Whose job is it? Gender differences in perceived role in heart failure self-care


  • Victoria Vaughan Dickson PhD, CRNP,

  • Linda Worrall-Carter PhD, RN,

  • Lisa Kuhn MHSc(Nur), RN,

  • Barbara Riegel DNSc, RN, FAAN

Victoria Vaughan Dickson
College of Nursing
New York University
726 Broadway, 10th Floor
New York, NY 10003, USA
Telephone: +212 992 9426
E-mail: vdickson@nyu.edu


dickson vv, worrall-carter l, kuhn l & riegel b (2011) Journal of Nursing and Healthcare of Chronic Illness3, 99–108
Whose job is it? Gender differences in perceived role in heart failure self-care

Aim.  To describe gender differences in perceived role in heart failure (HF) self-care and to explore how an individual’s perception of their role influences self-care behaviours. Specifically, two hypotheses were tested: (1) there are gender-specific differences in perceived self-care roles in HF and (2) there are differences in self-care secondary to a patient’s perceived role.

Background.  Gender differences in roles are ubiquitous in all societies. Rarely have these roles been examined as they contribute to performance of self-care in adults with chronic illnesses.

Methods.  Secondary analysis of three mixed methods studies (n = 99) of adults with chronic heart failure. Data were collected between 2006–2008.

Conclusions.  Two dominant perceived roles in self-care were identified: (1) active and (2) passive. These were further categorised according to the degree of independence described by participants in self-care decision making: (1) primary responsibility (27%), (2) collaboration (22%) and (3) reliant upon direction from others (51%).

Relevance to clinical practice.  Clinicians are encouraged to assess the individual’s perceived role in HF self-care as part of the self-care education process. Understanding patient perceptions of their role may help guide education, which may be particularly useful for those patients most likely to defer to others for HF management advice.


The incidence of heart failure (HF) continues to rise worldwide, particularly in developed nations (Schocken et al. 2008). Almost 6 million Americans currently have HF, with an incidence approaching 10 per 1000 population in persons older than 65 years of age (Lloyd-Jones et al. 2010). Men have higher incidence of HF, yet the prevalence is approximately equal because of women’s increased longevity (Stromberg & Martensson 2003). Despite advances in diagnosis, treatment, mortality and morbidity (Rosengren & Hauptman 2008), women and elders with HF still have poorer health outcomes and higher mortality compared with their male and younger counterparts (Jessup & Pina 2004).

Numerous factors contribute to poor outcomes in women with HF. Risk factors for HF are different in women and men. Men are more likely to develop systolic HF due to ischaemic heart disease (Martensson et al. 1998). Conversely, women are more likely to have diastolic HF secondary to hypertension and diabetes (Martensson et al. 1998). Diastolic HF is thought to be associated with fewer symptoms (Deswal & Bozkurt 2006). Several researchers have found patients with HF who are less symptomatic have poorer self-care (Rockwell & Riegel 2001, Lee et al. 2009a) and may only seek assistance when symptoms have fully escalated (Lehnbom et al. 2009).

In a prior study of HF self-care, we found distinct gender differences in decisions made both in interpreting and responding to symptoms (Riegel et al. 2009a). The men were better than women at interpreting their symptoms as being related to HF and in initiating treatment. These differences were associated with differences in self-care confidence, social support and mood. Building on these results in the current study, we sought to describe gender differences in perceived role in HF self-care and to explore how an individual’s perceived role influences their self-care behaviours. Specifically, we postulated that gender differences in HF self-care may be related to societal roles, which could influence how HF is experienced and the behaviours in which one is willing to engage.


Self-care in heart failure

Self-care is a model for illness control employed in numerous chronic diseases (Bayliss et al. 2003). Self-care behaviours are activities initiated and performed by patients on their own behalf (Artinian et al. 2002). The theory of self-care we use is based on naturalistic decision-making, which explains that in the real world setting, people make decisions that are individually meaningful and familiar, using their experience and the empirical information available to them at the time (Lipshitz et al. 2001). These real life decisions are influenced by the interaction between the individual, the problem and their current setting or environment. According to this framework, a decision made in a particular situation by one individual will differ from that made by another person because of differences in experience, knowledge, skill, values and the likely outcome.

