Do not believe what they say about nutrition / SurprizingFacts

Love for cabbage was associated with the shape of the navel in the form of a hollow

At the end of each year, millions of people give themselves Vows to change eating habits. Usually people share food on moralistic categories: good / bad, healthy / harmful, nutritious / delicious, for weight loss / fattening – but opinions on what kind of food belongs to which of these categories, diverge.

The US Nutrition Advisory Committee has recently released a new set of recommendations that define a healthy diet as a diet that emphasizes vegetables, fruits, raw grain, low-fat foods, seafood, beans and nuts, and reduces red and processed meat, Purified grain and sugar-containing foods and beverages. These recommendations immediately aroused a storm of controversy. In the editorial in the medical journal BMJ, a conclusion was drawn that there was a lack of rigorous evidence; This statement was actively challenged by members of the committee.

Some cardiologists recommend a Mediterranean diet rich in olive oil, the American Diabetes Association supports both low carbohydrate diets and low-fat diets, and the Committee of Physicians promotes vegetarianism for responsible medicine. Ask a smart fan of crossfire, and he will recommend a "paleodieta" based on what our ancestors supposedly ate in the Paleolithic. My colleague Walt Hickey stands for a ketogenic diet.

Who is right? Hard to tell. On the issue of nutrition, everyone has an opinion. But no one has solid evidence. Problems begin because of a lack of agreement about what makes a healthy diet. Reducing weight? Build muscle? Maintaining the strength of bones? Preventing heart attacks, cancer, dementia? Whatever you are worried about, we will not lack special diets designed to help you. Linking eating habits and addiction to products with health factors is easy to ridiculous – as you will soon see from a small experiment conducted by our editorial staff.

Presence in the family of a dog

Our invasion of the field of nutrition science demonstrated the depressing state of articles written about the effect of eating on health. As evidence we will take you backstage and show how these studies are conducted. The first thing to consider – nutrition researchers are studying an extremely complex problem, since it is impossible to know what exactly and in what quantities people eat, unless they are locked in a room under supervision. Therefore, almost all nutrition studies are based on measurements of food intake according to the records of the people themselves, who need to remember what they ate. The most common ways of doing this are keeping a diary of food, questioning for memory and questionnaires on the frequency of food consumption [food frequency questionnaire, FFQ].

There are several options for ready-made FFQ, but the technique is the same for everyone: you need to ask people how often they Eat certain foods and in what quantities. Do not always just remember everything that you ate, even if it was yesterday. People usually underestimate the amount and variety of foods eaten, they may not admit that they ate something or misinterpret the volume of food.

"As a result, it turns out that it is very difficult to conduct studies related to the diet," – Says Thorin Block, CEO of NutritionQuest, a company that conducts surveys. The company was founded by his mother, Gladys Block, a pioneer in this field, who began to develop FFQ at the National Cancer Institute. "You can not get rid of mistakes in any way". And yet, according to him, there is a certain hierarchy of similar questionnaires in completeness. Nutritional diaries are rated quite high, and with them – daily surveys, in which the administrator interviews the subject, and compiles a catalog of everything he ate in the last 24 hours. But, as Blok says, "it is necessary to conduct several such sessions in order to get a complete picture of the average diet of the subject." Researchers are usually not very interested in what people ate yesterday or the day before yesterday – they need to find out what they eat regularly. Studies using daily surveys usually underestimate or overestimate the impact of food that people do not eat every day, as they record a small and unrepresentative length of time.

When I tried to keep a diary of consumption, I discovered how much Blok was right – It is strikingly difficult to draw a picture of food habits from the data collected over several days. It so happened that I went to the conference that week, so I ate dry snacks and ate in restaurants, which is very different from home-based meals. My diary showed that one day before dinner I ate only a donut and two packages of chips. And what did I have for supper? It was a delicious curry from Indonesian seafood, but I could not bring here a list of all the ingredients.

Pages from food diaries

Another I have a lesson from brief diaries – the process of tracking consumption can affect what you eat. When I knew that I had to write this down, I was much more attentive to what I was eating, and sometimes as a result I did not eat something because I was too lazy to write it down, or because I realized that I did not I need a second donut (or I did not want to admit that I ate it).

