1 INTRODUCTION

Compromised bone health is often associated with depression and chronic pain. There is evidence to suggest a bidirectional relationship between bone health and depression. Specifically, researchers have suggested deleterious effects of depression on bone health through bone loss and increased fracture risk.¹ The causal mechanism is unclear but possibly due to depression altering concentrations of many hormones that affect bone formation and/or resorption, such as cortisol.¹ Medications taken for depression may also be detrimental to bone.^2-4 For example, serotonin can influence bone metabolism as serotonin receptors and transporters are present in osteoblasts and osteoclasts.⁴ At the same time, fragility fractures often precede the onset of depression.^{5, 6} In one cohort study, 10% of individuals reported depressive symptoms after a hip fracture.⁶

There is also evidence of a bidirectional relationship between bone health and chronic nonfracture pain. Medications taken for chronic pain, such as nonsteroidal antiinflammatory drugs (NSAIDs) and opioids, including codeine, can have negative bone health effects, including increased fracture risk, increased risk of falling and decreased bone mineral density.^{2, 7, 8} The mechanism underlying this effect is unclear.^{7, 8} Fragility fractures, such as vertebral fractures, can also result in long-term pain at the site of the fracture,^{9, 10} which is thought to be due to physiological changes in the spine and loss of height.^{11, 12} Other types of fractures have also been shown to result in reports of long-term pain at the site of the fracture,^{13, 14} possibly due to the development of arthritis at the fracture site.^{15, 16} Little is known about whether fragility fractures have an effect on chronic pain conditions unrelated to the fracture.

Few researchers have examined the co-occurrence of depression and chronic nonfracture pain in patients who have had a fragility fracture, and what factors may worsen these conditions. In one randomized controlled trial of patients with a wrist fracture, the most common reported comorbidities included osteoarthritis (34%–42%) and depression (11%–15%).^{17, 18} Although not determined in that study, it is expected that reports of chronic joint pain would have been common in most of the patients reporting osteoarthritis. Kelly-Pettersson and colleagues reported that approximately 22% of hip fracture patients reported depressive symptoms at baseline,¹⁹ but it is unclear whether the depressive symptoms existed at the time of hip fracture or occurred after the hip fracture was sustained.

Our purpose was to examine experiences of depression and chronic nonfracture pain in patients who presented with a fragility fracture and other comorbidities. A fragility fracture is one that occurs after a slip, trip or fall from standing height or less.²⁰ This is distinct from high-trauma fractures, which typically result from motor vehicle crashes and falls from greater than standing height.²¹ Fragility fractures are a sign of poor bone quality and a predictor of future fractures.^{22, 23} A Fracture Liaison Service is a model of care that identifies individuals over the age of 50 years who present to a hospital with a fragility fracture to ensure they receive a fracture risk assessment and bone health treatment according to current clinical practice guidelines.²⁴ Treatment includes pharmacological and nonpharmacological treatment, such as vitamin D supplementation, adequate calcium intake, and exercise.²⁵ Pharmacological treatment has been shown to reduce refracture^26-28 and mortality rates.²⁸ We anticipated that our investigation might partially contribute to our understanding of the modest uptake of bone health recommendations after a fragility fracture, such as taking bone active medication.^{29, 30} In one systematic review, less than 35% of individuals initiated medication after a postfracture intervention.³¹ Depression has been shown to be associated with poor adherence to bone active medication.³²

Using a qualitative approach and relying on patients' perspectives, we sought to understand the complexity and meaning of bone health in comparison to that of chronic pain and depression. Previous qualitative studies in bone health have demonstrated that patients have limited knowledge of bone densitometry and bone health³³ and are unclear about testing and treatment recommendations.^{34, 35} Studies have shown that the circle of care for those with fragility fractures is disrupted at vital communication junctures³⁶ and that patients perceive inconsistent messages within, and across, primary care providers and bone specialists.³⁷ One qualitative synthesis demonstrated that individuals create meaning of an osteoporosis diagnosis based on self-perceived risk, self-perceived severity of osteoporosis and self-perceived health.³⁸ Authors have also reported that patients do not connect their fractures to bone health,^{39, 40} that many patients classified as ‘high risk for future fracture’ do not believe they are high risk,⁴¹ and that having caregiving responsibilities affect the management of fragility fractures.⁴² Specifically, our objectives were to examine: (1) the experience of depression and chronic nonfracture pain in patients with a fragility fracture; and (2) the effects of the fracture on depression and chronic pain.

