Cholesterol intake as measured by unquantified and quantified food frequency interviews: implications for epidemiological research.

A number of dietary assessment methods have been developed for epidemiological investigations. The general tendency has been to strive for methods that are sufficiently simple to implement in a fairly large sample of respondents and yet are able to measure dietary intakes with an acceptable degree of reliability. Concomitantly, methodological studies are needed to determine whether the various methods can be used interchangeably. This study compares the unquantified and quantified food frequency interview methods for measuring cholesterol intake on 4638 subjects from a multi-ethnic population in Hawaii. The methods were only modestly related, with product-moment correlation of 0.36 for men and 0.30 for women. The unquantified frequency method tended to yield a higher cholesterol value than did the quantified frequency method among lower level consumers, whereas the opposite was true for higher level consumers. These findings suggest that unquantified frequency methods cannot be used interchangeably with corresponding quantified frequency methods in studies of association between dietary cholesterol and disease occurrence.