While self-care is believed to promote positive outcomes in patients with HF (Jovicic et al. 2006, Lee et al. 2007, 2010), overall, self-care is poor (Heo et al. 2008). Carlson et al. (2001) argue that this is not surprising because self-care is difficult in the context of subtle early symptoms and complex therapeutic regimens. Some patients with HF are more capable of undertaking self-care than others and this may be related to factors affecting their knowledge and confidence, such as past experience of managing symptoms (Dickson & Riegel 2009). Heo et al. (2008) found evidence of low self-care for both sexes, but higher levels of self-care in patients who were confident and had greater perceived control of their condition.

Gender differences

Gender differences in self-care dominate much of the HF self-care literature. It is a commonly held view that women tend to be better at self-care than men (Senay & Waters 2004). Women, for instance, adhere better to salt reduction diets for HF self-care than men (Chung et al. 2006). Yet many studies have failed to identify significant gender differences in HF self-care (Artinian et al. 2002, Jurgens 2006, Holst et al. 2007, Heo et al. 2008, Lee et al. 2009b, Riegel et al. 2010).

Gender roles

Gender differences in roles are ubiquitous throughout society. A role is a set of connected behaviours and obligations conceptualised by an individual as appropriate in a specific social situation (Biddle 1986). Roles consist of behavioural rules that can be applied to a variety of situations typically encountered in that role. Role behaviour is usually consistent with social norms and values. Perceived roles emerge out of interactions with other individuals or groups.

Few studies have examined perceived role in self-care of chronic illness and roles have not previously been examined in HF self-care. Sarkadi et al. (2005) evaluated the influence of perceived role on metabolic outcomes after a diabetes intervention and found a significant reduction of HbA1c after 24 months (p < 0·05) among those who perceived they had an active role in their diabetes management. Perceiving their role in diabetes management as ‘active’ was associated with better patient outcomes compared to them adopting a ‘passive’ role. Similarly, Paterson and Thorne (2000) in their grounded theory on developing expertise in diabetes self-care, describe how the process of becoming an expert includes taking an active role in self-management.

In addition to differing social roles for men and women with HF, perceptions of HF, its limitations and challenges vary according to patient gender (Stromberg & Martensson 2003). For this reason, we suspected there may be differences in the perceived role in HF self-care between men and women. Therefore, the purpose of this mixed methods study was to describe gender differences in perceived role in HF self-care and to explore how an individual’s perceived role influences their self-care behaviours. If gender differences in perceived role are found to exist, they may help explain some of the challenges that patients with HF have in performing self-care.


This study was a secondary analysis of three of our mixed methods studies (summarised in Table 1). Data were collected between 2006–2008. Qualitative meta-analytic methods (Noblit & Hare 1988) were used to analyse qualitative data from in-depth accounts of HF self-care and perceived role. Descriptive meta-analysis is a qualitative technique used to synthesise qualitative research findings and ultimately draw upon a richer data set in order to answer new questions. The process of qualitative meta-analysis is a rigorous process of creating ‘interpretations of interpretations’ (Mccormick et al. 2003, p. 938).

Table 1.   Summary of studies used in analysis
StudyOverview of design and sampleKey results
Factors associated with the development of expertise in heart failure self-care (Riegel et al. 2007)Design: Mixed methods
Purpose: To describe how expertise in self-care develops
Sample: N = 29
60% male, 60% Caucasian
Mean age 63·37 (±13·12) years;
Only 10·3% had adequate self-care practices (SCHFI >70)
Patients with poor HF self-care had worse cognition, more sleepiness were more depressed and had poorer family functioning
A typology of heart failure self-care management in non-elders (Dickson et al. 2008)Design: Mixed methods
Purpose: To examine the contribution of biobehavioural variables to HF self-care
Sample: N = 41
63·4% male, 68·3% Caucasian
Mean age 49·17 (±10·51) years; 58·5% NYHA III
A self-care typology was constructed from the integrated data: experts, novices and inconsistent. Variance in attitudes, self-efficacy and cognition contributed to significant differences (p = 0·001) in self-care practices among types
Symptom recognition among elders with heart failure (Riegel et al. 2009a, b, c)Design: Mixed methods
Purpose: To explore age-related differences in the ability to perceive symptoms of heart failure
Sample: N = 29
72·4% Male,
Mean age 68·69 (±14·40); 58·6% NYHA II
Older patients had more difficulty in detecting and interpreting shortness of breath than younger patients

In this study, qualitative meta-analysis was completed first by the principal investigator blinded to the quantitative results. Then the quantitative data was analysed using descriptive statistics, Student’s t-test and correlational analysis. The qualitative and quantitative data were integrated in the final analysis and interpretation stage of the project. Findings from the qualitative meta-analysis were used to generate hypotheses for testing. Specifically, we tested the hypotheses that there are differences in perceived roles related to gender and differences in self-care according to perceived role.

