Background

Rheumatoid arthritis (RA) is a common chronic inflammatory disease with a prevalence of 1% [1]. The disease has a significant negative impact at both individual and societal levels [2]. Early diagnosis and treatment reduce the risk of future joint damage and disability [3,4,5,6,7,8] and increase the chance of drug-free remission [9]. As a result, there is now a drive to identify and treat people as early in the disease process as possible, potentially before the onset of joint swelling [10].

Genetic factors contribute significantly to a person’s risk of developing RA [11]. Having a family history, especially having one or more first-degree relatives with RA, increases the risk of developing RA by approximately 3–5 fold [12,13,14]. Modifiable environmental factors such as smoking [15], alcohol intake [16, 17], and diet [18,19,20] also influence the risk of developing RA [21]. Interactions between these risk factors are likely [22]. Gathering information about an individual’s genotype, environmental exposures, systemic and joint related immune abnormalities may thus allow clinicians to predict future RA development in those who are currently asymptomatic [22]. Among those people with a close relative with RA, current risk models using both environmental and genetic factors have been shown to be highly discriminative for both seropositive and seronegative RA [23].

Relatives of RA patients are a prime target population for both risk stratification and preventive interventions [24, 25]. First-degree relatives of RA patients are currently in the focus of prospective observational cohort studies [26,27,28] and interventional trials [21, 25]. However, difficulties recruiting first-degree relatives to such studies have been reported [29]. As recruitment of first-degree relatives is usually dependent on the cooperation of RA patients themselves, it is also important to understand patients’ perceptions of RA risk, risk modification and the communication about these issues with their relatives. Qualitative studies have already started to explore the views of RA patients in this context [30].

In order to develop effective recruitment strategies and potential future preventive strategies for first-degree relatives we also need to have a good understanding of the perceptions this group, and their willingness to engage with preventive approaches. We have recently reported data from interviews with first-degree relatives in three European countries looking at their perceptions of being at risk of RA and of predictive testing. Relatives reported having  a range of concerns about both predictive testing and risk information related to RA, including the possibility that knowing ones risk might increase anxiety and reported concerns about the level of uncertainty associated with predictive testing. Those relatives who expressed positive views about predictive testing indicated that they would need support to understand risk information [22].

Recent qualitative research has suggested that uptake of preventive medication by first-degree relatives might be related to perceived baseline risk and experience of the disease through their relative(s) [31]. Furthermore, preliminary data from a choice experiment with first-degree relatives suggest that acceptance of such medication would depend on its effectiveness and side effect profile [24]. In relation to lifestyle changes, one recent trial looked at the effect of disclosure of first-degree relatives’ RA risk personalised with genetics, biomarkers, and lifestyle factors on their health behaviour intentions and actual behaviour [21] and found that those at risk were motivated to and actually changed certain behaviours after such personalised information. However, there is currently no information available about first-degree relatives’ perceptions around these lifestyle changes and pharmacological interventions aimed at reducing the risk of RA development. This paper provides an exploration of previously unreported data collected during our earlier interview study [22] with first-degree relatives examining these issues in more detail.

Methods

Participants

Eligible participants were the first-degree relatives of people with an existing diagnosis of RA. Patients with RA were approached during their routine secondary care clinic appointments in Birmingham (United Kingdom), Erlangen (Germany) and Vienna (Austria) and asked to consider contacting a first-degree relative about participation in an interview study about risk and predictive testing for RA. Full procedural details are presented elsewhere [22]. We restricted the selection of relatives to offspring and siblings and excluded parents. In order to be eligible to take part, first-degree relatives had to be aged 18 years or over and not diagnosed with inflammatory arthritis at the time of the study.

