Bone 13, S27-S31 (1992) Printed in the USA. All rights reserved.
Copyright
The Clinical Consequences
8756-3282/92 $5.00 + .OO o 1992 Pergamon Press Ltd.
of Vertebral Compression
Fracture
S.L. SILVERMAN.
The Osteoporosis Medical Center, Beverly Hills, California; Wadsworth VA Medical Center and the Rheumatology Division, UCLA School of Medicine, Los Angeles, California, USA. Address for correspondence and reprints: Dr S. Silverman, California, 90210, USA.
The Osteoporosis
Abstract
consequences of vertebral compression fracture. This paper reviews the relationship between VCF as a radiographic and clinical event and looks at the clinical consequences of VCF as it is known.
Vertebral compression fractures (VCFs) may be defined radiographically or as a clinical event. The prevalence of these fractures in women aged 50 and over has been estimated at 26% when defined as a reduction in vertebral height >15%. Retrospective reviews of case records have shown a clinical detection rate of VCF in white women of 153/100,000 person years. Of these clinically detected VCFs, 84% were associated with pain. VCF may be defined as a clinical event characterised by loss of height and acute pain. The pain of acute fracture usually lasts 4 to 6 weeks with intense pain at the site of fracture. Chronic pain may also occur in patients with multiple compression fractures, height loss and low bone density but is probably due to structural changes or osteoarthritis. Radiographic VCF may not be symptomatic. The greater the deformity, the greater the likelihood of pain and disability. As height is lost, patients experience discomfort from the rib cage pressing downward on the pelvis. Patients develop a thoracic kyphosis, a lumbar lordosis, and a protuberant abdomen with prominent horizontal skinfold creases. The reduced thoracic space may result in decreased exercise tolerance and reduced abdominal space may give rise to early satiety and weight loss. Sleep disorders may also occur. Patients lose self esteem. Self care may become difficult. They are often depressed. They become fearful of further fracture. They have distorted body image and poor health perception. Patients with one vertebral fracture are at increased risk of peripheral fracture and further vertebral fracture. The aims of acute management are to reduce symptoms and mobilise the patient as quickly as possible. Long-term goals are to maintain or increase skeletal mass and improve mobility and function. Key Words: Osteoporosis fracture - Disability
- Compression
Medical Center, 450 N Bedford Drive, Suite 111, Beverly Hills,
The Incidence Fracture
and Prevalence
of Vertebral
Compression
Vertebral compression fracture may be defined either as a radiographic event or as a symptomatic clinical event. Radiographic VCFs have been defined according to the presence of a crush or wedge fracture and, recently, according to objective criteria relating to percentage loss of vertebral height as assessed on X-ray (eg. reduction in vertebral height >15%). Vertebral deformity scores were then developed to summarise radiological observations in multiple vertebrae. Recently, population standards for specific vertebral heights have been collected and individual vertebral heights can be compared to these norms. The prevalence of radiographic VCF is high. Using a reduction in vertebral height of ~15% as a criterion for VCF, Melton et al reported a prevalence of VCF in Rochester, Minnesota in 1981-1983, as high as 26% in women aged 50 and over (Melton et al, 1989). The corresponding radiographic incidence of vertebral fracture rose from 5/1,000 person years (py) in women aged 50-54 years to 29.6/1,000 py in women aged 85 years and over. Most of these radiographic changes are, however, asymptomatic and are therefore not clinically diagnosed. Cooper et al, while studying Rochester men and women of all ages, found an age- and sex-adjusted incidence rate for clinically diagnosed vertebral fracture of 123/100,000 py for the years 1985-1989; an incidence that is approximately 35% of that estimated from the prevalence data (Cooper et al, 1992). The age-adjusted incidence in women (153/100,000 py) was almost twice that in men (81/100,000 py). The incidence in women is comparable to that for hip fracture, which has an overall ageand sex-adjusted rate of 114/100,000 py. Not all patients with clinically diagnosed vertebral fracture are symptomatic. In the study by Cooper et al, 16% of diagnoses were made incidentally in the course of radiographic investigations into unrelated disorders or as part of a workup in a high risk patient, on corticosteroids, for example. Of all the clinically diagnosed VCFs, 84% were associated with pain. When broken down further, it was shown that VCFs associated
fracture - Vertebral
Introduction The incidence of symptomatic vertebral compression fracture (VCF) is high and is probably similar to that of hip fracture. However, there is very little detailed knowledge of the clinical s21
S.L. Silverman: Clinical consequences of vertebral compresion fracture
S28
with severe trauma (traffic accidents or falls from greater than standing height) were always symptomatic, while only 81% of the moderate to minimal traumatic VCFs were associated with back pain. This included spontaneous VCFs that accounted for 46% of the latter total. Not all patients with symptomatic VCF are clinically diagnosed. The pain may be of short duration and responsive to analgesics, so it is likely that many vertebral compression fractures are not clinically diagnosed by physicians. Indeed some may be seen by chiropractors and masseuses, thus bypassing the attention of the physician (Ross et al, 1991). Fig. 1 shows the incidence of VCF.