Qualitative data

In each of the three studies, a series of similar open-ended questions were asked of participants (e.g. ‘Tell me about your heart failure’ and ‘What do you do on a daily basis to take care of your heart failure?’) to elicit in-depth accounts of self-care. To gain insight into how self-care was practiced, we had asked open-ended questions about who was involved in helping the participants and what they perceived their role in self-care to be. The qualitative data from these three studies yielded a rich description (more than 2000 pages of transcription) of self-care practices.

Following the qualitative meta-analytic steps described by Noblit and Hare (1988), we first reviewed the individual study findings of the three studies at the level of codes, focusing on the categories and themes related to the self-care role described by each individual. Next common themes about perceived roles were examined across the three studies. This within-study and across-study analysis was an iterative process involving a reanalysis of original raw data and reinterpretation of previous findings to answer additional questions about perceived role in self-care.

For the purpose of this study, comments about the individual’s role and the role of others (e.g. family, friends, healthcare providers) in self-care maintenance (e.g. diet adherence, daily weight monitoring, exercising) and self-care management (e.g. decision-making in response to changed symptoms) were re-examined and coded as reports of perceived role in self-care. The pattern of themes related to perceived role in self-care across studies was then reviewed to identify common characteristics about the emerging perceived roles. The qualitative evidence of perceived role was coded as categorical data for use in the quantitative analysis.

Methodological rigour of the qualitative analysis was maintained through an audit trail, periodic debriefing with the co-investigators and discussions with colleagues knowledgeable about HF self-care and gender roles. Reliability was measured by consistency of interpretation and coding (Byrne 2001). An audit trail of process and analytic memos and coding books was maintained to support the credibility of the study.

Quantitative data

Sociodemographic data were collected using an investigator-designed tool. Heart failure self-care was measured by the Self-Care of Heart Failure Index (SCHFI V4) (Riegel et al. 2004), a valid and reliable instrument with 17 items measured on a four-point Likert scale. These items form three scales: self-care maintenance (those monitoring and adherence behaviours done to prevent HF exacerbation), self-care management (the ability to recognise and respond appropriately to symptoms) and self-care confidence (perceived ability to engage in self-care maintenance and management behaviours). Scores on each of the SCHFI scales are standardised to 100; higher scores indicate better self-care. In this sample, the maintenance scale had a Cronbach alpha of 0·55, management 0·65 and confidence 0·86, which remained consistent with prior research. Adequate self-care was defined as a score of ≥70 on each of the SCHFI scales (Riegel et al. 2009b). Chi-square and Student’s t-test were used to examine differences in self-care by gender and by perceived role. The level of significance was set at p = 0·05.


In the three descriptive studies, patients were eligible for participation if they had documented evidence of symptomatic HF for at least 3 months and could speak and read English. Those with a history of a prior neurological event that could cause dementia or inability to perform tests (e.g. diagnosed dementia, unable to write or read in English, less than a fifth grade reading level, major visual or hearing impairment) were excluded. Two of the samples (total n = 70) were individuals with HF in the United States recruited from large urban medical centers (Riegel et al. 2007, Dickson et al. 2008). One sample (n = 29) was recruited from outpatient sites in Melbourne, Australia (Riegel et al. 2010).

Ethical considerations

In accordance with the Declaration of Helsinki, appropriate approvals were obtained for each of the two USA studies and in Australia, from the appropriate institutional human research ethics committees prior to subject recruitment.


Demographics and clinical characteristics of the sample of 99 are shown in Table 2. Overall, the sample was predominately Caucasian (74%) and male (65·7%). The ages ranged from 25–85 years (59·61 ± 13·42  years). Most (52·9%) were stratified as New York Heart Association (NYHA) functional class III with a mean length of illness of 6 years (75·33 ± 54·06  months). Most had systolic HF; the mean ejection fraction was 31·94 ± 16·08%. Idiopathic and ischemic cardiomyopathy were the most common causes of HF. Most participants had an implanted device to control heart rhythm.

Table 2.   Demographic and clinical characteristics of sample
 Males (n = 65)Females (n = 34)p
  1. **p < 0·05.