Data collection and analysis

The semi-structured interviews were conducted either face to face at the recruiting hospital sites or by telephone and were guided by an interview schedule (Table 1). The schedule was informed by a literature review [32, 33] and in consultation with an international team of healthcare professionals, researchers and patient research partners participating in the EuroTEAM (Towards Early diagnosis and biomarker validation in Arthritis Management) project [34]. The interviewers in the UK were conducted by RS (female research fellow with a psychology background) and KK (female researcher and nurse specialist)). Interviews in Germany were conducted by AH (male senior clinical research fellow) and interviews in Austria were conducted by EM (female occupational therapist and researcher). All interviewers had previous experience of conducting interviews and RS, KK, and EM have extensive experience of qualitative methods. Interviewees received information about EuroTEAM in the participant information sheet (i.e. that it is a multi-country research project funded by the EU) and were told the name of the person interviewing them, as well as their role in the project. All interviews were audio recorded and transcribed by a professional transcription service. Interviews conducted in Germany and Austria were translated from German into English following transcription by bilingual native speakers and the translations were checked by the respective research teams. All transcripts were analysed centrally at the University of Birmingham, United Kingdom.

Table 1 Outline Interview schedule
Full size table

Data collection and analysis were carried out in parallel to allow for an assessment of when thematic saturation had been achieved. Transcripts were analysed thematically [35] using NVivo (software programme for qualitative data analysis; [36]). One researcher (RJS) coded all the transcripts and three patient research partners blind coded three transcripts. The coding framework was discussed with the patient research partners and coding categories that lacked concordance were absorbed into the coding framework. The initial codes were then grouped into the most noteworthy and frequently occurring categories by RJS, KK and GS.

Results

Thirty-four first-degree relatives of RA patients were interviewed; 24 from the UK, three from Germany and seven from Austria. Participants were aged between 21 and 67 years (M = 39, SD = 10.8), and 76% (n = 26) were female. All but two British participants were white. Table 2 gives the individual participant characteristics including age, gender, relation to relative with RA and current musculoskeletal symptoms.

Table 2 Interviewee characteristics
Full size table

Interviews lasted between 30 and 90 min. The results presented here represent a new analysis of the resulting transcribed data. Findings related to an analysis of perceptions of risk and predictive testing are presented elsewhere [22]. The core themes presented here focus on perceptions of first-degree relatives about the possibility of making lifestyle changes and taking preventive medicine to modify their risk of developing RA. An overview of the organising themes and subthemes can be found in Table 3. Quotations referred to in the text can be found in Tables 4 and 5.

Table 3 Overview of themes related to the sub-analysis focused on modifying risk through lifestyle and preventive medicines
Full size table
Table 4 Modifying risk through lifestyle changes
Full size table
Table 5 Modifying risk through preventive medicine
Full size table

Modifying risk through lifestyle intervention

Many participants expressed positive views about undertaking lifestyle changes to reduce their risk of developing RA in the future (Table 4, Quote 1 (T4Q1)) or to delay the start of the disease (T4Q2). The main lifestyle factors identified by participants included healthy eating, increasing levels of exercise and smoking cessation. Some participants indicated that they were already trying to live as healthily as possible (although not necessarily focused on preventing RA), but that they would be willing to make additional changes (T4Q3–4). Diet in particular was highlighted as a modifiable risk factor for development of RA and one which patients could take control of (T4Q5). Some participants felt that being overweight and ageing were risk factors for the development of RA (T4Q6).

Participants felt it would be useful to know their personal risk of developing RA so they could modify their lifestyle as a preventive measure (T4Q7). Many participants indicated that they would need information about if and how certain factors such as exercise and diet might impact on the risk of developing RA before deciding what they would do (T4Q8–9).

Some participants expressed more negative views and indicated that they might not be prepared to make lifestyle changes as a preventive measure. Some were more concerned about the perceived negative consequences of making such changes and as a result of that might not want to engage with them (T4Q10). One person indicated that if the only information they have, is that they are at a heightened risk of RA they would not make a change to their diet (T4Q11).

