Illness Beliefs in African Americans With Hypertension

Illness Beliefs and Behavior

The common sense model (CSM) of illness representations (Leventhal, Meyer, & Nerenz, 1980; Leventhal, Nerenz, & Steele, 1984) provides a framework to understand how cognitions affect behavior. The CSM, which was the guiding framework for this study, posits that individuals create mental representations of their illness to “make sense” of the experience and to manage the problem. These representations are personal beliefs that contribute to the agency for engaging in health behaviors and lifestyle change (Leventhal et al., 1980; Leventhal et al., 1984). Illness beliefs are formed based on knowledge gleaned from the sociocultural environment (e.g., media, family, friends), from authoritative sources such as health care providers, and from the personal experience of the illness (e.g., symptoms or distress associated with the disease). The CSM illness representations are characterized by five dimensions including identity, cause, timeline, consequences, and cure/controllability. Identity refers to the beliefs a person holds regarding the label that is given to the symptoms and illness being experienced. Cause reflects the person’s beliefs about the cause of the illness. Timeline refers to beliefs about the expected course of the illness such as how long the illness will last, and whether the illness is acute, chronic, or cyclical in nature. The consequences of an illness reflect beliefs about possible effects the illness will have on one’s life, including physical, social, and economic effects. The cure/controllability of an illness refers to beliefs about whether an illness can be cured or kept under control. There is little research that examines the relationship of hypertension beliefs and self-care behaviors necessary for BP control.

Purpose

The purpose of this study, therefore, was to explore the relationship between beliefs about hypertension (conceptualized as illness representations) and BP self-care in a sample of African Americans diagnosed with hypertension. The specific research questions to be addressed included the following:

Method

Measures

Participants completed the revised Illness Perception Questionnaire (IPQ-R; Moss-Morris et al., 2002), BP Self-Care Scale (Peters & Templin, 2008), and demographic data sheet, and had their BP measured by the research team.

Illness representations

The IPQ-R (Moss-Morris et al., 2002) was used to measure hypertension representations. IPQ-R contains subscales that were derived from the CSM (Leventhal et al., 1980; Leventhal et al., 1984). The subscales examined in this study include beliefs about the following: timeline (duration of illness, acute or chronic), cyclical timeline (predictability of the illness), consequences (impact of illness on one’s life, including physical, psychological, social functioning), personal control (personal ability to control illness), treatment control (treatment’s ability to control illness), and illness coherence (reasonable understanding of the illness; Moss-Morris et al., 2002). The identity representations were not measured in this study as symptoms attributed to hypertension were not examined in this study.

The IPQ-R uses 5-point Likert-type responses ranging from 1 (strongly disagree) to 5 (strongly agree) with the items. High scores on the respective subscales reflect the beliefs that hypertension is chronic (timeline), unpredictable (timeline cyclical), profoundly affects quality of life (consequence), and is able to be controlled by either personal behaviors/characteristics (personal control) or by medical treatment (treatment control). Statements on the illness coherence subscale addressed participants’ beliefs about their level of understanding of the disease. Thus, high scores to statements such as “I do not understand high BP” indicate that the participant believes he or she has a limited understanding of hypertension. Cronbach’s alpha coefficient for the subscales using samples with chronic illnesses ranged from .79 to .89. Internal consistency for this study sample ranged from .59 to .83. Subscales with low performance included personal control .63 and treatment control .59 (Table 1).

OPEN IN VIEWER

Table 1.

Hypertension Representation Scale.

					Score
Scale/subscale	Cronbach’s α	Rating scale	No. of items	Study range	M (SD)
Timeline	.77	1-5	6	1.5-5.0	3.2 (0.87)
Consequences	.66	1-5	5	1.0-5.0	3.1 (0.85)
Personal Control	.63	1-5	6	2.5-5.0	4.0 (0.65)
Treatment Control	.59	1-5	5	2.2-5.0	3.9 (0.69)
Illness Coherence	.83	1-5	5	1.0-5.0	2.5 (1.2)
Timeline Cycle	.81	1-5	4	1.0-5.0	3.2 (1.0)

The causal attribution portion of the IPQ-R was a separate scale with 18 questions that measure participants’ beliefs about the causes of their illness (e.g., stress, chance, behavior) using 5-point Likert-type scale with responses ranging from 1 (strongly disagree) to 5 (strongly agree). Factor analysis was used to group causal items into categories specific to this sample.