Age58·92 (±15·56)58·85 (±14·16)0·983
Ejection fraction28·07 (±14·03)39·05 (±17·43)0·006**
Ethnicity (%)
 Caucasian78·5 (51)64·7 (22)0·238
 African-American16·9 (11)32·4 (11)
 Hispanic3·1 (2)0·0 (1)
 Asian1·5 (1)0 (0)
Marital status (%)
 Single23·4 (15)30·4 (10)0·330
 Married55·3 (36)39·3 (13)
Widowed10·1 (7)13·0 (4)
Divorced4·3 (3)17·4 (6)
Co-habitat4·3 (3)0 (0)

Self-care roles

Two perceived roles in self-care were identified: (1) active and (2) passive. These were further categorised according to degree of independence in self-care decision making: primary responsibility, collaboration and reliant upon the direction of others.

A significant proportion (27%) of participants perceived that they held primary responsibility for self-care, describing self-care as a job, ‘It’s my job. If I don’t do it, who will?’ These individuals participated actively in all aspects of their self-care. For example, they described taking responsibility for meal planning, grocery shopping and meal preparation, ‘It’s my life, I better know what I am eating’. They also reported independently managing HF symptoms when they occurred; most commonly by diuretic titration.

Other participants (22%) described self-care as a collaborative process with another person, usually a close family member with whom they made shared decisions or received help in performing self-care. These individuals also described collaboration with healthcare providers. The theme of shared responsibility emerged from the narrative accounts of self-care. This participant group did not act independently to manage symptoms, but rather sought out the guidance of a clinician and together made a decision about how to treat a symptom:

I started to have some swelling, so I called <doctor>; we decided to increase the <diuretic> for a few days then go back to the usual dose …That worked, and I did not have to go to the emergency room…

Most participants (51%) adopted a passive role in self-care. These individuals reported that they ‘just did what they were told’ whether it was by a family member or a healthcare provider. They reported that they would ‘absolutely not’ take any extra diuretic unless instructed by their healthcare provider, ‘she went to medical school not me…’. They did not perceive self-care was part of their responsibility as an individual with HF, but rather others ‘know best’. They relied on other people for making key decisions about what to do in the event of exacerbated symptoms. Some completely abdicated responsibility for meal planning, diet and medication dosing, describing their role as ‘I just do what I’m told…’Table 3 provides a summary of the perceived roles.

Table 3.   Summary of qualitative evidence of perceived roles
Perceived roleDescription
Primary responsibility
‘It’s my job’
• Assumes primary responsibility in self-care
• Describes active role in all aspects of self-care
including independent symptom management
(e.g. diuretic titration)
‘we talk about what to do… decide together’
• Collaborates actively with others (family, friends and healthcare providers)
• Describes self-care as a shared responsibility
• Seeks guidance of provider to manage
symptoms; reluctant to act independently
Relies on direction from others
I just do what I’m told’
• Takes passive role in all aspects of self-care
• Relies on others to monitor and manage symptoms

Quantitative results

The mean SCHFI score for self-care maintenance was barely adequate overall (72·75 ± 14·36), which was also the case with management (69·75 ± 19·72). Amongst those with symptoms in the three months prior to the study (67%), self-care management scores were adequate in only 58% of the sample.

Integrated data results

Self-care roles by gender

The themes were analysed across cases to look for patterns in perceived role and self-care in relation to gender. Men who took an active role in self-care described their perceived role as a primary responsibility. Men also described making independent decisions, which is a key component of self-care management. For example, they were more likely to describe diuretic titration as a self-care management action. Some discussed ‘rogue’ self-care decisions about symptom management that were made without any direction or instruction from their healthcare providers such as doubling a prescribed medication, ‘I figure if I feel a little better with one pill…two will be better…’.

Conversely, women who took active roles in self-care described collaborating with others, usually healthcare providers, in making decisions about self-care, ‘we will discuss if I need to increase the water pill; I want to hear what <doctor> thinks and then we decide’. None of the women in this sample described self-care actions that were not either prescribed (e.g. diuretic titration), or discussed and coordinated with their physician in advance.

