The role of partner support for health behaviours in people living with and beyond cancer: A qualitative study

Abstract Objective This study aimed to qualitatively explore how partner support for health behaviours is perceived, received, and utilised in people living with and beyond cancer (LWBC). Methods Semi‐structured audio interviews were conducted with 24 participants, 15 men and nine women, living with and beyond breast, prostate, and colorectal cancer. Inductive and deductive Thematic Analysis was used to analyse the data. Results Three key themes with six subthemes were identified relating to partner support for health behaviours: (1) Interdependence (Reciprocity, Overt Control, Influence & Motivation) (2) Concordance (Shared Attitudes & Health Beliefs, Shared Health Behaviour) and (3) Communal Coping (Communal Orientation towards Health and Decision Making, Co‐operative Action in Health Behaviour). Conclusions Partner support plays a unique and significant role in the health behaviours of people LWBC. Partners play a collaborative role in managing health and facilitating health behaviours, while the high level of concordance in couples may represent a potential barrier to change via the reinforcement of maladaptive health beliefs and behaviours. Implications for Cancer Survivors Overall, findings demonstrate that partners should be considered and included where possible when designing future behaviour change interventions for people LWBC.

There is strong evidence that symptoms, Quality of Life (QoL), and survivorship can be significantly improved by targeting multiple health behaviours, including physical activity (PA), diet, smoking and alcohol consumption. 4 Several meta-analyses have found increased PA post-diagnosis improved survival outcomes in 11 cancer types, 5 reduced breast cancer deaths by 34%, and reduced all cause mortality by 41%. 6 Furthermore, a higher intake of vegetables and fish was inversely associated with overall mortality, while a 'Western' dietary pattern was associated with overall mortality. 7 The strongest evidence for the efficacy of behavioural interventions is for breast, prostate and colorectal cancer and based on this empirical evidence, the World Cancer Research Fund has developed guidance for health professionals to help improve health behaviours in people LWBC. 8 Despite the benefits of adopting positive health behaviours, studies have shown that many people LWBC are not meeting health recommendations. 9,10 It is possible that clearer messaging and interventions from healthcare professionals could help improve health behaviours in this population and it has been suggested that a cancer diagnosis may present a 'teachable moment' whereby patients are open to making changes in lifestyle in response to a major health concern. 11 Moreover, a body of empirical evidence suggests that while long-term behavioural change can be difficult, it may be facilitated by concomitant support from the social environment 12 and that individuals attempting to make behavioural changes can be positively influenced by their significant others during the course of this process. 13 Social support is considered one of the major social influences on health behaviour. 14 The smallest network is of course the dyad and using one-to-one peer matched support has proven effective in breast cancer survivors and their daughters. 15 In addressing the impact of dyadic support on health behaviours of people LWBC, intimate partner relationships are of particular interest as partners have profound influence on one another, and health behaviour is often concordant across couples. 16 Moreover, partner support has been found to improve outcomes across a range of domains including smoking 17 and PA, 18,19 and improve cancer outcomes, lessen pain and lower mortality. 20 While few post-treatment interventions currently target cancer survivors and partners, positive findings from a recent scoping review indicate that there is potential for expanding this area of research 21 and several feasibility studies have shown promising results with couples-based behaviour change interventions including a PA intervention for breast and prostate cancer survivors, partnered strength training for prostate cancer survivors and a diet and exercise intervention for people living with and beyond breast, prostate and colorectal cancer. [22][23][24] However, there remains relatively little qualitative research exploring how partner support is experienced for health behaviours by people LWBC. The aim of the present research was to qualitatively explore the role of partner support for people LWBC and how this support may influence and facilitate their health behaviours.

| Design
This was a qualitative study using one-to-one semi-structured telephone interviews. The study adopted an interpretivist approach, suited to generating knowledge relevant for health and clinical practice. 25 This approach recognises the importance of situating the researcher in the context of that which is being studied, in order that they may offer an interpretive understanding of the meaning participants attribute to their own experiences. This study was part of the Advancing Survivorship Cancer Outcomes Trial (ASCOT), 26

| Recruitment and data collection
Participants who had received an initial diagnosis of breast, prostate, or colorectal cancer in 2012/2013 were recruited to ASCOT from 10 NHS Trusts across London and Essex between 2015 and 2019, randomised to receive a habit-theory based behaviour change intervention or control, and then assessed at 0, 3, 6 and 24 months.
Inclusion criteria for the trial were, adults (aged ≥18 years), diagnosed with non-metastatic breast, prostate, or colorectal cancer, not currently receiving anti-cancer treatment (except oral treatments taken at home), able to understand spoken and written English.
Exclusion criteria included individuals receiving anti-cancer treatment requiring hospitalisation, with metastatic cancer, or severe cognitive impairment. A follow-up survey was completed in 2020-2021 to understand the impact of COVID-19, where participants were given the option to consent to be contacted for qualitative interviews about factors influencing their health behaviours during the pandemic. Of the 788 survey respondents, 669 (85%) consented to interview, of which 573 indicated that they were married/living with partner. For the current study, participants were purposively sampled to ensure adequate representation of the three cancer types, gender, ASCOT intervention/controls, and rural/urban dwelling. It was important to hear the experiences of those who had received our behavioural intervention and those who had not, as well as participants living in both rural and urban areas, as location and access to commercial facilities and green spaces can have significant impact on health behaviour. Only participants who indicated they had a partner in the Covid-19 survey were invited for interview. The 1998sample size range was 15-25 participants, with data gathering to stop when thematic saturation was reached. This range was deemed adequately broad for one interviewer-one participant research, while maintaining the capacity to provide richly textured information. 28 Partner status was reconfirmed at the beginning of each interview. A topic guide was developed (see Supporting Information S1) covering areas of interest in relation to diet, PA, smoking and alcohol, with prompts to guide conversations to how partner support is perceived, received, and utilised in the relation to these domain. Each topic was explored sequentially and in-depth, giving participants every opportunity to reflect on, describe, and detail their experiences. One-toone interviews were conducted via telephone by a female Health Psychology Researcher (NG) with no prior relationship with any participants. Comprehensive notes were taken during and immediately following each interview. Interviews were audio recorded and transcribed verbatim by a professional transcription company.