Only a small proportion of the first-degree relatives were current smokers (five, one of whom smoked very rarely) and 2 were ex-smokers. Whereas one smoker described their willingness to give up smoking if a test result identified him as being a person at risk (T4Q12–13), another smoker indicated that a 50% baseline risk would not be enough to consider stopping (T4Q14). A third smoker indicated that they were only prepared to give up smoking if it was confirmed to them that this would definitely prevent them from developing RA (T4Q15). Participants who were not current smokers either identified or accepted that smoking is a risk factor for the development of RA and contemplated how smokers may react negatively to risk information (T4Q16). Another participant pointed out that it can be difficult to avoid passive smoking (T4Q17) and hence the exposure to this risk factor.

Willingness to take preventive medicines to modify risk

When discussing the possibility of taking preventive medicines if identified as being at high risk of developing RA, many participants highlighted that they would be worried about possible side effects of these medications (T5Q1). They indicated that they would need to think very carefully about the pros and cons of taking preventive medication before making a decision (T5Q2–3). Some participants further worried about the effect of the preventive medication on their existing medical conditions and the potential interaction with current medication (T5Q4–5). Other participants considered the long-term implications of taking medicines, and whether it would, for example, have an impact on family planning in the future (T5Q6). The immunosuppressive properties of disease modifying anti-rheumatic drugs were a particular concern for participants, especially those who had seen the side-effects of such medicines (e.g. an increased frequency of infections) experienced by their family members with RA (T5Q7).

For some participants, weighing up the side-effects of preventive treatment, which were often described as if they were highly likely or certain to occur, against the relative uncertainty of developing RA in the future, meant that they would most likely decline to have preventive medication unless it was shown to be highly effective (T5Q8). Many participants indicated that they would not be willing to take a preventive medication based on a “probability” provided by a predictive test and that they would need more “definitive” evidence (T5Q9). It was felt that this type of probability information (E.g. ‘You have a 50-50 chance to develop RA’) was too uncertain to inform a choice about medication. Participants reported that a preventive pharmacological therapy would be considered only if the medicine was highly effective in reducing their risk (T5Q10). There were further worries about potentially taking medication for the rest of their lives and the potential consequences of stopping medication (T5Q11). Participants suggested that if a test showed that there was a 50% risk that they would develop RA, this would not be enough for them to accept preventive medication, though they felt that this information might encourage them to monitor for the onset of potential symptoms of RA. Others suggested a relatively high predicted baseline risk might be high enough to convince them to accept preventive medication (e.g. one interviewee suggested this would be a 70–80% risk that they would develop RA; T5Q12).

In addition, several participants expressed negative opinions about taking medicines in general, in some cases as a result of past experiences with being prescribed medication that they perceived to be unnecessary,-and they anticipated that this attitude would stop them from taking medication to prevent the onset of RA (T5Q13–14). Others indicated that, based on the assumption of an equal level of risk reduction, they would rather prefer to make lifestyle changes as opposed to taking long term medication (T5Q15) in order to reduce their risk of developing RA.

There was also a group of participants who could understand why preventive medicine may be prescribed if they were found to be at risk of developing RA in the future and anticipated that they would be willing to accept such treatment (T5Q16). This was especially the case for those participants who had experienced the negative impact of the illness on their relatives (T5Q17).

Participants were supportive of the idea that predictive testing could be useful to alert those who are at high risk to be vigilant for the early symptoms of RA, so that treatment could be initiated as soon as symptoms appeared (T5Q18). Many participants suggested that they would prefer this strategy of early intervention during the symptomatic phase as opposed to taking a preventive medicine prior to the onset of symptoms (T5Q19).

Finally, for some participants the issue of engaging with a preventive intervention depended on an expectation that it would definitely be effective (T5Q20).

Discussion

Given the current impetus to identify early intervention points in the RA disease pathway, it is important to understand the perspectives of those at increased risk of developing RA about possible preventive interventions. The current study looked at one particular at risk group, first-degree relatives, and found that many anticipated being happy to make lifestyle changes such as losing weight, increasing exercise and changing diet to modify their risk of developing RA. However, there was less enthusiasm about a pharmacological approach, at either the pre-symptomatic or the symptomatic at risk stage [9, 24].