There were significant differences in role and self-care by participant gender. Among those who took an active role in self-care, men were more likely to assume primary responsibility and make independent decisions about self-care (p = 0·01). However, women had higher self-care maintenance scores (p = 0·04) on the SCHFI (Fig. 1). There was a trend towards better self-care management scores in the women (p = 0·06). Examining adequacy in self-care (defined as greater than a score of 70 on the SCHFI), there were significant differences in self-care management according to perceived role. More individuals who described their role as collaborative had adequate self-care (p = 0·05). Those who described their role as passive had poorer self-care (p = 0·01). Less than 30% of these subjects demonstrated adequate levels of self-care maintenance and management.

Figure 1.

 Gender differences in self-care among participants with active perceived role.


The results of this study suggest that perceived role in HF self-care vary by gender and that the roles are linked to quantitative evidence of self-care. Consistent with the work by Sarkadi and colleagues (2005), individuals in our study who reported they perceived their role in self-care as active had better self-care scores compared to those who took a passive role. A passive role in self-care, which emerged in our study as ‘relies on the direction of others’ has been associated with poorer self-care outcomes among people with diabetes (Sarkadi et al. 2005, Veg et al. 2006). In diabetes, self-care including diet adherence, glucose monitoring and symptom management is viewed as a mediating factor in achieving the desired endpoint of HbA1c. Individuals who described an active role in diabetes self-care had better HbA1c levels compared to those who assumed a passive role (Sarkadi et al. 2005).

In participants with HF, the relationship of self-care and outcomes is more complex. Self-care maintenance and management have been correlated to improved outcomes including a decrease in HF admissions (Riegel et al. 2009c) and reduced direct inpatient costs (Lee et al. 2010). For example, individuals who actively engage in self-care behaviours such as adherence to a low salt diet, symptom monitoring and appropriate management of symptoms, may effectively avert an exacerbation of HF due to fluid overload. Hence, these active patients may be able to avoid a hospitalisation and associated health care costs. Conversely, those who adopt a passive role and have inadequate self-care may be unable to avert an exacerbation. These individuals may be vulnerable to incorrect advice from others who lack essential HF knowledge or delays while waiting for direction from healthcare providers, which may lead to life-threatening cardiac decompensation.

Although our study is novel to the phenomenon of perceived role in HF self-care, our findings are consistent with Rodriguez et al. (2008), who examined preferred role and perceived involvement in medical decision making preferences. These authors found patients were not uniform in their desire to participate in their HF care despite the expectations of healthcare providers. In a sample of 90 adults being treated for HF by a Veterans’ Affairs provider, perceptions of involvement in HF care and preferred role in medical decision-making were assessed using the Perceived Involvement in Care Scale (Degnar et al. 1997) and Control Preferences Scale (Lerman et al. 1990), respectively. The majority of patients preferred to take the passive role (47%), fewer (31·1%) sought a collaborative role and a minority (21·1%) wanted to take an active role in the HF management (Rodriguez et al. 2008). Using multivariate regression analyses, researchers concluded patients’ preferred roles as measured by control preferences, is a strong, independent predictor of patients’ perceived involvement in HF care (β = −0·341; = 0·003).

Our results are unique in that we examined differences in perceived role by gender. Self-care role as an active and primary responsibility found mostly among men in our sample may be indicative of preference for control. This conclusion is consistent with the findings of Rodriguez et al. (2008) who reported that 17% of patient’s perceived involvement in care was explained by control preferences in a study of the male-dominated Veteran’s Affairs system. Generally, men have been found to have higher perceived control in HF self-care than women, which may influence perceived role in self-care. Heo et al. (2008) demonstrated that there are gender differences in factors affecting self-care, despite similar knowledge levels, physical, psychological and behavioural status. In men, higher perceived control was a significant determinant of self-care (p = 0·001). In our results, men’s preference for control emerged in the qualitative data through comments about primary responsibility.

Interestingly, although active self-care decision making was found mostly in men, the independent self-care role was not associated with better self-care. This finding was supported by the qualitative data accounts of incorrect self-care management of symptoms. These included increasing exercise when experiencing symptoms of chest pain or doubling medication without advice from a healthcare provider.

Our results contradict other studies (Sarkadi et al. 2005, Heo et al. 2008) that have reported active roles and perceived control were positively associated with self-care. The reason for this discrepancy is unclear and requires further investigation. It is possible that gender identity of men, often referred to as ‘masculinity’, affects how men manage their HF self-care. Gender identity is a key cultural factor that influences health-related behaviours (Jack 2005). According to experts, the masculine role requires that men are socialised to project autonomy, strength and physical aggression (Williams 2003). In an ethnography of African–American men with diabetes, masculine gender identity was found to influence diabetes self-management behaviours (Liburd et al. 2007). Participants described a desire to maintain control of diabetes through self-management actions regardless of clinical implications and even acted independently of physician direction (e.g. taking matters into their own hands). This was similar to our results.

