Attitudes and adherence to changes in nutrition and physical activity following surgery for prostate cancer: a qualitative study

Objectives Interventions designed to improve men’s diet and physical activity (PA) have been recommended as methods of cancer prevention. However, little is known about specific factors that support men’s adherence to these health behaviour changes, which could inform theory-led diet and PA interventions. We aimed to explore these factors in men following prostatectomy for prostate cancer (PCa). Design, setting and participants A qualitative study using semistructured interviews with men, who made changes to their diet and/or PA as part of a factorial randomised controlled trial conducted at a single hospital in South West England. Participants were 17 men aged 66 years, diagnosed with localised PCa and underwent prostatectomy. Interview transcripts underwent thematic analysis. Results Men were ambivalent about the relationship of nutrition and PA with PCa risk. They believed their diet and level of PA were reasonable before being randomised to their interventions. Men identified several barriers and facilitators to performing these new behaviours. Barriers included tolerance to dietary changes, PA limitations and external obstacles. Facilitators included partner involvement in diet, habit formation and brisk walking as an individual activity. Men discussed positive effects associated with brisk walking, such as feeling healthier, but not with nutrition interventions. Conclusions The facilitators to behaviour change suggest that adherence to trial interventions can be supported using well-established behaviour change models. Future studies may benefit from theory-based interventions to support adherence to diet and PA behaviour changes in men diagnosed with PCa.

posed in the objective.
Fifth, at what level of abstraction did the authors aim in identifying themes? As I read all of them, each is purely descriptive and thus generally best classed as categories and not themes. Refer to the editorial, "Confusing Categories and Themes" by J. Morse (2008). Might you revise your level of analysis? I also urge you to reconsider whether you have sufficient data to report certain themes. Most themes are described with a thin narrative and few quotes. The balance of narrative to data is typically indicative of the depth of development and representation of richness in the underlying data. Do the authors think they've adequately captured what the participants and data convey? I'd like to see a good deal more development here. Sixth, it is important to include in the method and the discussion section the information on how the rigor of the study was established. There are a number of references for this, including the seminal text by Lincoln & Guba (1985) and the article, "Critical Analysis of Strategies for Determining Rigor in Qualitative Inquiry" by J. Morse (2015). Seventh, there are multiple discrepancies between the statements (or the names of the categories) and quotes presented to support the statements/categories in the results section. For instance, the author wrote on page 6: "While other men described making extensive changes on starting the PBD within the trial." However, the quote that follows this statement is "As I say, I used to eat an awful lot of fruit and vegetables beforehand…" Another example: the authors put the following statement as an example for the "physical limitations" category: "Most men also relied on good weather. There was also little motivation to walk elsewhere when the weather was bad. "…there were some days where it was a total wash-out, and you think, "Well, there's no point in even trying," you know. "I'll make this my quiet day"…" However, it seems that the above-mentioned example is more appropriate for the "external obstacles category." Furthermore, the results section is difficult to understand as it lacks a clear picture overall. Therefore, the result section will benefit if the thematic map of the study is presented. Eighth, how are the authors constructing the discussion for this manuscript? Typically, a strong discussion recapitulates the main findings in brief and then goes on to argue original contributions, areas where findings corroborate extant literature, where the findings conflict, and finally limitations. This progression allows the authors to draw clearly defined conclusions in line with the method and results and to outline implications. I do not see this structure in your discussion. Critically, many statements in the discussion section are not supported by the study result. For instance, the following statements in the discussion section are not supported clearly by the study results: (1) "The findings suggest that a diagnosis of cancer can provide an opportunity for men to make changes to their nutrition and PA." (2) "While some men followed the intervention guidelines, others made quite extreme changes to their diet, such as eating well-over the recommended daily intake of fruit and vegetables." (3) "The facilitators to change identified in this study (i.e. … self-efficacy with taking lycopene and exercising)"; (4) "This qualitative study suggests that behaviour change models could help both inform interventions and promote long-term adherence to nutritional and PA behaviours." Ninth, the "strengths and limitations" section should be revised. First, it is unclear how the study's strength could be its "findings on the psychological, behavioural, and social factors associated with adherence to men's changes to their diet and PA". Usually, the strength of the study should be judged on the rigor of the study methodology; therefore, the researchers should specify on what quality criteria they based their research. Please see, Lincoln & Guba (1985) and the article on "Critical Analysis of Strategies for Determining Rigor in Qualitative Inquiry" by J. Morse (2015), as well as, Treharne GJ, Riggs DW. Ensuring quality in qualitative research. In: Rohleder P, Lyons AC (eds) Qualitative Research in Clinical and Health Psychology. Basingstoke: Palgrave Macmillan, 2014; pp. 57-73. Furthermore, although in the result section the authors mentioned that "men across all the intervention arms believed that they maintained a healthy diet before being diagnosed" and the authors provided several quotes that described the men's healthy behaviors before the study, they did not mention this fact in the limitation section. However, it is well known that the new health-related behaviors' is easier to achieve in people with previous healthy behaviors compared to those without previous health behaviors; therefore, the factors that supported and hindered men's adherence to the nutrition and physical activity could be different among those with and without previous healthy behaviors.

