PDF
ePub
Citation
Print
Introduction
Although Korea has achieved a high standard of living and health promotion during the last 40 years, hypertension is still a major public health problem. According to the Korean Health Survey (Ministry of Health and Social Welfare, 1999), 27.8% of adult Koreans have hypertension. It was 27.9% in the 2005 survey (Ministry of Health and Social Welfare, 2005). Strokes triggered by hypertension were one of the six leading causes of death in Korea. Twenty seven point three out of 100,000 deaths were caused by hypertension and the number contributed nearly 70% of deaths from all cardiovascular diseases (Kim et al., 2005).
There is general consensus that blood pressure is related with health behaviors of smoking, drinking, exercise and diet. Studies showed that blood pressure is positively related with smoking (Fogari et al., 1996; Geory et al., 1991; Imamura et al., 1996), alcohol consumption (Gruchow et al., 1985; Marmot et al., 1994; Puddey et al., 1985) and negatively related with physical exercise (Arroll & Beaglehole, 1992; Blair et al., 1984; Paffenbarger et al., 1983).
Desirable food habits for preventing hypertension included increasing consumption of whole grains, vegetables, and fruits (Appel et al., 1997; Midgley et al., 1996; Stamler et al., 1996) are well documented from European and American studies.
However, because dietary patterns of Koreans differ from people of other countries, the effect could be different. Moreover, many studies have examined the association between the intake of individual nutrients or foods and the risk of hypertension, but only recently has attention focused on the relationship of overall dietary patterns to health risk (Lin et al., 2003). In comparison to individual foods or nutrients, dietary patterns are more realistic in describing the relation of diet to health and disease (Maskarinec et al., 2000), because people do not eat isolated nutrients, but rather meals composed of a variety of foods. Complex combinations of nutrients from the food may interact (Randall et al., 1992). From 1998, the World Health Organization (WHO) had suggested that dietary guidance for populations should be based on foods instead of nutrients (World Health Organization, 1998).
In the present investigation, identification of the dietary patterns of Korean males which are related with hypertension was carried out using the national survey data.
Materials and Methods
Data source
The food frequency dietary data of 1,869 male subjects aged 20-65 years from the 2001 Korean National Health and Nutrition Examination Survey (KHANES) were used in the analysis. The 62 items food frequency questionnaire was employed for the dietary survey. Detailed information about the 2001 KHANES is described elsewhere (Kim et al., 2005; Ministry of Health and Welfare, 2005).
In brief, for the subject, they were selected based on stratified multistage probability sampling from the total Korean population; therefore, we can have the generalization from these findings to the general Korean population.
Assessment of hypertension
Among the data from the 2001 Korean National Health and Nutrition Survey (KHANES), data of systolic and diastolic blood pressure in a health examination survey were used for the analysis. Hypertension was defined as a systolic blood pressure above 140mmHg or diastolic blood pressure above 90 mmHg
Assessment of dietary patterns
Factor analysis (Gorsuch, 1983) was conducted to derive dietary patterns from food consumption frequency of 38 food items, using the Factor procedures in SAS (version 8; SAS Institute, Cary, NC). Table 1 shows 38 food items which were categorized from 62 groups to reduce the complexity of the data. The food groups were categorized based on the principles of similarity of nutrient profiles, or culinary usage of the foods, mainly according to the Standard Tables of Food Composition in Korea, 6th Revised Edition (National Rural Living Science Institute, 2001), and the classification of food groups used by the National Nutrition Survey.
The factors were rotated by an orthogonal transformation (Varimax rotation function in SAS) to achieve simpler structure with greater interpretability. We considered components with an Eigen value greater than 1.6 as a significant factor. Eigen value 1.6 were identified as a break-point in the Screen plot and an interpretable level from previous study done by Schulz et al (2003). This served to limit the number of factors, as well as to better identify more meaningful factors. After Varimax rotation, factor scores were saved from the principal component analysis for each individual. All data presented here are from the Varimax rotation. Factor scores were categorized into quintile based on the distribution of the study population.
Statistical analysis for association
To determine the association between dietary patterns and hypertension, the odds ratios (OR) estimated for each quintile compared with the lowest quintile of each dietary pattern using Logistic regression analysis. In this analysis, age, smoking habit, drinking habit and physical activity were used as covariates based on the results of initial analysis. But, in the analysis of drinker pattern, the drinking habit as covariate was excluded.
Results
The characteristics of the study subjects are shown in Table 2. The mean age of the subjects was 40.9 years. The residential areas of the subjects were categorized as metropolitan city (41.9%), middle or small city (35.9%) and rural (22.2%). Among the study subjects, 41.9% of them had education levels of more than collage education.
Mean systolic and diastolic blood pressures were 123.4 and 79.9 mmHg, respectively. Hypertension prevalence was 10.0% (20-29 years), 19.2% (30-39 years), 28.1% (40-49 years), 35.1% (50-59 years) and 46.8% (60-65 years).
