Heart Disease and The Mediterranean Diet(s)
Our nutrition voyage begins in the city of Lyon. It was here in France’s second largest city that a team of researchers, led cardiologist Michel de Lorgeril, recruited 605 people to participate in what would be a groundbreaking clinical experiment. The individuals in this study had their differences; they weren’t all the same age (although most were middle aged), they weren’t the same gender (although most were men), they had differences relating to education level, racial background, cholesterol levels, blood pressure, height, weight, and so and on. However, there was one similarity which all these people shared, one life experience which became the the reason for their inclusion in de Lorgeril’s study: they had all experienced a myocardial infarction, commonly known as a heart attack.
Before we continue this story in Lyon it seems worth briefly exploring the process of a heart attack. It’s an unfortunate fact that in the United States, starting around adulthood, the risk of experiencing a heart attack increases every single year we’re alive. This is likely because as we age we experience the continued development of atherosclerosis. Atherosclerosis, in excessively simple terms, results from the process of lipid (mostly cholesterol) loaded white blood cells being deposited under the surface of the artery wall. These deposits grow progressively bigger and less stable until, at a certain point, the wall of the artery can be weakened so much that it ruptures, releasing the contents. These contents cause blood cells to clump together, forming a blood clot which can travel through the blood stream to the heart. Upon reaching the heart this traveling blood clot, or embolus, can block up blood flow to the heart like a drain stopper in a bathtub. Without adequate blood flow, the heart is deprived of oxygen, and without adequate oxygen bad things happen.
I’ll try not to bore anyone with too much unnecessary biology here, but you need oxygen to live. If you deprive your cells of oxygen their ability to make energy is compromised and without energy your cells die. That’s why drowning kills you—no more oxygen to to your brain, so your brain dies. That’s why blood loss can kill you—without enough blood to carry oxygen your brain isn’t sufficiently oxygenated so it dies. And that’s why a heart attack can kill you—when the embolus blocks too much blood flow your heart is deprived of energy until it stops being able to pump blood to your brain. No blood to your brain, again, you die. Obviously people can survive a heart attack, our hearts can be resilient, but if you do have have a heart attack it probably means you have serious atherosclerosis and thus a greater chance of having and possibly dying from a future heart attack.
So we return to our group of heart attack survivors. Each of them, sometime after being admitted to the hospital, agreed to follow one of two diets designed to help prevent the occurrence of a future heart attack. One diet was designed by the American Heart Association (AHA) and called for a somewhat low fat diet, with each of the three types of fat (saturated, monounsaturated, and polyunsaturated) represented equally and a dietary cholesterol intake below 300 milligrams a day (a single egg yolk, by comparison, has about 180 milligrams of cholesterol). The other diet administered was a “Mediterranean Diet”. These days Mediterranean diets are considered by many to be healthy, but it was a lengthy journey to this recognition.
In the 1950’s a nutrition researcher from Minnesota named Ancel Keys started an observational study to look at the role of lifestyle factors (including diet) on cardiovascular disease risk in various countries throughout the world. One observation Keys’ made was that people living in countries bordering the Mediterranean Sea, including Spain, Italy, and Greece, had a surprisingly low incidence of cardiovascular disease compared to many other countries he looked at. Inhabitants of such countries also tended to follow a fairly similar dietary pattern, emphasizing vegetables, bread (“never more than a few hours from the oven” in the words of Keys), fruit, fish, beans, nuts, olive oil, potatoes and more white meat in place of red meat. Red wine was also a common component. Keys’ postulated that if people adopted this diet it might help prevent the occurrence heart attacks and strokes.
As with all such theories there were detractors. Opponents of Keys’ theory suggested other explanations might explain the cardiovascular benefits seen in Mediterranean countries. Such places, it was argued, are infused with a culture of physical activity, things like walking, bicycling, and swimming. These activities could very well promote a healthy arterial system independently of diet. The abundant sunshine of the Mediterranean was also suggested as an overlooked disease fighting agent. Finally, the practical implications of a dietary prescription which actually encompasses many different diets with some significant variations was seen as difficult. Needless to say, the “Mediterranean Diet” didn’t catch on immediately.
It is quite likely that the event most instrumental in pulling the linchpin from the Mediterranean diet grenade and demonstrating its healthful powers was the Lyon Diet Heart Study (LDHS), the experiment which is the main focus of this chapter. If you want to understand how the Mediterranean diet impacts health, this study is probably the most important. Heck, if you asked me what the most important study on nutrition and heart disease is I might say it’s this one.
Before getting to the results of the Lyon Diet Heart Study let’s get more into the specifics of the diet. Participants randomly assigned to the Mediterranean diet group were advised to eat bread, vegetables (particularly root and green vegetables), fruit, fish, and poultry in place of red meat. Moderate amounts of red wine at meals was allowed (but not required). Participants were also told to use olive oil and canola oil as their sole source of added fat, however this posed somewhat of a problem. Remember, this study took place in France. De Lorgeril and his colleagues worried that asking a French person to participate in a study which prohibited them from putting butter on their bread might not go over so well. Although I like to imagine cries of “Sacre bleu!” and angrily thrown berets the more likely response probably would have been polite declining. So the researchers came up with a possible solution. They developed a margarine made primarily of canola oil and gave it to the subjects to use in place of butter.