Radiographic
Clinically
VCF
diagnosed
1531100,000
VCF
py
\
/
(Nelson et al, 1991). They found that only 35% of new VCFs
(15% reduction in vertebral height) were painful, while only one in five were associated with a positive bone scan (positive bone scans were associated with ~25% reduction in vertebral height). Nelson et al. therefore, suggested that more stringent criteria be used to diagnose VCF and that this type of fracture be defined only as a clinical event. Further complicating the radiographic diagnosis of VCF in this study was the observation that vertebral height loss may be transient in patients in whom the percentage loss was less than 20-25%. Using a morphometric criterion of a 210% change in anterior, middle or posterior height, a ‘rebound’ phenomenon was observed in 54% of vertebrae with reduction in middle height, 28% with reduction in anterior height and 42% with reduction in posterior height. Lyritis et al studied the natural history of VCF in 210 patients (Lyritis et al, 1989), and reported two types of VCF. In the first, pain was acute and severe, improving gradually, with obvious radiographic deformity at onset. In the second, the pain was less severe and of shorter duration; however, a new attack of pain occurred within 6 to 16 weeks and often recurred for a period of 6-18 months. With the second type of pain, radiographic deformity was not evident during the first attack but wedging gradually developed over the next few months. This study suggests heterogeneity in the clinical manifestation of vertebral fractures. The diagnosis of symptomatic VCF is often delayed. Cooper et al, while studying the Rochester population, found that the median duration of symptoms before fracture diagnosis was 7 days, although in 25% of patients, symptoms had persisted for one month or longer before the diagnosis was made (Cooper et al, 1992). Fractures following severe trauma (a traffic accident or a fall from greater than a standing height) were diagnosed more rapidly. The Pain of Acute Vertebral Compression
Fig. 1. Incidence of vertebral compression
fracture.
The Relationship Between Vertebral Compression as a Radiographic and a Clinical Event
Fracture
Recent data from two large prospective studies have provided insight into the clinical correlates of new VCF as defined radiographically. Watts et al compared the radiographic and clinical manifestations of new VCFs in a US multi-centre study into the effects of etidronate (Watts et al, 1991). Four out of six patients (67%) with two or more new VCFs had height changes (>I cm) and five out of six patients (83%) had acute pain. All three of the patients with one new VCFs and greater than 40% reduction in vertebral height had height changes, while two out of the three patients (67%) had pain. Of the 10 patients with one new VCF and 25-40% reduction in vertebral height, seven (70%) had height loss and pain. Of the six pateints with one new VCF and 20-25% reduction in vertebral height, three (50%) had height loss and one (16%) had pain. This study suggests that the clinical measures of height loss and, to a lesser extent, acute pain are sensitive indicators of new VCF with moderate to severe deformity. In a large study investigating the effect of fluoride on VCF, Nelson et al reviewed the clinical significance of new VCF
Fracture
Acute VCF pain usually lasts from 2 weeks to 3 months and is associated with intense, deep pain at the site of fracture, with palpatory tenderness over the site. Leidig et al studied pain in patients hospitalised due to VCF (Leidig et al, 1990). The pain was described as intermittent or chronic, deep, worse with prolonged posture, bending, walking or standing. Pain was moderate to severe and sometimes intolerable. Pain was relieved by lying down, heat, diversion and pain relievers in that order. Pain was associated with physical limitation in bending, arising and dressing. Acute VCF pain is felt deeply and diffusely in the back, is worse with motion and reduced by rest. The patient is in discomfort while sitting or standing, and an acute episode may be so painful that the patient may lie on the floor unable to move. The spine is tender to deep palpation and to percussion, and paraspinal vertebral muscle spasm is often present. Radiculopathy is not uncommon with compression fractures and may cause either unilateral or bilateral pain, radiating anteriorly along the rib of the affected nerve root. Acute pain often resolves after two weeks of bedrest, but may persist if there is continued pressure on surrounding structures or if the episode is repeated at another site. Subacute pain, on the other hand, can last for several months. This pain appears to be mechanical in origin, with patients describing the pain as a mild to moderate backache. There is also evidence of muscle spasm and ligament tenderness on examination. Chronic pain may occur in patients with multiple compression fractures, height loss and low bone density.
