7
THE FRUIT & VEGETABLE SCREENER: CHIS 2000
fruits and vegetables calibrate the data to 24-hour recalls. The screener estimate of intake
represents what we expect the person would have reported on his 24-hour recall, given
what he reported on the individual items in the screener. As a result, the mean of the
screener estimate of intake should equal the mean of the 24-hour recall estimate of intake
in the population. (It would also equal the mean of true intake in the population if the 24-
hour recalls were unbiased. However, there are many studies suggesting that recalls
underestimate individuals' true intakes).
When describing a population's distribution of dietary intakes, the parameters needed are
an estimate of central tendency (i.e. mean or median) and an estimate of spread (variance).
The variance of the screener, however, is expected to be smaller than the variance of true
intake, since the screener prediction formula estimates the conditional expectation of true
intake given the screener responses, and in general the variance of a conditional
expectation of a variable X is smaller than the variance of X itself. As a result, the screener
estimates of intake cannot be used to estimate quantiles (other than median) or prevalence
estimates of true intake without an adjustment. Procedures have been developed to
estimate the variance of true intake using data from 24-hour recalls, by taking into
consideration within person variability [1, 2]. We extended these procedures to allow
estimation of the variance of true intake using data from the screener. The resulting
variance adjustment factor adjusts the screener variance to approximate the variance of
true intake in the population.
How did we estimate the variance adjustment factors?
We have estimated the adjustment factors in an external validation dataset available to us.
The results indicate that the adjustment factors differ by gender: 1.2 for men and 1.1 for
women. Under the assumption that the variance adjustment factors appropriate to the
California Health Interview Survey are similar to those in the Eating in America's Table
Study (EATS)[3], the variance-adjusted screener estimate of intake should have variance
closer to the estimated variance of true intake than would have been obtained from repeat
24-hour recalls. For a slightly different fruit and vegetable screener (7 rather than 8 items)
validated in the Observing Protein and Energy Nutrition (OPEN) Study, the variance
adjustment factors are quite similar, which gives us some indication that these factors
might be relatively stable from population to population. The OPEN Study screener is
available in NCI’s Register of Validated Short Dietary Assessment Instruments.
How do you use the variance adjustment estimates?
To estimate quantile values or prevalence estimates for an exposure, you should first adjust
the screener so that it has approximately the same variance as true intake.
Adjust the screener estimate of intake by:
multiplying intake by an adjustment factor (an estimate of the ratio of the standard
deviation of true intake to the standard deviation of screener intake); and
adding a constant so that the overall mean is unchanged.