The Strength of Evidence

Most of us love the taste of chocolates! 

In recent months we have been reading headlines that indicate chocolate may help prevent various diseases, and most recently, lower weight. To many this seems like a dream come true! Could it be that something as high in calories, fat, and sugar could really be good for us?

True, the cacao bean is from the vegetable kingdom. It naturally has a number of phytochemicals, flavonoids, and antioxidants–each of which has the potential of benefit when consumed by humans. There are also minerals, vitamins, protein and other nutrients in chocolate. But do all these things really make it a healthful food?

This brings us to two cardinal principles of evidence. We can never draw causal conclusions from a single study, and second, associations do not prove cause. It always takes many studies, and multiple types of studies to begin drawing valid conclusions.

One of the most extensive projects ever conducted on food and disease risk is an ongoing project of the American Institute for Cancer Research. Thirteen Centers of Excellence evaluated more than 7,000 peer reviewed articles, encompassing more than 100,000 studies, examining 20 different types of cancer. This study is continuously updated

The criteria used in evaluating this data is enlightening and helpful in learning how to draw conclusions from all research. This project judged the strength of the evidence as follows: 

  1. Convincing Evidence is so strong it is unlikely to be modified by future research. To fall in this category data has to be taken from large, well-designed studies and from several types of studies indicating a clear causal relationship between food and the disease.
  2. Probable Evidence suggests there is strong enough reason to support a causal relationship.  At least 2 independent cohort studies and 5 case control studies must support these conclusions.
  3. Limited Evidence means the direction of the evidence is generally consistent but there are some unexplained flaws in some of the studies. 
  4. No conclusions due to mixed evidence and inconclusive results.

Based on these criteria, where would we place the evidence for the benefits of chocolate? At this point in time there is nothing more than “limited evidence“.

Could it be that those who consume chocolate also eat other healthy foods, exercise more, stress less, or were simply born with lower risks, etc? Perhaps chocolate is only a marker for other good behaviors that really do convey health benefits.

We need to wait for much more definitive answers on this and many other claims.