all-cause mortality, they conveniently leave out a few lines following what they quote, which happened to be, “…5–10% from SFA (2.66 v 3.59 per 1000 person-years, 0.67 (0.62 to 0.73) compared with high (20% of energy) intake)…”, clearly emphasizing the benefit of reducing saturated fat intake to 5–10% of calories. Lastly, they declare that for dietary carbohydrate, higher consumption (from starch and sugar) is associated with higher CVD and mortality, and there is little need to restrict intakes of total or saturated fat for most populations in the context of contemporary diets, where reducing refined carbohydrates may be more relevant for decreasing the risk of mortality in individuals with insulin resistance and type 2 diabetes. While their final point is an important consideration, these conditions and refined carbohydrates’ contribution to adverse health outcomes are not being ignored. Oddly, they do not recommend reductions in refined carbohydrates and saturated fat, which is strange when considering the pattern with the lowest hazard ratio for all-cause mortality involves precisely that. Lastly, the dietary guidelines their entire review has set out to criticize do indeed suggest that refined carbohydrate and saturated fat intake should be reduced, and highlight healthy food patterns that will help achieve such reductions (35), leaving one questioning what issues they have with them.

The Women’s Health Initiative, PREDIMED, and Faulty RCTs

Following the discussion of PURE and the UK Biobank data is another common talking point among skeptics of the diet-heart hypothesis; that the evidence for the dietary guidelines recommendations to reduce SFAs had important methodological flaws, one being that they were small in size. The authors then bring up the Women’s Health Initiative (WHI), one of the larger and more recent trials on reducing CVD via lifestyle intervention, and briefly remark that the low-fat diet providing 9.5% of calories from saturated fat did not reduce the risk of heart attack or stroke. Again, numerous relevant details are being left out. The major one is that based on the failure of subjects to reach the studies adherence assumptions (that the intervention group would reduce the percentage of energy from total fat 13% and 11% compared to the control group at 1 and 9 years, respectively, and the control group incidence rate would be one third greater), the projected power to detect differences in CHD was only 40% (36). Based on this limited power, the researchers’ chance to detect a difference between the intervention and control group was already meager. This was further compounded by the less than moderate beneficial changes in the intervention group (and even some potentially detrimental ones) in comparison to the control group. Such paltry changes included a difference of weight loss of about 1 kg after three years, a less than 3% and 1% reduction in calories from saturated and trans fat, respectively, and a one-serving difference in fruit and vegetable intake. Harmful alterations were increased refined grain consumption and decreased intake of nuts. As such, the changes in biomarkers of interest in the intervention group, especially LDL-c, were similarly small, with the reduction being 3 mg/dl greater than the control. Given these factors, it is unsurprising that no significant reduction in risk was seen. However, despite the extremely limited power, when authors directed their attention towards subjects who achieved the lowest intake of saturated fat (<6% of energy), they remarkably still observed a marginally significant reduction in CHD risk. This once again demonstrates the value of reducing intake of saturated fat, even in a relatively healthy population with a lower baseline intake (~13% of energy) and a lower overall incidence of CHD (<1% of the subjects over about eight years of follow up). Moving onward, they bring up PREDIMED, saying, “Despite an increase in total fat intake by 4.5% of total energy (including slightly higher saturated fat consumption), major cardiovascular events and death were significantly reduced compared with the control group.” It is unclear how they even came to this conclusion, given that the supplementary data from the study itself shows that both the intervention groups reduced their intake of saturated fat as a percentage of calories from 10% to just about 9% (Table S9) (37). It is also worth noting that they reduced red meat and dairy consumption, both of which are foods this review’s authors seem to emphasize there shouldn’t be limits on consumption. They then conclude this paragraph by claiming that the six most recent reviews and meta-analyses of RCTs demonstrated that replacing saturated fat with polyunsaturated fat does not significantly decrease coronary outcomes or total mortality, citing three publications (Ramsden et al. 2016, Hooper et al. 2015, and Hamley 2017.). Two of these have major problems that have been previously addressed, and the highest quality review (Hooper) demonstrated a significant benefit of reducing saturated fat intake on CVD events. Also, in their updated 2020 review mentioned earlier, their meta-regression indicated that the extent of saturated fat reduction and the corresponding decrease in serum cholesterol was responsible for said effect. After the previous comment, authors then state, “Even if these analyses were to be challenged, for example, based on the criteria for study selection or other lines of evidence (42), an important possibility to consider is that an apparently lower risk of CVD with substitution of SFAs by polyunsaturated fatty acids could be attributed to a possible beneficial effect of polyunsaturated fatty acids and not necessarily to an adverse effect of SFAs”, followed by, “…the evidence from both cohort studies and randomized trials does not support the assertion that further restriction of dietary saturated fat will reduce clinical events.” Firstly, the initial statement is quite a speculation given that reducing saturated fat or replacing it with PUFAs, certain MUFAs (such as olive oil and nuts in PREDIMED), and whole grains consistently reduces CVD events (38). Second, even if this were true, why would one choose to consume a nutrient that does not lower the risk of the leading cause of death in the United States over almost all others (aside from refined carbohydrate)? After considering that almost everything the authors have cited up to this point has had notable flaws they failed to disclose, and that many of their sources cut against their position, the erroneous nature of the closing statement for this section is incredibly clear.

