Applying compositional data methodology to nutritional epidemiology.

Area of Science:

• Nutritional Epidemiology
• Biostatistics
• Public Health

Background:

• Dietary data often represent parts of a whole, posing statistical challenges in nutritional epidemiology.
• Understanding the impact of specific nutrients on disease requires methods that account for data composition and total energy intake.

Purpose of the Study:

• To apply a compositional data perspective using isometric log-ratio (ILR) transformation for statistical analyses in nutritional epidemiology.
• To evaluate the association between macronutrient intake (proteins, fats, carbohydrates) and metabolic syndrome using ILR variables.

Main Methods:

• Utilized isometric log-ratio (ILR) transformation for analyzing compositional dietary data.
• Employed logistic regression models to assess the odds of metabolic syndrome.
• Analyzed data from an Italian population-based study of middle-aged subjects.

Main Results:

• The ILR approach allows for robust statistical inferences on individual dietary components.
• The models successfully adjusted for total energy intake, isolating the effects of macronutrient proportions.
• Specific macronutrient intakes were linked to the odds of developing metabolic syndrome in the studied population.

Conclusions:

• Compositional data analysis, specifically using ILR transformation, provides a powerful framework for nutritional epidemiology.
• This method enhances the ability to investigate diet-disease relationships by accounting for the relative nature of dietary intake.
• The findings contribute to a better understanding of how dietary patterns influence metabolic health.