Patients

Eligible patients were identified from the Canadian Spine Outcomes and Research Network (CSORN) registry, all from the Saint John Regional Hospital, a tertiary care center. Inclusion criteria were: adult surgical candidates (≥19 years) with lumbar spinal stenosis or lumbar disc herniation who had not previously undergone lumbar spinal surgery. A randomised contact order was generated using an online randomization tool, and patients were contacted sequentially in that order to avoid selection bias. Of the 87 patients identified in the registry, 64 met the eligibility criteria, as some had already undergone surgery between registry entry and the time of contact. Of these, 18 consented to participate. The remaining 46 patients did not participate for the following reasons: 23 were not interested in the study, 19 did not respond to the call, and 4 were unable to complete the online interview or could not travel to the interview location. No systematic data were collected on non-participating patients, and therefore a formal comparison between participants and non-participants was not possible.

This study protocol was approved by the Research Ethics Boards from the Horizon Health Network (2024−3370) and the University of New Brunswick (2024−109). All patients provided signed informed consent, and the study was conducted in accordance with the Declaration of Helsinki.

Data sources

Demographic and clinical variables were obtained from the CSORN database. Back and leg pain intensity were assessed with separate 0–10 Numeric Rating Scales [27] with a similar scale implemented to quantify leg tingling, burning, and numbness. Pain-related disability was assessed with the modified Oswestry Disability Index (mODI), with scores ranging from 0 to 100 [28]. Health-related quality of life was evaluated using the EuroQol Visual Analogue Scale (EQ-VAS), with scores ranging from 0 to 100 [29].

Interview topic guide

We conducted Data-Prompted Interviews, a qualitative method in which personalised data are used during the interview to stimulate reflection and discussion [30]. In this study, the data prompts were based on each patient’s self-reported physical activity behaviour, collected at the beginning of the interview, thereby helping patients to recall, interpret, and elaborate on their experiences. Interviews were structured based on the TDF [19,20], which has been used to identify factors influencing physical activity in different settings, including patients following coronary artery bypass graft surgery [23] and heart failure [24].

To ensure consistent identification of barriers and facilitators in this population, we developed a semi-structured interview topic guide based on the TDF (see S1 File). Each patient participated in a single, individual interview. The topic guide included open-ended questions tailored to each patient’s personalised physical activity data, with additional prompts to explore each theoretical domain [30]. The target behaviour was defined as “the things you do that contribute to achieving 150 minutes of moderate-to-vigorous physical activity per week”, in line with the Canadian 24-Hour Movement Guidelines for Adults [31].

The topic guide was developed through consensus among several researchers (E.B., S.D., J.G., J.H.) and a patient representative (A.L.), a person with lived experience of degenerative lumbar spine conditions. A.L. contributed to the development of the interview topic guide by providing feedback on the relevance and acceptability of the questions from a patient perspective, participated in the consensus meetings during the analysis phase, reviewed the categorisation of interview extracts into the primary domains confirming their alignment with the patient perspective, reviewed the final themes upon completion of the analysis, and presented the preliminary results at a conference. The topic guide was piloted in two mock interviews to ensure clarity, understanding of the questions, and effective use of the prompts. Additionally, to stimulate discussion and enhance understanding, an A4 sized prompt card featuring images of different physical activity categories was explained to patients and displayed during the interview (see S2 File). Interviews were designed to be patient-centred, allowing for the modification of questions throughout the study to improve both the clarity and overall experience for the patients.

