Factors affecting self-care among hypertensive older adults dwelling in the community: A cross-sectional study


 Background: Hypertension is a prevalent health problem in older adults, with better outcomes expected through proper self-care. However, little is known about the effects of cognitive function level on self-care in older adults living in the community. Methods: This cross-sectional study, conducted from October 2019 to January 2020, analyzed the effect of cognitive function on self-care in elderly individuals aged > 65 years with hypertension who visited a local general hospital for the treatment of hypertension. The Korean versions of the Mini-Mental State Examination (K-MMSE) and Montreal Cognitive Assessment (MoCA-K) were used to assess cognitive function. The Hypertension Self-Care Behavior Scale (HBP-SC Behavior Scale) was used to analyze the subjects’ self-care, which was divided into diet behavior and health behavior (except diet). The general characteristics and degrees of self-care of the subjects were analyzed using descriptive statistics, and multiple regression analysis was used to analyze the factors affecting self-care. Results: Factors influencing HBP-SC diet behavior scores were religion (β =.27, SE = 0.69, p =.007) and MoCA-K scores (β =.31, SE = 0.08, p = .002). HBP-SC health behavior (except diet) scores were associated with comorbidities (β = −.20, SE = 0.60, p = .032), and the power of the model was 20%. However, there were no variables that significantly affected the total HBP-SC score, which included the diet behavior and health behavior (except diet) scores. Conclusions: Although there was no significant factor influencing the total HBP-SC score, religion, MoCA-K scores, and comorbidities were factors influencing diet behavior and health behavior (except diet). Therefore, tailored education takes into account religion, MoCA-K domains, and comorbidities is necessary to promote self-care in hypertensive older adults.

medication compliance, and stress management. Moreover, it is necessary for them to continuously engage in self-care behaviors [15,16]. Indeed, some previous studies have reported that self-care behaviors, including diet management, were adopted well by elderly patients with hypertension [17,18]. However, it is necessary to identify the factors essential for promoting self-care behavior in elderly patients with hypertension.
Factors affecting self-care behavior in patients with hypertension, as found in previous studies, include familial support, age, sex, economic activity, comorbidities, exercise, drinking, smoking, obesity, living with family, number of children, education level, and duration of hypertension [19][20][21].
However, studies elucidating self-care behaviors based on cognitive function in elderly patients with hypertension are scarce. This study investigated factors affecting self-care behavior in communitydwelling older adults with hypertension based on their cognitive function.

Study design
This cross-sectional study aimed to investigate the level of self-care behaviors and factors affecting self-care behaviors based on cognitive function in community-dwelling elderly patients with hypertension who were aged 65 years or older.

Setting
The participants in this study were recruited from S General Hospital, with an average of 800 outpatients per day located in S City, South Korea.

Participants
The participants in this study were older adults aged 65 years or older who visited the outpatient clinic at S General Hospital, received treatment for hypertension, and agreed to participate in this study after the purpose of this study was explained. The data collection period was from October 2019 to May 2020 and was conducted after explaining the purpose of this study to the hospital. Permission was obtained for data collection. After receiving participant written informed consent, the survey was conducted in a counseling room next to the internal medicine outpatient clinic. The survey was conducted one-on-one by the study researcher through a survey questionnaire that examined the participants' characteristics, cognitive function, and hypertension-related self-care behavior.

Inclusion criteria
-Older adults aged 65 years or older who were able to communicate verbally and nonverbally -Patients who had a systolic blood pressure ≥ 140 mmHg or a diastolic blood pressure ≥ 90 mmHg, had been diagnosed with hypertension by an internist, and had been taking antihypertensive drugs for more than 6 months.

Exclusion criteria
-Those with secondary hypertension were not included in the study.

Characteristics of the participants
The participant characteristics that were collected comprised a total of 12 items. Specifically, the general characteristics section included seven items regarding age, sex, education level, religion, marital status, living with family and primary caregiver, while the hypertension-related characteristics section consisted of four items regarding the duration of hypertension, the number of hypertensive drugs used, comorbidities, and body mass index (BMI).

