R. Kaaks et E. Riboli, VALIDATION AND CALIBRATION OF DIETARY-INTAKE MEASUREMENTS IN THE EPICPROJECT - METHODOLOGICAL CONSIDERATIONS, International journal of epidemiology, 26, 1997, pp. 15-25
The statistical power of prospective studies on diet in relation to ch
ronic disease risk can be improved by maximizing the variation in true
intake levels actually distinguished-or 'predicted-by dietary questio
nnaire assessments collected at baseline, This can be achieved by 1) d
eveloping a questionnaire method that provides measurements with the s
mallest possible random errors, thus maximizing the correlation of mea
sured with true habitual intake levels; and 2) increasing the between-
person variation in true dietary intake levels when combining multiple
cohorts in populations with diverse consumption patterns. The first a
pproach implies that, during the development or selection of the quest
ionnaire method, correlations between measurements and true intake lev
els can be monitored; the second approach requires adjustment for betw
een-centre differences in over- or underestimation of dietary question
naire measurements. Besides optimizing the statistical power, it is im
portant that the magnitude of the predicted variation in true intake l
evel is estimated accurately, so as to allow unbiased estimations of r
elative risks. To meet these various objectives, substudies must be co
nducted for the 'validation' or 'calibration' of dietary questionnaire
assessments, by comparison with additional measurements that have ind
ependent sources of error. This paper reviews the methodological consi
derations underlying the design and implementation of such substudies
in the EPIC project, a collaborative multicentre study in nine Western
European countries.