LDL, Insulin Resistance, Carbohydrate Condemnation, and Circulating Fatty Acids

Questioning the Utility of Serum LDL-c

In the next section, the authors attempt to instill doubt in the reader about the utility of LDL cholesterol as a biomarker for assessing the effect of saturated fat on cardiovascular risk. After declaring that it is quite clear LDL plays a causal role in the development of CVD, they stipulate that the reduction of LDL through diet cannot be inferred to result in CVD benefit without having the means to assess other biological effects that accompany this reduction. Whether or not there are means to assess other biological effects (which there are), as repeatedly displayed throughout this entire commentary, saturated fat reduction consistently and reliably reduces CVD incidence, so this point is entirely tangential. To follow up this stipulation, they note that postmenopausal estrogen plus progestin therapy and cholesteryl ester transport protein (CETP) inhibitors elicit no CVD benefit despite decreasing LDL cholesterol. They make mention of Mediterranean style interventions and pharmaceutical inhibition of sodium-glucose cotransporter type 2’s ability to reduce CVD risk while increasing LDL, suggesting that these supposedly unexplained deviations from the typical pattern somehow denigrate the relationship between dietary reductions of LDL and CVD risk. As for the increased risk associated with estrogen and progestin therapy, subgroup analysis showed that the risk was driven by those with the highest baseline LDL, for whom even the observed maximum reduction in LDL would not bring close to a normal value (Figures 1 and 4) (39). The reference they cite for CETP inhibitors clearly describes the reasons for discrepancies in the results of four different trials. It highlights the success of anacetrapib (which did significantly reduce CVD events) being due to the trials’ longer duration and its sustained effect on LDL-c (or non-HDL-C/apoB). The other trials were of substantially shorter duration, or resulted in less notable reductions in LDL-c, explaining the variance in observed results (40). Regarding the Mediterranean-style interventions reducing CVD risk, the two trials they cite are the Lyon Diet Heart Trial and a subgroup study of LDL oxidation biomarkers in PREDIMED participants. This is odd given that both demonstrated that significant reductions in LDL from reducing/replacing dietary saturated fat conferred protection against CVD, precisely the opposite of what they claimed these interventions show. Finally, given the positive benefit of SGCT2 on blood glucose and blood pressure, its ability to offer nephroprotection in CKD (which can compound the risk of CVD events in those with an established disease), and the “minimal changes in lipids” it causes, it is unsurprising they have also shown benefit concerning the risk of adverse cardiovascular outcomes (41). None of these change the existence of the relationship between LDL-c and CVD; they just highlight the multifaceted nature of cardiovascular disease and the existence of other risk modifiers. Therefore, it is inappropriate to suggest they somehow indicate that diet-induced reductions in LDL cannot be inferred to result in CVD benefit.

Honing in on sdLDL

The following paragraph brings up a few more common points of discussion. One is that because saturated fat restriction does not lower the more atherogenic sdLDL particle, and lowers HDL (with a negligible effect on the total: HDL ratio), the decrease in saturated fat intake cannot be inferred to yield a proportional reduction in CVD risk. While it has been consistently shown that small dense LDL and the total cholesterol to HDL ratio do indeed correlate well with CVD risk, in no way does this indicate that sdLDL is inherently more atherogenic or that HDL and total cholesterol are the central dictators of cardiovascular disease risk. Accordingly, most studies show that while sdLDL possesses a significant univariate association with CHD risk, it is seldom an independent predictor after multivariate adjustment for triglycerides and HDL. Therefore, Rizzo et al. suggest that “the increased risk associated with smaller LDL size in univariate analyses is a consequence of the broader pathophysiology of which small, dense LDL is a part (e.g. high triglycerides, low HDL cholesterol, increased LDL particle number, obesity, insulin resistance, diabetes, metabolic syndrome).” (42) Furthermore, a recent pre-print of a Mendelian randomization that observed the effects of various lipoprotein subfractions on CVD risk provides additional support that sdLDL is not inherently more atherogenic, with the authors concluding, “LDL and VLDL subfractions appear to have nearly uniform effects on CAD across particle size. Therefore, the results do not support the hypothesis that small, dense LDL particles are more atherogenic.”(43) While the point being raised here is important (i.e., that LDL cholesterol reductions are not necessarily the only metric that should be considered with respect to CVD risk reductions via dietary interventions), the other metrics are not being ignored. Even in most of the trials the JACC review authors claim changes in LDL did not correspond with reductions in cardiovascular events (PREDIMED and WHI); they did, as they do across a wide variety of therapeutic interventions (44).

Impact of Insulin Resistance on Atherosclerosis

The next paragraph begins with a description of insulin-resistant states, their increasing prevalence in the US, and a brief discussion of how they can increase atherogenesis. Authors then exclaim that individuals with insulin resistance experience impaired skeletal muscle glucose oxidation, increased hepatic de novo lipogenesis, and atherogenic dyslipidemia after a high carbohydrate meal. They also remark that these individuals have a higher propensity to convert carbohydrate to fat, further aggravating the insulin-resistant phenotype, including increases in circulating SFAs and lipogenic fatty acids such as palmitoleic acid. This incredible oversimplification implies that once an individual is insulin resistant, any carbohydrates will exacerbate obesity, hypertension, hyperglycemia, hyperinsulinemia, and numerous other adverse effects associated with this phenotype. Additionally, the statement that subjects experienced significant detriments following a high carbohydrate meal is grossly misleading, as the diets fed in the study the authors linked were both high in fat (35% of kcal) and carbohydrate (55% kcal), contained an additional 25% of subjects daily energy requirements in the form of sucrose, and the quality of the foods and nutrients provided was unclear (45). While refined carbohydrates are indeed the last thing that an insulin-resistant individual should center their diet around, diets including appreciable amounts of whole food complex carbohydrates such as legumes, whole grains, fruits, and vegetables have been repeatedly demonstrated to exert a host of beneficial effects on metabolic function. These include but are not limited to, improvements to insulin resistance, reductions in HbA1c, weight loss, decreased blood pressure, and improvements in lipids (46–58). This lack of acknowledgment of essential differences in food/macronutrient quality and their impacts on health again seems quite hypocritical given that one of the authors’ main contentions with dietary guidelines recommending reductions in saturated fat is that it fails to do so.

Serum Fatty Acids and Carbohydrate Calumny

The next two paragraphs put much effort into distinguishing between dietary saturated fat and circulating SFAs, starting by saying while some studies suggest increased saturated fat intake does not increase chronic disease risk, people with higher circulating even-chain SFAs have an increased risk of numerous chronic diseases. While the latter may be true, the former is most certainly not, and unsurprisingly they cite two studies mentioned previously that do not support such an assertion (Siri-Tarino et al. and Jakobsen et al.). They then say that circulating SFAs in the blood tend to track closer with dietary carbohydrate intake (again, no specification of quality here), and that changes in saturated fat intake of 2–3 fold have no effects on serum SFAs in the context of diets lower in carbohydrate. Adding on to this, they explain that the primary fatty acid product of de novo lipogenesis (DNL), palmitoleic acid, is a good proxy of DNL due to its low presence in the diet and that larger proportional increase when carbohydrates are converted to fat. Following this preface, they then discuss how multiple studies show a close link between increased dietary carbohydrate intake and increased serum palmitoleic acid levels independent of body weight changes and saturated fat intake (59–62). Thereafter, they exclaim how increased palmitoleic acid levels are associated with substantial increases in the risk of stroke, heart failure, and coronary artery disease. They do not mention that every one of these studies involves diets high in refined carbohydrates. Only one trial (62) reduced saturated fat intake below the limit recommended by the dietary guidelines and numerous other nutritional science organizations. Interestingly, this trial only looked at serum FA content and select inflammatory markers, and the low carbohydrate intervention involved a reduction in calories accompanied by an increase in MUFA, no significant change in SFA intake, and reduced refined carbohydrate. In contrast, the low-fat intervention also decreased calories, slightly reduced SFA intake, increased refined carbohydrate intake, and increased alcohol intake. Not only are results pertinent to actual metabolic function lacking, but the comparison seems unfair, and the potentially detrimental changes in the dietary patterns of the low-fat intervention are not representative of those recommended by any competent nutritional organization. Their attempt at portraying carbohydrates as inherently harmful to those with insulin resistance by referencing trials involving the consumption of highly processed sources, accompanied by other negligible diet changes, appears very disingenuous. Finally, they conclude that “Clearly, the impact of dietary SFAs on health must consider the important role of carbohydrate intake and the underlying degree of insulin resistance, both of which significantly affect how the body processes saturated fat. This intertwining aspect of macronutrient physiology and metabolism has been consistently overlooked in previous dietary recommendations.” As already acknowledged, it is certainly crucial to consider these factors. Nevertheless, it is unclear how the studies they brought up in this section, showing excess refined carbohydrate intake is harmful to health, change the well established harmful effects of excess saturated fat. Furthermore, the adverse effects attributed to carbohydrates are almost entirely insinuated to be in part via increases in DNL, reflected in increased serum palmitoleic acid. Such statements are very reductionist and ignore the complexity of changes in DNL commonly observed with metabolic syndrome (63). As repeated almost ad nauseam up to this point, the fact they mention intertwining aspects of macronutrient physiology and metabolism again is astounding given the continued indiscriminate vilification of carbohydrates displayed just prior.