Cognitive function
Cognitive function in older adults was measured using the Korean Mini-Mental State Examination (K-MMSE), which is useful in screening for dementia, and the Korean version of the Montreal Cognitive Assessment (MoCA-K), which is useful in screening for mild cognitive impairment.
The K-MMSE is a tool for screening dementia that was modified and translated into Korean by Kang et al. (1997) [22] from the Mini-Mental State Examination (MMSE) developed by Folstein et al. (1975) [23]. The K-MMSE consists of orientation to time (5 points), orientation to place (5 points), memory registration (3 points), attention and calculation (5 points), memory recall (3 points), language (8 points), and drawing (1 point). A score of 24 points or more (out of 30 points) indicates normal cognition, a score of ≥18 points to ≤ 23 points indicates mild cognitive impairment, and a score of ≤ 17 points indicates dementia. The reliability of the K-MMSE showed a Cronbach's α = .85 in a recent study [24], and its reliability revealed a value of Cronbach's α = .74 in this study.
The MoCA-K is a tool for screening mild cognitive impairment that was modified and translated into Korean to be suitable for Korean culture and linguistic characteristics by Kang et al. (2009) [25] from the Montreal Cognitive Assessment (MoCA) developed by Nasreddine et al. (2005) [26]. This tool consists of visuospatial ability/executive function (5 points), language abilities (3 points), memory/delayed recall (5 points), attention (6 points), abstraction (2 points), and orientation (6 points), with a total score of 30 points. One point is added to an individual's score if he/she has six years or less of formal education. A score of 23 points or higher was considered normal, and a score of 22 points or less was considered mild cognitive impairment. The reliability of the MoCA-K was Cronbach's α = .84 at the time of its translation and modification, and its reliability was Cronbach's α =.79 in this study.

HBP-SC behavior scale (hypertension self-care)
The Hypertension Self-Care Behavior Scale (HBP-SC Behavior Scale) was translated and modified for Korean older adults by An et al. (2017) [27] based on the Hypertension Self-Care Profile developed by Han et al. (2014) [28]. This tool consists of a total of 20 items, including 11 items on diet that assessed the subjects' habits regarding checking nutrition labels, sodium intake, total calories, and alcohol consumption, as well as nine except diet items regarding exercise, smoking cessation, antihypertensive medication compliance, and regular health screening. Each item is rated on a 4-point Likert scale, with scores ranging from a minimum score of 20 points to a maximum score of 80 points.
Higher scores indicate higher self-care behavior.
The reliability of the HBP-SC behavior scale had a Cronbach's α = .92 for HBP-SC behavior (total), a Cronbach's α = .91 for HBP-SC diet behavior, and a Cronbach's α = .85 for HBP-SC health behavior (except diet) in a study by An et al. (2017) [27]. The Cronbach's α value was = .73 for HBP-SC behavior (total), α = .77 for HBP-SC diet behavior, and α =.65 for HBP-SC health behavior (except diet) in this study.

Sample size
The number of participants required for this study was calculated to be 103 after performing a multiple regression analysis with a significance level of .05, a power of .80, an effect size of .15, and using 7 predictors. Considering a dropout rate of 10%, 110 participants were selected. Of them, five participants had incomplete information, and 105 participants were finally included in this study.

Statistical methods
The data collected in this study were analyzed using IBM SPSS Statistics for Windows version 22.0.
The reliability of the HBP-SC behavior scale, K-MMSE, and MoCA-K were analyzed using Cronbach's α values. Descriptive statistical analysis was used to investigate the general characteristics, hypertension-related characteristics, and cognitive function of subjects. The differences in their selfcare behaviors based on their characteristics and cognitive function were analyzed using a t-test and an analysis of variance. A post hoc test was performed using a Scheffé post hoc test. The correlations between the participants' characteristics and hypertension-related self-care behaviors were analyzed using Pearson's correlation coefficients. Multiple regression analysis was used to analyze the factors affecting the subjects' hypertension self-care behaviors.

Ethical considerations
This study was approved by the Institutional Review Board at K University located in D City, South Korea (40525-201906-HR-029-02), for the protection of the participants. After the purpose, procedure, and time required for this study were explained to the participants, they provided written informed consent to participate in this study. The participants were informed that the collected data would be used only in this study and not for any other purpose. They were also informed in advance that to protect their rights as participants, their anonymity would be guaranteed, that they could refuse to participate in this study at any time and that if they decided to withdraw their participation from this study or failed to complete the questionnaire during the survey, then they would not be disadvantaged and could do so without prejudice. All methods were carried out in accordance with relevant guidelines and regulations under Ethics approval and consent to participate