Self-care behaviors

The BP Self-Care scale (Peters & Templin, 2008) was used to measure participants’ frequency of engaging in the totality of self-care behaviors necessary for BP control. The self-care behaviors included diet (low fat, low salt, high fruits and vegetables), physical activity, stress management, weight control, alcohol and tobacco use, medication adherence, and doctor visits. The BP Self-Care scale is a 10-item questionnaire using 7-point Likert-type format with responses ranging from 1 (never) to 7 (always) engaging in the behavior, with higher scores indicating the practice of more self-care behaviors. Internal consistency reliability was .71 in a sample of adult African Americans (Peters & Templin, 2008). Internal consistency for this current study sample was .57.

Demographic data sheet

The demographic data sheet was used to measure demographic data, including gender, age, years of formal education, yearly income, and length of time with diagnosis of hypertension.

BP

BP was measured using the Omron HEM 907 automatic device. The reliability of the Omron measurement was established by comparing readings taken with a mercury sphygmomanometer.

Results

Hypertension Causal Attributions

Principal components analysis using varimax rotation was done to identify participants’ beliefs about the causes of their hypertension. The analysis extracted five factors, all with an eigenvalue > 1.0 and each explained more than 5% of the variance. The five factors combined, explained 68% of the variance in beliefs about the cause of hypertension (Table 2). The study sample believed that life stressors (e.g., emotional state, mental attitude, and family problems) were the primary cause of HTN, accounting for more than half of the explained variance. Following life stressors, participants attributed hypertension to factors outside of their control (e.g., germs, pollution, and chance). The etiology of hypertension that is espoused by the health care community (e.g., diet, medical care, alcohol use, and smoking) explained less than 10% of the variance in hypertension causal attributions in this sample.

OPEN IN VIEWER

Table 2.

Principal Components Analysis of Beliefs About Hypertension Causal Attributions.

	Factor 1	Factor 2	Factor 3	Factor 4	Factor 5
HTN causal factors	Stress-producing factors	Uncontrollable factors	Biomedical explanations	Risky behaviors	Heredity
Variance explained (%)	34.71	12.43	9.31	6.01	5.71
Emotional state	.795
Mental attitude	.785
Personality	.751
Family problems	.637
Overwork	.622
Aging	.567
Germ or virus		.794
Pollution		.771
Chance or bad luck		.708
Altered immunity		.584
Accident		.473
Personal behavior			.786
Diet or eating habit			.760
Stress or worry			.704
Poor medical care			.510
Smoking				.918
Alcohol use				.879
Heredity					.881

Causal attributions for hypertension were significantly different by demographic factors. Gender differences were noted in that more women attributed hypertension to stress than men, t(107) = 2.84, p = .01, while more men believed that hypertension was caused by chance, t(96) = 2.17, p = .03. More men than women also believed that risky behavior such as alcohol use, t(108) = 2.15, p = .03, and tobacco use, t(107) = 1.90, p = .06, caused hypertension. Participants with less than 12 years of formal education were more likely to believe that germs, t(31) = 3.51, p = .01, and smoking, t(107) = 2.53, p = .01, caused hypertension than participants with greater than a 12th-grade education. Causal attributions for hypertension were not significantly different by age, income, or length of time with a diagnosis of hypertension.

BP Self-Care Behaviors

Table 3 shows the mean frequency of engaging in self-care behaviors needed for BP control. Frequency is measured using responses ranging from 1 (never) to 7 (always). On average, the most frequently practiced self-care behaviors for BP control in this sample were keeping medical appointments and taking medication with over 64% of participants reporting that they do each of these behaviors always or most of the time. Forty percent of participants (n = 45) reported that they participated in at least 30 min of physical activity either most days or always. Less frequently practiced self-care behaviors were consuming a diet that is low salt, low fat, or high in fruits and vegetables, or maintaining a healthy weight as less than 20% of participants reported being able to do these behaviors on a regular basis. Twenty-five percent (n = 27) of participants reported most of the time and frequent tobacco use, and 6.5% (n = 7) reported most of time and frequent intake of greater than two alcoholic beverages each day.

OPEN IN VIEWER

Table 3.

BP Self-Care Behaviors.

	BP self-care behavior
Items	M (SD) responses on a scale 1 (never) to 7 (always)
Low-fat diet	3.62 (+1.88)
Low-salt diet	4.23 (+2.08)
Fruits and vegetables	3.60 (+2.00)
Physically active	4.66 (+2.04)
Maintain healthy weight	3.77 (+2.06)
Control stress	4.02 (+1.90)
Keeping doctor visits	5.79 (+1.77)
Medication adherence	5.80 (+1.90)
Tobacco use	2.75 (+2.54)
Alcohol use	1.92 (+1.70)

Note. BP = blood pressure.