Imperfections in Nutritional Science, Ancestry, Nutrigenomics, and Process Contaminants

No “One True Diet”

The review’s subsequent section begins with comments on the failure of the scientific community to determine “the one diet” that achieves optimal metabolic health for all. They then bring up the heterogeneity of dietary intervention outcomes, which they postulate to be a result of the fact that some individuals have better outcomes from specific diets than others. According to them, the objective should be to match each individual to their best, culturally appropriate diet. While it is true that for some individuals modifications may need to be made dependent upon their life stage, the presence of health conditions, and food availability (among other factors), results of well-controlled dietary intervention studies for improving health status have been relatively consistent. Heterogeneity is typically well explained by macronutrient quantity and quality differences, the degree of subject adherence to interventions, age, baseline disease presence, and other critical factors. Many of these are factors the authors have repeatedly highlighted the importance of considering, yet they seem to selectively ignore them when it is convenient. Individualizing nutrition is most definitely valuable, and it is odd to think someone within the field of nutritional science would dismiss this. That being said, they can’t possibly expect the creators of the guidelines to create such recommendations that will apply to each and every person, and it is not their main goal.

The Relationship between Nutrition and Genetic and Relevance to Recommendations for Saturated Fat

Moving onward, the authors begin to discuss nutrigenomics, which is a fascinating up and coming field of study; however, their discussion is very narrow and seems focused almost entirely on exonerating saturated fat. They preface this discussion by saying that “the once apparently tight link between dietary SFAs and CVD appears to be loosening as a result of mounting evidence that casts doubt on previously established belief,” which is once again, blatantly incorrect. Then they say that some of the debate centers on the role of variation in food sources of SFAs (more on this to come) and some on interindividual variation in biological and clinical effects of saturated fatty acids. Such variation is partly suggested to be a result of genetic variants that modulate the relationship between dietary SFA and CVD-related biomarkers. One of the variants discussed is the APOE4 allele of the apo E gene, which predisposes individuals to an increased risk of CVD, hypothesized to result due to more significant fasting plasma and postprandial responses to saturated fat. Further, another study that observed saturated fat interacts with a weighted genetic risk score for obesity to modulate body mass index is mentioned, along with an apo A2 promoter gene for which saturated fat is associated with higher average body mass in those homozygous for the T allele. Appropriately, they do not put too much weight on the latter associations, but they do state that current information suggests that genetic predisposition modulates the association between saturated fat intake. They then state this segment of the population, which they deem “SFA-sensitive,” may experience a benefit in reductions of saturated fat intake, so it could therefore be recommended for them specifically. There does not seem to be any other way to classify this position than utterly baseless. It is ridiculous to suggest that only a subgroup of people with genetic predispositions to an even higher CVD risk resulting from a higher intake of saturated fat should consider reducing their intake. This subgroup’s existence in no way suggests an absence of risk in those outside of it, as evidenced by the reduced risk of CVD elicited from reducing/replacing saturated fat in millions of people both in epidemiological studies and RCTs. That being said, this section is concluded with a few similarly problematic statements. The authors emphasize that type 2 diabetes and obesity are significant contributors to CVD risk and declare that the “optimal diet” (what happened to there not being such a thing as “the one diet” based on current research?) should be based on an individual’s “carbohydrate tolerance,” apparently determined by insulin resistance and insulin secretion capacity. Next, they claim that a diet higher in fat and fiber seems to be optimal for type 2 diabetics, only based on a single trial (64), and that a diet lower in total and saturated fat may solely be optimal for carbohydrate tolerant or insulin-sensitive individuals. While the dietary pattern they described as optimal can be beneficial to those with diabetes, it does not necessitate a higher intake of saturated fat, and individuals would serve to benefit from keeping it lower. Also, given the aforementioned success of high complex carbohydrate, low-fat diets in improving glycemia, lipids, blood pressure, and other disease risk factors in both healthy individuals and those with type 2 diabetes (55–58), the suggestion that only carbohydrate tolerant individuals would serve to benefit from a diet low in saturated and total fat is unwarranted. The end of this section echoes the author’s earlier points about the increasing prevalence of type 2 diabetes and the need for a more personalized and food-based approach to recommend levels of total and saturated fat in the diet. As stated earlier, this is undoubtedly something considered by health professionals who have expertise in providing nutritional advice. It does not indicate a need for any significant alterations in general dietary recommendations currently in place.

Dairy Saturated Fat: Uniquely Non-Atherogenic?