Results
Of the participants, 46 (43.8%) were men and 59 (56.2%) were women. In terms of age, the number of those in their 70s was the highest at 52 (49.5%), and the mean age was 76.17±6.54 years. In terms of education level, 78 (74.3%) had received no or primary education, and 27 (25.7%) had received more than secondary education. In terms of religion, 65 (61.9%) had religious affiliations, whereas 40 (31.8%) did not. In terms of marital status, 60 (57.1%) were married, 3 (2.9%) were separated, and 42 (40.0%) were widowed. In addition, 69 (65.7%) lived with their families. In terms of primary caregiver type, 36 (34.3%) and 46 (43.8%) reported their spouses and themselves, respectively, as their primary caregivers.
In terms of K-MMSE scores, 43 subjects (41.0%) were normal, 56 (53.3%) had mild cognitive impairment, 6 (5.7%) had dementia, and the mean K-MMSE score was 22.83±4.06 points. MoCA-K scores were obtained for 99 participants, which excluded the six subjects with dementia as identified by the K-MMSE. The results showed that 31 (29.5%) were normal and 68 (64.8%) had mild cognitive impairment; the mean MoCA-K score was 19.82±6.32 points (Table 1).   p=.001), married persons (F=3.81, p=.025), and those living with family (t=3.16, p=.002). In terms of comorbidities, the HBP-SC health behavior (except diet) score was higher in those with cerebrovascular disease (F=2.72, p=.008). In terms of cognitive function comparisons, the HBP-SC health behavior scores were higher in those with dementia based on the K-MMSE (F=3.57, p=.032) and in those with normal cognitive function on the MoCA-K (F= 3.25, p=.043).

