Limitation

Dietary pattern analysis used in this study considers the overall diet rather than individual food items. Therefore, we cannot investigate which particular nutrients are responsible for the observed difference in disease risk. For the same reason, dietary pattern analysis may provide less information about the biological relationship between specific food items and disease risks than a traditional single food item analytical approach (2). Also, the effect of specific food items associated with disease risk would be diluted in dietary pattern analysis.

This study is depending on a food frequency questionnaire. Therefore, it is inevitable to have recall bias. Factors such as socioeconomic status, ethnicity, and cultural background contribute to varying dietary patterns. However, in this study, only male Koreans were sampled. But it is also important to carry out the experiment in a diverse population who might have different eating patterns. Since the differences can be a result of various factors such as social and cultural background, the results would not disprove the authenticity of pattern analysis (2,9).