Procedure

Patients were recruited between 21 August 2024 and 12 February 2025. Interviews were conducted between the same period. All patients provided written informed consent. The interviewers received training from an experienced qualitative researcher with expertise in health behaviour change and data-prompted interviews (S.D.). The training included simulated interviews followed by a pilot interview with a patient representative. All interviews were conducted by two male co-investigators (D.F., J.G.). D.F. was an undergraduate student researcher with no prior experience in qualitative research but with experience in clinical research involving this patient population. J.G. was an experienced physiotherapist and post-doctoral researcher with clinical and research expertise in patients with degenerative lumbar spine conditions. Neither interviewer was involved in the clinical care of the included patients, minimizing the potential for bias or assumptions related to prior clinical responsibilities. Prior to data collection, the interviewers were made aware of how their professional backgrounds and assumptions might influence in their interviews and interpretation of findings, as part of their training with S.D. Specifically, J.G.’s background as a physiotherapist with clinical expertise in degenerative lumbar spine conditions may have led to assumptions about the barriers and facilitators patients experience, potentially influencing the formulation of follow-up questions during interviews and the interpretation of response during coding. To mitigate this, reflexivity was actively encouraged throughout the research process: field notes were used to document assumptions during interviews, coding decisions were made under the supervision of S.D. and discussed until consensus was reached, and the use of the TDF as a predefined theoretical framework helped to structure the analysis and limit the influence of subjective interpretation. All interviews were conducted either in consultation rooms or online, based on patient preference, and were audio recorded, transcribed verbatim and anonymised. Field notes were taken to record emotional responses and non-verbal communication, and were also drawn upon during the analytical process to provide additional context. Prior to the interview, patients were informed that the interviewers were members of the research team with an academic interest in understanding barriers and facilitators to physical activity among patients awaiting spine surgery. No interviews were repeated. Patients received a modest honorarium for their participation.

During the interview, patients first completed the Godin-Shephard Leisure-Time Physical Activity Questionnaire, a brief, three-item questionnaire to assess their weekly exercise behaviours. They were asked, “During a typical 7-day period (a week), how many times on average do you engage in the following types of exercise for more than 15 minutes during your free time?” The exercise categories offered were “strenuous exercise”, “moderate exercise”, and “mild/light exercise”. Patients provided numerical responses for each category. Based on their score, patients were classified as “active” if they accumulated 24 points or more, “moderately active” if they accumulated 14–23 points, and “insufficiently active/sedentary” if they accumulated fewer than 14 points [32]. The questionnaire results were then used as a data-prompt to contextualise and initiate the discussion, helping patients reflect on their activity level relative to the recommended target of 150 minutes of moderate-to-vigorous physical activity per week. The second part of the interview involved applying the topic guide, with questions and prompts adapted to each patient. Patients were informed of their right to access and review their transcriptions to verify, validate, and make any necessary corrections; however, no requests were made.

Although qualitative studies do not require a predetermined sample size, a final sample of 18 patients is consistent with empirical evidence indicating that saturation is typically achieved within 9–17 interviews in qualitative studies with relatively homogeneous populations and narrowly defined objectives [33]. Data saturation was determined following the approach recommended by Francis et al [34]: after an initial set of 10 interviews were conducted and analyzed, three additional interviews were conducted and analyzed to assess whether new barriers or facilitators emerged. If new themes were identified, three further interviews were conducted and analyzed. This iterative process was repeated until no new themes emerged across three consecutive interviews, at which point data saturation was considered reached. Recruitment and analysis were conducted simultaneously to allow ongoing saturation assessment. Consensus regarding data saturation was achieved among three researchers (S.D., J.G., J.H.). The Consolidated Criteria for Reporting Qualitative Research (COREQ) were used to ensure that all aspects of the qualitative research were reported [35] (see S3 File).

Analysis

Transcriptions were stored as digital documents and analyzed manually, without the use of software. Coding was conducted line-by-line directly within the interview transcripts. In line with published TDF guidance [20], the use of qualitative data analysis software is not mandated; analytical rigour derives from the systematic and transparent application of the analytic framework rather than the tools used to organise data. Data from the interviews were analyzed using the 2005 version of the TDF [19], which comprises 12 domains: knowledge, skills, social/professional role and identity, beliefs about capabilities, beliefs about consequences, motivation and goals, memory, attention and decision processes, environmental context and resources, social influences, emotion, behavioural regulation, and the nature of behaviours. Qualitative data analysis was conducted using a hybrid deductive/inductive approach, following established recommendations for TDF-based qualitative research [20] involving the following steps:

