Sex differences in hypertension prevalence and control: Analysis of the 2010-2014 Korea National Health and Nutrition Examination Survey

Although not fully understood, sex may affect both the prevalence and control rate of hypertension. The present study was designed to investigate factors associated with hypertension prevalence and control among Korean adults. We analyzed 27,887 individuals (12,089 males and 15,798 females) aged 30 years or older who participated in the fifth (2010–2012) and sixth (2013–2014) Korea National Health and Nutrition Examination Survey. Multiple logistic regression models were applied to delineate factors associated with the prevalence and control of hypertension separately for men and women. Overall, the prevalence of hypertension was higher in men (34.6%) than in women (30.8%). However, after the age of 60 years, hypertension was more prevalent in females than in males. Regardless of sex, the older the participants were, the more likely they were to have hypertension. Factors positively associated with hypertension prevalence were old age, low education, and high BMI in women (p<0.001) and increasing age, low income, alcohol intake, and high BMI in men (p<0.001). The overall control rate of hypertension was higher in women (51.3%) than in men (44.8%). However, after the age of 60 years, hypertension control rates were higher in men than in women. Factors decreasing hypertension control were white-collared women and young age, alcohol consumption in men. Sex differences in hypertension prevalence and control were discovered among Korean adults. After the age of 60, females were more likely to have hypertension and less likely to maintain hypertension control than males of the same age range. Accordingly, sex-specific approaches are recommended for effective blood pressure management.


Introduction
Hypertension is an essential public health issue, since it is a modifiable risk factor for cardiovascular disease, stroke, heart failure, and kidney failure [1,2]. Despite efforts to lower blood pressure, it remains as a problem due to increasing elderly population and unfavorable a1111111111 a1111111111 a1111111111 a1111111111 a1111111111 Measurements Blood pressure (BP) was measured three times from the right arm, after the participant had been seated for at least 5 minutes. If the first and second measurements differed more than 10mmHg for systolic or diastolic BP, then a third measurement was performed. We assessed the average value of the last two BP measurements. Body mass index (BMI) was calculated as the subject's weight in kilograms divided by the subject's height in meters (The Asian and Pacific perspective-World Health Organization) [15].
Risk factors. The health interview obtained details on the following: age, household income, education, occupation, alcohol consumption, and smoking history.
Alcohol intake was recorded as the frequency of consuming alcohol over the past year and more than once a week of alcohol consumption was considered as "drinking alcohol." For smoking, the participants were asked if they currently smoke or not.
In regards to grouping, age was divided into five groups: 30-39 years, 40-49 years, 50-59 years, 60-69 years, and 70+ years. Household income was split into quartiles. Education was categorized into three groups: <9 years, 9-11 years, or !12 years. Occupations were classified as white collar, blue collar, or unemployed and housewives. Alcohol intake was categorized as drinking alcohol or not. Smoking status was divided into current smokers and nonsmokers. A new categorization of BMI was made: 20 kg/m 2 or less as underweight, from 20 to <23kg/m 2 as normal, from 23 to <25kg/m 2 as overweight, and 25 kg/m 2 or more as obese.
Definitions for prevalence and control of hypertension. Self-reported hypertension prevalence, self-reported hypertension treatment and antihypertensive use was also obtained in the survey. Self-reported hypertension prevalence and self-reported hypertension treatment was obtained by asking the respondents if they were currently under hypertension and if the respondents were currently taking treatment for hypertension respectively. Antihypertensive use was asked if the respondents were taking medication now, and people who took medication pills over 20 days per month were considered as using antihypertensive. Hypertension prevalence was defined as people with an average SBP ! 140 mmHg or DBP ! 90 mmHg or whom taking medication for hypertension. Among people with hypertension, when having an average systolic BP < 140 mmHg and diastolic BP < 90mmHg is classified as controlled hypertension, regardless of medication use [16].

Statistical analysis
Multiple logistic regression models were used to identify factors associated with the prevalence and control of hypertension. Chi-square tests were conducted to compare categorical variables. All analyses were performed separately for men and women. In order to visualize sex differences according to age and BMI, a cubic spline was fitted using R software (version 3.2.5). Also, Hosmer-Lemeshow goodness of fit for logistic regression and C-statistics were evaluated. Other statistical analyses were conducted using SAS software (version 9.3; SAS Institute, Cary, NC, USA). Table 1 and Table 2 presents the characteristics of the male and female participants and compare variables among sex. Table 1 represents for all study participants, while Table 2 excluded respondents who had not hypertension until the study point: total included male participants, 4293; female participants, 5141. There were significant differences among sex for income, antihypertensive use, alcohol intake, and current smoking status. Men was more likely to be educated, have a job, and drink alcohol and less likely to receive hypertension treatment and use antihypertensive medications. Overall, 12.5% of the participants (12.2% of males and 12.7% of females) had a systolic BP > = 140mmHg, while 8.6% of the participants (12.2% of males and 5.8% of females) had a diastolic BP > = 90mmHg. Also participants self-reported hypertension, treatment and anti-hypertensive use was slightly higher among females: 92.7% (male), 94.6% (female); 88.6% (male.), 92.8% (female); 88.6% (male), 92.6% (female) each.

