Based on a study of 42 European countries, a recent scientific paper reported that, “the highest CVD [cardiovascular disease] prevalence can be found in countries with the highest carbohydrate consumption, whereas the lowest CVD prevalence is typical of countries with the highest intake of fat and protein.” And that, “The positive effect of low-carbohydrate diets on CVD risk factors (obesity, blood lipids, blood glucose, insulin, blood pressure) is already apparent in short-term clinical trials lasting 3–36 months (58) and low-carbohydrate diets also appear superior to low-fat diets in this regard (36, 37).” Basically, for heart health, this would suggest eating more full-fat dairy, eggs, meat, and fish while eating less starches, sugar, and alcohol. That is to say, follow a low-carb diet. It doesn’t mean eat any low-carb diet, though, for the focus is on animal foods.
By the way, when you dig into the actual history of the Blue Zones (healthy, long-lived populations), what you find is that their traditional diets included large portions of animal foods, including animal fat (Blue Zones Dietary Myth, Eat Beef and Bacon!, Ancient Greek View on Olive Oil as Part of the Healthy Mediterranean Diet). The longest-lived society in the entire world, in fact, is also the one with the highest meat consumption per capita, even more than Americans. What society is that? Hong Kong. In general, nutrition studies in Asia has long shown that those eating more meat have the best health outcomes. This contradicts earlier Western research, as we’re dealing with how the healthy user effect manifests differently according to culture. But even in the West, the research is ever more falling in line with the Eastern research, such as with the study I quoted above. And that study is far from being the only one (Are ‘vegetarians’ or ‘carnivores’ healthier?).
This would apply to both meat-eaters and vegetarians, as even vegetarians could put greater emphasis on nutrient-dense animal foods. It is specifically saturated fat and animal proteins that were most strongly associated with better health, both of which could be obtained from dairy and eggs. Vegans, on the other hand, would obviously be deficient in this area. But certain plant foods (tree nuts, olives, citrus fruits, low-glycemic vegetables, and wine, though not distilled beverages) also showed some benefit. Considering plant foods, those specifically associated with greater risk of heart disease, strokes, etc were those high in carbohydrates such as grains. Unsurprisingly, sunflower oil was a risk factor, probably related to seed oils being inflammatory and oxidative (not to mention mutagenic); but oddly onions were also likewise implicated, if only weakly. Other foods showed up in the data, but the above were the most interesting and important.
Such correlations, of course, can’t prove causation. But it fits the accumulating evidence: “These findings strikingly contradict the traditional ‘saturated fat hypothesis’, but in reality, they are compatible with the evidence accumulated from observational studies that points to both high glycaemic index and high glycaemic load (the amount of consumed carbohydrates × their glycaemic index) as important triggers of CVDs. The highest glycaemic indices (GI) out of all basic food sources can be found in potatoes and cereal products, which also have one of the highest food insulin indices (FII) that betray their ability to increase insulin levels.” All of that seems straightforward, according to the overall data from nutrition studies (see: Uffe Ravnskov, Richard Smith, Robert Lustig, Eric Westman, Ben Bikman, Gary Taubes, Nina Teicholz, etc). About saturated fat not being linked to CVD risk, Andrew Mente discusses a meta-analysis he worked on and another meta-analysis by another group of researchers, Siri-Tarino PW et al (New Evidence Reveals that Saturated Fat Does Not Increase the Risk of Cardiovascular Disease). Likewise, many experts no longer see cholesterol as a culprit either (Uffe Ravnskov et al, LDL-C does not cause cardiovascular disease: a comprehensive review of the current literature).