Results
Page 7, line 39: I suggest to specify only the first time that PBD is the Plant-based diet, and then use the acronym consistently.
Page 7, line 23: It could be of interest a description of some theme of the man included in the control group of the PrEvENT study.

Discussion
Page 11, line 32: In the literature the role of partner is considered a usefull support for men with PCa to adopt and maintain healthy behaviours. Partner may influence each other's diet and also exercise behaviours (PMID: 25807856). Your findings did not confirm the role of partners in supporting PA, so I suggest to discuss this point.

Edmunds, Kim
Griffith University, Centre for Applied Health Economics REVIEW RETURNED 08-Jan-2022

GENERAL COMMENTS
Thankyou for the opportunity to review this paper. Lifestyle compliance and adherence are important considerations for cancer patients and it is pleasing to see qualitative investigation happening alongside randomised clinical trials. I thoroughly enjoyed the paper and encourage you to extend your research in this important area as you propose in your protocol. While much research has been carried out on adherence, PCa patients with early stage disease are a unique group because they tend to be relatively healthy and the evidence for the impact of lifestyle change on their disease is strong. Any research that interrogates the why behind adherence or otherwise is always welcome. That said, there are areas in your paper I believe could be strengthened. Most importantly is that much of the research that backgrounds your study tends to be quite dated. Line 3 …another qualitative study… Interestingly, both the findings from these studies seem to suggest that these men prefer to blame things outside their control or not take responsibility for their actions (environment and work/life stress, vs overeating and being inactive?? Perhaps this could be used in devising motivational change…??? Just a thought I am not sure I agree with your tentative finding that men would adhere to healthy changes regardless of their perceived risk. Perhaps something else is happening that fits more with their preference not to take responsibility…if some one organises it and tells them what to do, they will happily do it??? Just a thought I think the small sample size is a limitation that should be mentioned.
Congratulations on a well conducted study on an important topic. My suggested changes are minor. However, I believe it is important in the background and in the discussion to incorporate other more recent studies against which to compare your own.