Significantly more men than women reported practicing self-care behaviors, including being physically active, t(102) = 2.52, p = .01; maintaining a healthy weight, t(103) = 3.58, p = .01; and maintaining a low level of stress, t(106) = 2.38, p = .02. Participants with greater than 12 years of education reported eating a low-fat diet significantly more often than those with less education, t(108) = 2.42, p = .02. Older participants (>than the median of 53 years) reported being able to manage stress more effectively than younger participants, t(106) = 2.58, p = .01, and persons who had a longer history of hypertension (>5 years) reported more frequent tobacco use, t(89) = 2.46, p = .02, than participants more recently diagnosed.

Hypertension Representations and BP Self-Care Behaviors

Hypertension representations were significantly associated with BP self-care behaviors (Table 4). Believing hypertension to be a chronic illness (timeline) was associated with keeping doctor appointments (r = .20, p = .03) and taking BP medication (r = .23, p = .02). Believing hypertension to be chronic was associated with using tobacco (r = .26, p = .01). Believing that one has the ability to control own BP (personal control) was associated with not being able to maintain low levels of stress daily (r = −.20, p = .04). Having limited understanding of hypertension (illness coherence) was associated with tobacco use (r = .22, p = .02) as well as with having two or more alcoholic beverages daily (r = .34, p = .01). No significant relationships were found between hypertension representations and other self-care behaviors, including eating a low-fat, low-salt, high fruit and vegetable diet; daily physical activity; and maintaining a healthy weight.

OPEN IN VIEWER

Table 4.

Correlation of Hypertension Representations and Self-Care Behaviors Needed to Control BP.

HTN representation	Low-fat diet	Low-salt diet	Fruits and vegetables	Physically active	Control stress	Maintain healthy weight	Alcohol use	Tobacco use	Visits doctor	Medication adherence
Timeline	−.04	−.01	−.06	.08	−.03	.09	−.03	.26**	.20*	.23*
Timeline cycle	−.11	−.08	−.04	.05	−.03	.03	.19*	.17	.06	−.13
Consequence	.03	−.09	−.02	−.01	−.08	.11	.14	.15	−.04	−.14
Personal control	−.03	.18	.01	.09	−.20*	−.09	−.10	−.13	−.02	−.13
Treatment control	−.01	.07	−.05	.15	.06	.15	−.07	−.08	.03	−.18
Illness coherence	−.10	−.08	−.09	−.14	.02	.09	.34**	.22*	−.12	−.06

Note. BP = blood pressure.

*

p < .05. **p < .01, two-tailed.

Causal Attributions and BP Self-Care Behaviors

A number of significant relationships were noted between causal attributions of hypertension and self-care behaviors (Table 5). Attributing the cause of hypertension to stress-producing factors (e.g., family problems, overwork, emotional state) was inversely associated with keeping doctor appointments (r = −.28, p = .01). As might be expected, believing hypertension was caused by factors outside of one’s personal control (e.g., germs, chance, pollution) was inversely associated with consuming a low-fat (r = −.26, p = 0.01) or low-salt diet (r = −.24, p = .02). Somewhat unexpected was the finding that attributing the cause of hypertension to risky behaviors such as smoking and alcohol use was actually associated with using tobacco (r = .42, p = .01) and drinking two or more alcoholic beverages daily (r = .37, p = .01). Also, attributing the cause of hypertension to biomedically accepted factors (e.g., eating habits, poor medical care, and personal behavior) was not significantly associated with engaging in any of the self-care behaviors needed for BP control.

OPEN IN VIEWER

Table 5.

Correlation of Causal Attributions and Self-Care Behaviors Needed to Control BP.

	Low-fat diet	Low-salt diet	Fruits and vegetables	Physically active	Control stress	Maintain healthy weight	Alcohol use	Tobacco use	Visits doctor	Medication adherence
Stress-producing factors	.08	−.03	.07	−.01	−.03	.06	.06	−.13	−.28**	−.17
Uncontrollable factors	−.26*	−.24*	−.09	−.06	.27**	.04	.11	.20	−.12	.07
Biomedical explanations	−.14	−.03	−.05	.09	−.07	−.10	−.17	.03	.10	.08
Risk factors	−.16	−.01	−.13	.19	.12	.13	.37**	.43**	−.03	−.01
Heredity	.07	.10	.02	.07	−.37**	−.15	−.26*	−.13	−.04	−.06

Note. BP = blood pressure.

*

p < .05. **p < .01, two-tailed.