In the following section, the authors begin by highlighting that the health effects of fats and oils may depend on their content of saturated and unsaturated fats, but are not only a sum of their lipid components. They state that they may also depend on the interacting effects of naturally occurring components and harmful components introduced by processing. They then reference the “trans-fat” story, which is a reasonably appropriate example, but just seems to be an aside that does not lead anywhere. Following a discussion of the history of suggestions to replace dairy fat with vegetable oils and the origin of the legislation that drove the saturated vs unsaturated debate, they mention that the major component of vegetable oils (polyunsaturated linoleic acid) was recognized to decrease plasma cholesterol concentrations by the 1950s. In contrast, saturated fat could raise it; therefore, the former was estimated to have a more favorable effect on atherosclerosis. Incredibly, they then jump to state that despite being high in saturated fat, dairy does not promote atherogenesis, with a single reference. The reference they cite happens to analyze the relationship between dairy fat and incident CVD from 3 pooled cohorts with over 5 million person-years of follow up. It showed that when compared to carbohydrates (excluding fruits and vegetables — so likely the refined grains the authors spent so much time using to paint carbohydrates in a bad light) that dairy fat was not associated with a significant increase in the risk of CVD for a 5% increase in energy by a minimal margin, as the RR was 1.02 with a 95% confidence interval of 0.98 to 1.05 (65). However, in their analysis on the effect of isocaloric replacement of dairy fat with other nutrients/foods, almost every single replacement elicited a significant reduction in the risk of CVD, including vegetable fat, omega 6, ALA, marine omega 3, and carbohydrates from whole grains, ranging from RRs of 0.9 (0.87 to 0.93) to 0.72 (0.69 to 0.75) for vegetable fat and whole grains respectively. The only exceptions were carbohydrates from refined grains/starches and other animal fat. Henceforth, the authors conclude, “The results suggest that, compared with dairy fat, vegetable sources of fat and PUFA are a better choice for reducing risks of CHD, stroke, and total CVD, although other animal fat (e.g., from meats) may be a less healthy choice than dairy fat. In addition, we showed that types of carbohydrates made a difference; the replacement of dairy fat with high-quality carbohydrates such as whole grains was associated with lower risk of CVD, but the replacement with refined starch and added sugar did not appear beneficial.” How this supports their assertion that dairy fat is not atherogenic or the larger overall point that saturated fat reduction is not warranted is unknown, as it strongly suggests the opposite.

Appeals to Ancestry

After this, they begin to bring up an even stranger point, that due to the fact the ability of humans to digest lactose in milk has evolved separately numerous times, it is unequivocal that humans “required continuous dairy consumption for survival to reproductive age.” It seems as if they’re attempting to suggest this “requirement” for dairy milk, which was based on the fact that it may have offered a survival advantage in previous years (unsurprising given it is a concentrated source of calories), extends to present day, which is just outright false. This is especially problematic because the prevalence of lactose intolerance worldwide is suggested to exceed 65% of the population (66). Expanding upon this, the authors also mention how bovine, goat, and sheep domestication coincided with the emergence of lactase persistence and that the meat from these species was likely a significant contributor of saturated fat to human diets, supplemented with some low polyunsaturated fruit oils (olive, avocado, and palm) where available. Continuing, they state that seed oil consumption would have been negligible back then and that these historical facts demonstrate that saturated fats were abundant, critical parts of the ancient human diet. Along with the lack of any research to support these final few assertions, there are numerous issues with this rhetoric. The main one is that any argument extending from this reasoning would be entirely founded on the logical fallacy “appeal to tradition.” Additionally, as they explicitly stated for milk consumption, these choices were likely made on the basis that they supported survival until reproductive age more than anything else. Therefore they would not necessarily have any relevance to constructing dietary patterns supporting a long (at least mostly) disease-free life, which is the goal of countrywide dietary guidelines. Interestingly, reviewing the best estimates we have on nutrient intakes of paleolithic populations, intakes of saturated fat were substantially lower than current-day populations, with estimates ranging from 7.5 to 12% of energy, indicating that a large fraction of the populations likely had intakes within the range recommended by dietary guidelines. Funnily enough, the JACC review authors specifically mention that domestication originated around 10,000 years ago with the advent of modern-day agriculture. The same publication mentions that some have hypothesized to be the origin of notable discordance between “older” and “newer” populations’ health (67). That being said, this is the farthest thing from what we should be using to formulate current day guidelines on, especially when we have decades of recent high-quality data available.

A Brief Digression on Coconut Oil

Next, the authors discuss the 1970s animal experiments that used coconut oil of “unspecified origin,” which caused dramatic increases in hepatic and blood cholesterol and were therefore deemed atherogenic. They mention these oils were usually highly processed and fully hydrogenated (without a source), and that “virgin” coconut oils produced in recent years using gentler preparation methods do not possess the same LDL cholesterol-raising properties, citing a study demonstrating this lack of effect in humans (68). While there was indeed no significant change in LDL following four weeks of consuming coconut oil daily, there are quite a few concerning factors that make this trial at least a bit suspect. First, the coconut oil group’s baseline saturated fat intake was already relatively high (~15% kcal). There was no information on post-trial intake, making it impossible to gauge the overall change in saturated fat intake, which is of critical importance. Given that the change in calories and total fat from baseline were only 71 kcal and 29 g when the coconut oil should have theoretically added about 430 kcal and 50 g of fat, it is very likely that significant alterations in dietary patterns were made, and that absolute saturated fat intake may have hardly changed. Regardless, another trial on virgin coconut oil (69) showed that it did indeed significantly increase LDL cholesterol over 30 days, and a recent meta-analysis (70) including these two trials along with 14 others demonstrated a significant increase with coconut oil consumption in trials over two weeks, so this matter is far from settled.

Process Contaminant Paranoia

After mentioning these trials, the authors discuss the recent realization that high-temperature treatment of oils in the presence of trace metals generates process contaminants. Moreover, they describe an in vitro trial demonstrating that direct administration of different coconut oil samples subject to various leveling of processing elicited different effects on cholesterol metabolism, with greater processing associated with greater increases in cellular cholesterol. It is strange that after criticizing others for not basing their decisions on high-quality evidence that the authors defer to mechanistic studies, one of the weakest forms of evidence to support speculation about the effect of potential oil contaminants. Subsequently, they bring up a study using (mostly non-enzymatic) oxidation-resistant linoleic acid and claim that it supports the hypothesis that oxidation products and not specific fatty acids cause plaque formation in mouse models (71). Such a statement is very misleading. The di-deuterated linoleic acid in this study only partially reduced atherogenesis of mice fed a high saturated fat and cholesterol diet, and alongside a reduction in oxidation products elicited significant decreases in LDL cholesterol. Pinning the entire process on oxidation products, especially when substituting common sources of PUFAs significantly reduces CVD incidence in humans, is unbecoming. Furthermore, placing so much weight on speculations stemming from rodent models such as this is a significant issue given the remarkable inconsistency in the predictive ability of said models for humans (72). There is a reason they reside near the bottom of the evidence hierarchy.