Another potential explanation may be that the manner in which men experience HF affects their perceived role and response to its symptoms. Stromberg and Martensson (2003) ascertain that the physical and social restrictions of HF symptoms affect men’s activities of daily living and are more burdensome for them than for women. Therefore, men may adopt a more active and independent role in order to maximise control as much as possible.

In our sample, women who reported an active role described collaboration in self-care. To our knowledge, this is the first study to describe a preference for collaboration in self-care decisions by women. There is evidence that women respond differently to cardiac and heart failure symptoms, often delaying help seeking behaviours (Nguyen et al. 2010). They have even been shown to prefer to first contact a family member when experiencing symptoms (Ashton 1999) or healthcare providers (Bowker et al. 2000). Differences in treatment seeking delay for cardiac symptoms between men and women may be related to recognition and interpretation of ambiguous symptoms or lack of knowledge (Finnegan et al. 2000). However, Turris (2009) posits that women use a variety of forms of knowledge (e.g. temporal, embodied, rational and relational) in making decisions about treatment seeking that encompasses the social processes that are considered important to them (Turris & Johnson 2008). Therefore, women may place daily self-care within their social and familial roles, and seek collaboration with other family members on decisions about daily HF self-care.


The results of this study must be judged within the context of its limitations. Although each of the three original studies used similar general questions to gather accounts of self-care and perceived role in self-care, interpretation of the findings relied on interview data obtained from the primary studies. Contradictory accounts that were not clarified by the investigator at the time of the original interview required the investigators to resolve discrepancies within the analysis process by determining the predominant theme through content analysis of the transcribed data.

In addition, although this descriptive meta-analysis used a sample size of 99, which is robust for a qualitative study, it was a homogenous group. In each of the three original studies, efforts were made to obtain a diverse sample based upon sociodemographic and illness characteristics; yet the final sample was predominately Caucasian and male. In this meta-analysis of the combined samples there was good variability in age, illness characteristics, length of time with HF and self-care practices, allowing for richness in the qualitative data analysis. However, a sample of HF patients with more women, African, American and Hispanic patients is needed in order to understand the interaction of gender and these cultures on perceived role. By definition, roles are constructed by societal norms and individual values (Biddle 1986). It is likely that additional perceived roles may emerge from a more diverse sample comprised of individuals from other cultures and may influence self-care in different ways.

Participants in this study lived across wide geographic regions including urban, suburban and rural regions of three neighbouring states in the mid-Atlantic region of the United States and the urban city of Melbourne, Australia. However, all three samples were recruited from HF specialty settings and participants had established patient–provider relationships. These individuals may have been better clinically managed than a population of patients in more disadvantaged settings (McAlister et al. 2001) without access to healthcare providers. Further exploration is needed in patients with HF who are managed outside of specialty HF programmes and in more remote settings with limited healthcare services to determine if perceived role in self-care differs.

Relevance for clinical practice

The implications of this research include the need for clinicians to assess an individual’s perceived role in HF self-care as part of the self-care education process. Understanding perceived role may help guide patient education, especially for those who are likely to defer to others. For those who do not perceive an active role in self-care, it is essential to include key support people in self-care education and to understand this group may delay acting on symptoms until directed by a healthcare provider. Or, they may rely on the direction of others even when that direction is wrong.

These findings also suggest the need to develop and test coaching interventions to help women assume independent responsibility for self-care, as well as to understand how men make decisions related to self-care and improve those decisions made by men to ensure that they are appropriate.


This exploratory study provides preliminary insight into the importance of perceived role in self-care as a variable that may influence the day to day decisions that individuals make about self-care including diet, medication adherence and symptom management. Importantly, our findings suggest the need to further study the relationship of perceived role and self-care in order to support the development of independent and correct self-care practices and optimise self-care related outcomes.


The researchers gratefully acknowledge that this study was supported by the following awards: Sigma Theta Tau Doris Bloch International Research Award and Penn Institute of Aging Award, University of Pennsylvania, Philadelphia, PA, USA.


Study design: VVD, BR; data collection: VVD, BR, LWC, LK; data analysis: VVD, BR and manuscript preparation: VVD, BR, LWC, LK.