It's hard to fool an instinct that requires lying about food being eaten, but the FFQ seeks to overcome the unrepresentative short food records, estimating that people consume more Long periods. When you read an article with a text like "blueberry prevents memory loss," the evidence is likely to have been taken from some FFQ. The questionnaire is usually asked what the interviewee used to eat during the last three, six or 12 months.

To understand how these polls work and how reliable they are, we hired Blok to try out a six-month questionnaire from their company on me , My colleagues Anne Barry-Jester and Walt Hickey, as well as a group of volunteers.

Some questions – how often do you drink coffee? – were quite simple. Others have puzzled us. Take at least tomatoes. How often did I eat them for six months? In September, when my garden was full, I ate cherry tomatoes as a baby candy. Perhaps I ate 2-3 tomato cherokee tomatoes with vinegar and olive oil a day. But at the same time from November to July, I could not eat tomatoes at all. And how do I answer this question?

Questions about the volume of food puzzled everyone. In some cases, the survey gave us unusual, albeit useful, advice – for example, it showed the approximate volume of half a cup, a whole cup and two cups of yogurt with photos of cups filled with sawdust. Other issues seemed absurd. "Who even knows what a cup of salmon meat looks like, or two cups of pork ribs?" Walt asked.

Although the questionnaire had to simply measure how much food we consume, it sometimes seemed that there was criticism in questions-drink Do we have fatty milk, low-fat milk or fat-free milk? I noticed that when choosing from three varieties of meals I always tried to choose the average, regardless of the actual size of my portions.

Despite these difficulties, Anna and Walt have done everything possible to answer honestly and most Fully. After that, we compared the results. The questionnaire showed that fat cheese and different types of alcohol were our main sources of calories.

Then our diets split. Walt lost 25 kg on a ketogenic diet, Anna eats quite a bit of protein, and I, according to FFQ, absorb almost twice as many calories as each one.

Can these results be true? Anna and I are almost the same height and weight; We, probably, could share clothes with each other. How can I eat twice as many calories as she? The bloc acknowledged that it is difficult to accurately count calories, especially without food records over a long period of time, and if we start to understand individual nutrients, it is still more confused. He referred to the 1987 study, which requires a daily average of 27 days for men and 35 days for women to be collected for a full assessment of average caloric intake. And some nutrients to track even more difficult – for example, to assess the intake of vitamin A, women need 474 days. This indicates that our reports can be true, or they can contain a bunch of mistakes.

Love for chips was associated with good results in mathematics

Of course, measurements that rely on memory have their limitations, "says Brenda Davy, a professor of human nutrition from the Virgin Technological Institute. "But most of us, nutrition experts, believe that they have value." Calories are the most difficult to measure, she says, noting evidence that people underestimate the amount of food they absorb, which they consider unhealthy, such as fatty foods or sweet foods. "But this does not mean that they underestimate all the indicators. This does not mean that there are problems with the measurement of the eatable fiber or calcium. "

Questioners understand that the answers are not ideal, and try to correct them with confirmatory studies that check data from FFQ with data obtained by other methods – usually a survey About the food taken for the last day or the diary of consumption for a longer period. The results of confirmatory studies, according to Blok, allow researchers to take into account the volatility of daily consumption.

Critics of FFQ, for example, Edward Archer, a specialist in computational psychology from the Research Center for Nutrition and Obesity at the University of Alabama in Birmingham, says that these confirmatory Research – just a reasoning that does not come out of the logical circle. "You take one type of subjective report and confirm it with another type of subjective report," he says.

It's harder to write down everything you eat than it may seem, says Tamara Melton, a nutritionist and official representative of the Academy of Nutrition And dietology in Atlanta. Among other things it is almost impossible to measure the ingredients and serving sizes when you are not eating at home. "It is not comfortable. If you are at a business lunch, you will not get your measuring cup. "

When Anna and I compared the consumption of calories according to our FFQ to what we calculated with our weekly consumption diaries, the results did not match. We had difficulty figuring out the sizes of our portions for FFQ, and who knows which of the results was more accurate?

Difference in calorie intake according to FFQ (green) And diaries of consumption (blue)

Although doubts about the accuracy of the self-assessment of food consumption have existed for decades, recently this debate has intensified, according to David Allison, director of the Research Center for Nutrition and Obesity at the University of Alabama at Birmingham. Allison was the author of the expert report of 2014, compiled by a working group on measuring the energy balance, in which the use of "definitely inaccurate" measurement methods was called "unacceptable" for developing health strategies, research and clinical practices. "In this case," the researchers wrote, "the saying 'anything better than nothing' needs to be changed to 'anything worse than nothing'."

