My grandmother came to the US from Italy in 1921. She ate the typical Mediterranean diet: olive oil, wine, lots of salad, nothing processed, homemade bread, homemade spaghetti sauce, some meat but not much, walnuts. She ate in moderation, her weight never varied by more than a few pounds. She never took a prescription drug, and I only saw her a few times take an aspirin. She never felt deprived. Every once in a while I'd see her eat some ice cream (always with espresso dumped on it). She did like Jell-O, which I never figured out. Died at age 101.
26
Genetics play an enormous role in who lives a long time and who doesn't. I have a friend of Greek ancestry, who is now in his mid-fifties, looks much younger, has the libido of a 20-year-old, and can eat anything he wants. He never gains an ounce despite eating a whole pizza or four hamburgers at a sitting, and his blood work is all fine. This despite having lived the most unhealthy lifestyle possible when younger (alcoholism and IV drug use, got sober in his 30's). His family has a history of extreme longevity -- males in his father's line commonly live to be over 100. They've got something that seems to repair damage and prevent changes to energy metabolism (i.e. diabetes). So for some in the Mediterranean, it's not the diet, it's the genes. Choose your parents wisely. For those of us who didn't, this type of diet can't hurt, but don't bet on making it to 100.
20
There is only ASSOCIATION for the benefits with the diet. There is no CASUAL basis for the benefits of the diet. It has now been shown with the new study that there is no proof of CAUSE and EFFECT which does not seem to be stated by the media.
7
OTHER FACTORS Need to be included in the studies of diet, which must expand from not only olive oil and nuts but to include the effects of biotics and probiotics. Meaning that the vegetables, fruits and whole grains increase the variety and effectiveness of the intestinal biome, as do fermented foods such as yogurt, cultured milk (i.e., buttermilk and kefir). Of course doing a long-term double blind study is very difficult. Were the study accompanied by epidemiological studies based on rates of cardiovascular disease, longevity and other factors, across generations.
7
Are they sure they have found all the flaws in the original study? If not, then we cannot be sure these results are any better than the original ones.
4
I like olive oil as much as the next person, but more than half-a-cup a day seems like a stretch (one liter per week is ~140 mL or 0.6 cups). Bottoms up, oil lovers. Putting aside the randomization issues that have surfaced, this diet, as proposed in the original NEJM paper, is simply not doable. Further, it's expensive to follow. Note that the study participants got their extra-virgin olive oil (and their walnuts - thank you California Walnut Commission) for free.
6
Science is about the doing. It is disappointing that human errors during the implementation of clinical trials are unavoidable. Disregard how rigorous and robust the study design can be, if one person fails to follow the protocol correctly and fail to report it, the damage has been made.
The key msg here is a healthy diet pattern lowered the risk for CVD, and similar results have been reported in previous lyon heart study. I think we can reach a rough consensus here that a healthy diet pattern is no harm!
4
Despite the fact that official medical authorities think they understand the causes of obesity, type 2 diabetes and heart disease these chronic conditions have increased exponentially over the past 40 years, and show no signs of responding to the kinds of diet advice they've been giving us. That's why it's imperative to get the science right before dispensing advice. Randomized, double-blind, placebo- controlled clinical studies are the gold standard of science for a reason: they enable researchers to exclude all variables except the one they are testing for and definitively proving or disproving their hypothesis about the causality of physical phenomena. Just because this methodology is more difficult to implement in the field of human health than in the study of physics or chemistry is no excuse for not trying harder to do so, since in this case shoddy science has resulted in the sickness or death of millions, if not billions of people from preventable chronic diseases.
5
Wow. Very surprised Cleveland Clinic cardiologist recommends Mediterranean diet to his patients. Is it possible he is unaware of famous Cleveland Clinic cardiologist Caldwell B Esselstyn? Dr. Esselstyn recommends no oil or animal fat. His recommendations are based on scientific facts not diet studies. See his website and YouTube lectures.
5
Low-fat diets have gone the way of the dinosaur. They don't work for most people. This was discovered after years of people gaining weight on low-fat diets. If you're gaining weight with little to no oils in your diet what good does "scientific evidence" do for you?
19
Most of the weight gain in low fat diets came with the excess of sugar added in processed products, where even jelly beans were considered a low fat food. People knew no boundaries!