Ostensible Problems with the AHA’s Presidential Advisory

In the following paragraph, authors state that human studies assuming all foods high in saturated fat are similarly atherogenic in many cases stem from an era before the recognition of process contaminants, which seems as if they are suggesting this is the only atherogenic characteristic of saturated fat-rich foods, and if so is an incredibly disturbing conjecture. Next, they claim that the recent Presidential Recommendation to avoid saturated fats from the American Heart Association is based on studies in the 60s and 70s, 3 in Europe, and 1 in America. This is an extraordinarily uncharitable and blatantly incorrect characterization of the AHA’s Presidential Recommendation. In reality, it analyzed these four trials, six smaller, lower quality “non-core trials,” the Lyon Heart study, PREDIMED, additional observational studies on populations following similar Mediterranean diets, RCTs reducing saturated fat intake and decreasing LDL-c (and vice versa), multiple meta-analyses of observational studies on SFA intake and CVD incidence, large scale cohorts observing the effects of replacing SFA on CVD incidence, and additional supplementary genetic and mechanistic evidence (73). These considerations aside, the JACC review’s authors discuss that the 3 European trials are confounded because they used customary diets as comparisons, and supposedly partially hydrogenated fish oils were major constituents of European margarine during the 1960s and 70s, for which they cite a book titled “The Story of Margarine”. Unfortunately, this book was inaccessible. Alas, they declare that The Oslo study explicitly estimated the partially hydrogenated fish oil intake at 40 to 50 g/day. Whether this pertains to the control or experimental group is unclear. However, the only information given in the publication is for the experimental diets’ composition, which was the following: “In an analysis of the experimental diet as consumed by 17 selected dieters, the mean daily intake was: protein, 92 gm.; fat, 104 gm.; carbohydrates, 269 gm.; and cholesterol, 264 mg. Daily intake of calories was 2,387. Calories derived from fat constituted 39 percent of the total. The sources of fat were: soybean oil (72 percent), fish fat (11.6 percent), animal fat (8.8 percent), cereal fat (5.0 percent), and fat from other sources (2.6 percent).”74 Given these numbers, and assuming “fish fat” is entirely partially hydrogenated fish oil, 11.6 percent of 104 g fat is still only 12 grams, way off from what they claim, so this bit is pretty confusing. Continuing in this vein, they assert that since the three European trials used customary diets as comparisons, it can be inferred that they were tests of polyunsaturated fats against trans-plus-saturated fats, and the effects cannot be ascribed to saturated fat alone. Henceforth, excluding these trials, as they see fit, only the (smaller and underpowered) US trial (75) remains, which did not show a significant difference between the control and intervention group for its primary endpoint. However, results were significant for all endpoints combined and fatal atherosclerotic events despite the relatively low total event rate and consequential low statistical power. A few things merit discussion here. First, the entire reason they suggest the European trials should be excluded is an inference they are confounded that is not conclusively proven. Second, even if this were the case, the large body of evidence from other RCTs and observational studies discussed in the AHA’s Presidential Report is sufficient to underscore the importance and benefit of reducing/replacing dietary saturated fat.

Food Matrix, Saturated Fat, and their Connection to Health

Finally, they conclude this section by saying, “…these observations strongly support the conclusion that the healthfulness of fats is not a simple function of their SFA content, but rather is a result of the various components in the food, often referred to as the “food matrix.”. As discussed in the following section, ample evidence is available from research on specific foods that other food components and the food matrix likely dominate over saturated fat content. Recommendations should, therefore, emphasize food-based strategies that translate for the public into understandable, consistent, and robust recommendations for healthy dietary patterns.” Once again, the evidence they have brought forth does not seem to strongly suggest that the saturated fat content of foods is negligible with regard to a food’s healthfulness and impact on CVD risk. As for the food components and matrix, their actual relevance and implications for a foods’ impact on health will be addressed accordingly. Finally, as mentioned a few times previously, the current (and hopefully future) dietary recommendations are arguably already understandable, consistent, and robust, although there is always room for improvement.

Safe Sources of Saturated Fat?

Full Fat Dairy

In some of the review’s final paragraphs, the authors specify their reasons as to why they feel a few specific foods with high saturated fat content are unfairly discouraged. These foods include yogurt and cheese, dark chocolate, and red meat, discussed in that order. For yogurt and cheese, authors begin by describing that dairy is the primary source of SFAs in most diets and that major dietary guidelines recommend low or fat-free versions to limit SFA. However, that food-based meta-analyses consistently find the two are inversely associated with CVD risk, citing a cohort (76), a literature review (77), another cohort on type 2 diabetes (78), and one actual meta-analysis (79). This seems as if it was just a citation error, but regardless the meta-analysis they cite actually provides weak, if any, evidence suggesting an inverse association of cheese and yogurt with CVD. The inverse association was only found for cheese and was attenuated when a single large study was removed in a sensitivity analysis. Furthermore, the subgroup analyses demonstrated the strongest inverse association with subjects below 50 years of age, a group in which CVD incidence would be lower and therefore possess far less of an ability to effectively gauge the effect on CVD risk. Even more concerning, numerous trials adjusted for hyperlipidemia, serum cholesterol, or saturated fat, which would also strongly impair their ability to pick up on an increased risk of CVD. One final consideration is that they chose to exclude one of the largest, and perhaps best quality studies (Hu et al. 1999), from their analysis, which demonstrated a significantly increased risk of CVD within the highest quintile of full-fat dairy consumption, and with a higher ratio of full to low-fat dairy consumption.

To the author’s benefit, another earlier meta-analysis (80) showed a similar neutral association of yogurt and cheese with CVD, and another published this year demonstrated a significant inverse association (81), however no detail on the fat content, amount of intake, and overall diet composition of subjects was provided. This is problematic given that the details on the amounts, what the two are replacing, and the actual fat content are incredibly important to consider and could meaningfully influence the observations. The cohort they cite even exemplifies these points, showing that a higher intake of saturated fat from dairy (equivalent to 1–2 servings in the highest quintile depending on the source) is associated with a slight reduction in CVD risk, and that replacement of an equivalent amount from meat elicited a substantial RR of 0.75 (76). This gives some merit to their statements about the saturated fat content of foods not being the only determinant of its overall effect on health, yet it does not eliminate the possibility that cheese and yogurt contribute to CVD risk as they seem to suggest.