The problems with the questionnaires are even deeper. They are not just unreliable, they give out huge amounts of data with a lot of variables. The resulting cornucopia of possible combinations of variables makes it too easy to fit them to beautiful and incorrect results, as we learned, inviting readers to pass FFQ and answer a few questions about themselves. We received 54 full answers and looked for links between the data in them, just as researchers are looking for links between food and dangerous diseases. Finding such connections was ridiculously easy.

Our shocking new study found that

There are no translations available.

Consumption of such products as Related to P-value
Raw tomatoes Judaism <0.0001
Egg roll The presence of a dog <0.0001
Energy drinks Smoking <0.0001
Potato chips Good grades in mathematics 0.0001
Aromatherapy Strange rash during the past year 0.0002
Molluscs Right-handedness 0.0002
The Lemonade Confidence that the C / F "Collision" was to receive a reward for the best film 0.0004
Fried fish in breading Membership in the Democratic Party 0.0007
Beer Frequent smoking 0.0013
Coffee The presence of a cat 0.0013
Salt A good impression of the work of the Internet provider 0.0014
Low-fat steak Non-Religiousness 0.0030
Tea with ice Confidence that the C / F "Collision" did not deserve a reward for the best film 0.0043
Bananas Good grades for reading 0.0073
Head of cabbage Belly button in the form of a hollow 0.0097

Our FFQ issued 1066 variables, and additional questions were sorted by the respondents on 26 possible characteristics (for example, right-handed or left-handed). This large amount of data allowed us to build 27,716 regressions in just a few hours (full results can be viewed on GitHub). With such a set of possibilities, we were guaranteed to find "statistically significant" correlations that are not relevant to reality, says Veronica Wieland, a statistician who runs the Center for Mathematical Medicine Batell in the National Children's Hospital in Columbus, Ohio. Using a p-value of 0.05 or less to measure statistical significance is equivalent to a 5% error, says Wieland. And with 27,716 regressions, 1386 false positive results can be expected.

But false-positive results are not the only problem. According to Wieland, the probability was great that we would find real correlations, useless from a scientific point of view. For example, our experiment found that people cutting off fat from a steak were more likely to be atheists than believing that fat is a gift from God. It is possible that this correlation is real, says Wiland, but this does not mean that it expresses the connection of cause and effect.

A preacher advising parishioners not to cut off fat from meat so as not to lose faith, may laugh, but Specialists in the epidemiology of nutrition often give recommendations based on such unreliable evidence. Several years ago, Jorge Chavarro, a specialist in the epidemiology of nutrition at the Harvard School of Public Health, recommended that women attempting to conceive switch from fatty foods to fatty foods, such as ice cream, based on data from FFQ during nursing research. They and their colleague Walter Willet also wrote a book promoting a "fertility-boosting diet" based on these data. When I contacted Chavarro to ask how confident he was of the connection between diet and fertility, he said that "of all the connections we found in this we were the least sure". And of course, it was she who got in the headlines.

Almost any product you can imagine has been associated with any health impact in scientific papers that have been expertly evaluated and used tools such as FFQ "Says John Ioannidis, an expert on the reliability of research results from the Center for Innovation in Meta-Research at Stanford. In an analytical work of 2013 published in the American Journal of Clinical Nutrition, Ioannidis and colleagues randomly selected 50 common products from a cookbook and found studies evaluating the relationship of each of them with the risk of cancer. It turned out that the research found a link between 80% of products – including salt, eggs, butter, lemons, bread and carrots – and cancer. Some of them indicated an increased risk of cancer, others decreased, but the effect of these products on health was "exaggerated to the impossibility of exaggeration," says Ioannidis, with a weak evidence base.

Продукты, из-за которых увеличивается, или уменьшается риск рака. Оценки даны для 20 продуктов, каждый из которых изучался, по меньшей мере, в 10 исследованиях.

