Diet Screener in the 2005 CHIS: Validation Results

Risk Factor Assessment Branch (RFAB) staff have assessed the validity of the CHIS Diet Screener in the Eating at America's Table Study (EATS) and the Observing Protein and Energy Nutrition Study (OPEN). In these studies, multiple 24-hour recalls in conjunction with a measurement error model were used to assess validity. The screeners used in these studies were similar but not identical to that used in CHIS 2005. We have constructed similar questions from the data available and the scoring algorithms developed specifically for CHIS. For added sugar, questions answered on the Diet History Questionnaire administered in both studies were used as proxies for some of the screener items.

Estimates of total Pyramid Servings of fruits and vegetables and teaspoons of added sugar in EATS and OPEN are shown in the table below.

  Men (median intake) Women (median intake)
Recall Screener Recall Screener
Total fruits and vegetables (Pyramid servings)
EATS 5.8 5.5 4.3 4.6
OPEN 6.2 5.3 5.2 4.7
Total fruits and vegetables (cup equivalents)
EATS 3.38 3.24 2.43 2.66
OPEN 3.60 3.13 3.09 2.75
Added sugar
EATS 17.5 18.1 12.3 11.7
OPEN 17.0 17.0 13.0 11.8

These validation results suggest that dietary exposure estimates computed for the CHIS may be useful to compare subgroup means, especially for populations consuming mainstream diets. The estimates may be less useful for populations with ethnic diets, including Asian and possibly Latino populations.

At the individual level, correlations between the screener and estimated true intake for servings of fruits and vegetables and added sugar are shown in the table below.

  Men Women
Total fruits and vegetables (Pyramid servings)
EATS 0.67 0.49
OPEN 0.58 0.73
Total fruits and vegetables (cup equivalents)
EATS 0.70 0.52
OPEN 0.62 0.70
Added sugar
EATS 0.59 0.66
OPEN 0.69 0.66

Overall, about 25 to 50 percent of the variability in the true intake in fruit and vegetable and added sugar will be captured by the screener questions. Thus, although significant error may be associated with these estimates of diet, we believe the exposure estimates still substantially reflect what individuals are actually consuming.

Validation results for the Multifactor Screener, which includes a similar fruit and vegetable component, are reported in detail in:

National estimates based on the 2000 NHIS Multifactor Screener are presented and compared with other national data in:

Return to Top