1. Interviews were read multiple times by J.G. to ensure familiarity with the data.
2. The first two interviews were manually coded by J.G. under the supervision of S.D. to develop a shared coding strategy and resolve initial uncertainties. A content analysis approach was used to allocate responses deductively to relevant theoretical domains (multiple-domain allocation was possible).
3. The remaining interviews were manually coded by J.G., consulting S.D. whenever uncertainty arose regarding domain allocation. Coding was performed by a single researcher (J.G.). In all cases of uncertainty, consensus was reached through discussion between J.G. and S.D. This supervised approach and ongoing consensus process are consistent with established recommendations for TDF-based qualitative research [20], where coding decisions are guided by a predefined theoretical framework and consensus-based resolution of disagreements is considered an appropriate alternative to independent dual coding [36].
4. Data saturation was reached, as the final three transcripts revealed no additional barriers or facilitators beyond those already identified.
5. Summaries of domain codings were inductively grouped into subthemes by J.G. and reviewed in an initial meeting with S.D., J.H., and J.G. A second meeting was then held with E.B., A.L., D.F., J.H., and J.G., using the outcomes of the first meeting to further refine and finalise the subthemes. The patient representative (A.L.) participated in this process to ensure both the categorisation of interview extracts into the primary domains and the resulting subthemes aligned with the patient perspective.
6. Relevant theoretical domains were identified through consensus discussions among the aforementioned co-investigators (J.G., D.F., S.D., A.L., E.B., J.H.). Domain relevance was determined based on the following criteria when the domain acted as a barrier: (i) high frequency of specific beliefs, assessed by comparing the number of coded quotes across domains (see S1 Table), and (ii) presence of clear beliefs, defined as statements that were explicitly and unambiguously related to physical activity behaviour and expressed consistently across multiple patients. Frequency was used alongside belief clarity and consistency, in line with recommendations that domain relevance in TDF-based research should reflect the convergence of multiple indicators rather than quote count alone [20]. The (iii) existence of conflicting beliefs was also documented for information but was not used as a criterion for relevance. Facilitators were documented and reported within each key domain but were not used as criteria for determining domain relevance. The criteria used to classify TDF domains as key barriers are summarised in S4 File, which also provides additional details on domains that functioned as facilitators. Interviewed patients were not asked to provide feedback on the study findings.

Key and other domains

Domains were classified as key or other to prioritise those areas with the strongest perceived influence on physical activity behaviour in this population, which may inform both clinical practice and the development of future interventions. The criteria used to classify domains as key are described in the Methods section and summarised in S4 File. All 12 theoretical domains were identified. Within each domain, specific subthemes were identified and described as barriers, facilitators, or both, reflecting the context-dependent nature of each factor. Illustrative quotes are marked with (b) for barriers and (f) for facilitators.

Principal findings

This study aimed to identify the barriers and facilitators influencing engagement in physical activity among patients with degenerative lumbar conditions awaiting surgery. These patients typically experienced moderate to severe pain and substantial disability, shaping their daily functioning and perceptions of physical activity. Overall, patients described their experiences as a continuous negotiation between limitations and facilitators, reflecting the complex and dynamic interplay of personal, emotional, and contextual factors that characterise physical activity behaviour in this population.

Physical restrictions appeared to be a dominant barrier. Back and leg pain, together with leg tingling, burning, or numbness, often shaped daily life and influenced activity levels [37]. Patients typically engaged in physical activity when symptoms were less intense, yet pain remained unpredictable and often overwhelming. No meaningful differences in perceived barriers or facilitators emerged between patients with lumbar stenosis and those with lumbar disc herniation, suggesting that the behavioural mechanisms identified here may apply broadly across degenerative lumbar conditions. Fear of symptom exacerbation frequently led to avoidance or modifications [38], creating a cycle where symptoms restricted participation and inactivity further contributed to deconditioning. This pattern reflects a self-reinforcing experience where multiple TDF domains coexist. Specifically, pain and functional limitation co-occurred with diminished beliefs about capabilities, negative beliefs about consequences, an erosion of the physical skills needed to remain physically active, and persistent emotional distress. Impaired balance, reduced endurance, and comorbidities further limited engagement. At the same time, patients expressed clear benefits of physical activity, such as better mood, sleep, weight control, and strength [39], yet these were often overshadowed by the fear of aggravation, suggesting that interventions need to address not only physical capacity but also patients’ beliefs about the safety and consequence of physical activity. Guidance from health professionals played a pivotal role: lack of instruction generated uncertainty and fear, whereas appropriate guidance increased confidence and safety.