3
Wait a sec. The article says the studies issues were addressed by dropping some data points, and changing the values of others. It then quotes a bunch of US academics who dismiss these changes off hand. Why? Without details, they just sound arrogant.
7
Those not convinced can go back to the standard American diet. Problem solved.
10
It's refreshing to read an article on scientific methods. Yet the writer raises a central point she does not unwrap: Why are some like Berry still unconvinced by the reanalysis? In not pursuing this, she loses a chance to show how day to day scientific inquiry works.
It might be that Berry doesn't like the way the reanalysis dealt with the non-random subjects. Or it might be that he's simply suspicious that if these sampling errors were found, there will be others; "smoke and fire." He might disagree with the exact math used to adjust and reanalyze. Or he might have had philosophical disagreements with large scale experimental diet studies. Or, frankly, he might be in a competing group with a grant application that's pushing its own hypotheses.
See, science is about the doing, not the exact results. Questions change, results change. We scientists pick at each other's work incessantly, for many reasons. But in doing so, we evolve a better picture of how the world works. Eventually, faster or slower, we move toward a rough consensus and pick at the details. Gravity and evolution come to mind. Consensus in nutrition and medicine lags a touch. We humans are a tough study. Asteroids don't require informed consent.
But in 2018 America, the general public needs to know this caution about the doing is normal science. It does not mean particular findings are "fake." Nor does it mean that we should toss our olive oil. A rough consensus is emerging, we just need more nitpicking.
35
Any diet that is not the American diet is going to be better for your health. It was unusual to see obese people in most of the world, until American chain food and packaged junk reached them. Now they're fat too, and suffering from the same health problems (diabetes, cardiovascular, etc.) that were pretty much unheard of before. Healthy local diets are giving way to junk. How ironic that Americans would flock to the very diets they are destroying around the world. And by that I don't mean the average people, of course, but Big Food and its incessant push to conquer the world. But, by placing no restrictions on them and allowing them to rule unchallenged, the average people are ultimately at fault. For allowing Big Food to ruin their own health, as well as the health of the world at large.
13
Right, blame America. People in other countries aren't being fed the American diet at gun-point -- they're choosing it themselves. The local, healthier diets are still available. Healthy eating just takes coomon sense.
8
If you compare a not very good diet (Mediterranean) to a terrible diet (Western diet), the not very good diet will shine by comparison. That does not make it a good diet.
If you add not very good ingredients (oil) to a very good diet (plant based), and people on that diet do well, it may be because the plant based diet is good enough to mitigate the use of oil and not that the oil is good for you. Scientists and doctors design the studies so that they get the outcome they want.
7
@plsteiner:
Yes, vegans, please save us from the horrific effects of olive oil.
But wait -- oils are all plant-based!
13
The statisticians aren't sure the results of analysis are clear, the cardiologist is...I'll go with the statisticians.
However, like with all studies, it's best not to make a decision on the outcome of just one, but to weigh the evidence of multiple well done studies before reaching a decision.
5
An easy and very WRONG choice. Talk to some older statisticians about that.
2
Anyone who has actually spent a few days around the Mare Nostrum should notice that the components of the diet are part of a very different lifestyle.
Epidemiologists try to pick out the one thing among many that makes The difference.
Statisticians are very good at making detailed calculations about things they don’t understand.
But, a recent JAMA paper concluded that the best diet is the one you will follow.
A friend who worked for Nabisco retorted that they don’t force anyone to buy their products.
9
In the 1980s and 90s, while visiting Crete numerous times, I noticed that the dates on gravestones testified to people living into their 80s and 90s then. A younger date was rare.
This diet is a no brainer to a healthy life. However, there are also other positive contributing factors such as a low stress life style, filled with strong ties to families and friends. For instance, the wedding list? Just invite the whole village and no one is offended.
People are also more active with gardening or farming and housekeeping activities and walk from place to place. I never recall seeing a fat person. The Mediterranean diet is a lifestyle, not just a list of foods.
58
It was rare to see a fat person in most parts of the world. Until American chain food and packaged junk reached them. Now, they're fat, too.
8
An additional correlation - the advent of refrigerators in households - obesity follows.
They could be right. They could be wrong. It's still my favorite diet by far.
8
If it is true that their diet is responsible for better health in the Mediterranean population then is it also valid for the French population whose diet is centred on fatty meat, full-fat dairy and minimal amounts of olive oil overall.
