Consistency and Generalizability of Dietary Patterns in a Multiethnic Working Population.

BACKGROUND

Dietary pattern analysis is a complementary method to nutrient analysis in evaluating overall diet-disease hypotheses. Although studies have been conducted to derive dietary patterns among Malaysians, their consistency across subgroups has not been examined.

OBJECTIVE

The study aimed to derive dietary patterns empirically and to examine the consistency and generalizability of patterns across sex, ethnicity, and urban status in a working population.

DESIGN

This was a cross-sectional study using data from the Clustering of Lifestyle Risk Factors and Understanding its Association with Stress on Health and Well-Being among School Teachers in Malaysia study collected between August 2014 and November 2015. Dietary intake was assessed using a food frequency questionnaire, and dietary patterns were derived using factor analysis.

PARTICIPANTS/SETTING: Participants were teachers from selected public schools from three states in Peninsular Malaysia (n=4,618).

MAIN OUTCOME MEASURES

Dietary patterns derived using factor analysis.

STATISTICAL ANALYSES PERFORMED

Separate factor analysis was conducted by sex, ethnicity, and urban status to identify dietary patterns. Eigenvalue >2, scree plot, Velicer's minimum average partial analysis, and Horn's parallel analysis were used to determine the number of factors to retain. The interpretability of each dietary pattern was evaluated. The consistency and generalizability of dietary patterns across subgroups were assessed using the Tucker congruence coefficient.

RESULTS

There was no subgroup-specific dietary pattern found. Thus, dietary patterns were derived using the pooled sample in the final model. Two dietary patterns (Western and Prudent) were derived. The Western dietary pattern explained 15.4% of total variance, characterized by high intakes of refined grains, animal-based foods, added fat, and sugar-sweetened beverages as well as fast food. The Prudent dietary pattern explained 11.1% of total variance and was loaded with pulses, legumes, vegetables, and fruits.

CONCLUSIONS

The derived Western and Prudent dietary patterns were consistent and generalizable across subgroups of sex, ethnicity, and urban status. Further research is needed to explore associations between these dietary patterns and chronic diseases.