Observational studies had suggested that vitamin E reducesthe risk of heart dise

Question

Observational studies had suggested that vitamin E reducesthe risk of heart disease. Careful experiments, however, showedthat vitamin E has no effect, at least for women. According to acommentary in the journal of the American Medical Association:

    Thus, vitamin E enters the categoryof therapies that were promising in epidemiologic and observationalstudies but failed to deliver in adequately powered randomizedcontrolled trails. As in other studies, the "healthy user" biasmust be considered i.e. the healthy life style behaviors thatcharacterize individuals who care enough about their health to takevarious supplements are actually responsible for the better health,but this is minimized with the rigorous trial design.

a. What is the difference between observationalstudies and experiment?

b. What is "randomized control trial"?

c. How does "healthy user bias" explain how people whotake vitamin E supplements have better heath in observationalstudies but not in controlled experiments?

Explanation / Answer

A) In an experiment investigators apply specifictreatments to subjects and measure the effects of thetreatments on the subjects. In an observational studyinvestigators observe and measure variables of interest, but do notactually assign treatments to the subjects. B) A study in which people are allocated at random toreceive one of several treatments. One ofthese treatments is the standard of comparison or control(control is usually no treatment at all, a placebo, etc). C) In observational studies the "healthy user bias" is that ofthe people observed, the ones that care enough about their bodiesto take supllements (such as vitamin E in this case) are alsolikely to eat healthier, excercise more frequently, and in generaltake better care of themselves than people who do not take suchsupplements. In an experiment this is not a problem becausesubjects are assigned to either take the supplement or not (if theassignment is done randomly then lifestyle will be about the samein each group, as long as the sample size is relativelylarge).

Related Questions

Navigate

• Browse (All)
• Browse O
• Subjects

---