2 METHODS

This study was part of a larger study examining bone health management in patients with a fragility fracture who reported one or more comorbidities. It involved a 2-year qualitative investigation guided by phenomenology as conceptualized by Giorgi and Wertz (the ‘Duquesne school’).^43-45 Phenomenology was relevant to our study design as it emphasizes the importance of direct experiences, perceptions and actions.^{46, 47} Ethical approval for the study was received by Unity Health Toronto (REB#: 14-301).

We recruited patients from a Fracture Liaison Service serving approximately 430 patients annually in a Canadian urban hospital.^{48, 49} Consistent with phenomenology, we employed criterion sampling⁵⁰ where eligible individuals were English-speaking men and women, 45+ years old, who self-reported currently taking prescription medication for at least one additional chronic health condition (bone health could not be the only chronic condition for which patients were taking medication). To aid with recruitment, we created a list of eligible chronic health conditions, such as arthritis and high blood pressure, from two Canadian sources^{51, 52} and expanded it to include other conditions such as those causing secondary bone loss. Participants were not eligible if they only reported conditions on the list of excluded conditions (see Table 1). Patients who exhibited cognitive difficulties that might compromise their ability to give informed consent or participate in an interview were not approached.

A fracture prevention coordinator (V. E.-G.) identified potentially eligible participants and determined their interest in participating in our study. The coordinator screens patients with fragility fractures and facilitates treatment with bone active medication according to the Canadian Clinical Practice Guidelines for the Diagnosis and Management of Osteoporosis.²⁵ A female study coordinator (L. F.) with 13 years of qualitative research experience and a Bachelor of Arts degree telephoned patients to obtain research consent, confirm eligibility based on patients' self-report of existing chronic health conditions, and schedule baseline interviews (60–90 min). Follow-up interviews (approximately 30 min) were scheduled approximately 1 year after the baseline interview. We conducted the majority of baseline interviews in patients' homes within 6 weeks of their fracture and follow-up interviews by telephone. Topic areas in the baseline interview guide addressed the recent fracture and patients' experiences with bone health and their other health conditions, such as depression and conditions that might be associated with chronic pain, including the medications taken for these conditions. For example, we asked patients about the condition(s) of concern to them at the moment and about the importance of bone health in comparison with their other health conditions. We also asked how bone health recommendations might affect the management of these other conditions. The second interview focused on changes, if any, to patients' experiences with bone health and their other health conditions as well as any changes in medication regimens (see Table 2).

In phenomenological studies, the outcome is a description of the essence, or structure, of what is perceived and experienced across individuals.⁴⁴ Iterative analyses of the data began after the first two baseline interviews were conducted. Preliminary codes were identified and then revised as more interviews were conducted. Two individuals (J. E. M. S., L. F.) with qualitative expertise analysed the transcripts independently and met regularly to develop and finalize a coding template, which was then applied to all transcripts using NVivo.⁵³ To promote a comprehensive examination of the data,⁵⁴ all transcripts were coded by the two coders. The analysis was guided by Giorgi's procedures.⁴³ To organize the data, meaning units, or codes, were documented in the margins, and codes relevant to our objectives were grouped together. We reflected on the relationships among the codes and developed themes that were supported by direct quotations from participants.⁵⁵ Discussions regarding analysis and interpretation of the data were reviewed by the study team as data collection and analysis progressed. Multiple thematic possibilities were considered by the team (a critical appraisal strategy referred to as imaginative variation^{44, 45}) and a consensus was sought on the final themes based on the relevance,⁵⁶ novelty and clinical significance of the findings. Cases that did not fit with our general findings were discussed and reported to promote transparency in the research process.⁵⁷

Consistent with GRIPP2-SF,⁵⁸ a public contributor was a formal part of the research team from the beginning of the study. He was involved in the grant-writing process, which included the drafting of the interview guide and refining the focus of the research questions. Along with all team members, this individual was involved in discussions about the data and the findings reported (see Table 3).