They also refer to a meta-analysis on circulating biomarkers of dietary fat intake, specifying that data from 4 cohorts showed those in the top vs bottom third of plasma measures of C17:0 (heptadecanoic acid) had a decreased risk of CHD (82). In addition to the issues present in the previously discussed meta-analyses, it has been acknowledged that C17:0 (along with C15, pentadecanoic acid) are poor indicators of dairy intake, and caution should be taken in interpreting any findings related to observations in epidemiological studies (83). Alongside their discussion centered around cheese/yogurt and CVD, authors briefly mention that some studies indicate higher whole-fat dairy consumption is associated with a lower risk of diabetes. Again providing an odd selection of citations; two literature reviews (84,85), the de Souza meta-analysis discussed earlier (which observed implications of serum Ct16:1n7 concentrations), and a cohort on circulating fatty acids (C15:0, C17:0, and Ct16:1n7) associated with dairy consumption and their relation to type 2 diabetes risk (86). While the latter two did suggest an inverse association between serum values of these fatty acids and T2D, and although it is much more likely that Ct16:1n7 correlates with dairy intake, others have raised concern that it is also unreliable (87), and these studies both fall victim to the numerous problems just discussed for those on CVD.

As an addendum, it is only fair to note that one of the sources the authors cited to support their CVD claims was a cohort pertinent to the assessment of diabetes risk changes resulting from substitutions between subgroups of dairy (78). This particular study had more data that allowed for further investigation, which revealed a few interesting details. Most striking was that the consumption of full-fat dairy products was incredibly low for yogurt, and a little less so for cheese. For yogurt and cheese, mean intakes were 0.05/0.08 and 1.38/1.47, and ranges were 0.01–0.88/0.01–0.60 and 0.48–3.39/0.46–3.44 servings per day for the two in men/women, respectively. No notable associations were found for cheese for all the substitutions tested; however substitution of low for whole-fat yogurt increased the risk of T2D, whereas substitution of whole fat yogurt for low-fat milk, whole-fat milk, and buttermilk all significantly reduced the risk. Given the low mean and ranges of intake, it seems highly likely that these findings are spurious at best, and further details only increase this likelihood. Looking over the radar charts provides insight as to the differences in the dietary profiles of those in the lowest and highest intakes of full-fat dairy. They reveal that a higher intake of full-fat dairy was associated with reduced intakes of red and processed meats, sugar-sweetened beverages, butter, and greater fruit and vegetable intake. Taking this into account, although variables related to dietary factors are considered, given the extent of the differences in the low vs high full-fat dairy intake groups in conjunction with the very low range, it is unreasonable to make any firm conclusions from this particular study. A few additional meta-analyses had discordant results, with some showing inverse associations for cheese (88,89), another demonstrating no effect (90), and one showing an increase in risk (91). Unfortunately, none of these was stratified by fat content, although three of four showed no significant association when dividing dairy foods up between full and low-fat. Results for yogurt were similarly inconsistent, showing an inverse association with T2D in three meta-analyses (90–92), and no association in two (88,89). Like those on cheese, none of these was divided up into low and high-fat content sources. Factoring in all of these studies, it’s quite clear that the association is not nearly as strong as the authors make it out to be, being questionable at best.

The authors wrap up this paragraph by concluding, “Cheeses and yogurts consist of complex food matrices and major components include different fatty acids, proteins (whey and casein), minerals (calcium, magnesium, phosphate), sodium, and phospholipid components of the milk fat globule membrane (115). Yogurt and cheese also contain probiotics and bacterially produced bioactive peptides, short-chain fatty acids, and vitamins such as vitamin K2. The complex matrix and components of dairy may explain why the effect of dairy food consumption on CVD cannot be explained and predicted by its content in SFAs.” One immediate point that warrants attention is that the authors make no effort to provide evidence that any of the components discussed here have relevance to dairy’s influence on health, CVD, or other notable outcomes. As such, any insinuation regarding these nutrients or other components is purely speculative. Also, quite ironic is that most of them are also present in low-fat yogurt/cheese (many even in larger amounts, i.e., whey and casein, calcium, magnesium, phosphate, etc.). Even if the saturated fat content was irrelevant, they make no real case for consuming full fat over low-fat/fat-free alternatives, and from the evidence discussed here, the latter seems to be associated with far more favorable outcomes.

Despite these points, it is only fair to acknowledge the authors do indeed appear to be correct that, at the very least, cheese and yogurt cannot be directly equated to other foods containing similar amounts of saturated fat. This is shown by the fact that feeding trials have consistently demonstrated that when compared to a diet lower in SF and higher in MUFA, PUFA, or carbohydrate, cheese raises LDL cholesterol to a lesser degree than butter providing an equivalent amount of saturated fat (93,94). However, this discrepancy seems to be present only in subjects with a high baseline LDL, as discussed by Brassard et al. in their 2017 publication. Furthermore, in a meta-analysis (95) observing comparisons to other foods, including reduced-fat cheese, tofu, and egg white, full-fat cheese tended to elicit significant increases in LDL cholesterol. Unfortunately, data on the impact of full-fat yogurt in isolation on LDL is pretty scarce. However, two trials, including a three or four-week intervention with high-fat cheese, yogurt, and milk, resulted in a significant increase in LDL95, or mitigation of the decrease observed following a lower fat intervention, even when coinciding with a massive increase in overall fiber intake and reduction in sugar intake compared to the low-fat intervention (96). If the food matrix or nutrient content of full-fat dairy such as cheese and yogurt affects their impact on cardiometabolic risk factors, the magnitude appears to be small and would likely not be of substantial importance. Taking into account inconsistent associations of full-fat cheese and yogurt with CVD/T2D, their ability to negatively influence LDL-c, the similar or superior nutrient profile of low-fat alternatives, and findings from pooled cohorts showing replacement of dairy fat with virtually all other sources of fat and carbohydrate reduce the risk of CVD, it is abundantly clear that current recommendations to limit SFAs from these two foods are appropriate.

Dark Chocolate

After cheese and yogurt, the authors direct their attention to dark chocolate, stating that it contains stearic acid, which has a neutral effect on CVD risk and contains other nutrients that may be more important for CVD/type 2 diabetes. They note that it possesses potential preventative effects on the two, supported by experimental and observational studies. While they provide no source for the stearic acid claim, the three meta-analyses they link (97–99) do offer consistent evidence confirming a small protective effect of chocolate with respect to CVD and type 2 diabetes, in agreement with a meta-analysis (100) on dark chocolate and cocoa powder’s effects on serum lipids and an RCT (101) of its impact on insulin sensitivity.