S.L. Silverman: Clinical consequences of vertebral compresion fracture The Natural History Fracture Pain
of Acute
Vertebral
Compression
The natural history of acute vertebral compression fracture pain has only recently been studied in clinical trials. Filliponi found that spontaneous VCF pain at rest was more severe than pain on movement or with pressure (Filliponi, 1990). The pain of acute VCF may not decrease significantly during the first 7 to 10 days. Montagnini et al, in a IO-day study, found no significant change in the reduction of pain, as assessed by the patient or physician, and no significant change in analgesic consumption (Montagnini et al, 1989). In a study assessing pain in 56 hospitalised women with acute VCF, pain (as assessed by the physician) had decreased by only 22% at day 7, and by only 33% at day 14 (Lyritis et al, 1990). Analgesic consumption fell by 16% at day 5 and by 33% at day 14. Gennari et al found that spontaneous pain, measured with a visual analogue scale, was only significantly decreased by day 15, but at day 30 it had decreased by approximately 40% (Gennari et al, 1991). In the same study pressure-provoked pain only began to decrease by day 20 and had decreased by 50% at day 30. There are few long-term studies of the pain of acute VCF. Ringe found, when studying patients with corticosteroidinduced osteoporosis and back pain, that pain in the control patients persisted throughout the 6-month study period (Ringe, 1987). The aim of analgesic use in VCF is not only to reduce pain but also to maximise function to a level that will allow patients to mobilise. In one study by Filliponi, NSAID consumption was variable over the first four weeks following fracture (Filliponi, 1990). Although NSAID consumption tended to decrease by the second week, more NSAID consumption was needed as activity increased. Long-term Fracture
Consequences
of Vertebral
Compression
As height is lost, there is a reduction in the size of the thoracic and abdominal cavities with resulting long-term consequences (Kaplan, 1983). Patients develop thoracic kyphosis and lumbar lordosis, and may have reduced exercise tolerance as a result of restricted lung volume due to the loss of support for the rib cage and stooped posture. Patients develop a protuberant abdomen with horizontal skinfold creases near the pelvic rim. The abdomen may be flat on awakening but becomes more distended as the day progresses. Early satiety is often noted, with secondary weight loss. Patients notice increasing waist size, and may feel bloated after eating. As the rib cage approaches the pelvic rim, patients may experience discomfort and difficulty with sleeping. Vertebra1 compression fracture may be associated with chronic back pain, which may be caused by structural changes or unrelated degenerative osteoarthritis of the spine. Patients may experience chronic back pain and muscle fatigue when sitting or standing for long periods and often find they are most comfortable in bed. They have been found to have reduced spinal extensor muscle strength. They may have difficulty getting to sleep or have fitful, non-restorative sleep. The pain and subsequent sleep disorder may result in the development of fibromyalgia (Silverman & Wallace, unpublished observations). Patients experience difficulties when travelling in cars. Car doors are often heavy, and height loss may result in driving difficulties. Exercise is difficult unless the back is supported, and lifting and bending is avoided. Patients lose self esteem and have distorted body image
S29 (Linnel et al, 1991). They are dissatisfied with their current health and are pessimistic about their future health. Since self care and dressing often become difficult, patients become dependent on family members for help. They are often depressed. Patients are often afraid of being pushed over in crowds or climbing stairs. Some patients may become dependent on analgesics including narcotics. The Mechanism and Causes Compression Fracture
of Bone Pain in Vertebral
The mechanism of bone pain in vertebral compression fracture is complex and has been reviewed by (Gennari et al, 1991). The sensation of bone pain primarily depends on the action of nociceptors in the periosteum and of the joint capsule; the cortex and bone marrow are insensitive to pain. Bone pain in VCF arises from two mechanisms: the direct action on bone nociceptors; or a secondary mechanical effect. Vertebral compression fracture can stimulate nociceptors directly by structural damage or by direct compression of tissue, or indirectly by the release of chemical mediators associated with tissue injury. Mechanical pressure on insensitive areas of bone may also trigger pain by altering the shape of a nearby joint, thereby stimulating distant nociceptors. Using a visual analogue scale (VAS), Gennari et al examined the relationship between acute pain and VCF in postmenopausal patients with osteopenia (Gennari et al, 1991). They studied 54 patients with no VCF and 53 patients with VCF. No pain was experienced in 3 1% of non-fracture patients, while only 6% of patients with fracture had no pain. Mild to moderate pain, or a score (VAS) of ~10, was reported in 66% of the non-fracture patients and 56% of the patients with fracture. However, only 2% of the notiracture patients had a score of greater than 10 and none had scores greater than 14, while 40% of patients with fracture had scores of greater than 10 and 8% had scores greater than 15. Thus, while postmenopausal osteopenic patients with or without fracture may experience pain, in general, severe pain was experienced only in patients with fracture. Hospitalisation for Vertebral Compression
