Send to

Choose Destination
See comment in PubMed Commons below
Prev Med. 1989 Sep;18(5):653-60.

Human dietary assessment: methods and issues.

Author information

Division of Cancer Prevention and Control, National Cancer Institute, Bethesda, Maryland 20892.


It is possible that a "good" diet may enhance response to a chemopreventive agent, or that a diet deficient in some nutrients may weaken host defenses or host response to the agent. Thus, dietary intake should be assessed in chemoprevention as well as dietary studies. The advantages and disadvantages of several methods are presented. In addition, several principles and issues involved in dietary studies are discussed. Studies should attempt to assess the whole diet, not one or a few nutrients. Interpretation of the results of dietary studies can be seriously flawed and conclusions incorrect if only one or a few nutrients are assessed. Studies that ask about vegetable products but calculate only a fiber index are forced into drawing conclusions about fiber, when the true effective component may be elsewhere. Investigators should be cautious in the analysis and interpretation of results involving correlated nutrients. If two variables are quite highly correlated, either positively or negatively, a nutrient which has no effect may appear to have one, merely by being correlated with the other. Some examples are presented. Misclassification or measurement error is a serious problem for all dietary methods, although for different reasons. For most frequency or few-day-record methods, sample sizes must be multiplied several-fold over the sample size which standard formulas produce, to retain power in the face of measurement error.

[Indexed for MEDLINE]
PubMed Commons home

PubMed Commons

How to join PubMed Commons

    Supplemental Content

    Loading ...
    Support Center