Dr. Svetlana Doubova, Mexican Social Security Institute
Comments to the Author: The manuscript entitled "Attitudes and adherence to changes in nutrition and physical activity following surgery for prostate cancer: a qualitative study" aimed at investigating factors that increase men's adherence to the nutrition and physical activity interventions. I find it difficult to understand the methods, results, conclusions, and scientific contribution of this study.
First, the introduction section should specify if the study was informed by a theory or conceptual framework? If so, describe it and how it was used. If not, say so and why. It is confusing that the authors mentioned the Theory of Planned Behaviour only in the discussion section.
Response: We apologise from this confusion. This qualitative study was not informed by a theory or conceptual framework. We have clarified in the introduction that we used a qualitative study to explore adherence factors in this patient population, which could inform theory-led interventions on page 3. Forth, there are discrepancies between the study objectives and study methodology. The study objectives focus on the factors that supported and hindered men's adherence to the nutrition and physical activity; yet, in the method section on page 4 the authors specified that the participants "were asked mainly about their diet and PA before participation in the trial".
We apologise for this confusion, which was due to grammatical error. We amended the respective paragraph on page 5 to clarify that we had one participant, who was in the control group, and only used data relating to his diet and PA before participation in the trial in the analysis: "One man in the control group was included in the sampling. He received no intervention aside from standard publicly available nutrition and PA information, if requested. Data from response asked about his diet and PA before participation in the trial was only used for analysis." At the same time, the patient interview guide that the authors presented focuses mostly on the participants' experiences with the intervention, asking about the positives and negative elements of the intervention arm and participant's opinions about the associations between diet, physical activity, and cancer. There is unclear how the questions in the interview guide as shared in the paper helped answer the questions posed in the objective.
Response: Our objective was to identify factors that support behaviour change to the nutrition and physical activity interventions. As the interviews were semi-structured, men were able to elaborate on their responses to questions asked from the interview topic guide. The flexibility of this approach led to many of the men's responses extended beyond their experiences of performing the interventions from a trial perspective. As the analysis was data-driven, we were able to identify data that supported our study objective in the context of the men's daily life.
Fifth, at what level of abstraction did the authors aim in identifying themes? As I read all of them, each is purely descriptive and thus generally best classed as categories and not themes. Refer to the editorial, "Confusing Categories and Themes" by J. Morse (2008).
Response: Thank you for providing the reference to the editorial. Our analysis was descriptive in that we identified semantic (surface level) themes within the data. We analysed the data in line with Braun and Clarke (as referenced in the manuscript) who use the term 'theme'. We, therefore, believe that we should continue using this term for consistency.
Might you revise your level of analysis? I also urge you to reconsider whether you have sufficient data to report certain themes.
Response: As mentioned above, our analysis was descriptive in that we identified surface level themes within the data. We acknowledge that the interviews were not in-depth as they were conducted as part of feasibility trial, which used the interviews to obtain data on trial processes as well as intervention adherence. We are, therefore, unable to perform an in-depth analysis on the data. We also acknowledge that data was not as rich for some themes and we have mentioned this as a limitation on page 10: "Data analysis was limited by the lack of depth in responses from the interviews. This is likely due to the interviews being part of data collection for a feasibility RCT, which assessed trial processes as well as intervention adherence." Most themes are described with a thin narrative and few quotes. The balance of narrative to data is typically indicative of the depth of development and representation of richness in the underlying data. Do the authors think they've adequately captured what the participants and data convey? I'd like to see a good deal more development here.
Response: As mentioned in our previous comment above, we acknowledge that the data was not indepth and we have mentioned this in the discussion section on page 10. We chose the quotes which best supported each of themes. This led to some quotes not being as 'rich in qualitative data' as others. There were additional quotes that we could have included but did not do so for space reasons.
Sixth, it is important to include in the method and the discussion section the information on how the rigor of the study was established. There are a number of references for this, including the seminal text by Lincoln & Guba (1985) and the article, "Critical Analysis of Strategies for Determining Rigor in Qualitative Inquiry" by J. Morse (2015).
Response: Thank you for providing these reference by Lincoln & Guba (1985) and Morse (2015). We did perform negative case analysis and have amended the data analysis section to include this on page 5: "Negative case analysis (i.e., identifying contradictory data) were used to broaden or confirm the interpretation of the themes and were resolved through discussion between the researchers and revisiting the transcripts." However, we were unable to perform other strategies to increase the rigour of the methods and have included the following text in limitations section on page 10: "…this study has several limitations. Data analysis was limited by the lack of depth in responses from the interviews. This is likely due to the interviews being part of data collection for a feasibility RCT, which assessed trial processes as well as intervention adherence. Therefore, further assessments of rigour would not have benefitted the data analysis." We believe it is more appropriate to state this in the discussion rather than the methods section.
Seventh, there are multiple discrepancies between the statements (or the names of the categories) and quotes presented to support the statements/categories in the results section.