The illness representations were examined as predictors of BP self-care behaviors. Multiple regression analysis using stepwise entry of the seven illness representation subscales (timeline, consequences, personal control, treatment control, illness coherence, timeline cycle, and causal attributions) were used to predict the totality of BP self-care behaviors needed for BP control (i.e., using the mean composite score). Results of the regression analysis showed that illness coherence was the only illness representation that was a statistically significant predictor of BP self-care behaviors (F = 7.10, p < .01), accounting for 7% of the variance. This result indicates that those who had reasonable understanding of their hypertension were more likely to engage in BP self-care behaviors.

Discussion

The purpose of this study was to examine the relationship of beliefs about hypertension (illness representations) and self-care behaviors for BP control. This study was the first to examine the relationship of hypertension representations and the totality of lifestyle behaviors for BP control in African Americans. Results from this study illustrate the importance of understanding patients’ hypertension representations and behavioral responses to hypertension and highlights points of intervention that may improve patient outcomes. The key findings from this study include the following: (a) The primary causal attribution of hypertension was stress-producing factors, (b) attributing hypertension to stress-producing factors was negatively associated with keeping doctor appointments, and (c) believing that hypertension is a chronic condition was positively associated with self-care behaviors, including keeping doctor appointments, and medication adherence. These findings are important given the role of self-care behaviors in the management and control of hypertension.

African Americans in this study did not strongly endorse biomedical explanations as the cause of hypertension but instead endorsed stress-producing factors (e.g., family problems, overwork, emotional state) as the predominate cause of hypertension. This finding is consistent with other studies investigating hypertension beliefs with African American samples (Lukoschek, 2003; Webb & Gonzalez, 2006; Wilson et al., 2002). Belief in stress-producing factors as causes of hypertension is inconsistent with current medical understanding of hypertension, and in this study sample, was associated with less engagement in behaviors known to reduce or control BP. Attributing hypertension to stress-producing factors was significantly associated with missing doctor appointments in this sample of hypertensive African Americans. This finding is supported by Hekler and colleagues (2008) who found that African Americans who believed their hypertension was caused by life stressors did not participate in lifestyle modification behaviors for BP reduction. Another study found that hypertensive African Americans believed that stress-producing factors such as dealing with perceived or real racism, work pressures, and multiple role responsibilities were the major cause of their hypertension. African Americans believe that their hypertension should be treated with stress-reducing behaviors such as relaxation, solitude, spirituality, and exercise (Kronish, Leventhal, & Horowitz, 2012; Webb & Gonzalez, 2006). Similarly, another sample of hypertensive African Americans believed that their BP medication was treating their ability to handle stress rather than their actual BP (Lukoschek, 2003). Combined, these study findings underscore the difference between patient and provider explanations about the etiology of hypertension and suggest that health care providers who address stress within their treatment strategy may potentially increase the probability of achieving better BP control in African Americans.

Only one hypertension representation was associated with behaviors that are endorsed by the medical community as necessary for BP control. Believing that hypertension is a chronic condition was associated with keeping doctor appointments and taking antihypertensive medication. These two behaviors also were reported as the most frequently practiced behaviors for BP control in this sample. These findings are not surprising given that two thirds of the sample had been living with hypertension for more than 5 years and would have some understanding about the chronic nature of the disease. Finding that illness coherence was the only significant predictor of self-care behaviors in this study may indicate that this sample has a degree of health literacy concerning hypertension. Health literacy, being broader than reasonable understanding of the illness (illness coherence), is defined by the National Library of Medicine as the degree to which individuals have the capacity to obtain, process, and understand basic health information and services needed to make appropriate health decisions (Ratzan & Parker, 2000). These services may be obtained from the participant’s medical doctor; thus, keeping appointments would be a very important component of health literacy. In addition, Okonofua, Cutler, Lackland, and Eagan (2005) found that hypertensive African Americans were more likely to perceive medication as the only way to control BP compared with Caucasians who tended to believe lifestyle modification behaviors were beneficial. Okonofua and colleagues (2005) further suggested that hypertensive African Americans may be less aware of the benefits of lifestyle modification behaviors for their BP and may consider providers who emphasize lifestyle modification behaviors to be less concerned with their BP control. However, given that on average, our sample was fairly well educated, a lack of awareness of the benefits of lifestyle change is unlikely. Yet, other factors such as limited access to stores that have fresh fruits and vegetables and limited safe areas for regular physical exercise may be more relevant reasons for our sample’s limited practice of these behaviors (Larson, Story, & Nelson, 2009; Lee, Mama, Medina, Ho, & Adamus, 2012; Lopez & Hynes, 2006; Robert & Reither, 2004). It is plausible that acceptance of such limitations may serve to reinforce the belief that the totality of lifestyle modification behaviors add little value to BP control.

