Wednesday, January 27, 2010

Tea, Weight Loss, and Control Studies

The Linus Pauling institute at Oregon State University has an outstanding page summarizing recent scientific research on the health effects of tea. This page makes an interesting comment about tea and weight loss:
"It is currently unclear whether tea or tea extracts promote weight loss. Large-scale clinical trials that control for energy intake and expenditure are needed to answer this question."
Are control studies really what we want here?

Western science is based around the idea of controlled studies that isolate one variable and examine the effect of changing that variable when others stay constant. But do controlled studies actually capture what happens in a practical setting? Not at all.

Drinking tea has many mechanisms by which it can influence weight loss through influencing energy intake:

  • Tea is a calorie free drink. Tea can be consumed as a substitute for calorie-rich drinks such as soft drinks, milk, fruit juice, or energy drinks. In this case, drinking tea could directly reduce someone's calorie intake.

  • Tea can be quite bitter, and is generally not very sweet. Drinking unsweetened tea on a regular basis could change people's tastes, making them more accustomed to foods that are less sweet. This could reduce their calorie intake through reducing refined sugars in their diet.

  • Tea contains caffeine, which is known to reduce appetite. Tea also contains a myriad of other chemicals, many of which are poorly understood. It is reasonable that tea could have other effects on appetite and food preferences beyond those caused by caffeine as well. Either way, drinking tea is likely to lead people to reduce their calorie intake overall.

Controlling for energy intake is missing the point. People want to know whether drinking tea will help them lose weight. The vast majority of these people, even those who count calories some of the time, do not effectively control their energy intake the way a scientific study would. People interested in weight loss in the practical setting of their own lives will be more interested in how drinking tea will influence their caloric intake through appetite, food preferences, and any other means. Control studies are missing the point!

Oh, and one last observation--tea extracts do not have the first and second benefit above. Actually drinking tea is necessary to get the full weight-loss potential, since part of the way tea can help you lose weight is through the act of drinking and tasting it, not just through chemicals that you can pop in a pill.


  1. Actually, I think you're talking about something slightly different than what they are. They are addressing a specific question, and you are addressing a broad question.

    Controlling for energy intake is exactly their point. What they are saying is needed is to compare, say, 2000-calorie WITH tea (or its extracts) to 2000-calorie intake WITHOUT tea to see if there's a different result. That will answer the specific question of whether it's the presence of tea that makes a difference (if there is one) in weight.

    You, on the other hand, are asking a broader question--are there ANY influences of tea that might lead to weight loss. The question you are asking includes not only the biochemical question they're trying to answer, but also ones of perception, behavior, psychology, motivation, etc. To answer the question you want answered, a great number of different studies need to be done.

    Part of the issue is that the only real science on tea consumption has been done in the last 10 years or so. We know SO LITTLE about tea, that people's expectations of what we ought to know is so far ahead of what's actually been researched that they go making assumptions, wanting defined answers to things that haven't even been well defined, and so forth.

  2. Yes, in the end it would be valuable to know both what could be gained by a controlled study, and what could be learned by broader studies to get at what I'm talking about in this post.

    However, I still take issue with the statement made on the Linus Pauling Institute's homepage, and with those types of statements in general. It's a common sort of statement that scientists make--thinking, writing, saying, or implying that their particular mode of researching is more broadly applicable or more practical for real-world conclusions than it actually is.

    The original statement (read it carefully!) is: "It is currently unclear whether tea or tea extracts promote weight loss. Large-scale clinical trials that control for energy intake and expenditure are needed to answer this question."

    This statement isn't really true. The question of whether "tea promotes weight loss" is a broad, general question. Not only are such trials unnecessary to answer that question (in a practical setting) but they could be misleading (i.e. finding no effects or only weak effects in a control setting whereas the effects in a real-world setting could be large and significant).

    Is this making sense?

    If scientists are going to do science well and not mislead people by making overly general statements, they need to be more precise in their use of language. The statement about control studies could be made true and more precise by making the first sentence "It is currently unclear whether tea or tea extracts promote weight loss through altering metabolism" or "through any mechanisms other than effects on caloric intake".

    Am I splitting hairs? I think this stuff is important because people glance at it without thinking, trust it because the source looks authoritative, and end up getting the wrong impression of how the scientific process works.