Estimating usual food intake distributions by using the multiple source method in the EPIC-Potsdam Calibration Study.

Area of Science:

• Nutritional Epidemiology
• Statistical Modeling
• Dietary Assessment

Background:

• Accurate estimation of usual food intake is crucial for dietary assessment, especially for infrequently consumed foods.
• Traditional methods struggle with large proportions of nonconsumers in short-term dietary measurements.
• The Multiple Source Method (MSM) was developed to address these challenges in statistical modeling.

Purpose of the Study:

• To evaluate the applicability and effectiveness of the Multiple Source Method (MSM) for estimating usual food intake distributions.
• To assess MSM's performance with food groups characterized by a high percentage of nonconsumers.
• To demonstrate MSM's utility using real-world dietary data.

Main Methods:

• The Multiple Source Method (MSM) was employed, combining probability and amount of consumption.
• Covariates and habitual consumption frequency were incorporated into the statistical model.
• MSM was applied to a dataset from the EPIC-Potsdam Calibration Study (2004) with 393 participants, using two 24-h recalls and one FFQ.

Main Results:

• Usual intake distributions were successfully estimated for 38 food groups with >70% nonconsumers in 24-h recalls.
• MSM-derived intake estimates closely corresponded with observed values, including group means.
• The method effectively handled external information on nonconsumer proportions.

Conclusions:

• The Multiple Source Method (MSM) is a robust statistical technique for estimating usual food intake distributions.
• MSM is particularly valuable for food groups with a significant proportion of nonconsumers.
• The study confirms MSM's utility when at least two repeated dietary measurements per participant are available.