1. Introduction
Hypertension is a major public health issue worldwide [
1], and it is closely associated with increased risks of morbidity and mortality in cardiovascular and kidney diseases [
1,
2]. According to the Korea National Health and Nutrition Examination Survey data, the prevalence of hypertension in Korean adults has increased from 26.5% in men and 23.8% in women in 2007, to 35.1% in men and 29.1% in women in 2015 [
3].
Diet is an important determinant of hypertension development. A prudent dietary pattern has shown an inverse association with elevated blood pressure. In particular, the Dietary Approaches to Stop Hypertension (DASH) eating pattern, which emphasizes higher consumption of whole grains, legumes, nuts, fruits, vegetables, and low-fat dairy products, and lower consumption of fat, sweets, and refined grains, has been shown to have a favorable effect on blood pressure in epidemiologic studies [
4,
5,
6,
7]. In contrast, a Western-style dietary pattern, which consists of refined grains, red/processed meat, and high fat foods, has been associated with an increased risk of hypertension [
8,
9,
10].
A few studies regarding dietary patterns associated with hypertension have been conducted in Asian populations [
11,
12,
13]. In a cross-sectional study, a typical traditional diet, characterized by higher intakes of pork, egg, soybean, vegetables, fruits, and nuts, was inversely associated with hypertension in Chinese adults [
13]. Among Japanese women aged 40–69 years, a diet pattern with higher consumption of vegetables, potatoes, fruits, beans, and seaweeds was related to lower blood pressure levels [
11]. Recent studies have also suggested gender differences in the relationships between dietary factors and risks of chronic diseases, such as metabolic syndrome and hypertension [
14].
As dietary patterns and their relationships with hypertension vary according to gender, age, and ethnicity, investigating population-specific dietary patterns associated with hypertension is important. In addition, gender differences in the association between dietary pattern and hypertension has not been investigated in a prospective study. Therefore, the aim of this study was to identify the associations of dietary patterns with the risk of incident hypertension in Korean adults and examine gender differences in the associations, using a large community-based cohort dataset.
2. Materials and Methods
2.1. Study Population
The Korean Genome and Epidemiology Study is a prospective, community-based cohort study, which has been conducted to explore lifestyle and genetic risk factors associated with chronic diseases, such as diabetes mellitus and hypertension, in Koreans. The cohort included 10,030 Korean adults, aged 40–69 years, who lived in the Ansan and Ansung areas. Baseline examinations were performed between 2001 and 2002 and follow-up examinations were performed every 2 years over a 6-year period (2001–2006). All participants completed questionnaires on demographic information, lifestyle, medical history, and health conditions at baseline; anthropometric and blood pressure measurements were conducted biennially.
Of these 10,030 participants, 3271 participants who had hypertension at baseline, 799 participants who refused to participate in the follow-up surveys, 110 participants who had cardiovascular disease or cancer, 202 participants who had implausible energy intakes (<500 kcal or >6000 kcal per day), and 425 participants who did not complete the food frequency questionnaire (FFQ) were excluded from the study. In addition, 133 participants were excluded because of missing data. Finally, a total of 5090 participants (2457 men and 2633 women) were included in the data analysis. The study protocol was approved by the Institutional Review Boards of the Korea Centers for Disease Control and Prevention and Kyung Hee University, and informed written consent was obtained from all participants.
2.2. Dietary Assessment
Dietary intake was assessed using a validated, 103-item, semi-quantitative FFQ [
15] at baseline. In the FFQ, participants were asked how often and how much they consumed each food, on average, during the past year. Frequency was classified into nine categories: never/seldom, once/month, 2–3 times/month, 1–2 times/week, 3–4 times/week, 5–6 times/week, once/day, 2 times/day, or ≥3 times/day. The answer for portion size had three categories: small (1/2 serving), standard (1 serving), or large (≥2 servings).
To identify dietary patterns, food items were categorized into 26 food groups based on common food groups classified in the Korean nutrient database [
16]. Since the grain and grain product intake of Koreans is very high, this group was divided into seven subgroups: white rice, whole grains, noodles, rice cakes, breads, pizza and hamburgers, and cereals and snacks. The vegetables group was divided into two subgroups: vegetables and kimchi (traditional fermented salted vegetables). Kimchi was presented as a single group because it is commonly eaten as a traditional side dish in Korea, but it has high sodium levels, unlike other vegetables. Coffee was also separated from the beverage group, because the consumption of coffee is very common between meals or after meals among Korean adults. Nutrient intake was calculated using the seventh edition of the Korean food composition table [
17].