Responses: We have made amendments to the results section in accordance with your comments below. We believe this has resolved the discrepancies.
For instance, the author wrote on page 6: "While other men described making extensive changes on starting the PBD within the trial." However, the quote that follows this statement is "As I say, I used to eat an awful lot of fruit and vegetables beforehand…" Response: This quote referred to a participant who ate significant more portions of fruit and veg a day than advised. We considered this an extensive change to his diet even though he ate a lot of fruit and veg before the trial.span style="font-family:Calibri; font-style:italic; -aw-import:spaces"> We understand that the initial part of the quote confuses the point we were trying to make. Therefore, we have amended the quote as follows on page 6: "…I found that I was really, sort of, [toning] myself up almost on fruit and veg. I think it said you had to eat 5 more portions of fruit and veg a day than normal, so I was getting up to, at some stages, about 20, I think, a day." P17, PBD and brisk walking Another example: the authors put the following statement as an example for the "physical limitations" category: "Most men also relied on good weather. There was also little motivation to walk elsewhere when the weather was bad. "…there were some days where it was a total wash-out, and you think, "Well, there's no point in even trying," you know. "I'll make this my quiet day"…" However, it seems that the above-mentioned example is more appropriate for the "external obstacles category." Response: On review of the themes, we agree that this statement would be more appropriate in the external obstacles subtheme. We have, therefore, moved this statement to the external obstacles subtheme.
Furthermore, the results section is difficult to understand as it lacks a clear picture overall. Therefore, the result section will benefit if the thematic map of the study is presented.
Response: Thank you for the suggestion of a thematic map. We have now included one in the manuscript, which we believe offers a clearer picture (Figure 1).
Eighth, how are the authors constructing the discussion for this manuscript? Typically, a strong discussion recapitulates the main findings in brief and then goes on to argue original contributions, areas where findings corroborate extant literature, where the findings conflict, and finally limitations. This progression allows the authors to draw clearly defined conclusions in line with the method and results and to outline implications. I do not see this structure in your discussion.
Response: We agree that restructuring the discussion section would be beneficial. We have, therefore, rearranging some of the paragraphs in line with the order of themes presented in the results section. We have also included two additional subheadings (asterisked below) in the existing structure of the discussion listed below: • Summary of findings* • Support with other studies*

• Strengths and limitations • Main implications and future research
Critically, many statements in the discussion section are not supported by the study result. For instance, the following statements in the discussion section are not supported clearly by the study results: (1) "The findings suggest that a diagnosis of cancer can provide an opportunity for men to make changes to their nutrition and PA." Response: We agree that this statement does not fit well with the study results. We have, therefore, removed it from the discussion section.
(2) "While some men followed the intervention guidelines, others made quite extreme changes to their diet, such as eating well-over the recommended daily intake of fruit and vegetables."

Response:
We have commented about this point in the discussion section that men could benefit from more education on eating practices on page 9.
(3) "The facilitators to change identified in this study (i.e. … self-efficacy with taking lycopene and exercising)"; We also amended a sentence in the conclusion of the abstract on page 2: "Future studies may benefit from theory-based interventions to support adherence to diet and PA behaviour changes in men diagnosed with prostate cancer." Ninth, the "strengths and limitations" section should be revised. First, it is unclear how the study's strength could be its "findings on the psychological, behavioural, and social factors associated with adherence to men's changes to their diet and PA". Usually, the strength of the study should be judged on the rigor of the study methodology; therefore, the researchers should specify on what quality criteria they based their research. Please see, Lincoln & Guba (1985)  Response: We have amended the strengths and limitations to relate specifically to the study methodology on page 10: "The strength of this study is that it has provided a thematic analysis of men making diet and PA changes soon after prostatectomy, which included a negative case analysis to support the rigour of the study." Furthermore, although in the result section the authors mentioned that "men across all the intervention arms believed that they maintained a healthy diet before being diagnosed" and the authors provided several quotes that described the men's healthy behaviors before the study, they did not mention this fact in the limitation section. However, it is well known that the new healthrelated behaviors' is easier to achieve in people with previous healthy behaviors compared to those without previous health behaviors; therefore, the factors that supported and hindered men's adherence to the nutrition and physical activity could be different among those with and without previous healthy behaviors.
Response: Thank you for highlighting this. We have added the following sentence to the limitations section on page 10: "In addition, men in all the intervention arms discussed that they were already maintaining a healthy diet and engaging in regular physical activities before their diagnosis. This could suggest that the current findings are limited to men more willing and able to perform these health behaviours." Reviewer: 2  8,10,11,12,20,21,22).
Thank you for incorporating a PPI group in your research. I comment on the paper as and where such issues arise.
Abstract Page 4 Line 22-expression and accuracy Participants were 17 men with a median age of 66 years, ….,who underwent surgery Why don't you use prostatectomy rather than surgery, as you did in you feasibility study? I would think it more appropriate for your readership at BMJ.

