A Comparison of Cluster and Factor Analysis to Derive Dietary Patterns in Korean Adults Using Data from the 2005 Korea National Health and Nutrition Examination Survey.
- Author:
Yoon Ju SONG
1
;
Hee Young PAIK
;
Hyojee JOUNG
Author Information
1. School of Human Ecology, The Catholic University of Korea, Bucheon, Gyeonggi, Korea.
- Publication Type:Original Article
- Keywords:
dietary patterns;
cluster analysis;
factor analysis;
Korean Health and Nutrition Examination Survey(KHANES)
- MeSH:
Adult;
Aged;
Carbohydrates;
Cluster Analysis;
Energy Intake;
Factor Analysis, Statistical;
Fruit;
Humans;
Korea;
Male;
Nutrition Surveys;
Ovum;
Vegetables;
Surveys and Questionnaires
- From:Korean Journal of Community Nutrition
2009;14(6):722-733
- CountryRepublic of Korea
- Language:Korean
-
Abstract:
The purpose of this study was to explore dietary patterns and compare dietary patterns using cluster and factor analysis in Korean adults. This study analyzed data of 4,182 adult populations who aged 30 and more and had all of socio-demographic, anthropometric, and dietary data from 2005 Korean Health and Nutrition Examination Survey. Socio-demographic data was assessed by questionnaire and dietary data from 24-hour recall method was used. For cluster analysis, the percent of energy intake from each food group was used and 4 patterns were identified: "traditional", "bread, fruit & vegetable, milk", "noodle & egg", and "meat, fish, alcohol". The "traditional" pattern group was more likely to be old, less educated, living in a rural area and had higher percentage of energy intake from carbohydrates than other pattern groups. "Meat, fish, alcohol" group was more likely to be male and higher percentage of energy intake from fat. For factor analysis, mean amount of each food group was used and also 4 patterns were identified; "traditional", "modified", "bread, fruit, milk", and "noodle, egg, mushroom". People who showed higher factor score of "traditional" pattern were more likely to be elderly, less educated, and living in a rural area and higher proportion of energy intake from carbohydrates. In conclusion, three dietary patterns defined by cluster and factor analysis separately were similar and all dietary patterns were affected by socio-demographic factors and nutrient profile.