I am aware that for the past 50 years this phenomenon is discounted as “the French paradox”. Anyone with a scientific mind acknowledges that paradox is not a valid scientific concept. I.E., if facts do not support a hypotheses then the hypotheses is invalidated.
16
I found the Mediterranean diet through my own weight-loss efforts and lifestyle management, and I believe in its health benefits (especially in contrast with the stereotypical "American diet").
A mistaken impression that some people might have (certainly I had, years ago) is to look at these diets as a sort of elixir or other quick-cure to a weight and/or other health problem. It's certainly not that: a careless lifestyle and diet lived during the work week, followed by one healthy meal on the weekend, isn't enough to do the trick.
But a general lifestyle change, on multiple levels (diet, exercise, stress management, etc.), is conducive to an improvement in overall health.
Unfortunately, no diet or exercise regimen will hold the Grim Reaper at-bay indefinitely.
But you may just get a few more good years for visiting Machu Pichu, Tuscany, or just holding your grandchild in your arms.
For those reasons alone, I pass on the french fries...
13
How remarkable that there’s so much concern that this study of common-sense healthy eating be “random.” To insist that members of one household follow different diets or the results are invalid seems to insure that “valid” results will be impossible to generate.
Like the correlation between melting glaciers and CO2 emissions, or banking deregulation and financial crises, or fat-free/sugar-free gluttony and obesity, I suspect the planet will perish before evidence that requires altering comfy habits is “conclusive.”
4
But the concerns are valid scientifically. This is one reason why nutrition studies are very difficult.
11
A quote from the NYT article that reported the original study:
"The participants stayed with the Mediterranean diet, the investigators reported. But those assigned to a low-fat diet did not lower their fat intake very much. So the study wound up comparing the usual modern diet, with its regular consumption of red meat, sodas and commercial baked goods, with a diet that shunned all that."
I also read the original journal article.
I believe the study was a good faith attempt to investigate a meaningful problem, and as with all studies in the "real world" (not the laboratory) it is beset with problems. Nevertheless, all of these imperfect studies move us forward. So I don't want my comment to be too critical.
However, I believe the findings of the study can be summarized as "participants who followed a diet plan did better on some outcome measures than did participants to did not follow a diet plan."
From my own "real world" study (i.e., what I, with my own biases observes) what appears to me to be the general finding from the research and reports of research I have read is that it matters little what type of diet you follow. Instead, the main controlling variable affecting health is whether or not you are following a plan.
What do people who follow low-fat, or low carb, or Mediterranean, or Paleo, or Vegan, or vegetarian, etc., diets have in common?
They are following a plan. Their eating is intentional. It is mindful.
The particular plan is less important.
17
Sadly much of the olive oil we have on our shelves in our US supermarkets has been adulterated with inferior oils and who knows exactly what. I am only guessing that those in the Med have access to the real, good stuff. That and they are generally not working their selves to death and prioritize proper dining habits/hours (lengthy, sit down lunches) and family/leisure time. It is not about the food alone.
21
I don't think it's that hard to find decent olive oil in the US (and I lived in France and have travelled around the French & Italian Mediterranean), but you do have to some idea of what you're looking for, and you have to use your sense of smell and taste. It probably also depends on where you live and what your grocery options are. I guess I've always lived in places with good grocery options in the US. I also suppose that if you're not sure what's good, you don't know what you're supposed to be smelling or tasting. But I agree it's not food alone.
6
I don't need a randomized study to know that the Mediterranean diet is far healthier than the average American diet.
67
While the researchers workout their demurs, I have little doubt he Mediterranean Diet remains the best option for healthful eating.
24
It's certainly the least "fad-free" way of eating.
3
I'm going to continue with the diet because it's tasty, it makes me feel healthy and, according to my doctor,I am healthy.
25
The main problem with this article is that it does not review the literature. The PREDIMED was not the first Randomized Controlled Trial to show that a Mediterranean diet reduces death from cardiovascular disease-- that honor goes to the Lyon Diet Heart Study. The Lyon Heart Diet study was randomized Mediterranean compared to American Heart Association Diet which was resoundingly worse.
The quotes are obviously biased and selected, not randomized themselves. There are zero high quality randomized controlled diet studies that prove superiority of one diet to all others.
