A Semiparametric Transformation Approach to Estimating Usual Daily Intake Distributions

Abstract

bstractThe distribution of usual intakes of dietary components is important to individuals formulating food policy and to persons designing nutrition education programs. Usual intake of a dietary component for a person is the long run average of daily intakes of that component for that person. Because it is impossible to directly observe usual intake of an individual, it is necessary to develop an estimator of the distribution of usual intakes based on a sample of individuals with a small number of daily observations on each individual. Daily intake data for individuals are nonnegative and often very skewed. Also, there is large day-to-day variation relative to the individual-to-individual variation and the within-individual variance is correlated with the individual means. We suggest a methodology for estimating usual intake distributions that allows for varying degrees of departure from normality and recognizes the measurement error associated with daily dietary intakes. The estimation method contains four steps. First, the original data are standardized by adjusting for weekday and interview sequence effects. Second, the daily intake data are transformed to normality using a combination of power and grafted polynomial transformations. Third, using a normal components-of-variance model, the distribution of usual intakes is constructed for the transformed data. Finally, a transformation of normal usual intakes to the original scale is defined. The approach works well for a set of dietary components selected from the [1985][1986]

Unique patterns of dietary adequacy in three cultures of Canadian Arctic indigenous peoples

“…The Software for Intake Distribution Estimation (SIDE) software (version 1, 1996, Iowa State University) for SAS version 8.0 (SAS Institute) was used to estimate usual nutrient intake distributions. With a series of transformations and statistical manipulations [34][35][36] , SIDE uses observed nutrient intakes and repeat observations to estimate distribution of usual nutrient intakes by partitioning the observed inter-individual and intra-individual variance which, in effect, removed intra-individual variance in the data. Variance of measurement error, scaled fourth moment of measurement error and number of individuals with repeat observations in an 11% subsample of Inuit recalls were used to transform 1-day distributions to distributions of usual intakes for the three cultural groups, thus permitting these intakes to be assessed in an identical manner.…”

Brazilians’ experiences with iron fortification: evidence of effectiveness for reducing inadequate iron intakes with fortified flour policy

“…The use of methods to remove within-person variance and estimate usual nutrient intake is widely recommended and has been implemented in several studies worldwide. To do so, at least one replication of the 24HR is needed in a subsample of the study population (12) . In the post-fortification period we administered two non-consecutive 24HR, the first was in person and the second a telephone-based interview (in a sample of 50·06 %).…”

“…Therefore, to correct the distribution of Fe intake in the pre-fortification period, we applied the variance components derived from the post-fortification period. To estimate within-person variance components by each age and sex group and the prevalence of inadequate Fe intake we used the Software for Intake Distribution Estimation (PC-SIDE) that implements the method proposed by Nusser et al (12) .…”

Estimating the distribution of usual dietary intake by short-term measurements

“…Almost simultaneously, four different methods were proposed by Slob (1993), Wallace et al (1994), Buck et al (1995), and Nusser et al (1996), where the last method can be considered as an extension of the suggestions of the National Research Council (1986) of the USA. Recently, Gay (2000) suggested a further method, also discussed in this paper.…”

“…Naturally, it is impossible to have such a large number of repeated measurements. (Buck et al, 1995) Symmetric b None Variance decomposition ANOVA (SAS, SPSS) Nusser method (Nusser et al, 1996) Normal after transformation Power=grafted polynomial Variance decomposition SIDE (SAS), C-SIDE (UNIX) Gay method (Gay, 2000) Normal …”