So what was the effect of diet on these heart attack survivors? The researchers planned to stop the study after 5 years to find out. However after only 27 months it was pretty clear what was going on. Compared to the AHA diet group the group eating the Mediterranean diet had experienced about 70% less heart attacks, 70% less cardiovascular disease deaths (deaths from heart attacks and strokes), and 70% less total deaths. If you find yourself unimpressed by such a reduction in death and disease I suggest you go looking for another diet study where such a reduction was surpassed or even equalled. When you ultimately turn up nothing you may find yourself a bit more awed by the numbers presented here.
This is far from the first and certainly not the last clinical diet study carried out but none have matched up to it in terms of risk reduction. So the pertinent question is certainly: why? What elements of the LDHS diet made it so protective? There’s no reason to think every single part of the diet was important, the diet could have been healthful in spite of certain aspects. So what parts of the LDHS diet were the most important? There are many contenders, so let’s see if we can siphon them out.
Fruits, Vegetables, and Antioxidants
The increase in fruit and vegetable meant that the Med diet group had a greater intake of vitamin C, vitamin E, and other plant derived antioxidants. These compounds have received a fair bit of attention as potential heart disease preventing agents, so they represent a possible mechanism explaining the results of the LDHS. Vitamin C, for example, was recommended as a key component of a heart healthy lifestyle by the Nobel Prize winning scientist Linus Pauling. Pauling suggested that atherosclerosis might be at its core partly, even primarily, a disease caused by inadequate vitamin C. Unfortunately, nearly every clinical study to date has observed no significant benefit of vitamin C supplements on heart disease risk. Vitamin E has shown some benefits depending on the dosage used, but such benefits are generally small and often inconsistent. Beta carotene, an antioxidant and vitamin A precursor found in fruits and vegetables, has also produced no beneficial effects on heart disease risk in clinical trials. Whether taken alone or in combination these antioxidants are no magic bullet preventing heart disease.
At this point it would fair to point out that isolated supplements may not adequately reflect the complex interactions of all the potentially beneficial compounds in fruits and vegetables. Thus, let us examine a somewhat controversial study known as the Women’s Health Initiative. A dutiful student of nutrition will likely be aware of at least some aspect of the dietary modification trial of WHI. The study was a massive undertaking utilizing more than 50,000 older women in one of the largest randomized clinical trials ever conducted. One group was told to eat their usual diet while the other, the experimental group, was advised to eat less fat and increase their intake of fruits, vegetables, and grains. However, unlike in the LDHS no reduction in cardiovascular disease risk was observed. This by no means proves a diet higher in fruits, vegetables doesn’t reduce cardiovascular disease risk, but the results of the WHI trial seem to suggest such dietary modifications cannot fully explain the benefits seen in the Lyon Diet Heart study. The effects could be plant specific, as some evidence has found leafy green vegetables, which were specifically recommended in the LDHS, are often the vegetable most strongly associated with a lesser risk of cardiovascular disease.
Saturated Fat and the Lipid Hypothesis
Another possible explanation might be in the fat. It has been proposed by some that the results of the LDHS were due to a reduction in saturated fat, a type of fat which the CDC, the American Heart Association, and the Department of Health and Human Services recommends avoiding. To explain the reasoning behind this viewpoints I fear I must go on another prolonged tangent, so bear with me. Since as far back the 1800’s it has been observed that people with heart disease (and the accompanying atherosclerosis) often have greater amounts of cholesterol in their blood than what is typical. However, back then it was unclear whether this cholesterol caused the atherosclerosis and heart disease or whether it was simply a byproduct of the underlying disease process. To paraphrase the poet Ogden Nash one must be cautious not to conclude from simple observation that shaking tree branches cause wind.
This started to change when Russian scientist Nikolai Anitschkov produced convincing evidence that elevated blood cholesterol preceded the development of heart disease. In a series of experiments which would make a member of PETA cry Anitschkov attempted to produce cardiovascular disease in rabbits by constricting, cutting, pulling, and burning their blood vessels, by severing their nerves, and by injecting them with a kitchen sink of toxins, including digitalin, strophanthin, adonidin, ergotin, theocin, barium chloride, hydrastin, nicotine, and formalin (side note: these are all good Scrabble words). While many of these methods resulted in arterial damage and thickening none of them resulted in the type of arterial plaque seen in human atherosclerosis. However, in 1913 Anitschov did produce atherosclerosis resembling that seen in humans by feeding rabbits large amounts of cholesterol. Unlike humans, when rabbits ingest cholesterol they experience a massive increase in their blood cholesterol and subsequently develop atherosclerosis. Future experiments have noted that raising blood cholesterol levels in baboons, cats, chickens, chimpanzees, dogs, goats, guinea pigs, hamsters, monkeys, mice, parrots, pigs, pigeons, rabbits and rats (by way of diet change, thyroid gland removal, and other means) quite often produces atherosclerosis.