Yet one other odd association was discovered: “In fact, our ecological comparison of cancer incidence in 39 European countries (for 2012; (59)) can bring another important argument. Current rates of cancer incidence in Europe are namely the exact geographical opposite of CVDs (see Fig. 28). In sharp contrast to CVDs, cancer correlates with the consumption of animal food (particularly animal fat), alcohol, a high dietary protein quality, high cholesterol levels, high health expenditure, and above average height. These contrasting patterns mirror physiological mechanisms underlying physical growth and the development of cancer and CVDs (60). The best example of this health paradox is again that of French men, who have the lowest rates of CVD mortality in Europe, but the highest rates of cancer incidence. In other words, cancer and CVDs appear to express two extremes of a fundamental metabolic disbalance that is related to factors such as cholesterol and IGF-1 (insulin-like growth factor).”
That is an argument people have made, but it’s largely been theoretical. In response, others have argued the opposite position (High vs Low Protein, Too Much Protein?, Gundry’s Plant Paradox and Saladino’s Carnivory, Carcinogenic Grains). It’s true that, for example, eating meat increases IGF-1, at least temporarily. Then again, eating in general does the same. And on a diet low enough in carbs, it’s been shown in studies that people naturally reduce their calorie intake, which would reduce IGF-1. And for really low-carb, the ketogenic diet is specifically defined as being low in animal protein while higher in fat. A low-carb diet is not necessarily a high-animal protein diet, especially when combined with intermittent fasting such as OMAD (one meal a day) with long periods of downregulated IGF-1. Also, this study didn’t appear to include plant proteins in the data, and so we don’t know if eating lots of soy, hemp protein powder, etc would show similar results; although nuts were mentioned in the report as being similar to meat in correlating to CVD health but, as far as I know, not mentioned in terms of cancer. What would make animal proteins more carcinogenic than plant proteins or, for that matter, plant carbohydrates? The hypothetical mechanism is not clear.
This anomaly would’ve been more interesting if the authors had surveyed the research literature. It’s hard to know what to make of it since other studies have pointed to the opposite conclusion, that the risks of these two are closely linked, rather than being inversely associated: “Epidemiologically, a healthy lifestyle lessens the risk of both cardiovascular disease and cancer, as first found in the Nurses’ Health study” (Lionel Opie, Cancer and cardiovascular disease; see Rob M. Van Dam, Combined impact of lifestyle factors on mortality). “Research has shown there are interrelationships among type 2 diabetes, heart disease, and cancer. These interrelationships may seem coincidental and based only on the fact these conditions share common risk factors. However, research suggests these diseases may relate to one another in multiple ways and that nutrition and lifestyle strategies used to prevent and manage these diseases overlap considerably” (Karen Collins, The Cancer, Diabetes, and Heart Disease Link).
Yet other researchers did find the same inverse relationship: “We herein report that, based on two separate medical records analysis, an inverse correlation between cancer and atherosclerosis” (Matthew Li et al, If It’s Not One Thing, It’s Another). But there was an additional point: “We believe that the anti-inflammatory aspect of cancer’s pan-inflammatory response plays an important role towards atherosclerotic attenuation.” Interesting! In that case, one of the key causal mechanisms to be considered is inflammation. Some diets high in animal proteins would be inflammatory, such as the Standard American Diet, whereas others would be anti-inflammatory. Eliminating seed oils (e.g., sunflower oil) would by itself reduce inflammation. Reducing starches and sugar would help as well. So, is it the meat that increases cancer or is it what the meat is being cooked in or eaten with? That goes back to the healthy and unhealthy user effects.
As this confounding factor is central, we might want to consider the increasingly common view that inflammation is involved in nearly every major disease. “For example, inflammation causes or is a causal link in many health problems or otherwise seen as an indicator of health deterioration (arthritis, depression, schizophrenia, etc), but inflammation itself isn’t the fundamental cause since it is a protective response itself to something else (allergens, leaky gut, etc). Or as yet another example, there is the theory that cholesterol plaque in arteries doesn’t cause the problem but is a response to it, as the cholesterol is essentially forming a scab in seeking to heal injury. Pointing at cholesterol would be like making accusations about firefighters being present at fires” (Coping Mechanisms of Health).
