Rohrmann Sabine, Klein Gisela
Abteilung Klinische Epidemiologie, Deutsches Krebsforschungszentrum, Heidelberg, Germany.
Eur J Public Health. 2003 Sep;13(3):262-8. doi: 10.1093/eurpub/13.3.262.
The aim of this study was the development and validation of a short list of food items to assess the intake of total fat, saturated, mono-unsaturated, polyunsaturated fatty acids, and cholesterol. The short list should be able to correctly classify persons according to their intake.
A short list of 20 food items was selected out of 1009 seven-day food records by means of the variance-based method Max_r. This list was validated using data from a further 479 persons who completed seven-day food records (validation sample 1, VS1) as well as a food frequency questionnaire (validation sample 2, VS2). The intake of total fat, different fatty acids, and cholesterol from the complete VS1 (VS1(complete)) and from the complete VS2 (VS2(complete)), respectively, was computed. Further, the intake in VS1 (VS1(short)) as well as in VS2 (VS2(short)) using only the 20 food items on the short list were calculated.
Pearson correlation coefficients between the intake calculated from the items on the short list and the nutrient intake calculated from the full instrument in VS1 and VS2, respectively, were r = 0.81-0.91. In a quartile's cross-classification 53.4-64.1% of the participants were assigned to the same quartile. When comparing VS1(complete) with VS2(short), neither correlation coefficients nor the cross-classification differ much from the comparison of VS1(complete) with VS2(complete).
The short list shows good results in both validation samples. Thus, the short list can assess the variability of fat intake and classify persons according to their intake.