3 RESULTS

We recruited 26 patients (5 men, 21 women) aged 45–84 years old with the hip (n = 5), wrist (n = 15), spine (n = 1), shoulder (n = 1), other (n = 2) or multiple fractures (n = 2). This sample size met recommendations for phenomenological studies.^{44, 47} Thirteen reported a history of previous fractures (11 women, 2 men). According to the Canadian clinical practice guidelines, 16 were deemed high risk for future fracture due to previous fractures and/or a current hip or spine fracture.²⁵ Four participants could not be located 1 year later for a follow-up interview. In the majority of cases (n = 17), baseline interviews were conducted in patients' homes. The remaining nine interviews were conducted by telephone (n = 4), in the interviewer's office (n = 3) and at a location suggested by the participant (n = 2).

Participants presented with one to seven chronic health conditions in addition to compromised bone health. The most common chronic health conditions reported were depression (n = 12), high blood pressure (n = 11), arthritis (osteoarthritis and rheumatoid arthritis combined) (n = 10), high cholesterol (n = 9) and hypothyroidism (n = 7). Sixteen participants reported conditions resulting in chronic nonfracture pain, such as arthritis, ‘nerve pain’, sciatica and spinal stenosis (7 were currently taking prescription medication for the pain, including analgesics, opioids, NSAIDs and anticonvulsants for neuropathic pain) and 12 participants reported depression (9 were currently taking prescription medication for depression, including selective serotonin reuptake inhibitors, serotonin and norepinephrine reuptake inhibitors, norepinephrine and dopamine reuptake inhibitors, benzodiazepines, anticonvulsants and atypical antipsychotics). In total, 21 of the 26 patients reported depression and/or chronic nonfracture-related pain, of which seven reported having both depression and chronic pain. The majority of participants with depression or chronic pain reported giving up physical and social activities due to these conditions.

Two themes were consistent, based on our analysis: (1) depression and chronic nonfracture pain overshadowed attention to bone health; and (2) the fracture exacerbated reported experiences of existing depression and chronic pain. We did not observe any differences between men and women, or in participants having different comorbidities or other characteristics, related to these two themes.

4 DISCUSSION

In our study of patients with a fragility fracture and other health conditions, most patients reported depression or chronic nonfracture pain and several lived with both depression and chronic pain. These two conditions were considered to be all-consuming and were prioritized over bone health. Patients also indicated that the fragility fracture had worsened their depression and chronic pain. Our results overall suggest that depression and chronic pain are often present in the setting of fragility fracture, that they have important consequences, and that they may be under-recognized and undertreated, especially chronic pain.

Our qualitative study was not designed to assess prevalence, and may have inadvertently selected for individuals at risk for mental health conditions (to be eligible, participants had to report one or more comorbidities in addition to compromised bone health). Nevertheless, the high proportion of patients in our study reporting depression is worrisome because depression has been associated with poor adherence to osteoporosis medication³² and increased falls.⁵⁹ The experience of sustaining a fracture has been documented to precede depression,^{5, 6} so it is not surprising that existing depression may be exacerbated by the fracture. However, our results suggest that the potential implications of these relationships may be overlooked in clinical settings and may have negative consequences for bone health. The relationship between depression and bone health is clearly complex and the themes identified in our research need additional attention in future studies. For example, future studies should explore more explicitly individuals' perceptions of the effort and energy devoted to bone health in the face of depression and chronic nonfracture pain, as well as the emotional implications of dealing with all three.

Older adults with chronic pain often alter or reduce their social and physical activities in some way to avoid pain.⁶⁰ This is a rational response to prevent further pain but it might paradoxically result in deconditioning and social isolation.⁶⁰ A reduction in physical activity has implications for postfracture recovery, which often involves physiotherapy exercises. As demonstrated by our findings, social isolation may also aggravate existing depression. Various safe medication treatments exist for chronic pain in older adults but less than one-half of patients with chronic pain in our study reported taking medication for their pain. This is perhaps not surprising, given that older adults are reluctant to take pain medication for conditions such as osteoarthritis, and when they do, they take the pain medication at a lower dose or frequency than prescribed.⁶¹ We did not ask participants if they had declined offers of pain medication or if they had not been offered medication for their chronic pain so cannot comment on why less than one-half reported taking medication for their pain. Future studies are needed to examine in detail changes in pain and pain medication. Chronic pain is not easily treated and treatments may become less effective over time, such as in the case of opioid-induced hyperalgesia.⁶²