Но проблемы не только статистические. По словам Иоаннидиса, многие из результатов исследований были биологически маловероятны. К примеру, в исследовании 2013 года обнаружили, что у людей, употреблявших орехи три раза в неделю, смертность снижается на 40%. И если бы орешки действительно так сильно снижали риск смерти, это было бы революционное открытие, но это число почти наверняка сильно преувеличено, как сказал мне Иоаннидис. И без контекста она не имеет смысла. Получит ли 90-летний человек такие же преимущества, как 60-летний? Сколько дней или лет нужно поддерживать ореховую диету, чтобы включился её эффект, и как долго он будет действовать? Именно на эти вопросы люди хотят получить ответы. Но как показал наш эксперимент, очень легко использовать опросы по питанию для связи продуктов с какими-то результатами, но очень сложно понять, что означают эти связи.

FFQ «не идеальны», — сказал Чаварро, но на сегодня других вариантов почти нет. «Возможно, мы достигли ограничений текущей методологии оценки питания, и для улучшения результатов потребуется серьёзный сдвиг».

У текущих исследований есть ещё одна фундаментальная проблема: мы ждём от них слишком многого. Мы хотим отвечать на вопросы типа, что лучше для здоровья – масло или маргарин? Поддержит ли мой ум в хорошем состоянии поедание голубики? Будет ли у меня рак кишечника из-за бекона? Но исследования используют измерения, основанные на памяти, и эти инструменты слишком грубы для того, чтобы отвечать на такие конкретные вопросы.

Иоаннидис сказал, что одна из причин может заключаться в том, что отдельные продукты типа насыщенных жиров или антиоксидантов могут очень мало влиять на риск возникновения заболевания. Это его заключение следует из более тщательных проверок с рандомизацией. Именно поэтому в заголовках часто пишут об относительных рисках – сколько людей заболели раком в группе, потреблявшей больше бекона, по сравнению с теми, кто его не ел. Относительные риски почти всегда выше абсолютных, но на самом деле нас интересует абсолютный риск (к примеру, ваш риск заболеть раком при потреблении бекона). Если, допустим, один из 10000 человек, евших больше всех бекона, заболеет раком, то по сравнению с тремя людьми из 10000 человек, не евших бекон, разница будет тройной. Но разница абсолютных рисков – шанс в 0,01% против шанса в 0,03% — крохотная, и её будет недостаточно для того, чтобы изменить свои пищевые привычки.

Тенденция представлять результаты исследований более точными и важными, чем они есть на самом деле, также объясняет, почему мы сталкиваемся с такими противоречивыми заголовками по поводу таких продуктов, как кофе. «Большие данные низводят поддельные точные результаты до уровня шума», – писал Иоаннидис в своём анализе от 2013 года.

Так что мы возвращаемся к нашему изначальному вопросу: что такое здоровая диета? Основы мы знаем – достаточное количество калорий и протеина, чтобы не дать телу умереть. Нам нужны питательные вещества вроде витамина С и железа. А вот сверх всего этого мы, возможно, придаём слишком большое значение продуктам, говорит Арчер, врач из Исследовательского центра питания и ожирения. «Есть культуры, воздерживавшиеся от фруктов и овощей, и прекрасно существовавшие тысячи лет», – говорит он. Некоторые популяции сегодня прекрасно живут, употребляя крайне мало овощей, а другие выживают почти полностью только на растительной пище. Итог всего этого, по словам Арчера, состоит в том, что наше тело хорошо адаптируется и прекрасно может сообщать нам о том, что ему необходимо, если только мы научимся его слушать.

И даже если это так, я сомневаюсь, что мы перестанем искать тайные эликсиры здоровья в наших кладовых и холодильниках. Причина огромной востребованности таких исследований со стороны СМИ и общества та же, по которой учёные тратят миллиарды на проведение этих исследований. Мы живём в мире, в котором страшные заболевания постоянно поражают окружающих нас людей, и иногда это происходит внезапно. Естественная реакция на сердечный приступ или рак знакомого человека – поиск способов защитить себя от такой судьбы. Мы обращаемся к еде, чтобы обрести хоть толику контроля. Мы не можем влиять на происходящее внутри наших клеток, но мы можем влиять на то, что мы вкладываем в наши тела. Науке ещё только предстоит найти волшебный витамин или питательное вещество, позволяющее нам оставаться здоровыми, но мы явно настроены на продолжение наших поисков.