2.3. Definition of Hypertension
After resting in a seated position for at least 5 min, participants’ blood pressures were measured by trained technicians using a mercury sphygmomanometer (Baumanometer Standby; W.A. Baum Co. Inc., New York, NY, USA). Two readings of systolic blood pressure (SBP) and diastolic blood pressure (DBP) were recorded, and the average of the two readings was used for the analysis. Hypertension was defined as SBP ≥ 140 mmHg or DBP ≥ 90 mmHg or the use of antihypertensive medication, based on the 2003 guidelines of the Joint National Committee on Prevention, Detection, Evaluation, and Treatment of High Blood Pressure [
2]. SBP and DBP were calculated at Korotkoff phase I and Korotkoff phase V, respectively.
2.4. Covariates
Demographic factors, socioeconomic status, and lifestyle factors of participants were determined using questionnaires by trained interviewers. Educational level was categorized into three groups: ≤6 years, 7–12 years, or >12 years. Monthly household income was categorized into four groups: less than 1 million Korean Won (approx. less than 850 US
$ in 2016), 1 to less than 2 million Korean Won (approx. 850 to less than 1700 US
$), 2 to less than 3 million Korean Won (approx. 1700 to less than 2500 US
$), and greater than or equal to 3 million Korean Won. Smoking status was classified into three groups: never-smoker, former smoker, or current smoker. Alcohol consumption was classified into three groups: never-drinker, former drinker, or current drinker. Physical activity was assessed using metabolic equivalent of task (MET) hours per day. To obtain METs, participants reported hours spent on sleep and five types of activities, classified according to activity intensity, including sedentary, very light, light, moderate, and heavy activity. Total MET-hours per day were calculated by multiplying the reported hours spent on each type of activity during a day by the MET values that were determined based on the type or intensity of activity [
18]. Height was measured to the nearest 0.1 cm using a stadiometer (Aluminum anthropometer; Samhwa Instrument, Seoul, Korea) and body weight was measured to 0.1 kg in light clothes, without shoes, using a bioelectrical impedance analysis (Inbody 3.0; Biospace Corp., Seoul, Korea) by trained research staff. Body mass index (BMI) was calculated as weight divided by height squared (kg/m
2).
2.5. Statistical Analysis
Dietary patterns were derived for men and women separately using factor analysis. For the adjustment, the contribution of each food group to dietary pattern extraction was calculated as the energy intake consumed from each food divided by total energy intake consumed in a day. They were entered into the factor analysis using the FACTOR procedure of SAS software. To identify the number of factors to be retained, we used the eigenvalues > 1.3 criterion, the most widely used in factor analysis, as a first step. The factors were rotated by orthogonal transformation to achieve a simpler structure with greater interpretability. Post-rotated factor loadings revealed four factors that described the distinct dietary patterns of the study population. The scree plot showed a small break in the eigenvalues after four factors, which suggested that retention of four factors was appropriate for rotation. After varimax rotation, factor scores were saved from principle component analysis for each individual. The factor scores for each dietary pattern and for each individual were determined by summing the intake of each food group weighted by factor loading. The factor scores of each dietary pattern were categorized into tertiles and used for comparisons of nutrient intakes and other lifestyle variables.
Continuous variables were expressed as means and standard deviations, and categorical variables were expressed as percentages. Differences in baseline characteristics were examined by Chi-square tests or generalized linear model tests, as appropriate. We calculated the hazard ratios (HRs) and 95% confidence intervals (CIs) for incident hypertension across the tertiles of dietary pattern scores by time-dependent Cox proportional hazards models. In multivariable adjusted models, Model 1 was adjusted for age and BMI. Model 2 was adjusted for age, BMI, physical activity, educational level, household income, smoking status, and alcohol intake. All data analyses were conducted using SAS software, version 9.4 (SAS Institute, Cary, NC, USA). All p-values < 0.05 were considered statistically significant.
4. Discussion
In a large scale, prospective study, we found different dietary patterns according to gender among middle-aged and older Korean adults. The dietary patterns for men were coffee, fat, and sweets, prudent, whole grains and legumes, and traditional patterns, whereas women had a Western pattern instead of the traditional pattern. Among these dietary patterns, the whole grains and legumes pattern was associated with a 23% lower risk of incident hypertension in women, after adjustment for potential confounders, such as age, BMI, income, education, alcohol intake, smoking, and physical activity. This significant association was not observed in men. Other dietary patterns were not associated with hypertension in either men or women.