5
The PREDIMED design protocol for EVOO was "Participants in the group assigned to a Mediterranean diet with extra-virgin olive oil received 1 liter of the oil per week per household, with the recommendation to consume at least 4 tablespoons per day of extra-virgin olive oil per person." So randomization for the EVOO treatment was to households not individuals. Further, Med diet guidelines mislead by citing 4tbsp/d dosage, instead of the recommendation to use EVOO as primary family cooking oil. Even further, the authors fancy that the higher polyphenol content of certain brands of EVOO from Spain or Portugal was responsible for the effect, but that is unlikely since other studies do not confirm the substance presence, and those ingredients were not tested separately, only that bottles of EVOO were supplied free to one arm and the others including controls could buy it. The result should be correlation not causation and that might not move us to such radical changes in diet. I happen to like EVOO and use it anyway, to placate my physician. But genome-wide association studies will likely soon lead to personalized medicine for nutritional advice for those of us with varying responses and that will supercede PREDIMED.
6
Have fully randomized studies been made to prove with almost absolute certainty that junk food is bad for health, or for specific aspects of health like heart disease? Why such a high bar on the Mediterranean diet?
47
Thanks to the power of the sugar industry (like big tobacco), the bar for healthy living evidence remains unattainable.
15
@tdb:
Unfortunately, there are several competing definitions of what is or isn't "junk food." The past 30 years of official dietary guidelines consider steak, eggs, and butter to be junk food, but promote processed grain products, soybean oil, and margarines as "heart healthy."
7
I don’t see how anyone can ever determine the “perfect” diet, no matter how rigorous the scientific study.
A high protein, high vegetable (these two are not mutually exclusive) low sugar, whole food vegan diet is probably the best way to go, but that’s an expensive, culturally unpopular path in today’s America.
And then there are genetics. As a never-smoker vegan with a BMI of 21.5, I should theoretically be at low risk for heart disease/stroke; and yet, as the fourth generation of victims of massive stroke, I was reliant on three different pharmaceuticals by age fifty to control my sudden, mystifying (to my physician) high blood pressure.
22
A healthy diet can tilt the odds of avoiding a chronic condition in your favor, but the risks never go to zero -- or to 100%. And genetics or injury or disease may require medications to control conditions they cause. But most of us (roughly 80%) do not have these underlying factors adversely affecting our health. I wish you good health and long life, and lovely vegan food! (or as I put it, plant based whole foods -- delicious!)
28
Many times I have heard it said that a Mediterranean diet is too expensive for many. I will counter that. I have followed a Mediterranean diet for probably 40 years —and I do occasionally indulge. I eat organic fruits and vegetables as much as possible, eat whole grains, seek out grass fed, pastured meats and seafood that is wild or at least sustainably raised. At age 72 I look many years younger and am fit. I still work and fool most of my peers. And yes, this is an expensive diet regimen — initially. But I love to cook (this is key) and our plates are 3/4 vegetables, often roasted, whole grains (rice, farro, lentils, beans etc) and the rest is quality protein (meats, fowl, or fish), usually 2 to 4 ounces only. We also eat less than the average person, eating slowly and stopping when full. I have taught myself how to create in quantities so that after a busy day at work our dinner is “fast food.” (I dream of a 3rd career teaching busy people how to cook full flavored meals this way). I also understand how difficult this diet can be to establish. I have yet to see, anywhere, the counter of long term expense/cost of cheap food. The offset of my upfront expense of food is good health and no costly medical care. Exercise (not obsessive) and yoga keep me bone healthy and in full enjoyment of life with my family and grandchildren. By the way these two grandchildren also crave my food, loving kale and chard as much as good meat and desserts. It can be done.
18
I have been conducting public health research for nearly 2 decades and have published dozens of analytical papers in peer reviewed journals. My particular field does not use randomized designs as much as some other fields, but we do use them when we can. Randomized experimental designs are preferred when they are available, but they are not the silver bullet that many MDs and statisticians make them out to be. Much has been learned and sound public policy has been based on the evidence provided by quasi-experimental and observational study designs. I am always a bit irritated when I hear an MD say that no conclusions can be made from a study just because it was not randomized, or as in this case, there were problems with how the randomization was carried out. When people do that, they are letting the perfect become the enemy of the good.
I have not read the study reported on in this article but it sounds like a good faith effort and the authors took all reasonable steps to perform and report a detailed and thorough analysis. Surely, there is some information to be gleaned from this and the results probably do provide some justification for claims about the health benefits of the Mediterranean diet. Clearly there is more work to be done and much of it will be imperfect. However, at the end of the day, reasonable people will assess the preponderance of evidence and make reasonable conclusions. Waiting for the "perfect" study before doing so is waste of time and money.