Our findings have important implications for clinical practice. Treatment for fractures involves a combination of nonoperative and surgical management, pain management and rehabilitation. A fracture event is typically not considered an opportunity to investigate mental health issues or other sources of chronic pain. Yet the perspectives of our patients encourage health care providers to ask initial questions about depression and pain, and taking other steps as appropriate, including referral to mental health care. This might add time and complexity, but we anticipate that it would allow exploration of issues faced by patients who have a fracture, and also promote more successful recovery from the fracture and ongoing bone health management. Further, patients who have had a fragility fracture may not recognize that fractures, especially those that result in challenges to mobility, might worsen their mood and pain. From a holistic perspective, we suggest that health care providers help patients appreciate that bone health treatment prevents subsequent fractures, which, in turn, may prevent the worsening of unpleasant subjective states that matter to patients.

The main limitation of our study is that the downstream effects of the fragility fracture on depression and pain remain unclear as we did not follow patients for longer than 1 year. We did not collect information on adherence to bone health recommendations so cannot comment on whether pain and depression affected the uptake of these recommendations. We also did not collect information on body mass index so cannot comment on the link between it and bone health. Further, we did not ask whether patients were depressed or had chronic pain before previous fragility fractures. A diagnosis of depression and chronic pain was based on self-report. We did not measure depression (or depressive symptoms) or pain and we did not determine the course of symptoms so cannot comment on the severity of these conditions. However, the reported prescriptions for medication in patients with depression and chronic pain suggest these conditions were diagnosed by a health care provider and the quotes we have provided demonstrate the apparent impact of these conditions. We conclude that these quotations support the findings and promote that the study maintains a patient perspective. Another strength of our study was the use of a phenomenological approach, which was well-suited to our objectives to collect data on patient's experiences. Finally, two independent qualitative researchers coded and analysed the data and we drew on the expertise of the research team, some of whom have previously collaborated in investigating the relationship between depression and pain (J. E. M. S., M. G., S. T., G. H.).^{14, 61, 63, 64}

5 CONCLUSION

Pre-existing depression and chronic pain appear to be common in individuals with a fragility fracture. Patients expressed that experiences with depression and pain were prioritized over bone health and in fact, worsened as a result of the fracture. Health care providers who treat fragility fractures might ask their patients about existing depression and chronic pain and take appropriate steps to address patients' more general emotional and physical state. Interventions might emphasize that better bone health can promote not only physical but also psychological well-being.

AUTHOR CONTRIBUTIONS

Joanna E. M. Sale made substantial contributions to conception and design and analysis and interpretation of the data, drafted and revised the manuscript critically for important intellectual content, approved the version of the manuscript submitted and agreed to be accountable for all aspects of the work. Monique Gignac made substantial contributions to conception and design and analysis and interpretation of the data, revised the manuscript critically for important intellectual content, approved the version of the manuscript submitted and agreed to be accountable for all aspects of the work. Lucy Frankel made substantial contributions to the acquisition of data and analysis and interpretation of the data, revised the manuscript critically for important intellectual content, approved the version of the manuscript submitted and agreed to be accountable for all aspects of the work. Stephen Thielke made substantial contributions to the analysis and interpretation of the data, revised the manuscript critically for important intellectual content, approved the version of the manuscript submitted and agreed to be accountable for all aspects of the work. Earl Bogoch made substantial contributions to conception and design and analysis and interpretation of the data, revised the manuscript critically for important intellectual content, approved the version of the manuscript submitted and agreed to be accountable for all aspects of the work. Victoria Elliot-Gibson made substantial contributions to conception and design and analysis and interpretation of the data, revised the manuscript critically for important intellectual content, approved the version of the manuscript submitted and agreed to be accountable for all aspects of the work. Gillian Hawker made substantial contributions to conception and design and analysis and interpretation of the data, revised the manuscript critically for important intellectual content, approved the version of the manuscript submitted and agreed to be accountable for all aspects of the work. Larry Funnel made substantial contributions to conception and design and analysis and interpretation of the data, revised the manuscript critically for important intellectual content, approved the version of the manuscript submitted and agreed to be accountable for all aspects of the work.

CONFLICT OF INTERESTS

The authors declare that there are no conflict of interests.