This study identified factors affecting HBP-SC behavior based on cognitive function in older adults
with hypertension, analyzed overall HBP-SC behavior, and stratified these behaviors into HBP-SC diet behavior and HBP-SC health behavior (except diet) categories. The factors affecting HBP-SC diet behavior in elderly adult hypertensive patients were religion and MoCA-K performance, and that affecting HBP-SC health behavior (except diet) was having a comorbidity.
Religion was a factor affecting HBP-SC diet behavior, similar to a study that reported that older adults who frequently attend religious events were more likely to practice health-promoting behaviors [29]. This study revealed that lower MoCA-K scores were associated with lower HBP-SC diet behavior scores, which is comparable to the results of a study by Harkness et al. (2014) [30] showing that lower MoCA scores indicated poorer self-care behavior in older adults with cardiovascular disease.
However, the aforementioned study did not compare self-care behavior after evaluating cognitive function using the MoCA in patients, unlike in the present study, and thus, it is difficult to make a direct comparison between the two.
With respect to having a comorbidity as a factor affecting HBP-SC health behavior (except diet), the results of a previous study reported that the presence of comorbidities in old adult patients with heart failure and cognitive impairment negatively affected self-care behavior, which was similar to the results of this study, but the present study differed in terms of the participants by including only older adults with hypertension [31]. Nevertheless, when comparing the aforementioned results, it is thought that the presence of comorbidities (including hypertension) in older adults is a factor affecting selfcare behavior.
In this study, the K-MMSE and MoCA-K were used to assess HBP-SC behavior in relation to the participants' cognitive function. The results showed that the mean K-MMSE score in the participants was 22.83 points, indicating that the participants in this study had mild cognitive impairment on average. In terms of self-care behavior based on K-MMSE scores, lower K-MMSE scores were associated with lower levels of self-care behavior assessed with both HBP-SC health behavior (except diet) and HBP-SC behavior (total). There are a limited number of previous studies that evaluated cognitive function using the MMSE or MoCA and then compared HBP-SC behaviors based on cognitive function status. However, a study involving patients with heart failure reported that lower MMSE scores were associated with lower medication adherence [32], and another study involving patients undergoing hemodialysis reported that lower MMSE scores were associated with lower levels of self-care behavior [33], suggesting that older adults with chronic diseases primarily need to have their cognitive function assessed and then have their self-care behavior analyzed based on cognitive function status. In addition, the mean MoCA-K score was 19.82 points, indicating that the participants had mild cognitive impairment on average, which was similar to the K-MMSE results. In this study, those who were judged to have normal cognitive function from the K-MMSE results but were at the boundary of mild cognitive impairment were identified as having mild cognitive impairment through the secondary measurement, the MoCA-K, which is a distinctive feature of this study.
The results of the item analysis of the HBP-SC diet behavior domain showed that the score for the item 'Read nutrition labels to check information on sodium content when purchasing food' was the lowest, indicating that it is necessary to enhance awareness among Korean older adults to check nutrition labels when purchasing food. However, the results of a study by Han et al. (2014) [28] found that the score for the item 'Read the nutrition labels when purchasing food' was the highest, indicating a difference in this item between Korean and American older adults. Considering a study showing that the use of nutrition labels developed in an easy format for older adults had positive effects on label knowledge, usage, and attitudes [34], it is necessary to conduct various studies regarding nutrition labeling for older adults with hypertension in South Korea.
The results of the item analysis of the HBP-SC health behavior (except diet) domain revealed that the score for the item 'Check my blood pressure at home' was the lowest. In a study by Buis et al.
(2020) [35], continuous blood pressure monitoring at home with data entry into a mobile machine and the provision of personalized interventions were effective in controlling blood pressure levels. Thus, a national health care approach that provides home blood pressure monitors can be explored.
With respect to MoCA-K scores, which affected contextual HBP-SC diet behavior, lower MoCA-K scores were found to be associated with lower levels of HBP-SC diet behavior, and a tablet PC using IoT can be installed in the kitchen to help those with impaired cognitive function prepare meals [36].
In addition, self-care behavior was found to be low in older adults with hypertension and comorbidities.
Previous studies have reported that self-care behavior was poorer in hypertensive patients with cardiovascular disease or diabetes with more severe comorbid states [31,37]. In light of these findings, comorbidities in older adults with hypertension should be taken into consideration when attempting to improve self-care behavior [38] In the multiple regression analysis, significant influencing variables were not found, but variables that were found to have a difference in hypertension self-care behavior included education level, marital status, living with family, and the K-MMSE score. In other words, a higher education level and married persons living with family were associated with higher self-care behavior, and lower K-MMSE scores were associated with poorer levels of self-care behavior [13,39,40]. Therefore, replicating this study and including a large number of participants to examine variables such as education level, marital status, living with family, and K-MMSE score will be meaningful.
The significance of this study is as follows. First, cognitive function was assessed with both the K-MMSE and MoCA-K, which are validated tools for the assessment of older adults with hypertension, and the factors affecting hypertension self-care behavior based on cognitive function status were analyzed. Those who were found to have normal cognitive function using the K-MMSE were identified as having mild cognitive impairment using the MoCA-K. Then, these participants were classified as having mild cognitive impairment to analyze factors affecting their hypertension self-care behaviors. Previous studies have shown that hypertension in older adults not only causes complications but also reduces cognitive function [40] and that those with mild cognitive impairment have higher rates of progression to dementia than those with non-mild cognitive impairment [41][42][43][44].
Therefore, the management of hypertension in older adults is related to chronic disease management and deterioration of cognitive function in normal older adults, indicating that cognitive function in older adults is an important factor for improving self-care behavior in older adults with hypertension.
Therefore, tailored intervention programs that carefully consider cognitive function status in older adults are needed for the management of hypertension in older adults.
Second, HBP-SC behavior was divided into two factors: HBP-SC diet behavior and HBP-SC health behavior (except diet). This was based on the results of a study [27] regarding the validity of the Korean version of the HBP-SC behavior scale [27], in which the Korean version was validated by dividing it into two categories of self-care diet behavior and self-care behavior except diet. In the multiple regression analysis, no significant variable was identified for the HBP-SC behavior total score. However, when it was divided and analyzed into HBP-SC diet behavior and HBP-SC health behavior (except diet), the results derived significant variables such as religion and MoCA score in the HBP-SC diet behavior category and having a comorbidity in the HBP-SC health behavior except diet category. In the sense that individuals' active religious activities indicate socialization, they may affect their hypertension diet management, with those with higher MoCA scores or better cognitive function able to pay greater attention to their diet management. The presence of comorbidities in the HBP-SC health behavior (except diet) domain was found to be a factor affecting self-care because those with more chronic diseases face greater difficulties in practicing healthcare behaviors [39].
This study had the following limitations. First, the Cronbach's α value of the HBP-SC behavior scale used in this study was lower than that at the time of the development of the original tool [27] and lower than that observed in another study that employed the same scale (Kim, 2019) [45]. At the time of development of the original scale, the study participants were generally older adults aged 65 or older, and their cognitive function statuses could not be identified as they were not measured. Another study (Kim, 2019) [45] that involved older adults with normal cognitive function included participants with a K-MMSE score of 25 points or higher, which was higher than that of the participants in the present study, who obtained a mean K-MMSE score of 22.83 points. Therefore, the differences in Cronbach's α values may be due to the differences in the cognitive function of the participants between the two studies.
Second, the levels of self-care behavior in those diagnosed with dementia based on the K-MMSE results were found to be higher, which may be because the participants completed the questionnaire together with their respective primary caregivers and usually received direct help from their caregivers in treating their hypertension. In addition, future studies examining depression, self-efficacy, and social support, which have been found to be variables affecting self-care behavior, are needed. Because this study collected data randomly from hypertensive older adults who visited a single general hospital located in S City, it is difficult to generalize the results of this study.

Conclusions
Factors affecting self-care behavior in older adults with hypertension based on cognitive function included religion, MoCA-K scores, and comorbidities. Having a religion was associated with higher levels of self-care, and a lower MoCA-K score was associated with poorer HBP-SC diet behavior.
HBP-SC health behaviors (except diet) were poorer in those with comorbidities. Therefore, to improve self-care behavior in older adults with hypertension, it is necessary to consider their religious practices and comorbidities, assess their cognitive function, and provide tailored education to them.