So the hypothesis that elevated blood cholesterol (in the form of a cholesterol carrier called LDL) causes atherosclerosis and heart disease was born, eventually termed the lipid hypothesis. Further support for the lipid hypothesis came from genetic studies. It was observed that people with gene mutations resulting in heavily raised LDL cholesterol in the blood (a condition referred to as familial hypercholesterolemia) had a much higher risk of developing heart disease. Mutations to genes (such as PCSK9) which resulted in lower cholesterol levels, on the other hand, were associated with a lower risk of developing heart disease.
Anyway, my intention is not to belabor the arguments for the lipid hypothesis, merely explain that there is some evidence behind it. This is important in understanding why many groups consider saturated fat a health threat. It has been consistently observed that saturated fat raises LDL cholesterol levels in the blood when ingested. This observation led to the development of the diet-heart hypothesis, which suggested that because saturated fat raises LDL and increased LDL seems to promote cardiovascular disease saturated fat will therefore increase cardiovascular disease risk. However, regardless of the legitimately of the lipid hypothesis or the diet-heart hypothesis (both have been met with criticisms I won’t delve into here) both are not explanatory factors in this scenario because in the Lyon Diet Heart Study LDL cholesterol levels were the same in both groups. The LDHS is an interesting study because it shows that diet can powerfully protect against heart disease without affecting cholesterol levels.
This isn’t to say saturated fat doesn’t have harmful effects independently of its effects on LDL (and some have suggested this is the case), which would mean saturated fat isn’t entirely off the hook. However, if saturated fat were entirely determining the results of the LDHS then one might expect to have seen a reduction in cardiovascular disease in the WHI trial where saturated fat was similarly reduced. Clearly other factors were at play.
I must begin this section by discussing some of the different types of dietary fat. Fatty acids can be thought of as chains of carbon molecules attached to each other. Dietary fats are often designated into groups based on the number of double bonds (rather than single bonds) between carbon molecules contained in their molecular structure. Saturated fats have no such double bonds, monounsaturated fats have one double bond, whereas polyunsaturated fats have more than one double bond, thus the poly- prefix. Various types of polyunsaturated fats exist. One such type is linoleic acid (LA), an omega-6 fatty acid (named because it has its first double bond on the 6th carbon down the carbon chain from its omega end). This fat was reduced in the diet of the Med group of the LDHS. Although linoleic acid rich foods and oils have been frequently recommended for their cholesterol lowering effects a significant body of research suggests LA may negatively effect cardiovascular disease risk.
In 2013 clinical researcher Christopher Ramsden and associates published a paper in the British Medical Journal wherein he looked at randomized, clinical trials in which linoleic acid based oils (like safflower oil and corn oil) replaced saturated fat based oils (like butter) and calculated that the risk of death from coronary heart disease was 33 percent higher when people consumed the linoleic acid based oils. So it seems the reduction in linoleic acid in the LDHS Med diet may have been responsible for some of its benefits.
Further evidence for this comes from the Oslo Diet Heart Study (ODHS). The ODHS study was similar in many ways to the LDHS: it recruited heart attack survivors and asked half of them to follow a diet designed to prevent future heart attacks. This experimental diet called for increased fruits, vegetables, and brown bread, and meat was discouraged in favor of fish (particularly shellfish and oily fish like sardines). However, in the ODHS soybean oil was provided to subjects in the experimental group instead of the olive oil/canola oil combo as in the LDHS. For those unaware soybean is fairly rich in linoleic acid and as a result the ODHS diet’s main difference from the LDHS diet was an increase in linoleic acid (instead of an increase in oleic acid, the main fat type increased in the LDHS diet). After five years those subjects following the ODHS’s experimental diet had died from heart attacks 56 percent less than the control group, an impressive result. However, they did not experience a particularity impressive reduction in non-fatal heart attacks or total cardiovascular disease related deaths compared to the subjects in the LDHS. Given this I think the Oslo Diet Heart Study provides compelling evidence that eating less linoleic acid was an important reason for the Lyon Diet Heart Study’s benefits. However, linoleic acid reduction was clearly not the sole factor or the ODHS wouldn’t have observed the benefits it did. Furthermore, if linoleic acid was the only thing that mattered there probably should have been similar benefits seen in the WHI trial which also reduced dietary linoleic acid, but there wasn’t.
