With breakfast and chocolate against heart disease

"Fruhstucksmuffel have an increased heart attack risk" – So or similarly, it’s the last days almost all media. But the study on which the reports are called, let this end hardly too

The first news released last Tuesday Fruh, inspired by a message of the German Press Agency. Over the week, almost all the media reported about: Manner who waiver on the breakfast, dangerous their hearts. The risk of cardiovascular diseases rang drastically to 27 percent.

At this end, a study that had published the trade magazine Circulation. A team of US scientists had observation data from 27.000 manner over 20 years evaluated. The problem: Such observation studies are virtually unable to prove a causal relationship between individual factors.

The difficulties will quickly become clear when you take a look into the study itself. According to this, manners who waive her breakfast, fatherless full-time jobs, were rarely athletic, fatherless smokers and stronger alcohol consumers. Everything known risk factors for heart disease. Scientists speak here of so-called confounders. You can – which the study authors have also tried – to eliminate all known such effects. However, it is always about to ame that there are more confounders who are unknown.

"From this structure of inseparable factors now to pitch the breakfast, is of course adventurous", explains Gerd Antes from the German Cochrane Center in Freiburg. "The study is again a perfect example of how association or correlation is abused as a causal context." Gerd Antes is considered one of the leading professionals in evidence-based medicine in Germany.

Printed differently: You also have to claim that those who waive their breakfasts have a high risk of becoming smokers or a bad athlete. The statement of goods due to the existing data just as good or badly escortable.

"However, the study has much more fundamental problems", explains Antes. "So a group of employees in medical institutions was examined here. People who often have overtime hours and otherwise increased risk factors are exposed. It can hardly be justified from this on the overall booklet."

The problem with donors

The study here is an example of the difficulties to examine the impact of dating habits on later diseases. Medicines are usually tested in so-called randomized double-blind tests. In doing so, comparatively comparable patients are split in groups and compares a drug to be tested with a placebo, ie an ineffective pill.

If the allocation of patients to the groups is really random, confounders are excluded, because there is no connection between the life habits of the student participants and the question of whether they land in the placebo or the drug group. Similar is impossible with dance habits. It already fails because no one can eat a placebo-breakfast without remembering it. It is also difficult to move from randomized people to follow a long-range period of dating a dating habits.

This is very principle that it is very difficult to come into solid study results in donors. Of the many recommendations on healthy eating are very few with good evidence. The difficulty is also evident in that many people the impression arises, science did not know what she wanted. Recommendations of once today are considered ubobt and contradictory recommendations are not infrequently.

In the Guardian Aubert, Student Author Eric Rimm to criticize the study results. He is aware of the limits of an observation study, but nevertheless ames that this is a causal connection. Especially because there are already a rough amount of studies that indicate similarly. In her study, the authors refer to early work, which showed a connection between missing breakfast and other risk factors such as overweight, hypertension or diabetes.

Publication bias

But Gerd Antes do not convimate that: "In this area there is a coarse problem with the so-called publication bias. You do not know how many studies of this kind have already been carried out."

Publication Bias refers to a phanomen in science, in which many studies carried out on a topic, but only a few later will be published. This creates a distorted impression. A study examining the context of the lack of freshness stuck with heart disease and results that there is no context, was probably not appearing in a journal because the result is completely negative.

In medication studies, it is now considered good practice to register studies prior to execution at a public study register. As a result, it is at least possible to understand how many studies there were a topic that were not published. That too works only partially and does not have all pharmaceutical studies are previously registered. Date studies are the prior registration but so far the absolute exception.

Gerd Antes’s wife also report, because he directs the German register for clinical trials. "Studies on dining matters can also be filed with us", explains Antes. "However, the hitherto almost no one makes."

Studies that have a connection between certain dating habits and positive health effects are frequently picked up by media. Especially if actually as unhealthy foods have positive effects, it is often reported.

"Last year there was a similar case: dark chocolate should allegedly protect against heart attacks", explains Gerd Antes. "That went so far that the German company for internal medicine discussed whether health insurances should reimburse patients with dark chocolate. Unfortunately, not only journalists, but often also medical societies that spread and interpret fraghous study results."