Response: We have changed surgery to prostatectomy.
Line 28-expression relationship of nutrition and PA with PCa…OR the impact of nutrition and PA on PCa… Response: We have changed relationship to impact.

Page 5 Line 5-expression
There is also evidence that high intake of… Response: We have amended this sentence on page 3: "High intakes of dairy products is also associated with increased prostate cancer risk [8,9]." Line 9ff updated references Many of the references cited are somewhat dated and while Kenfield is a seminal article, systematic reviews by internationally recognized exercise physiologists (e.g. Cormie et al. 2017 The impact of exercise on cancer mortality…)have updated these and are worth including.

Page 12
Line 3 …another qualitative study… Response: This has been corrected.
Interestingly, both the findings from these studies seem to suggest that these men prefer to blame things outside their control or not take responsibility for their actions ( I think the small sample size is a limitation that should be mentioned.
Response: Thank you for this suggestion. We have added the small sample size as a limitation on page 10.
Congratulations on a well conducted study on an important topic. My suggested changes are minor. However, I believe it is important in the background and in the discussion to incorporate other more recent studies against which to compare your own.
Response: Thank you for your helpful comments. We believe we have strengthened the background and discussion sections using your comments.
All the best with your future research. interest your manuscript which, in my opinion, gives interesting information regarding the perspectives of men with prostate cancer regarding the benefit of adopting a healthy lyfestyle. Although I have limited experience in qualitative research, I understand that the research has been conducted with rigor. I did not review the work the first time but it seems to me that you have made numerous changes that make the manuscript clearer and more complete for the readers. The only thing I ask of you is a very minor revision, that is to integrate into the discussion the results of a very recent study that has several points in common with your results, and can therefore support your arguments under discussion. this is the study by Bressi and colleagues published in February 2022 (PMID: 35194723), the results of which say that a high proportion of men are insufficiently active when diagnosed with PCa. More that 60% are obese. However, even when exposed to behavioral risk factors, they are not willing to change their lifestyle. This seems to confirm the incomplete awareness of the role that lifestyle has in the genesis of cancer and on health in general. I would suggest to add this very recent reference in the paragraph "Support with pother studies", because it can support the statement that you write in that paragraph. the same study also report barriers and facilitator to exercise, confirming that lack of time and bad weather act like barriers while the perception of a psychological and physical benefit associated to exercise are facilitaors. Thus, I believe this results (obtained in a cohort of 40 man with PC) can make your results stronger. Authors' response: Thank you for your positive comments.

VERSION 2 -AUTHOR RESPONSE
Reviewer comment: I did not review the work the first time but it seems to me that you have made numerous changes that make the manuscript clearer and more complete for the readers. The only thing I ask of you is a very minor revision, that is to integrate into the discussion the results of a very recent study that has several points in common with your results, and can therefore support your arguments under discussion. this is the study by Bressi and colleagues published in February 2022 (PMID: 35194723), the results of which say that a high proportion of men are insufficiently active when diagnosed with PCa. More that 60% are obese. However, even when exposed to behavioural risk factors, they are not willing to change their lifestyle. This seems to confirm the incomplete awareness of the role that lifestyle has in the genesis of cancer and on health in general. I would suggest to add this very recent reference in the paragraph "Support with other studies", because it can support the statement that you write in that paragraph.