Variations in sensitivity, specificity, and predictive value of a dietary fat screener modified from Block et al.

OBJECTIVE

To evaluate the ability of a fat screener modified from Block et al to discriminate persons whose diets consist of 38% of kilocalories or more from fat from the remainder of the population.

DESIGN

Sensitivity, specificity, and positive predictive value of the adapted screener were calculated. Percentage of kilocalories from fat was assessed by means of a semi-quantitative food frequency questionnaire. Three cutoff points from the fat screener were used to examine which best identified those whose diets consisted of 38% of kilocalories or more from fat.

SUBJECTS

Nine hundred ninety-seven persons responded to a food frequency questionnaire sent to a random sample of 2,000 members of a health maintenance organization.

RESULTS

Sensitivity and specificity varied depending on which cutoff point from the fat screener was used. Sensitivity reached a high of 83.3% and specificity reached a high of 92.1%, but the screener was never highly sensitive and specific simultaneously, and the results did not vary considerably by race. The screener had low rates of gross misclassification into quintiles (< or = 2.7%) and was more effective at classifying respondents into quintiles of total fat intake (64.9% to 85.5% classified in the same quintile) than into quintiles of percentage of kilocalories from fat (43.2% to 60.5% classified in the same quintile).

CONCLUSIONS

The adapted fat screener may be used in conjunction with other dietary evaluation methods, but it exhibits insufficient sensitivity and specificity to be used as a single assessment method.