Omega-3 Fatty Acids: The Alpha and the Omega 3
The next thing we need to cover is omega-3 fatty acids. Most people are at least somewhat familiar these fats, considering that omega-3 containing fish oil pills are, by some estimates, the most widely taken nutritional supplement in the United States. Like linoleic acid, omega-3 fats are polyunsaturated (containing multiple double bonds) but unlike linoleic acid omega-3 fats have their first double bond only 3 carbons down their carbon chain from their omega end. There are various omega-3 fatty acids, differing primarily by the length of their carbon chain. ALA has 18 carbons, EPA has 20 carbons, and DHA has 22 carbons. EPA and DHA are found in fish (and are almost always together as a pair) while ALA in found in things like flax seeds, chia seeds, and canola oil.
In the 1970’s and ’80s it was observed that Inuit from Greenland had an extremely low incidence of heart disease. This led some researchers to propose that their large intake of omega-3 rich fat from marine animals played a role in this apparent protection. Observational evidence would later report that eating more omega-3 fatty acids (EPA and DHA) was often associated with a decreased risk of heart disease. Ultimately this led to large, randomized clinical trials testing the effects of ingesting omega-3’s from fish oil on cardiovascular disease risk. Generally, fish oil performed quite well in this respect. Let’s touch on two of the largest of such trials.
The Diet and Reinfarction Trail (or DART) recruited 2033 subjects who had previously had a heart attack and had half of them eat more fatty fish or take a fish oil supplement. After 2 years this group experienced a 25% reduction in cardiovascular events and a 29% reduction in mortality. The GISSI trial also randomized heart attack survivors into different groups, 2828 into a control group and 2836 into an intervention group. For about 3 and a half years the intervention group took a 1 gram omega-3 supplement while the control group took a placebo. This study found that taking the omega-3 supplemented group experienced 11% less cardiovascular events and 14% less deaths than the control group.
Given that part of diet advice for the Med group in the LDHS was to increase fish intake it’s possible, even likely, that some of the benefit came from increased EPA and DHA intake. However, the amount of the omega-3’s wasn’t reported, so their significance to the overall effect is difficult to determine. However, the other omega-3 fat, ALA, was reported in the LDHS. Compared to the other diet group the Med diet group was ingesting about 1.8 grams of ALA, about three times more.
ALA hasn’t received the same amount of attention as it’s fish oil bound cousins EPA and DHA but some evidence does suggest it has benefits. Observational evidence has often found a moderate association between ALA intake and a decreased risk of coronary heart disease risk. A meta analysis published in the American Journal of Clinical Nutrition in 2012 reported people with a higher ALA intake were 20 percent less likely to die from coronary heart disease. ALA also appears to reduce inflammation, a quality it shares with fish oil. I couldn’t hope to succinctly summarize the arguments for and against the role of inflammation in atherosclerosis (although I could point out that a variety of anti-inflammatory drugs, including aspirin, show some ability to reduce heart disease risk) but needless to say this anti-inflammatory action of ALA might be indicative of it’s possible benefits. However, ALA’s true test should be randomized, controlled trials. There are far less of these than those on fish oil, but there are a few.
Soybean oil is the oil extracted from soybeans. “Wow I never would have guessed” you might be sarcastically thinking. Okay, well let me tell you something less well known; soybean oil is a prominent source of ALA. It’s not overflowing with ALA, but it has a decent amount. With that in mind let’s consider two studies using soybean oil. The first is the LA Veterans trial, a double blind study which included 846 men and had half of them replace dairy fat (e.g. butter) in their diet with soybean oil. After 8 years rates of death from cardiovascular disease was lower in the soybean oil group, although some criticisms of this study exist. For one thing the soybean oil group had less heavy and moderate smokers. Additionally, due to repeated heating the researchers suggested the butter in the dairy fat group was depleted of vitamin E and may have influenced the results. Nonetheless, this study is consistent with a protective effect of ALA considering the soybean oil group was ingesting more linoleic acid, a dietary factor which we’ve seen seems to have harmful effects on cardiovascular health.
The other study using soybean oil was the Finnish Mental Hospital Study. It reported a reduction in the occurrence of cardiovascular disease events when soybean oil replaced dairy fat in the diet of patients in a mental hospital but this study also has criticisms. The soybean oil group was possibly advantaged by eating less refined sugar and being less commonly prescribed thioridazine, an anti-psychotic drug which was later taken off the market due in part to its ability to cause heart problems. But, since the soybean oil group was again consuming more linoleic acid this group wasn’t free from disadvantage.
The last study to consider was carried out in India. It compared the effect of a supplement of fish oil, an ALA rich mustard oil supplement, and a placebo in 360 people who had experienced a suspected heart attack. After a year the risk of a non-fatal heart attack was about 40 percent less in the fish oil and mustard oil groups compared to the placebo group. However, risk of death from cardiovascular disease related causes was lower in the fish oil group only. Nonetheless, this study strongly supports the hypothesis that ALA protects against heart disease.