Contextual circumstances also shaped engagement. Access to safe spaces and natural environments facilitated activity, while financial limitations, inadequate facilities, and reduced healthcare access acted as barriers. Household and work responsibilities had a dual role, serving as the main source of activity for some but leaving others with little time or energy for it. Weather conditions were similarly ambivalent, as poor weather discouraged outdoor activity and, for patients already managing balance difficulties, heightened the perceived risk of falls and injury.

Emotional burden was closely tied to symptoms. Living for years with the condition was often accompanied by reports of persistent sadness, frustration, and fear, which discouraged participation [39]. Nevertheless, many patients described a marked improvement in mood, satisfaction, and well-being following physical activity, highlighting the tension between emotional barriers that preceded physical activity and the benefits experienced once it was completed.

Given the chronic nature of the condition, patients developed personalised, experience-based strategies to cope with symptom fluctuations and remain active within their physical limits [40], largely unaware of formal physical activity guidelines. Planning and monitoring activities further promoted adherence and fostered motivation by helping patients track their progress and strive for improvement over time. Patients tended to schedule activity at moments when symptoms were lower, integrating symptom management into their daily routines in a way that reflects the close relationship between behavioural regulation and habit formation.

Social influences generally acted as facilitators. Encouragement from family, friends, and healthcare providers supported engagement, although overprotection from relatives and conflicting advice from health professionals sometimes created barriers. Motivation and personal goals facilitated many patients to engage in physical activity despite their limitations [41], and this persistence is itself a clinically relevant finding, suggesting that the desire to remain active often endures even when capability is reduced. Daily routines and habits were seen as facilitating the integration of physical activity into everyday life, whereas disruptions to routines made re-establishing activity seem challenging, underscoring the importance of continuity and habit formation in this population.

Strengths and limitations

This study had several strengths. It provided a patient-centered perspective on engagement in physical activity in individuals with degenerative lumbar spinal conditions awaiting surgery, including patients regardless of personal or contextual circumstances. Semi-structured interviews enabled a comprehensive exploration of barriers and facilitators [30], and the use of the TDF ensured a structured and theory-driven analysis [19]. The inclusion of both lumbar stenosis and lumbar disc herniation captured a broader and clinically meaningful range of experiences, enriching the findings. A patient representative contributed to the development of the topic guide and assisted with the interpretation of the findings, and the topic guide was piloted twice before recruitment.

Some limitations should be acknowledged. First, the study was conducted at a single center, which, combined with the lack of racial and ethnic diversity, may restrict the transferability of the results to different healthcare systems and more diverse populations. The recruitment process may also have been subject to self-selection bias, as patients more interested in physical activity might have been more likely to volunteer. The inclusion of two different diagnoses could have introduced variability in experiences, and the imbalance between lumbar stenosis and lumbar disc herniation cases may have influenced the relative weight of perspectives. Reliance on self-reported data also entails a risk of recall or response bias [42], and the absence of objective measures of physical activity, such as accelerometers, means that reported activity levels could not be verified. The study was based solely on patient interviews, without triangulation from other sources such as healthcare providers or observational data, which limits the depth and corroboration of the findings. No systematic data were collected on non-participating patients, precluding a formal comparison between participants and non-participants and limiting our ability to assess the representativeness of the sample. Finally, formal inter-rater reliability metrics were not calculated. Although coding decisions were guided by the prespecified categories of the TDF framework, the absence of a formal index is acknowledged as a limitation.

Relation to other studies

Our findings are consistent with previous qualitative research in patients with lumbar stenosis and lumbar disc herniation. As in earlier work [17], symptoms, particularly pain, emerged as a key barrier, shaping both participation in structured rehabilitation and daily activities. The self-reinforcing cycle identified here, whereby pain drives inactivity, which in turn worsens deconditioning and comorbidities, extends previous observations by illustrating the mechanisms through which this cycle operates across multiple behavioural domains simultaneously. Mixed perceptions of physical activity were also consistent with prior studies: while some patients experienced physical and mental benefits [17,43], others considered it of limited personal value, especially when presented as an alternative to surgery [43]. By broadening the focus from prehabilitation to physical activity in general, our study extends these observations and shows how barriers and facilitators operate not only in hospital-based programs but also in everyday life.

