medical research studies often seem to conflict

Why medical research studies often seem to conflict


When a person hears a new medical research study, the reaction might be “Wait, this is the opposite of what I heard before!”  Yes, it’s true, often new research comes out that seems to contradicts earlier studies, leaving many people confused and frustrated. Why?

This apparent conflict has arisen recently in subjects such as saturated fat, coconut oil, eggs, coffee and butter, among others. Today we explain why this happens.

Limited medical research studies

The first issue is that many research studies are not well-funded, and can only recruit a relatively small number of participants. Each person added to the study adds expense, and many studies have under 100 participants. That means that the research does not have much “power”, and can give skewed results. It could be the people recruited for that study had characteristics that gave results different from previous studies on the same question.

Large or “Meta” studies

The way that some researchers deal with small studies is to do a “meta-analysis”. That means they combine the results of multiple smaller studies to get an average result. The few studies that gave strange or unexpected results are “averaged out” in the meta-analysis. The average result from the majority of smaller studies gives a consensus result in a meta-analysis. (Many studies we report on here in this blog are meta-analyses.)

Most studies not randomized

Research studies fall into two broad types: randomized controlled studies and correlational studies. The first—randomized controlled—is much better than a correlational study, but takes much longer and is much more expensive to carry out. So unfortunately, most studies are the correlational type.

The randomized study

In a randomized “prospective” controlled study, typically one half of the study participants are randomly assigned to receive one treatment and the other half undergo a different treatment. Then at the end of the study the researchers see which treatment (or diet or whatever is studied) had better results.

The best type of randomized controlled study is also “double-blind”, meaning neither the researchers nor the participants know what treatment they are receiving. For example, half of the participants receive the “real” medication being studied, and the other half receives a “placebo” that looks and tastes like the real medication. During the study, the researchers themselves also do not know who is getting the real medication, and the “code” is only revealed at the end. This minimizes bias from the researchers.

Research bias

The researchers themselves may have a bias and desire a particular result, and that can influence the study. An example might be a study sponsored (paid for) by a drug company. Even though the research is conducted by university investigators, the researchers might feel pressured to get the results the drug company desires.

Media loves surprising stories

The media tends to report studies that show surprising results; different from previous studies on the subject. Surprising results get bigger headlines and more clicks and readers. There may have been many studies that confirmed the “conventional wisdom” but are not reported by the press because the expected is more boring.

What can you do

Try to not look just at the results of one study. If you are really interested in a topic, find other studies about the same subject. Put more faith in randomized controlled studies than in correlational studies. Look at the number of participants. Select meta-analysis studies over small studies. Sometimes you can tell who paid for the study.

If you know which institutions and medical journals have the best reputations, rely more on their studies. Here at ProcuraMed we do our best to choose the best studies to report, and if the study has limitations, we try to tell you. Finally, realize that science and medicine are constantly evolving, and the “accepted wisdom” changes over time!

To find a doctor, of any specialty, anywhere in Brazil, check out our website:

Esta postagem também está disponível em: Portuguese (Brazil)

Recommended doctors

This post is related to specialty Clinical Medicine. Below is a list of some physicians related to this specialty.