Red Meat and Poultry
Part of the Lyon Diet Heart Study was the recommendation to replace red meat with poultry. Unfortunately, there is no experimental evidence examining how red meat alone affects heart disease incidence. Observational evidence on the matter is mixed. Although many studies have shown red meat is associated with a greater risk of heart disease, these studies often group fresh red meat with processed red meat (like bacon) despite some obvious differences between the two (salt content being one such differences). Furthermore, in these population studies a higher intake of red meat quite often correlates with health behaviors thought to be potentially damaging, like more alcohol consumption, cigarette smoking, and physical inactivity. When studies attempt to account for such factors fresh red meat is typically found to have no association with heart disease risk. However, even if such associations (or lack thereof) were definitive this doesn’t mean meat from poultry isn’t better than red meat. There will be no clear answer to the question of which color of meat is healthier here, but if you’ll indulge me I’d like to cover a speculative reason why red meat might differ from white meat in its effects in health.
One difference between meats like beef and meats like chicken is iron. Red meat has a bit more iron than poultry. What is the significance of iron to heart disease? I’ve covered the lipid hypothesis which states that higher LDL levels increases atherosclerosis and cardiovascular disease. However, despite the biological probability of the lipid hypothesis it appears to quite often fall short in its ability to predict cardiovascular disease in various populations, even when other CVD risk factors are accounted for. This oddity seemed explainable however, after a series of experiments came out in the late ’70’s and early ’80’s. It appears that in the process of atherosclerosis white blood cells called macrophages take up LDL and this process is largely responsible for the plaque build up responsible for heart disease. However, It was found that macrophages appear to take up LDL when it has been oxidatively modified. This led some to propose an amendment to the lipid hypothesis stating that oxidized LDL, not normal LDL, promotes heart disease. This would explain why increased LDL causes atherosclerosis—higher LDL indicates a slower transit time of LDL through the bloodstream and thus a greater exposure to the oxidative stress.
A number of studies appear to support the oxidative lipid hypothesis, finding that levels oxidized LDL in the blood better predicts heart disease than levels of total LDL in the blood. This hypothesis also explains why linoleic acid lowers LDL but may increase heart disease risk. Polyunsaturated fats, like linoleic acid, are more sensitive to oxidation and when consumed are incorporated into LDL making LDL in the blood more susceptible to oxidation. The oxidized lipid hypothesis does fail to explain the protective effects of omega-3 fatty acids (which are also sensitive to oxidation), although the oxidized lipid hypothesis is unlikely to be the only factor relevant to the complicated condition that is heart disease (inflammation is another such candidate).
Iron may have some relevance to the oxidized lipid hypothesis. Iron is a reactive metal, prone to oxidizing various substances. Supplementing large amounts of iron has been shown to increase the oxidation of LDL as well as its susceptibility to oxidation. Various observations also seem to suggest iron may promote heart disease in some contexts. First: women, ladies, females. Whatever you call them they have a much lesser risk of heart disease than men, at least until they reach menopause. For many years the major explanation for this observation was the idea that female hormones, like estrogen, were protective. However, after a number of clinical trials failed to find such hormones protected against heart disease risk the popularity of this idea has faltered. An alternative theory is that pre-menopausal women, through their monthly menstruation, are losing iron and this is at least partially responsible for their reduced heart disease risk compared to males. On the other side of the coin is hemochromatosis, a disorder in which iron levels accumulate excessively. Hemochromatosis is often associated with an increased risk of cardiovascular disease, supporting the role of excess iron in this disease.
However, the FeAST trial was a clinical trial using blood draws to reduce iron stores in men and older women with peripheral artery disease and failed to find a reduction in cardiovascular events, although their were some apparent benefits in specific subgroups. Furthermore, while a number of studies have reported associations between bodily iron levels and heart disease many have not. This contradiction may be explained by considering the interaction of LDL and iron. When researchers analyzed data from the NHANES II population study they had a curious finding. After controlling from possible confounding factors high LDL (above 160 mg/dL) was not significantly associated with cardiovascular disease. High transferrin saturation, a marker of iron levels in the body, was also not significantly associated with cardiovascular disease. However, people with both high LDL and high iron levels were 5 and a half times more likely to have died from cardiovascular disease. Wow! This interaction may be explained by the fact that people with low LDL and high iron avoided heart disease by having LDL that didn’t spend enough time in the blood stream to be oxidized by the iron. Alternatively, those with high LDL and low iron didn’t have a significant source of oxidative stress so their LDL wasn’t significantly oxidized and they didn’t develop heart disease.
So, because red meat can contribute to increased iron levels it might be more harmful than white meat like poultry in certain contexts (like when bodily iron levels are too high). However, whether this is necessarily true is far from clear and likely requires further study because we can say for sure. Furthermore, red meat is not without its potentially beneficial nutrients. It contains about 10 times more carnitine than most white meats like chicken. Although carnitine has been the subject of some criticisms recently which I won’t be delving into, carnitine shows a number of health benefits. In randomized clinical trials on heart attack survivors carnitine has reduced cardiovascular disease events and reduced the death rate. Carnitine has also demonstrated similar mortality reductions in a clinical study on people with heart failure. So to suggest red meat is entirely without mechanisms which could make it beneficial compared to poultry is a dubious claim.