Unprocessed Red Meat

The final food the author’s remark has insufficient evidence to suggest reducing intake based on saturated fat content is (unprocessed) red meat. They give four references to back up this claim: a meta-analysis of cohort studies (102), two meta-analyses of RCTs (one on surrogate biomarkers (103) and one on actual outcomes (104)), and a small cohort (105). Aside from the fact that these publications do not even represent a modicum of the evidence on red meat intake and CVD, type 2 diabetes, and cancer, the three they gave to justify their claims are weak at best. The first meta-analysis they cite only includes 4 and 5 cohorts observing the effect of red and processed meat intake on CHD and type 2 diabetes incidence. Furthermore, three of the four observing the effect of red meat intake on CHD made adjustments for serum cholesterol, which, as discussed previously, is incredibly problematic given its role as a causal intermediate. Finally, although the association with diabetes was non-significant, it was by an incredibly small margin. Due to the limited number of studies considered and the range in intakes observed, it would be ill-advised to conclude this from this publication alone.

The first of the two meta-analyses of RCTs they refer to investigated how one half or more servings per day of red meat impacted CVD risk factors, particularly serum lipid and blood pressure values. This publication concludes that red meat did not significantly impact LDL, HDL, triglycerides, systolic blood pressure, and diastolic blood pressure. This is a perfect example of the importance of not just taking results from RCTs at face value simply because it is referred to as the “gold standard” for nutritional science research. First, many of the trials intentionally used lean red meat for their comparison, which is not representative of consumers’ typical choices and detracts from the ability to determine the impact of naturally occurring saturated fat on a subject’s lipid and blood pressure values. Second, well over half of the trials involved adding red meat to one’s habitual diet or a diet designed to elicit weight loss. While the former is not necessarily an issue, failure to consider the baseline diet (energy intake, saturated/trans fat intake, cholesterol intake, and refined carbohydrate intake, among other factors) would prevent a reliable assessment of the impact of adding red meat from being made. If subjects maintained or decreased intake of these nutrients in the intervention, or lost weight, this would bias the results towards a null or even positive outcome given that weight loss and changes in these nutrients can affect blood pressure and lipids. Third, the comparators (food given to the control group) varied significantly, another potential source of issues since different foods or overall dietary patterns could impact lipid and blood pressure parameters. Finally, both groups’ mean final LDL concentrations were above the normal range (3.18 and 3.13 mmol/L for intervention and control). This gives merit to the theory that initial nutrient intake was subpar and had already given rise to lipid values, indicating that the conclusion red meat supposedly does not increase LDL is of limited value in this context. In a recent and far more thorough meta-analysis of RCTs on the same topic (106), many of these issues and their impact on observations were well-documented. This publication’s main finding was that replacing high-quality non-meat protein sources with red meat significantly increases LDL cholesterol, even considering diets designed to decrease lipid values. Another important note was that the total saturated fat content of the intervention diet containing red meat modulated the results, with any differences between LDL becoming non-significant from other comparator diets when total saturated fat intake was matched or lower in the intervention.

The second meta-analysis cited by the authors is shockingly even more concerning. The goal was to observe the effect of reducing unprocessed red meat consumption in RCTs of over six months duration on chronic disease morbidity and mortality (specifically cancer and CVD). First of all, an RCT is far from an appropriate model for assessing the impact of dietary habits on chronic disease incidence and mortality, as these are lifestyle diseases that take decades to develop and manifest. It is extraordinarily naive to believe that a randomized controlled trial would include enough participants or be carried out long enough in order to have the power to observe a significant change in these metrics. Further, the feasibility of doing so is reflected in the fact that this analysis includes a single trial, the Women’s Health Initiative trial. As expected, no significant impact of reducing red meat on the outcomes considered was observed. That being said, despite the immense limitations of using this approach, a difference of a single serving (~1.4) per week of red meat between groups, and adjustment for cholesterol in their analysis, the results were still borderline significant: “…all-cause mortality 0.99 [95% CI, 0.95 to 1.03]), cardiovascular mortality (HR, 0.98 [CI, 0.91 to 1.06]), and cardiovascular disease (HR, 0.99 [CI, 0.94 to 1.05])…little or no effect on total cancer mortality (HR, 0.95 [CI, 0.89 to 1.01]) and the incidence of cancer, including colorectal cancer (HR, 1.04 [CI, 0.90 to 1.20]) and breast cancer (HR, 0.97 [0.90 to 1.04]).” Once again, this fails to provide convincing evidence, if any at all, that red meat does not contribute to chronic disease risk.

Finally, the authors briefly mention that in a recent analysis of pooled prospective cohort studies, unprocessed red meat was associated with a small, significant increase in all-cause mortality and incident CVD, along with chicken and processed red meat. They did not disclose that this “small” association was for only two servings of red meat a week, under the average intake of all Americans, on top of almost 200 g processed meat and around 300 g chicken according to data for 1999–2016 from NHANES (107).

After bringing up this study, they move on to make another point and close the paragraph, as if this is all the evidence that exists on the subject. Red meat consumption has consistently been shown to significantly raise the risk of all-cause mortality, CVD, diabetes, colorectal cancer, and stroke in numerous meta-analyses and pooled results from large prospective cohorts (108–127). The extent to which the existing literature pertinent to this discussion was ignored is inexcusable, and the conclusion drawn by the authors was highly misleading. In an attempt to reconcile the results of the one publication they did mention suggesting it may be harmful, they reason that it is a major source of protein, bioavailable iron, minerals, and vitamins that may comprise an essential part of the diet for the elderly and low-income populations. Although considerations for specific demographics are important, the recommendations the authors are criticizing are meant to apply to the general population, which mostly is not composed of these individuals, so it seems odd to bring them up without any sort of prompting. Furthermore, focusing on nutrient content and the cost of a food while ignoring strong, consistent associations with increased risk of the most prevalent, life-threatening, or debilitating chronic diseases is myopic. What is also strange is that red meat is not typically considered a cheap food, and that there other nutrient-rich, low-cost foods with substantial health benefits, such as legumes, whole grains, and select nuts and seeds. A few studies on older populations underscore the protective effects of these legumes and the potential harm of higher meat intake (128–131). Henceforth, with this in mind, their points appear even less convincing.