HTN illness perceptions have been examined in other populations. Studies from the United Kingdom (Ross, Walker, & MacLeod, 2004) and Taiwan (Chen, Tsai, & Lee, 2009) examined illness perceptions and their relationship to medication adherence and adherence to self-care management in participants with hypertension. Unlike the African American participants in our study, the participants from the United Kingdom and Taiwan believed that traditional risk factors for hypertension along with psychological factors were primary causes of their hypertension (Chen et al., 2009; Ross et al., 2004). Moreover, unlike our study findings, personal effort (personal control) was a significant predictor of adherence to self-care management behaviors (Chen et al., 2009) and believing in medical treatment (treatment control) was associated with medication adherence (Ross et al., 2004). Similar to our study findings, the Taiwanese participants also reported a significant association between believing that hypertension was a chronic illness (timeline) and adherence to self-care management behaviors (Chen et al., 2009). Findings from HTN illness perception studies suggest the need for health care providers have an understanding of patients’ hypertension beliefs as they may have a significant impact on self-care behaviors leading to BP control.

As with all studies, there were limitations in the current study that should be considered when reviewing our results. The study design was cross-sectional which limits the ability to show causal relationships. Theoretically, the CSM purports that illness representations influence behavior. However, the reverse may also be valid as behavior may influence beliefs. Longitudinal studies are needed to clarify the direction of hypertension beliefs and behaviors and also to examine how illness representations evolve and change over time. Second, our sample was a relatively small nonprobability sample of predominantly low-income African Americans that limits the generalization of our findings to other groups. However, a number of our key findings were consistent with other studies that build knowledge regarding the hypertension beliefs of African Americans as well as demonstrating the association of these beliefs on the self-care behaviors necessary for BP control. Our data were self-reports, so there was no way to verify the frequency of engaging in self-care behaviors. However, self-report data are consistently used to examine behavior in research as well as relied upon in clinical practice. Future research may consider an objective method of assessing actual self-care behaviors. In addition, some of the significant relationships found between variables that were not intuitive, such as the association of hypertension chronicity with tobacco use, may be a spurious relationship that needs more investigation in future studies. The relationships between hypertension beliefs and risky behaviors such as tobacco and alcohol use also warrant further investigation.

Another limitation of this study was low internal consistencies on several subscales on the IPQ-R (Moss-Morris et al., 2002) and the BP Self-Care Behavior scale (Peters & Templin, 2008). Low internal consistencies on IPQ-R subscales including personal control (.63), treatment control (.59) and consequences (.66). The control items have had low internal consistencies when used to assess other chronic illnesses (Moss-Morris et al., 2002). The authors of the IPQ-R (Moss-Morris et al., 2002) suggested that more work is needed to unravel the control items. In addition, more consequence items may be needed to improve the consequence subscale internal consistency. Although the BP Self-Care Behavior scale (Peters & Templin, 2008) had low internal consistency, this, in part, may be explained by problems with the wording of two items related to tobacco and alcohol use. In addition, the scale addressed the totality of behaviors needed for BP control that indicates items were measuring different aspects of BP control which reduces internal consistency. More research using the IPQ-R (Moss-Morris et al., 2002) and BP Self-Care Behavior Scale (Peters & Templin, 2008) needs to be done in a similar population to further assess these subscales. Finally, due to the exploratory nature of our data analysis, no corrections were made for multiple comparisons for the HTN beliefs and individual self-care behavior scores.

Despite these limitations, results from this study have important clinical implications. Findings from this study suggest that health care providers should inquire about hypertension beliefs from their hypertensive patients, particularly those whose BP is uncontrolled. Understanding the patient’s beliefs may provide important information regarding barriers to engaging in behaviors leading to BP control. Our findings also suggest that health care providers should inquire about life stressors facing patients that may interfere with them engaging in needed BP self-care. Finally, findings from this study suggest that self-care behaviors may be better supported by hypertensive African Americans if the behaviors are put forth as stress relieving activities along with BP control.

This study found that a sample of community-dwelling African Americans with hypertension did not strongly endorse biomedical explanations of hypertension, but more strongly endorsed life stressors and factors outside of their personal control as major causes of their hypertension. These beliefs had a significant impact on their engaging in self-care behaviors necessary to achieve BP control. These findings suggest that health care providers should carefully assess patients’ beliefs about hypertension to more effectively treat and control BP.