Monounsaturated Fat, Olive Oil, and Statistical Chance
One of the main aspects of the Lyon Diet Heart Study diet touted as beneficial is the increased amount of monounsaturated fat. Monounsaturated fat is often recommended as a healthier alternative to saturated fat by groups like the American heart Association due largely to its ability to reduce LDL cholesterol levels. However, while a number of observational studies have reported a greater intake of monounsaturated fat is associated with a less cardiovascular disease such results are far from consistent. Many animal experiments have also failed to demonstrate a beneficial effect of monounsaturated fat on atherosclerosis development when compared to saturated fat. Furthermore, a clinical trial published in 1965 reported no effect on cardiovascular disease events when a monounsaturated fat rich diet (from olive oil) was compared to a saturated fat rich diet, although this study had barely over 2 dozen people per group and the monounsaturated fat group also reduced their consumption of carbohydrates and protein. Nevertheless, the evidence demonstrating monounsaturated fat as beneficial for heart disease is somewhat weak.
However, consider this: the chemical differences between a living person and a recently deceased person are actually quite small, but these differences seem to matter quite a bit. Similarly, refined olive oil and virgin olive oil are both primarily composed of monounsaturated fat but if you think this means they’re indistinguishable in their health effects I might ask if you’d also enjoy having coffee with a corpse (I acknowledge there are some misanthropes out there who prefer people laconic rather than loquacious, in which case a cadaver might be a preferable companion). The point is, virgin olive oil contains a considerable quantity of chemicals with beneficial effects which are less abundant in refined olive oil produced using high heats and hydrocarbon solvents. Likely for this reason virgin olive oil has been shown to reduce LDL oxidation and lower inflammation markers compared to refined olive oil, other monounsaturated fats, and some saturated fats.
Consider the PREDIMED trial. I’ll be heavily simplifying the methods of this study for the purpose of brevity. There were three diet groups, each with over 2 and a half thousand people, with one group differing primarily due to an increase in virgin olive oil consumption. Compared to a control group (which consumed slightly more calories from carbohydrates instead of olive oil) the virgin olive oil experienced 30 percent less strokes 28 percent less cardiovascular events. A quick note on statistical significance: although the virgin olive oil group also experienced roughly 30 percent less cardiovascular death and 15 percent less total death than the control group the researchers calculated that there was a 15 percent and 11 percent chance these results could have been due to statistical chance, respectively. In most scientific fields, a result must be less than 10 percent likely to have occurred due to chance to be considered statistically significant (often a percentage even less than this is required).
Confused by this concept? Consider this: someone asks you to determine if their penny has a picture of Abraham Lincoln on both sides. They agree to flip it as many times as you ask them to. Obviously as soon as you see a tail side come up you know it isn’t a double headed penny. However, if you keep seeing heads come up you really can’t say for sure the penny is two headed, but each time it lands heads up decreases the likelihood of there being a tails side. So how many times would you have to see the coin land with honest Abe’s face up before you felt confident saying there was no tails side? In science the answer to this question is often about 4 or 5 times because this corresponds to a ≤10% chance of there being a tails side (a coin with only one head side has a 50% chance of landing heads up when flipped. Therefore said coin has a 6.25% chance of landing heads up 4 times in a row and a 3.125% of landing heads up 5 times).
Nevertheless, I think the PREDIMED study is decent evidence that virgin olive oil consumption could have have played some part in the outcome of the LDHS. The role of monounsaturated fat itself, however, remains unclear, even if the hypothesis that it is protective would provide another explanation for the apparent superiority of the LDHS diet over the diet of Oslo Diet Heart Study (which didn’t increase monounsaturated fat intake) and the WHI study (which decreased monounsaturated fat intake).
Next on our list of foods and nutrients is whole grains. Among the many beliefs about nutrition, the idea that whole grains are healthy, or at least healthier than refined grains, is without a doubt one of the most popular. The recommendation to eat whole grains was also part of the Lyon diet. Even though we don’t know for sure if participants really followed this recommendation, is it possible they did and benefited from it?
First, what is a whole grain? A grain consists of a bran, a germ, and an endosperm. The bran is a structure on the outer layer of the grain which plays a part in protecting the rest of the grain. For this reason the bran is often rich in fiber to make it tough and resilient to the elements. The bran is also rich in magnesium. Next is the germ, which as the name implies is the reproductive part of the grain (think of it as the egg yolk of the grain). Due to this biological function the germ is richer in many vitamins and minerals than the other parts of the grain. Finally, we have the endosperm, the typically starch rich center of the grain. Whole grains contain all three of these portions of the grain, while refined grains (like white bread and white rice) have the bran and endosperm removed, leaving the endosperm. As a result, whole grains contain greater amounts of several nutrients. This is typically the basis for recommending whole grains. However, a few small studies have suggested that some of the nutrients in whole grains appear to be less absorbable and certain components in whole grains may actually promote the loss of certain nutrients, including calcium, magnesium, zinc, phosphorus, and vitamin D.