Results and discussion
The number of people with hypertension, unadjusted odds ratio (OR), adjusted OR, and its confidence interval for men and women is represented in Table 3. Adjusted OR was adjusted for seven variables in Table 3: age, income, education, occupation, alcohol intake, smoking, and BMI. The percentage of hypertension was 34.6% (4188 of 12,089) in men and 30.8% (4861 of 15,798) in women. For the adjusted model in men, increasing age, low income, alcohol intake, and high BMI were associated with increased odds for having hypertension, while young age, smoking, and low BMI were negatively associated with hypertension. In women, hypertension was significantly associated with old age, low education, and high BMI only. In an unadjusted model for women, hypertension was positively associated with low income and blue-collar work and negatively associated with alcohol intake and low BMI, although these did not remain significant after adjusting.
In Table 4, the number of people with controlled hypertension, unadjusted odds ratio (OR), adjusted OR, and its confidence interval for men and women is shown. Adjusted OR was adjusted for seven variables in Table 4. Among the 4188 men and 4861 women with hypertension, 1871 men (44.7%) and 2496 women (51.3%) kept their blood pressure controlled (defined as SBP <140mmHg and DBP <90mmHg). Among hypertensive men in the adjusted model, control rates were positively significant in individuals of older age and negatively associated with alcohol drinkers. In the unadjusted model for male hypertensives, aging and low education was related to increase in control rates, while negatively significant for alcohol drinking, white and blue collar jobs, and current smokers. In adjusted model, for hypertensive women, only one age group (60-69) was positively associated hypertension control rates and white collar jobs were negatively associated. However, control rates were higher in hypertensive women with high age, low education, and high BMI in the unadjusted model. Fig 1 presents the logit proportions of hypertension prevalence by sex and deciles of age and BMI with spline fits. Overall, hypertension tended to increase with increasing age and BMI, although patterns differed according to sex. Under age 60, hypertension was prevalent in men than in women; however, beyond the age of 60, hypertension in women was more common. Interestingly, the log odds of hypertension prevalence for respondents older than 60 has shown different trend among sex: female increased compared to male. Meanwhile, hypertension in men was prevalent than in women of relatively low BMI status, although there was no sex-difference in the prevalence of hypertension among individuals of higher BMI status (!25 kg/m 2 ). Fig 2 presents the logit proportions of hypertension control by sex and deciles of age and BMI with spline fits. Hypertension control rates increased according to age until around the age of 70 years in both sexes, and decreased thereafter. Among younger individuals, male hypertensives showed lower control rates than their female counterparts. Additionally, hypertension control rates tended to decrease among hypertensive men with relatively high BMI; there was no distinct relationship between BMI and hypertension control among women.
Our study examined sex difference in factors affecting prevalence and control of hypertension. Aging and high BMI were associated with prevalence in both male and female. Alcohol consumption was negatively associated with hypertension control in men. However, socioeconomic status (low income) and behavioral factors (alcohol drinking and nonsmoking) were positively associated with the presence of hypertension in men but not in women. Also, less education was strongly associated with hypertension prevalence in women, while the association was only modest in men. Kautzky-Willer A. provided evidence that the relationship between hypertension and education differ between sexes: education was more closely related to hypertension and overall health status in females than in males [17].
When combining incomplete datasets in KNHANES V and VI, we ignored the weighted value for each section. Residential area and marital status were excluded from analysis, since these variables have not been found to have a major impact on hypertension in Koreans. Since this study is a cross-sectional design, which captures a specific point in time, it was unable to reflect cause and effect relationships. Also, because hypertension was recorded after only one   checkup, white-coat hypertension and masked hypertension could not be ruled out. Another limitation is that socioeconomic factors and lifestyle behaviors might involve measurement errors, since information were collected through interviews. Lastly, mechanisms of the noted sex differences could not be found. According to our study, hypertension is common among females over the age of 60. Thus, there might be an additional variable, which is not included, affecting hypertension. Variables selected in this study analysis were selected according to guidelines reported in other studies [12,18]. Future studies accounting other potential important variables may prove to be of considerable value.
Respondents with hypertension totaled 30.4% overall (34.2% in males and 26.9% in females) in this study ( Table 2). These rates are below those in the United States, other European countries and south Asian countries [19][20][21][22][23][24]. While the rates were high compared to other east Asian countries [25]. Our study suggest age-and sex-specific strategies to prevent and control hypertension among Korean adults.

Conclusions
This study examined sex differences in hypertension prevalence and control among Korean adults. Sex disparities in hypertension status was shown: females are more likely to be hypertensive than male, after the age 60. Also, factors associated with hypertension prevalence and control differed by sex. Our findings suggest that sex specific approach is critical in improving hypertension and its' control.