National Cancer Institute Home at the National Institutes of Health | www.cancer.gov

Five-Factor Screener: Validation Results

RFMMB staff have assessed indirectly the validity of parts of the Five-Factor Screener in two studies: NCI's Observing Protein and Energy (OPEN) Study and the Eating at America's Table Study (EATS). In both studies, multiple 24-hour recalls in conjunction with a measurement error model were used to assess validity.

Table 1. Estimated mean F&V servings*, Fiber, Calcium, Dairy Servings, and Added Sugar from 24HR and screener and de-attenuated Pearson correlation coefficient between true intakea and screener, by gender: OPEN
Dietary Factor N Mean (95% CI) De-attenuated Pearson
Correlation Coefficient (SEE)
24HR Screener Difference
Total Pyramid servings of F&V* (square root)
Men 260 2.48 2.28 -0.212 0.58
(0.066)
Women 221 2.29 2.14 -0.162 0.73
(0.078)
Pyramid servings of F&V* -- Fries (square root)
Men 260 2.43 2.22 -0.212 0.61
(0.067)
Women 221 2.25 2.11 -0.152 0.74
(0.070)
Grams of Fiber (cube root)
Men 260 2.78 2.56 -0.222 0.52
(0.59)
Women 221 2.55 2.36 -0.182 0.54
(0.70)
Mgs of Calcium (fourth root)
Men 260 5.44 5.20 -0.252 0.59
(0.066)
Women 221 5.18 4.87 -0.312 0.44
(0.080)
Tsps of Added Sugar (cube root)
Men 260 2.61 2.58 -0.04 0.68
(0.039)
Women 222 2.39 2.29 -0.102 0.66
(0.045)
Servings of Dairy (square root)
Men 260 1.13 1.13 -0.00 0.64
(0.041)
Women 221 1.03 0.99 -0.04 0.64
(0.44)


Table 2. Estimated mean F&V servings*, Fiber, Calcium, Dairy Servings, and Added Sugar from 24HR and screener and de-attenuated Pearson correlation coefficient between true intakea and screener, by gender: EATS
Dietary Factor N Mean (95% CI) De-attenuated Pearson
Correlation Coefficient (SEE)
24HR Screener Difference
Servings of Total F&V* (square root)
Men 184 2.44 2.36 -0.081 0.70
(0.058)
Women 247 2.07 2.11 0.04 0.54
(0.060)
Servings of F&V* -- Fries (square root)
Men 184 2.34 2.29 -0.05 0.72
(0.054)
Women 247 2.01 2.06 0.05 0.55
(0.058)
Grams of Fiber (cube root)
Men 184 2.73 2.59 -0.121 0.60
(0.059)
Women 247 2.41 2.35 -0.062 0.55
(0.054)
Mgs of Calcium (fourth root)
Men 184 5.42 5.26 -0.042 0.60
(0.061)
Women 247 5.01 4.93 -0.07 0.56
(0.053)
Tsps of Added Sugar (cube root)
Men 446 2.64 2.67 0.03 0.59
(0.037)
Women 519 2.32 2.35 0.03 0.66
(0.032)
Servings of Dairy (square root)
Men 446 1.15 1.17 0.03 0.74
(0.32)
Women 519 0.95 1.00 0.052 0.73
(0.029)


Table 3. Median intakes of fruits and vegetables (Pyramid servings*) and added sugar (teaspoons) for NHANES 2001-02, NHIS 2005, and CHIS 2005 by gender and race/ethnicity
Gender and
Race/Ethnicity
Pyramid Servings* of Fruits and Vegetables Teaspoons of Added Sugar
NHANES 2001-02
(24HR)
NHIS 2005
(9-item screener)
NHANES 2001-02
(24HR)
NHIS 2005
(4-item screener)
Men
Non-Hispanic White 4.7 5.2 20.5 19.0
Non-Hispanic Black 4.2 5.3 23.6 21.4
Hispanic 5.2 5.8 21.6 21.4
Women
Non-Hispanic White 4.2 4.4 14.0 12.8
Non-Hispanic Black 3.5 4.3 19.3 15.5
Hispanic 4.0 4.6 16.3 14.5


Table 4. Median intakes of fiber (gm) and calcium (mg) for NHANES 2003-04 and NHIS 2005, by gender and race/ethnicity
Gender and
Race/Ethnicity
Fiber (gm) Calcium (mg)
NHANES 2003-04
(24HR)
NHIS 2005
(18-item screener)
NHANES 2003-04
(24HR)
NHIS 2005
(2-item screener)
Men
Non-Hispanic White 15.7 17.9 1061 876
Non-Hispanic Black 12.7 17.4 787 753
Hispanic 17.2 22.3 917 864
Women
Non-Hispanic White 12.5 13.9 688 653
Non-Hispanic Black 9.7 13.4 542 584
Hispanic 13.3 16.8 703 691

These validation results suggest that dietary exposure estimates computed for the 2005 NHIS Cancer Control Supplement (CCS) may be useful to compare subgroup means, especially for populations consuming mainstream diets. The estimates may be less useful for populations with more ethnic diets, including Asian and possibly Latino populations. Although significant error may be associated with these estimates of diet, we believe the exposure estimates still substantially reflect what individuals are actually consuming.


* Using 1992 Food Guide Pyramid definitions of servings.
a Estimated from multiple 24-hour dietary recalls in a measurement error model.
1 p < 0.05
2 p < 0.01

Last Modified: 11 Apr 2014