What exacerbates or moderates inflammation will be pivotal to overall health (Essentialism On the Decline), especially the nexus of disease called metabolic syndrome/derangement or what used to be called syndrome X: insulin resistance, diabetes, obesity, heart disease, strokes, etc. In fact, other researchers point directly to inflammation as being a common factor of CVD and cancer: “Although commonly thought of as two separate disease entities, CVD and cancer possess various similarities and possible interactions, including a number of similar risk factors (e.g. obesity, diabetes), suggesting a shared biology for which there is emerging evidence. While chronic inflammation is an indispensible feature of the pathogenesis and progression of both CVD and cancer, additional mechanisms can be found at their intersection” (Ryan J. Koene et al, Shared Risk Factors in Cardiovascular Disease and Cancer). But it might depend on the specific conditions how inflammation manifests as disease — not only CVD or cancer but also arthritis, depression, Alzheimer’s, etc.
This is the major downfall of nutrition studies, as the experts in the field find themselves hopelessly mired in a replication crisis. There is too much contradictory research and, when much of the research has been repeated, it simply did not replicate. That is to say much of it is simply wrong or misinterpreted. And as few have attempted to replicate much of it, we aren’t entirely sure what is valid and what is not. That further problemetizes meta-analyses, despite how potentially powerful that tool can be when working with quality research. The study I’ve been discussing here was an ecological study and that has its limitations. The researchers couldn’t disentangle all the major confounding factors, much less control for them in the first place, as they were working with data across decades that came from separate countries. Even so, it’s interesting and useful info to consider. And keep in mind that almost all official dietary recommendations are based on observational (associative, correlative, epidemiological) studies with far fewer controls. This is the nature of the entire field of nutrition studies, as long-term randomized and controlled studies on humans are next to impossible to do.
So, as always, qualifications must be made. The study’s authors state that, “In items of smaller importance (e.g. distilled beverages, sunflower oil, onions), the results are less persuasive and their interpretation is not always easy and straightforward. Similar to observational studies, our ecological study reflects ‘real-world data’ and cannot always separate mutual interactions among the examined variables. Therefore, the reliance on bivariate correlations could lead to misleading conclusions. However, some of these findings can be used as a starting point of medical hypotheses, whose validity can be investigated in controlled clinical trials.” Nonetheless, “The reasonably high accuracy of the input data, combined with some extremely high correlations, together substantially increase the likelihood of true causal relationships, especially when the results concern principal components of food with high consumption rates, and when they can be supported by other sources.”
This data is meaningful in offering strong supporting evidence. The finding about animal foods and starchy foods is the main takeaway, however tentative the conclusion may be for real world application, at least in taking this evidence in isolation. But the inverse correlation of CVD risk and cancer risk stands out and probably indicates confounders across populations, and that would be fertile territory for other researchers to explore. The main importance to this study is less in the specifics and more in how it further challenges the broad paradigm that has dominated nutrition studies for the past half century or so. The most basic point is that the diet-heart hypothesis simply doesn’t make sense of the evidence and it never really did. When the hypothesis was first argued, heart disease was going up precisely at the moment saturated fat intake was going down, since seed oils had replaced lard as the main fat source in the decades prior. Interestingly, lard has been a common denominator among most long-lived populations, from the Okinawans to Rosetans (Ancient Greek View on Olive Oil as Part of the Healthy Mediterranean Die, Blue Zones Dietary Myth).
This study is further support for a new emerging understanding, as seen with the American Heart Association backing off from its earlier position (Slow, Quiet, and Reluctant Changes to Official Dietary Guidelines). Fat is not the enemy of humanity, as seen with the high-fat ketogenic diet where fat is used as the primary fuel, instead of carbohydrates (Ketogenic Diet and Neurocognitive Health, The Ketogenic Miracle Cure, The Agricultural Mind). In fact, we wouldn’t be here without fat, as it is the evolutionary and physiological norm, specifically in terms of low-carb (Is Ketosis Normal?, “Is keto safe for kids?”). Instead, that too many carbohydrates are unhealthy used to be common knowledge (American Heart Association’s “Fat and Cholesterol Counter” (1991)). Consensus on this shifted a half century ago, the last time when low-carb diets were still part of mainstream thought, and now we are shifting back the other way. The old consensus will be new again.