The interviewees’ reactions to the possibility of making lifestyle changes might be in part due to the social desirability of giving such a response and it is unclear how many would actually adopt these behaviours. One would expect that using well trained interviewers who, with one exception were non-Rheumatologists, went some way to avoiding such responses. Further, the results from the trial conducted by Sparks and colleagues [21] indicate that first-degree relatives were both motivated to modify their behaviour and reported making actual lifestyle changes (e.g. increasing their fish intake; quitting smoking) especially after receiving personalised information about their risk status and the impact of health behaviours on their risk status. This implies that first-degree relatives might indeed follow through on making lifestyle changes after receiving relevant (and personalised) information. Indeed, some interviewees in the current study indicated that they would appreciate further information about the impact of such behavioural changes on their risk of developing RA. Others said that they would require definitive confirmation that it would lead to a significant reduction in their risk of developing RA before contemplating making such changes. Interviewees were more reluctant to contemplate making changes in their lifestyle if they considered there to be negative effects of making such changes. Stopping smoking can be a difficult lifestyle change to make due to the addictive nature of tobacco [37]. In our sample, the perceived benefits of stopping smoking in relation to RA would make some participants more likely to consider smoking cessation if they were found to be at risk of developing RA. Others would need more certainty that this lifestyle change would eliminate their risk of developing RA altogether. These findings are in line with research in other at risk groups for chronic conditions such as diabetes or cardiovascular disease (CVD). In a study where individuals were given their actual risk score for CVD, some indicated that they had made or intended to make lifestyle changes such as diet change or stopping smoking, whereas others would not make those changes, for example, because they disbelieved the risk score or downplayed their risk, or because they actively resisted making such changes and for example ‘could not be bothered’ to stop smoking [38]. A person’s motivation to be healthy prior to being tested for a chronic condition, will impact on how they perceive their risk and their willingness to make behavioural changes once they receive their results [32]. Further research is needed to comprehensively assess predictors of responses to information about disease risk.

The finding that first-degree relatives were more hesitant about preventive medication is in agreement with how patients with RA in previous research felt that their family members might react if given the option of preventive medicine [30]. In line with findings from previous research in RA (e.g [27, 39]), potential side effects, long-term implications of taking the medication and possible interactions with existing conditions and medications were a major concern for many of our interviewees and they indicated that they would weigh up perceived benefits and potential harms before making a decision about therapy. Personal experiences of RA in family members influenced decision making about the value of preventive medication for several interviewees. On the one hand, those who had observed the side effects of medications such as Disease Modifying Anti-Rheumatic Drugs (DMARDs) experienced by their relatives indicated that they might be more reluctant to take the medication as a preventive measure. On the other hand, some interviewees, who had witnessed the negative effects of RA on their relative’s daily life, were keen to use preventive medicine to reduce the risk of developing musculoskeletal symptoms. Acceptance of preventive medication was further associated with the issue of certainty of future RA development. Many felt that uncertainty around future development of RA would discourage them from accepting medications, which they associated with negative side effects. Some interviewees indicated that they would only consider taking medication after the onset of symptoms, and the main value of predictive screening would be to identify the extent to which they should be on the alert for the start of RA symptoms – at which point they would consider therapy which is also in line with findings from a recent focus group study with first-degree relatives [39]. Interviewees in the current study were further not directly asked what they perceived their current risk of developing RA to be, but it is likely that different levels of perceived risk might have had an impact on their views on preventative measures. Future research should explore this relationship further.

Interviewees in this study were not provided with detailed information about the likely duration, frequency of method of administration of preventive treatments. Some of our interviewees appeared to assume that treatment to reduce their risk of RA would involve long-term medication and associated extended risk of side effects. This could account for negative viewpoints towards such treatment. However the trial that is currently evaluating preventive therapy (hydroxychloroquine) for this group (asymptomatic individuals) involves a 12 month course [25], which could be considered as relatively long term treatment.