Table 3 showed dietary patterns identified from the factor analysis in terms of the factor-loading matrixes for the 3 major dietary patterns. The 3 major dietary patterns were retained in Screen plot Eigen values for males. In the model, the greater the loading of a given food item to the factor, the larger the contribution of that food item to a specific factor. And a negative loading designates negative association with the factor, while a positive loading designates positive association with the factor.
Dietary pattern 1, which was loaded heavily on vegetables, brown seaweed, fish, mushroom, modestly loaded with pork, chicken, egg with cereal, could be labeled the 'traditional' pattern.
Dietary pattern 2 was heavily loaded with fast food, bread, ham, modestly loaded with dairy product, chicken, fish paste and negatively loaded with cabbage (radish) could be labeled the 'western' pattern. Dietary pattern 3 which was heavily loaded with beer and soju, modestly loaded with pork and then negatively loaded with beans, cereal, and dairy products could be labeled as the 'drinker' pattern.
Table 4 shows mean blood pressure according to quintile of dietary patterns among subjects.
Diastolic blood pressure were increased with quintile of pattern 1 (traditional) up to 4th quintile and then decreased at the 5th quintile.
Both systolic and diastolic blood pressure did not show any significant variation with quintile of dietary pattern 2 (western).
Systolic blood pressure increased at quintile 2 of dietary pattern 3 (drinker pattern) and then decrease from the quintile 3.
Table 5 showed the results of second stage of analysis based on logistic regression analysis. It showed hypertension risk (in terms of odd ratio) according to the quintile of three dietary patterns identified by 4 analytical models. Crude model was built without adjusting covariance. Model 1 was built with adjusting for age, education, smoking, and drinking status. Model 2 was built with additionally adjusting for physical activity on model 1. Model 3 was built with further adjustments for energy intake on model 2.
In the crude model, hypertension risk was reduced with increasing consumption of traditional dietary pattern (pattern 1) except 2nd quintile). Models 1, 2, and 3 did not show any significant relationship between hypertension and traditional dietary pattern (pattern 1). Even though it was not statistically significant, there was a tendency that hypertension risk increased (odd ratio>1) with increasing quintile (up to the 4th quintile) of pattern 1 (traditional pattern) at models 1, 2, and 3. It may imply that the traditional dietary pattern which was rich in vegetables did not have any protective effect on hypertension risks among Korean males.
However, any of the relationships between hypertension and dietary patterns (1, 2 and 3) were not statistically significant at that model with adjusting covariance.
Discussion
From the factor analysis, three major Korean dietary patterns were identified. Pattern 1 was more or less similar to the 'Korean traditional', 'Healthy' and 'Prudent' pattern observed among Japanese (Sadakane et al., 2008), U.S.A. (Newby et al., 2004) and Norwegian (Konstantinova et al., 2008) population.
Pattern 2 appeared to be the western dietary pattern from the Korean point of view. Pattern 3 seemed heavily loaded with alcoholic food item such as beer, soju (Korean liquor) and meat products, thus quite an imbalanced dietary pattern from the nutritional heath point of view.
Unlikely observation from Norwegian and American studies (Konstantinova et al., 2008; Newby et al., 2004), the hypertension risk did not decrease with vegetable rich dietary pattern 1 (traditional) among Koreans. It may be explained by the dietary practice of Koreans who consumed the salted vegetables (Kimchi: fermented cabbage) instead of fresh vegetables (Kim & Lee, 2002). It may imply the sodium consumption from salted vegetable had played an important role in developing hypertension among Koreans. The ages of study subjects could be another factor contributing to the finding. As shown in the table 1, the prevalence of hypertension increased with age. The prevalence was 10%, 19.2%, 28.1%, 35.1% and 46.8% among the study subjects of aged 20-29, 30-39, 40-49, 50-59 and 60-65, respectively. The tendency of consuming a more traditional Korean diet among the aged population than the younger population (Song et al., 2005) might have been a contributed to the findings. According to the another recent study in Korea, the rapid adoption of the westernized dietary pattern has occurred especially among the younger generation (Song et al., 2005). The results were similar to other studies carried out in Japan (Kim et al., 2004).
In summary, unlikely findings from other countries, the vegetable rich traditional Korean diet did not show any protective effect on hypertension among Korean males. The prevalence of hypertension with the aging seemed to play an important role on this finding. And the dietary practice of Koreans consuming salted vegetables (as a form of Kimchi) instead of fresh vegetables might have been another contributing factor on these findings.
However, a full explanation of the findings remained to be answered with further investigation since none of the dietary patterns showed any statistical significance with hypertension. The statistical weakness on the model may also suggest that other variables than the variables included in the present study might have played a more important role in developing hypertension among Korean males.
XML Download