Comments on Gaps in Research and Potentially Distracting Dietary Guidelines

Before wrapping up their review, the authors comment on what they feel to be gaps in current research and essential considerations for future investigations. They initiate this paragraph by claiming that recommendations to reduce SFA intake without considering specific fatty acids and food sources are not aligned with current evidence. As a result, they suggest that these recommendations: “may distract from more effective food-based recommendations, and may also cause a reduction in the intake of nutrient-dense foods (e.g., dairy, unprocessed meat) that may help decrease not only the risk of CVD, type 2 diabetes, and other non-communicable diseases, but also malnutrition, deficiency diseases, and frailty, particularly among “at-risk” groups.” This is both false and incredibly hyperbolic. As demonstrated throughout this entire commentary, with the exception of dark chocolate, they continually failed to substantiate any of these claims. Regarding nutrient intake, full-fat dairy products hardly offer any advantage, if at all, over low-fat. Elimination of red meat does not necessitate decreased nutrient intake, especially given healthy substitutions such as fish, other lean meats, whole grains, legumes, and nuts/seeds are made. The full body of evidence shows that the effects of full-fat dairy are inconsistent at best, and low-fat dairy products appear to be favorable. Red meat is continually shown to have adverse effects on the risk of the diseases they mentioned, so the intention of including it in their sweeping claims is unknown. Next, they suggest that a focus on SFAs has had an unintended consequence of misleadingly guiding government, consumers, and industry towards foods low in SFA but rich in refined starch and sugar, and that guidelines should consider the types of SFAs, and more importantly, the foods containing them and their diverse effects on health outcomes. In addition to the fact that these choices were almost definitely motivated by factors other than the actual dietary guidelines, the consideration of different SFAs and foods containing them only seems to hold water to a very limited extent. Moderate amounts of dark chocolate and potentially whole coconut (which the authors interestingly chose not to discuss) are likely the only exceptions to the strong relationship between saturated fat intake and adverse health outcomes.

Closing Remarks

Processed Foods and Saturated Fat Bias

Finishing up this section and transitioning into their final paragraph, they strongly recommend a more food-based translation of guidelines to achieve a healthy diet, reconsidering the recommendations to reduce intake of total SFAs, and caution regarding the incorporation of processed foods “until much more is known about the health effects of specific process contaminants so that their levels can be minimized.” Carrying on, they claim the long-standing bias against foods rich in saturated fat should be replaced with a recommendation for diets consisting of healthy foods. Immediately following, they state, “We suggest the following measures: 1) enhance the public’s understanding that many foods (e.g., whole-fat dairy) that play an important role in meeting dietary and nutritional recommendations may also be rich in saturated fats; 2) make the public aware that low-carbohydrate diets high in saturated fat, which are popular for managing body weight, may also improve metabolic disease endpoints in some individuals, but emphasize that health effects of dietary carbohydrate — just like those of saturated fat — depend on the amount, type and quality of carbohydrate, food sources, degree of processing, etc.; 3) shift focus from the current paradigm that emphasizes the saturated fat content of foods as key for health to one that centers on specific traditional foods, so that nutritionists, dietitians, and the public can easily identify healthful sources of saturated fats; and 4) encourage committees in charge of making macronutrient-based recommendations to translate those recommendations into appropriate, culturally sensitive dietary patterns tailored to different populations.”

Missing the Mark

There are some excellent points here; however, their relevance is questionable. Regarding their recommendations encouraging a food-based translation of a healthy diet/a focus on diets consisting of healthy foods, it is almost sure that anyone interested in constructing policies for designing and implementing dietary guidelines would agree wholeheartedly. The panel in charge of creating the guidelines in 2015 was in such agreement that is exactly what they did. On the United States Department of Agriculture/Health and Human Service’s website it is explicitly stated, “A healthy eating pattern includes: A variety of vegetables from all of the subgroups — dark green, red and orange, legumes (beans and peas), starchy, and other, fruits, especially whole fruits, grains, at least half of which are whole grains, fat-free or low-fat dairy, including milk, yogurt, cheese, and/or fortified soy beverages, a variety of protein foods, including seafood, lean meats and poultry, eggs, legumes (beans and peas), and nuts, seeds, and soy products, and oils” (35), which leaves one questioning precisely what the JACC review’s authors are even contesting.

Moving on to their postulation about a “long-standing bias” against foods with saturated fat and the subsequent plea for the creators of the guidelines to reconsider their suggestions to reduce total saturated fat intake, there is just no substantial evidence for either of these. The suggestion there exists of some sort of bias is entirely unsubstantiated, and the totality of the evidence indicates that no such reconsideration is needed. In addition to the robust evidence from numerous meta-analyses already discussed, the Scientific Advisory Committee on Nutrition’s recent 2019 report concluded that based on 47 systematic reviews, meta-analyses, and pooled analyses on saturated fat intake (mainly from desserts, full-fat dairy, and meat/meat products), there is no need to modify the recommendations to limit saturated fat to less than 10% of calorie intake. Furthermore, a reduction beyond this amount would impose significant population-level health benefits (132). Next, their claim that many foods that play a role in meeting dietary and nutritional recommendations may also be high in saturated fat is just as baseless, especially considering that low-fat and fat-free versions of the products they are speaking about (dairy and meat specifically) are just as rich in nutrients, if not more so than their higher fat counterparts. Furthermore, numerous other nutrient-dense and health-promoting foods, many recommended by the current guidelines, are available. Their second central point about low carbohydrate diets high in saturated fat being useful for managing body weight and improving metabolic disease endpoints, and that health effects of dietary carbohydrate depend on the amount, type and quality of carbohydrate, food sources, degree of processing, etc., is of even less relevance. Not only did they provide absolutely no evidence in support of low carbohydrate diets eliciting weight loss or improvements in metabolic disease endpoints “in some people,” but the latter points are also subsumed in the current recommendations for healthy eating patterns provided by the USDA/DHHS. Their third and fourth points once again falsely characterize the guidelines, both asserting that the current paradigm only emphasizes the saturated fat and macronutrient content of foods and fails to include culture-sensitive dietary patterns, which as highlighted numerous times at this point, is just untrue.

While the authors of this state of the art review make some bold claims, including that saturated fat limits are arbitrary, that numerous foods rich in SFAs have no association with CVD or diabetes, and that the guidelines should emphasize food-based recommendations for healthy dietary patterns, they fall incredibly short of corroborating them. The evidence they provide in an attempt to do so is weak, inconsistent, and many times even contradicts their claims. They continually misrepresent the studies being presented and the current dietary guidelines. Additionally, they put forth speculative claims stemming from animal or in vitro/vivo models and make misleading statements that will create more confusion amongst the public. Current evidence makes a consistent and robust case for reducing saturated fat, especially from red meat, to decrease morbidity and mortality from diabetes, cancer, and cardiovascular disease. Although they do raise the important point of factoring the quantity and quality of foods into dietary guidelines intended to minimize disease and support long, healthy life, such considerations are already made by those currently in place. While the current guidelines may not be perfect, they are much better than this review’s authors make them out to be, and do not require most of the adjustments they suggest.

If you have enjoyed this article, please consider supporting the site by donating here and subscribing to our email list below (we don’t spam!).