| Analysis
Initially, 15 interviews were conducted by NG, audio files were listened to twice to become familiar with broad themes of conversations and to facilitate early identification of patterns, before being coded and analysed using Thematic Analysis, with extensive referral to methods employed by Braun and Clarke. 29 Six stages of familiarisation, initial code generation, searching for themes, reviewing themes, defining themes, and writing up were followed. Analysis of the transcripts was informed by inductive methods to derive themes from the data and deductive methods to situate those findings within a theoretical model. To reduce single method, single-researcher bias, three transcripts were second-coded independently by RC, while emerging ideas were regularly discussed with the research team.
Once an initial coding framework was agreed upon, the remaining interviews were conducted in sets of three until thematic saturation was reached. Inductively derived codes were mapped to central constructs of Social Support theory 30 Emotional, Tangible, Informational and Appraisal support. 31,32 Interdependence theory 33 and Theory of Communal Coping 34 were introduced to support the analysis. Data was managed using an Excel spreadsheet and organised methodically within a participant-led theme matrix to ensure clarity, to avoid losing context and tone of conversations, and to demonstrate clear justification of the pathway from coding to conclusions drawn. The process was iterative, and codes and themes were continuously adapted to ensure data was accurately reflected in the findings. Illustrative quotes are provided stating participants gender, age, and cancer type.

| RESULTS
In total, 15 (62.5%) men and 9 (37.5%) women were interviewed, including seven participants with a diagnosis of prostate cancer, seven with breast cancer and seven with bowel cancer. One participant had prostate and colorectal cancer, one had prostate and skin cancer and one had breast and anal cancer. See Table 1 for participant characteristics.
Participants ranged in age from 53 to 85 years, and all were married, with the length of marriage ranging from 26 to 64 years.
There were nine living in a village/small town, while 15 lived in a large town/city. Fourteen participants had been assigned to the ASCOT intervention group, 10 were in the control (usual care) group. Interviews were between 30 and 90 min long. Three themes with six sub-themes were identified in the analysis, summarised in Figure 1, and discussed below.

| Interdependence
Understanding the role of partner support for health behaviours revealed the importance of these dynamics in the everyday lives of people LWBC. 'We' was the pronoun of choice for participants when discussing their health behaviour. This early observation set the scene for the overarching theme of interdependence that was identified. This was evident in all relationships and took differing forms depending upon the existing dynamics of the relationship.

| Communal coping with health
All participants described a collaborative approach towards their health and health behaviour. Partner's health was considered a joint concern, with overt expressions of communal coping. Participants stated that they had managed to cope with most things together, pulled together and managed to get through. Coping reflect this interdependence in action.
Within the present study, it was notable that participants were in highly compatible and supportive relationships and direct partner effects on health behaviours could be seen often, via partners overt control and influence. The findings suggest that partners are not only a 'source' of support but are actively engaged in coping with all aspects of each other's health, in a reciprocal and collaborative process.
Interdependence theory further separates processes of influence into 'joint' and 'mutual joint' effects. 35 The present data provides clear indication of joint effects, where the actions of the self and the partner impact upon the health of the individual. But perhaps most significantly, there was considerable evidence of mutual joint effects.
That is, partner support was associated with partners engaging in health protective behaviours together, and in some cases initiating Our results are in line with previous findings which suggest that attitudes and health beliefs are highly concordant across couples. 16 While some of this may be attributable to assortative mating, previous longitudinal research suggests that spousal influence itself is associated with health enhancing behaviours. 40 The concept of 'social contagion' within couples is well documented across behaviours 41 and is supported by studies attempting to change one partners risk behaviours, resulting in 'behavioural diffusion' which positively benefits the non-participating partner. 35 In a large-scale study examining the influence of marital status on attendance at colorectal cancer screening, it was shown that married adults were more likely to attend screening than non-married, and that inviting both members of a couple together further increased screening uptake. 42 It is thought that partners monitor and regulate each other's behaviour in ways that influence health behaviour by means of 'social control'. 43 However, partners can also reach concordance through mutual reinforcement of unhealthy behaviours, 44 as demonstrated in our study in relation to episodic binge eating and drinking.
From the outset of analysis, it was notable that 'We' was the first-person pronoun of choice for all participants when answering questions about their health. Previous research which analyses We/I ratio scores, found a higher score was significantly associated with relationship quality and predicted positive changes in heart failure symptoms at follow up. 45 Moreover, a study examining how couples describe coping with breast cancer found that resilient couples coordinate their coping efforts by defining the cancer experience as a dyadic stressor, or 'we-stress', that affects both, and this is evident in the 'we-language' surrounding the descriptions of their experiences. 46 48 it would be interesting to discover the extent of this transformation in younger couples and to explore the possibility of enhancing relationship-centred motivations as part of future behaviour change interventions, which has shown some success in dyadic intervention studies. 49

| Study limitations
To our knowledge this is the first study to qualitatively explore the role of partner support for health behaviours in people LWBC and was conducted during the Covid-19 pandemic, when participants were primed to consider their health behaviour closely. We

| CONCLUSION
The study offers unique insights into how people manage health