Evidence for the protective role of whole grains usually comes from the massive amount of observational studies showing whole grain consumption is typically correlated with less heart disease. However, because whole grains have such a good reputation and because they are not as tasty as refined grains (ok, this is a subjective claim, whatever) those who eat more whole grains are usually very health conscious. Whole grain eaters smoke less, exercise more, eat more fruits and vegetables, and so on. Thus, whole grains eaters may less heart disease because of their other healthy behaviors not because of the type of bread they eat.
There are experiments showing whole grains, compared to refined grains, produce beneficial effects such as a reduction in blood pressure. However, some experiments have shown negative effects, including a raising of LDL as well as potentially increased LDL oxidation. Looking at such studies doesn’t exactly show one grain type is the clear winner in terms of effects on risk factors for heart disease.
There is one randomized, controlled study on whole grains called the DART trial. If you recognize that name it’s because I already mentioned it as a study on omega-3’s. The DART trial was actually several independent studies wrapped up as one. The study I’ll be talking about here asked 1017 male heart attack survivors to eat whole grains, while a similar group of heart attack survivors were to continue eating refined grains*. After 2 years the number of men who died from coronary heart disease in the whole grain eating group was 28 percent higher than the other group. Although the likelihood this result was due to chance didn’t reach the 5% threshold the researchers were looking for it was very close. So while there is some debate as to whether this study found whole grains to be harmful, they certainly weren’t helpful.
The effect of whole grains on nutrient status may be an interesting area of future study. A bettering and worsening of the status of certain nutrients might mean whole grains could have different effects on health (good or bad) in different contexts. Nutrition researcher Sir Robert McCarrison carried out experiments where he found wheat germ could promote tissue calcification (considered an important driver of atherosclerosis) in rats but this effect could be ameliorated by supplying the rats greater amounts of certain nutrients (like vitamin A).
While the idea whole grains could actually be harmful might be an improbable notion to some, there is still insufficient evidence to suggest whole grains are better than refined grains. Future research is needed to determine how nutrient status, the type of whole grains (most research is on wheat rather than oats or rice), and other factors influence how whole grains affect hearth disease.
*It’s worth noting that the DART study was conducted before it was required that refined flour be fortified with certain vitamin and minerals (for example iron), which could further change the health effects of refined grains in the time since the DART study was conducted.
Ahh salt, most delicious of the minerals (although I’ve heard good things about chalcanthite). Salt is the savior of unsavory soup, it perfectly pairs with potatoes, and is tasty with tequila. What other alliterations are there to add about salt? Oh, how about this: considered a concern for cardiovascular disease. Easily one of the most well known theories about nutrition is the idea that salt is bad for the cardiovascular system. Because of this I felt it important to address salt, even though salt intake was not examined in the Lyon Diet Heart Study. However, many looking to emulate the Mediterranean diet may find themselves eating a fair amount of salt, due to the many foods from said culture containing a decent amount of salt. This includes pickled vegetables, olives, salted fish and meat, salty cheese, and tomato sauce with a fair amount of salt. Might this undermine some of the benefits of a Mediterranean diet?
The main reason salt came to be considered a risk factor for heart disease came from its effect on blood pressure. Beginning in the 1950’s, two major observations were being popularized. The Framingham study, along with other epidemiological studies, was noting an association between high blood pressure and cardiovascular disease. Meanwhile a number of researchers, particularly Lewis K. Dahl, were noting a seemingly causal connection between salt intake and higher blood pressure. This led to the hypothesis that eating more salt would increase heart disease risk. However, as suggested previously many people in Countries along the Mediterranean consumed fair amounts of salt but such areas weren’t suffering from rampant heart disease. Likewise, residents of Japan consume more salt than nearly every other developed country yet have, by some estimates, the lowest risk of heart disease mortality.
However, such observations are a poor basis for rejecting a hypothesis, so eventually clinical studies were conducted. In a study in Taiwan retired people were given diets with common table salt (sodium chloride) or half table salt and half potassium chloride, meaning one group was eating a lower sodium, higher potassium diet (3.8 versus 5.2 grams of sodium per day in the KCl and NaCl groups respectively). This group had a statistically significant reduction in cardiovascular events and a non-statistically significant reduction in total death. However, because potassium chloride was utilized in the sodium restricted group, this study may be demonstrating the benefits of increased potassium, rather than salt restriction.
The TOHP I study and its sequel, TOHP II, both tested the health effects of salt restriction on subjects with and without high blood pressure (4). Both studies observed a statistically non-significant reduction in the rates of cardiovascular disease events and deaths. At follow up the TOHP I study reported a significant 52 percent reduction in cardiovascular events in the salt restricted group.
