The use of only self-report measures as criteria regarding validity may be considered as a second limitation of our study. Theoretically, differences in response styles between ethnic groups that are consistent across questionnaires might also lead to a similar association of responses on the ADL index and the self-reported health measures that we used as criteria