Modeling Usual Intake Using Dietary Recall Data
For simplicity, dietary recommendations intended to achieve nutrient adequacy and promote health are often expressed in terms of daily targets. However, because nutrients are stored in the body, it is unnecessary to achieve those targets every day. Furthermore, because dietary intake varies from day to day, “usual” or long-term average intake is a key concept in dietary assessment.
Unfortunately, all self-report dietary assessment tools are prone to error. The types of errors vary by instrument, and are described in Module 12 of the Dietary Tutorial. (We recommend that you review Module 12 before beginning this module.) Due to the presence of these errors, statistical methods are needed to estimate usual intake of dietary constituents.
This module provides essential background for Modules 19-22. It describes the history of usual intake estimation, statistical methods that have been used to estimate the distribution of dietary intake, and the development of a unified framework for estimating usual dietary intakes (“the NCI method”). It also illustrates the use of balanced repeated replication (BRR) to estimate standard errors; BRR is used in Modules 19-22.
Task 1: Describing Measurement Error
Measurement error may have a large impact on estimating usual intake of dietary constituents. In this task, different types of measurement error and the impact that this error can have on estimates of usual intake are described.
Task 2: Describing Statistical Methods that Have Been Used to Estimate the Distribution of Usual Intake with a Few Days of 24-hour Recalls
Statistical methods have been developed to estimate the distribution of usual intake when 24-hour recalls are used to assess dietary intake. This task describes these methods, including similarities and differences.
- Key Concepts about Statistical Methods that have been used to Estimate the Distribution of Usual Intake with a Few Days of 24-hour Recalls
Task 3: Using a Unified Framework to Estimate Usual Dietary Intakes
In collaboration with colleagues from numerous institutions, the National Cancer Institute (NCI) developed a unified framework to predict usual dietary intakes of episodically-consumed or ubiquitously-consumed dietary constituents. This method requires that two or more 24-hour recalls are available for at least a subset of a sample. It can be used for a variety of general applications.
Task 4: Using Balanced Repeated Replication to Estimate Standard Errors
Balanced repeated replication (BRR) is used to estimate variance in succeeding modules.
- Key Concepts about using Balanced Repeated Replication (BRR)
- How to Estimate Standard Errors with Balanced Repeated Replication (BRR) Using SAS
- Download Sample Code and Datasets