* * *
- Cholesterol levels were tightly correlated to the consumption of animal fats and proteins – Countries consuming more fat and protein from animal sources had higher incidence of raised cholesterol
- Raised cholesterol correlated negatively with CVD risk – Countries with higher levels of raised cholesterol had fewer cases of CVD deaths and a lower incidence of CVD risk factors
- Carbohydrates correlated positively with CVD risk – the more carbohydrates consumed (and especially those with high GI such as starches) the more CVD
- Fat and Protein correlated negatively with CVD risk – Countries consuming more fat and protein from animal and plant sources had less CVD. The authors speculate that this is because increasing fat and protein in the diet generally displaces carbohydrates.
Food consumption and the actual statistics of cardiovascular diseases: an epidemiological comparison of 42 European countries
Pavel Grasgruber,* Martin Sebera, Eduard Hrazdira, Sylva Hrebickova, and Jan Cacek
We found exceptionally strong relationships between some of the examined factors, the highest being a correlation between raised cholesterol in men and the combined consumption of animal fat and animal protein (r=0.92, p<0.001). The most significant dietary correlate of low CVD risk was high total fat and animal protein consumption. Additional statistical analyses further highlighted citrus fruits, high-fat dairy (cheese) and tree nuts. Among other non-dietary factors, health expenditure showed by far the highest correlation coefficients. The major correlate of high CVD risk was the proportion of energy from carbohydrates and alcohol, or from potato and cereal carbohydrates. Similar patterns were observed between food consumption and CVD statistics from the period 1980–2000, which shows that these relationships are stable over time. However, we found striking discrepancies in men’s CVD statistics from 1980 and 1990, which can probably explain the origin of the ‘saturated fat hypothesis’ that influenced public health policies in the following decades.
Our results do not support the association between CVDs and saturated fat, which is still contained in official dietary guidelines. Instead, they agree with data accumulated from recent studies that link CVD risk with the high glycaemic index/load of carbohydrate-based diets. In the absence of any scientific evidence connecting saturated fat with CVDs, these findings show that current dietary recommendations regarding CVDs should be seriously reconsidered. […]
Irrespective of the possible limitations of the ecological study design, the undisputable finding of our paper is the fact that the highest CVD prevalence can be found in countries with the highest carbohydrate consumption, whereas the lowest CVD prevalence is typical of countries with the highest intake of fat and protein. The polarity between these geographical patterns is striking. At the same time, it is important to emphasise that we are dealing with the most essential components of the everyday diet.
Health expenditure – the main confounder in this study – is clearly related to CVD mortality, but its influence is not apparent in the case of raised blood pressure or blood glucose, which depend on the individual lifestyle. It is also difficult to imagine that health expenditure would be able to completely reverse the connection between nutrition and all the selected CVD indicators. Therefore, the strong ecological relationship between CVD prevalence and carbohydrate consumption is a serious challenge to the current concepts of the aetiology of CVD.
The positive effect of low-carbohydrate diets on CVD risk factors (obesity, blood lipids, blood glucose, insulin, blood pressure) is already apparent in short-term clinical trials lasting 3–36 months (58) and low-carbohydrate diets also appear superior to low-fat diets in this regard (36, 37). However, these findings are still not reflected by official dietary recommendations that continue to perpetuate the unproven connection between saturated fat and CVDs (25). Understandably, because of the chronic nature of CVDs, the evidence for the connection between carbohydrates and CVD events/mortality comes mainly from longitudinal observational studies and there is a lack of long-term clinical trials that would provide definitive proof of such a connection. Therefore, our data based on long-term statistics of food consumption can be important for the direction of future research.