58
This viewpoint of yours is why the field has so many problems and is often not well respected by other scientists in different fields. "Randomized experimental designs" are in fact the silver bullet for very good reasons. One doesn't get to redesign research protocols just because they aren't convenient.
This is not the perfect becoming the enemy of the good, but shoddy research leading people down potentially false paths.
A "good faith effort" is not enough.
13
"However, at the end of the day, reasonable people will assess the preponderance of evidence and make reasonable conclusions."
That put a big smile on my face; it sounds so casual, and in fact is so very critical.
It was one of two epiphanies provided to me by my chief mentor that completed my transformation into a scientist. Grad school - does not do that by itself.
The two of us were working on a specific problem in genetics. He was one of the genuine greats- one of the founders of the entire field of cytogenetics; vast knowledge. We had a probable conclusion - then I began picking nits; yes, but what we know does not rule out this; or this, or- . I was not trying to show off my knowledge of genetic arcana, I was trying to show how meticulous I could be- basically good.
Eventually he became exasperated with me, and resorted to "Because it's not REASONABLE!" Politely leaving the "you puppy!" unsaid.
Can you make that into a statistic? You cannot. What he was saying was that the entire collection of facts, knowledge, and wisdom in his head spoke against my nits.
He would be the first to acknowledge- he could still be wrong. But the reality of science is; his head was worth 20 perfect statistical studies, in a heartbeat. Summed up as - this is reasonable; that is not.
25
The reason "randomized" is NOT a silver bullet has nothing to do with the design.
It has to do with the execution. We can, and probably should, write books on the myriad ways beautiful randomized experiments were totally messed up - by people - in the process. Not installed correctly; aberrations not reported (as in this case). Data taken in non-randomized fashion. Analysis, ditto. Human bias and human needs to go to the bathroom NOW can creep in at any point.
The other major failing of randomized studies is that they give many participants an unwarranted comfort that they now have a solid experiment, so they don't have to be on their guard for human whoopsies.
Whoopsies happen, even to the randomized.
2
"Despite serious problems in the way the study was conducted" the Mediterranean diet is still one of the best life style changing approaches to healthier eating around. Any reasonable food plan "rich in fruits, vegetables, nuts and olive oil, with moderate levels of fat" coupled with daily exercise is bound to be effective in reducing weight and keeping one in decent health. There is no full proof, magic formula which will prevent someone from getting a heart attack or stroke. I keep thinking of Jim Fixx and his early demise at the age of 52. But the Mediterranean approach is a delicious and easy option to consider.
21
As usual, people with no knowledge of the facts bring up the death of Jim Fixx, when it is not relevant to the study in the article. Jim Fixx was an early proponent of exercise, specifically running, to help prevent heart attacks. He did have a heart attack while out on a run and died at age 52. However, his was special case. He had a genetic disorder, familial hypercholesterolemia, which causes people to have extremely high levels of cholesterol no matter what their diet. His father and grandfather died of heart attacks in their 30s and early 40s. The only way to control it is statins and he died before statins became available. The fact that he added at least 10 years (about 20%) to the average lifespan of the men in his family is a tribute to the efficacy of a healthy lifestyle.
55
@Scientist
I appreciate the history lesson of Jim Fixx and the background surrounding his health and life you supplied. However, I am always at a loss as to why corrected information is given in such a condescending manner. Arrogance and rudeness is never warranted and not necessary.
so, critically: choose your parents carefully.
The study's design should have been seen as deeply flawed from the beginning. Any researcher should know how important it is to maintain the randomization of the experimental groups. And if there was any oversight on the study before it began, it failed to exercise scientific common sense. Furthermore the peer review prior to publication should have flagged the problems before it went into print. And now that this news has broken, the scientists who have 're-analyzed" their data are disingenuous, to say the least.
Scientific study of human health related to diet or any other aspect of behavior is highly challenging in that experimental controls are so hard to establish and maintain. But still, it is critical to do just that. How else are we supposed to make good decisions about what we eat? There is already plenty of skepticism about what research really means, because different studies on the same topics yield different results. The least that researchers can do is make sure their admittedly hard work is built on a solid foundation.
8