Consistent with our findings, previous studies support an inverse relationship between dietary patterns rich in whole grains and legumes and hypertension. In the Women’s Health Study, women who consumed ≥4 servings/day of whole grains had a 23% lower risk of hypertension relative to those who consumed <0.5 servings/day of whole grains, after adjustment for lifestyle, clinical, and dietary factors, over a follow-up period of 10 years [
19]. A randomized clinical trial reported that women consuming a whole grain and legume diet containing barley products, brown beans, and chickpeas for 4 weeks showed a significant reduction in blood pressure relative to a control group (−3 mmHg vs. 0 mmHg) in Swedish women aged 50–72 years [
20]. A cross-sectional study of Iranian adults aged ≥19 years showed that the prevalence of elevated blood pressure was significantly lower among frequent legume consumers (≥3 times/week) compared with those who did not consume legumes (31.3% vs. 43.5%) [
21].
The favorable effect of the whole grains and legumes pattern on hypertension might be due to combined and synergic effects between various nutrients and food components [
22,
23]. Nutrients included in whole grains and legumes, such as dietary fiber, calcium, potassium, and magnesium, have been shown to be helpful for reducing blood pressure [
24,
25,
26,
27]. A randomized, controlled trial showed that a diet rich in high-fiber plant foods, including dietary fiber (54 g/day) and beta-glucan (4.9 g/day), for 6 weeks significantly reduced SBP by 6.6 mmHg compared with a control diet, among Swedish adults aged 25–65 years [
28]. Potassium or magnesium in diets directly affects blood pressure levels by controlling the contraction or vasodilation of vascular cells [
25,
29].
The components of whole grains and legumes might have a beneficial effect on hypertension by mediating vascular endothelial functions [
30]. Jenkins and colleagues reported that a whole grain diet was associated with a lower reactive hyperemia index, a surrogate of peripheral vascular function, relative to a control diet [
31]. In the Coronary Artery Risk Development of Young Adults study, a diet characterized by a higher intake of whole grains and legumes was inversely related to specific cellular adhesion molecules, which are known as biomarkers of endothelial dysfunction [
32]. In addition to this, whole grains and legumes may moderate blood pressure through improving insulin sensitivity [
33] and/or inducing an increase in anti-inflammatory markers, such as adiponectin [
22]. Recently, gut microbiota and its metabolites have been known to influence the regulation of blood pressure. Short chain fatty acids (SCFA), derived from the gut microbial fermentation of dietary fiber, may regulate blood pressure via SCFA receptors. For example, the stimulation of G protein-coupled receptor 41, a SCFA receptor, decreases blood pressure by inducing vasorelaxation [
34,
35].
Interestingly, the inverse association of the whole grains and legumes pattern with hypertension was found only in women. The gender difference in the association between dietary pattern and hypertension has not been clearly explained. Diet might influence hypertension between men and women differently due to the renin-angiotensin system, sex hormones, or sex-specific genes, and increased inflammatory markers, which are related to the control of blood pressure. The response to stimulation of the renin-angiotensin system may differ between men and women [
36]. This may be related to sex hormones. For example, the decline in estrogen activates the renin-angiotensin system and increases angiotensin II levels; these alterations are related to elevated blood pressure [
37]. Likewise, the effects of diet or nutrients on blood pressure may be greater in women than men. In a study of the Korean National Health and Nutrition Examination Survey, women who consumed fried foods frequently (≥2 times/week) had a 2.4-fold higher risk of hypertension compared with those who rarely consumed fried foods, whereas this significant association was not shown in men [
38]. The reduction in blood pressure with a diet of reduced sodium, potassium supplementation or dark chocolate intervention was greater in women than men [
39,
40,
41]. A modified lifestyle approach involving a healthy diet was more effective in alleviating risk factors of hypertension in women than in men [
37]. Furthermore, there are many different pathophysiological and social reasons why the results may have been only significant in women, which could not be specifically determined in the present study. Therefore, the present study suggests that gender-based differences should be considered in the prevention and management of hypertension. Further studies are needed to elucidate underlying mechanisms, to explain gender differences in the association between diet and hypertension.
This study has several limitations. The study participants may not be representative of the entire Korean population, as this study was only conducted among middle-aged and older Korean adults. In addition, the procedures of factor analysis require several arbitrary decisions to identify dietary patterns. Despite these limitations, this is the first prospective study to evaluate the role of dietary patterns in the development of hypertension among Korean adults using large-scale cohort data. These findings suggest that a specific dietary pattern may be an independent risk factor for developing hypertension in Korean adults, and also indicates a gender difference in the association between dietary patterns and hypertension risk.