A recent discrete choice experiment looking at first-degree relatives’ preferences related to pharmacological interventions, suggested that method of administration may be an important determinant of preventive treatment acceptability [40]. In contrast a related best worst scaling pilot study found that the efficacy and risks of treatment were more important than method of administration in decision making about preventive treatments [24]. Further quantitative evidence is needed to clarify the relative importance of treatment related and other attributes in the choices made by at risk individuals about preventive therapy for RA.

Collectively these findings suggest that educational approaches will be needed to support preventive strategies, and should include information about the benefits and risks of lifestyle changes and pharmacological interventions as well as details about the nature of the intervention itself. This supportive information should be given in the context of personalised risk estimation with clear (and where possible quantitative) information about the potential for risk reduction associated with such interventions [41]. As predictive algorithms for RA continue to evolve and improve, the interplay between genetic and environmental factors could be highlighted and it is possible for interactive tools to give people personalised information about their risk of developing RA given their genetic make-up and current lifestyle, demonstrating how RA risk can be changed by making certain lifestyle changes. The recent trial of such a tool shows a positive impact on both intended and actual risk related behaviour change [21]. However, there remains a misalignment between public expectations of risk assessment (e.g. being able to predict with 100% certainty whether or not an individual will develop a disease) and the realistic possibilities of predictive models – which provide probabilistic risk estimates with varying confidence intervals [30]. Levels of confidence in estimates of risk or efficacy may impact on preferences for preventive treatments [30, 40, 42], therefore informational resources accompanying preventive interventions should therefore also manage expectations around the accuracy of risk information [42].

Our results further highlight the need to communicate risk information in a way that is sensitive to the personal/family context or life situation of each individual. One could also provide information that is specifically tailored for different groups with varying experiences of RA in order to provide a richer, more personalised reflection of their own experience. First-degree relatives are a distinct at risk group and their views are likely to be influenced by their personal experiences with RA. As such, their information needs are likely to be different from those of the general population and of other at risk groups, such as patients with clinically suspect arthralgia [43] or undifferentiated arthritis. In ongoing research we are exploring the perspectives of other at risk groups and the wider public, and predictors of perceptual variability within each of these groups.

A number of limitations need to be considered when interpreting the findings of the current research. Firstly, the proportion of men in the sample was relatively small. Although this does reflect the fact that fewer men than women develop RA [1], important viewpoints may have been missed in the current study and future research should aim to include a larger sample size of men. Similarly, although we interviewed relatives in three different European countries, the current sample is not ethnically diverse and as such does not comprehensively represent the population of each of the participating countries. The sample size and qualitative approach used here do not allow for comparative analysis between the three countries and assessment of cross-cultural differences are not possible on the basis of the current data. Future research should therefore specifically address these omissions in order for screening and intervention programmes to be tailored for as wide a target population as possible and to provide related information in a way that is gender or culturally appropriate. Further quantitative evidence is needed to explore psychosocial and cultural predictors of preference heterogeneity and the degree to which at risk individuals trade off positive and negative aspects of preventive interventions for RA.

Since access to first-degree relatives is usually indirectly through patients with existing RA, future research needs to extend and quantify findings of previous research [30] related to if and how RA patients are likely to communicate with their first-degree relatives about their increased risk of developing RA and risk reduction options. This information will help to facilitate the development of efficient preventive approaches and recruitment of participants to preventive studies. The current research discussed participants’ willingness to make (hypothetical) lifestyle changes. However, although there is consensus about some lifestyle factors which influence a person’s personal risk of developing RA (such as smoking), more research is needed to understand the degree of risk reduction, in the context of different genetic backgrounds, associated with risk related lifestyle changes [21]. Only when such data are available can those at risk make informed decisions about preventive intervention.

Conclusion

The current research gives an indication of first-degree relatives’ perceptions of lifestyle changes and preventive medication to reduce their risk of developing RA and the factors associated with acceptability of such preventive measures. It identifies factors which should be highlighted in informational materials aimed at first-degree relatives who are considering participating in related research studies or preventive approaches as well as highlighting factors which warrant further investigation.