Professional guidance emerged as a key facilitator across studies. One study highlighted the importance of structured rehabilitation [17], whereas our patients and another study in a rehabilitation setting [43] emphasised the need for consistent and coordinated information across different professionals and contexts. The finding that patients relied primarily on personal experiential knowledge rather than formal guidelines adds a new dimension, suggesting that improving guideline awareness and professional communication may be particularly impactful in this population.

Environmental and motivational factors were similarly with prior work [17]. Lack of time, financial limitations, and restricted access to safe environments hindered physical activity, and our study further highlighted the ambivalent role of household and work responsibilities. Motivation fluctuated over time, and our findings add that its persistence despite physical barriers is a clinically meaningful finding that interventions should seek to harness.

Findings from broader reviews in musculoskeletal [44,45] and chronic low back pain [46] populations were also largely consistent with ours. Pain aggravated by comorbidities was a prominent barrier, and both fear of injury and perceived benefits were reported across reviews. Relatives and healthcare professional emerged as important facilitators, and the role of positive emotions and habit formation was similarly underscored.

Implications and future research

Beyond identifying individual barriers and facilitators, the findings suggest the presence of distinct conceptual profiles of engagement in physical activity, based on patterns observed across interviews. Three broad groups can be described: patients with low physical fitness and high perceived barriers who are unlikely to change their behaviour prior to surgery; patients with higher fitness and strong facilitators who can maintain activity despite symptoms; and patients with low fitness but openness to change if barriers and facilitators are adequately addressed. The identification of these profiles was not a formal aim of the study. These profiles are not intended as fixed categories but as a heuristic to illustrate population heterogeneity and inform the starting point for individualised clinical assessment, recognising that any intervention must ultimately be tailored to the individual, as discussed below. Future research should explore whether these profiles can be validated and whether they predict outcomes or response to intervention.

Our study suggests that personal, emotional, and contextual factors must be considered when promoting physical activity in patients awaiting surgery for degenerative lumbar spinal conditions. Given the multifactorial nature of the barriers and facilitators identified, prehabilitation interventions should be tailored to each patient’s perceived barriers and facilitators, their beliefs about physical activity and its consequences, and their emotional relationship with it. This initial conversation would also benefit from addressing symptom-specific concerns, distinguishing between lumbar and leg symptoms, and to explore patients’ own views on what kind of support would be most helpful. Such an approach aligns with the TDF domains identified as key in this study, and would allow clinicians to tailor recommendations to each patient’s specific profile. Specific behaviour change techniques and general approaches could then be mapped onto each patient’s profile. Beliefs about capabilities and consequences may be addressed through motivational interviewing, which incorporates strategies such as reframing physical activity as safe and beneficial [47]. Emotional barriers suggest the value of acceptance-based strategies to manage fear and frustration [48]. Skills-based limitations support the use of supervised, graded physical activity programs that rebuild confidence through progressive practice [49].

More broadly, intervention should aim to address the self-reinforcing cycle of pain, inactivity, and deconditioning by supporting patients in identifying safe and tolerable activity levels and providing clear, consistent guidance from healthcare professional [50]. Given the prominence of environmental barriers, home-based or hybrid delivery formats should be considered alongside strengthening individual facilitators such as social support [51]. Building and maintaining daily routines and habits appears particularly important, given how difficult patients found it to re-establish activity once disrupted. Structured action planning could support habit formation during the preoperative period, while self-monitoring tools such as activity trackers may help patients during the initial stages before behaviours become automatised [50].

These findings can inform surgeons, family doctors, physiotherapists, and relatives in supporting patients, and may guide healthcare managers in prioritizing strategies to promote physical activity in this population. Future research should include qualitative studies exploring the perspectives of surgeons, family doctors, and physiotherapists on preoperative physical activity. Feasibility studies are needed to test interventions that address the barriers and facilitators identified here, ideally including a control group and objective measures of physical activity such as accelerometers. Given that this population is typically older, future studies should test age-adapted programs that account for comorbidities and functional limitations while promoting safety and adherence. Interventions should also aim to strengthen outcomes highly relevant in older adults, such as independence in daily activities, fall prevention, reduction of frailty, and enhanced social engagement. Future research with larger, diagnosis-specific samples should also examine whether the barriers and facilitators identified here differ between lumbar stenosis and lumbar disc herniation, given that the present study was not powered to detect such differences.