So is the evidence that salt is bad settled? Not quite. Excessive salt restriction may actually have risks to cardiovascular disease as well. In Italy a randomized, controlled trial was carried out to see if a salt restricted diet could benefit people with congestive heart failure (5). The results weren’t pretty: the salt restricted diet led to more hospitalizations for heart failure and subjects in this group died more than the group not restricting their salt. Rates of death were more than 150% higher in the salt restricted group, a statistically significant result.
How do we interpret these seemingly conflicting results? By understanding that the answer to the question “is salt good or bad?” is; yes. I think people get to hung up declaring things like salt either good or evil, healthy or unhealthy, cats or dogs (wait, what?), but it’s highly likely there is a middle ground. Back in 2011 a study was published which caused a bit of a stir. It reported on the risk of cardiovascular disease in 28 thousand people with different levels of urinary excretion of sodium (a marker of the intake of this mineral). A low excretion of sodium was associated with an increased risk of cardiovascular disease while large amounts of sodium excretion was also associated with an increased risk of cardiovascular disease, leading to a “J- shaped” mortality curve for sodium intake. And this is not the first time such a finding has been reported. Such a J curve probably explains why studies on salt show such variability; while some observational studies have found salt intake to be associated with cardiovascular disease, some have found salt to be protective against cardiovascular disease.
This is simply observational evidence, prone to problems, but it fits a hypothesis which is also consistent with the clinical trials previously discussed. Insulin resistance, or worsened blood sugar control, seems to be enhanced when diets contain 1700 mg of sodium per day or less (this may be the result of the potentially harmful hormone aldosterone which is increased when salt is restricted). 1700 mg was also the level shown to increase death in the Italian heart failure study discussed earlier. This may suggest that a sodium level at or below 1700 mg can bring about unhealthy physiological changes. Alternatively, the TOHP studies suggest that lowering salt intake from 3600 to about 2600 mg per day may be beneficial to cardiovascular disease. 2600 mg of sodium per day was also roughly the amount taken by the subject in the heart failure study who died less. This means two studies found 2600 mg was more beneficial than a sodium intake about 900 mg less or more.
Although future research may elucidate this more adequately, I believe a dietary intake of around 2600 mg of sodium (equivalent to a heaping teaspoon of salt), give or take about 500 mg, is close to the healthiest intake, accepting of course that specific individual factors may make this amount either higher or lower. Exercise, for example, dramatically increases sodium loss meaning working out increases a person’s salt needs.
Conclusion: So What Should I Eat?
In my opinion, certain blood tests may make determining the optimal diet for heart disease prevention. These include LDL and HDL and markers of iron levels (including serum iron and transferrin saturation).
High iron levels, a situation more likely to be found in men and somewhat more likely in post-menopausal women, might be addressable by limiting red meat and certain iron rich shellfish, such as oysters and clams. Furthermore, because vitamin C increases iron absorption from plant foods an additional dietary method of treating high iron might be separating vitamin C rich foods (like citrus fruits and berries) from iron rich plant foods (like beans, lentils, and refined grains with added iron). Such iron rich plant foods can instead be consumed with iron absorption inhibiting foods, like coffee, tea, chocolate, and dairy products (e.g. cheese). However, the need for these iron lowering treatments is probably context dependent.
As was mentioned in the section on red meat, iron is more problematic when there’s lots of LDL around for it to oxidize. Such is the case with high LDL levels. Because HDL helps protect against LDL oxidation, high LDL appears to be less problematic when HDL is similarly raised. On the other hand, low levels of HDL in the context of high LDL truly represents a bad cholesterol profile. In this scenario high iron creates a great milieu for atherosclerosis and cardiovascular disease. Therefore, bothering with iron reducing might not be that important if the ratio of LDL to HDL is low (2:1 is considered quite good).
Beyond that, here’s a basic dietary framework:
- Eat seafood. Include the omega-3 rich stuff like Anchovies, Herring, Sardines, Salmon, Mackerel, Crab, and Smelt.
- Eat vegetables. In particular, eat salads made up of green vegetables and put virgin olive oil on it.
- Eat nuts and seeds with lots of ALA and not too much linoleic acid. The best such foods are flaxseeds and chia seeds. Walnuts or hemp seeds are alright if the first two can’t be used.
- Eat fruits. There is no clear “healthiest fruit” so just eat a variety of fresh fruit you enjoy.
- Avoid high linoleic acid, low ALA oils; corn oil, safflower oil, grapeseed oil, cottonseed oil. These are found commonly in fried foods, packaged foods, and at many fast food restaurants.
- Don’t worry too much about eating whole grains. Eat the grains that makes you feel good, whether it be white bread, brown rice, oatmeal, or even no grains.
- Look at all the food you eat in a typical day, adding up the sodium from the labels. If your sodium intake is noticeably below or above the RDA of 2300 mg/day consider modifying your salt intake to end up closer to this amount. If above 3000 mg/day, consider strategies to lower sodium to below this amount.