In fact, our ecological comparison of cancer incidence in 39 European countries (for 2012; (59)) can bring another important argument. Current rates of cancer incidence in Europe are namely the exact geographical opposite of CVDs (see Fig. 28). In sharp contrast to CVDs, cancer correlates with the consumption of animal food (particularly animal fat), alcohol, a high dietary protein quality, high cholesterol levels, high health expenditure, and above average height. These contrasting patterns mirror physiological mechanisms underlying physical growth and the development of cancer and CVDs (60). The best example of this health paradox is again that of French men, who have the lowest rates of CVD mortality in Europe, but the highest rates of cancer incidence. In other words, cancer and CVDs appear to express two extremes of a fundamental metabolic disbalance that is related to factors such as cholesterol and IGF-1 (insulin-like growth factor).
Besides total fat and protein consumption, the most likely preventive factors emerging in our study include fruits (particularly citrus fruits), wine, high-fat dairy products (especially cheese), sources of plant fat (tree nuts, olives), and potentially even vegetables and other low-glycaemic plant sources, provided that they substitute high-glycaemic foods. Many of these foodstuffs are the traditional components of the ‘Mediterranean diet’, which again strengthens the meaningfulness of our results. The factor analysis (Factor 3) also highlighted coffee, soybean oil and fish & seafood, but except for the fish & seafood, the rationale of this finding is less clear, because coffee is strongly associated with fruit consumption and soybean oil is used for various culinary purposes. Still, some support for the preventive role of coffee does exist (61) and hence, this observation should not be disregarded.
Similar to the “Mediterranean diet”, the Dietary Approaches to Stop Hypertension (DASH) diet, which is based mainly on fruits, vegetables, and low-fat dairy, also proved to be quite effective (62). However, our data indicate that the consumption of low-fat dairy may not be an optimal strategy. Considering the unreliability of observational studies highlighting low-fat dairy and the existence of strong bias regarding the intake of saturated fat, the health effect of various dairy products should be carefully tested in controlled clinical studies. In any case, our findings indicate that citrus fruits, high-fat dairy (such as cheese) and tree nuts (walnuts) constitute the most promising components of a prevention diet.
Among other potential triggers of CVDs, we should especially stress distilled beverages, which consistently correlate with CVD risk, in the absence of any relationship with health expenditure. The possible role of sunflower oil and onions is much less clear. Although sunflower oil consistently correlates with stroke mortality in the historical comparison and creates very productive regression models with some correlates of the actual CVD mortality, it is possible that both these food items mirror an environment that is deficient in some important factors correlating negatively with CVD risk.
A very important case is that of cereals because whole grain cereals are often propagated as CVD prevention. It is true that whole grain cereals are usually characterised by lower GI and FII values than refined cereals, and their benefits have been documented in numerous observational studies (63), but their consumption is also tied with a healthy lifestyle. All the available clinical trials have been of short duration and have produced inconsistent results indicating that the possible benefits are related to the substitution of refined cereals for whole grain cereals, and not because of whole grain cereals per se (64, 65). Our study cannot differentiate between refined and unrefined cereals, but both are highly concentrated sources of carbohydrates (~70–75% weight, ~80–90% energy) and cereals also make up ~50% of CA energy intake in general. To use an analogy with smoking, a switch from unfiltered to filtered cigarettes can reduce health risks, but this fact does not mean that filtered cigarettes should be propagated as part of a healthy lifestyle. In fact, even some unrefined cereals [such as the ‘whole-meal bread’ tested by Bao et al. (32)] have high glycaemic and insulin indices, and the values are often unpredictable. Therefore, in the light of the growing evidence pointing to the negative role of carbohydrates, and considering the lack of any association between saturated fat and CVDs, we are convinced that the current recommendations regarding diet and CVDs should be seriously reconsidered.