Fracture
About one third or more of patients with VCF may seek medical attention, usually for pain. Cooper et al studied 1012 women aged between 48 and 81 years (Cooper et al, 1990). Seventynine (7.8%) were found to have one or more vertebral fractures (defined as a 20% reduction in vertebral height, as viewed on lateral spine X-ray). Back pain was reported in 35.4% of the fracture cases. However, 36.1% of the non-fracture cases also reported back pain, thus questioning whether the back pain was due to the vertebral compression fracture alone. In 1984, we studied the incidence of hospitalisation for vertebral compression fracture in California, finding an incidence in white females of 34.6/100,000 person years. Using a correction factor for attribution of diagnosis (Phillips et al, 1988) and the estimates of Cooper et al, 1992) for the incidence of vertebral compression fracture in white females of 153/100,000 person years, one may estimate a pre-DRG implementation hospitalisation rate as high as 16% for clinically diagnosed fractures. The hospitalisation rate for radiographic VCF would be considerably lower (less than 5%). A similar incidence of hospitalisation of 38/100,000 py was obtained from the National Hospital Discharge Data in 1984, which found 83,000 hospitalisations (Holbrook et al, 1985).
S.L. Silverman:
s30
The Cost of Vertebral Compression
Fracture
The costs of vertebral fractures are reviewed elsewhere (Norris, Kanis, this volume) but are difficult to quantitate. Back pain has been estimated to cost over $1.5 billion per year in the United States (Holbrook et al, 1985) but most of this is due to pathologies unrelated to VCF. Using the California Discharge Data, the cost in California, in 1984, for hospitalised VCF, in white females, was $13.2 million. Correcting for ethnic factors, the estimated cost for hospitalised VCF rises to between $119.8 and $143.4 million in the same year. This compares with the total cost of hospitalised hip fracture in California of $151.7 million and between $1.4 billion and $1.65 billion in the USA. The indirect costs, after hospitalisation and of non-hospitalised vertebral fracture, are not known. The Disability of Vertebral Compression
Fracture
There are little data on the disability that results from VCF. Using the United States Health Interview Survey Handbook, Holbrook et al have estimated that a recognised VCF results in, on average, two weeks of bed rest and one month of restricted activity (Holbrook et al, 1985). Ettinger et al found that 8.4% of patients with moderate to severe vertebral deformity required help at home (Ettinger et al, 1988). When populations are screened for evidence of vertebral deformity, Ettinger et al found that it is not uncommon for women with moderate to severe vertebral deformities (>25% reduction in anterior vertebral height) to be asymptomatic (Ettinger et al, 1988). However, in a larger cross-sectional study, Ettinger et al found that when vertebral height loss was more severe, vertebral deformities were associated with substantial pain and disability and/or loss of height (Ettinger et al, 1991). Leidig et al reported that pain intensity, physical limitations and mood significantly correlated with a spinal deformity index (Leidig et al, 1990). Ross et al re-analysed the original Ettinger data, using risk analysis and odds ratio rather than correlation analysis, and found that significant vertebral deformity was associated with back pain, increased number of doctor visits, and additional expense (Ross et al, 1991). Significant vertebral fracture (235% loss of area in any single vertebra) was associated with back pain, activity limitation due to back pain, medical consultation for back pain, and additional expense. Furthermore, loss of height was a strong predictor of disability and adverse psychosocial outcomes (poor health, mastery, embarrassment due to appearance, and fear of falling). There is little information about the effects of VCF on function and quality of life. Linnel et al compared women with VCFs to independent elderly women (Linnel et al, 1991). They found that women with fractures had greater depression and anxiety. In addition, these women had distorted body image, stooped posture and their general appearance was a major concern. In addition, women with fractures consistently scored lower on measures of well-being and general health. Management
of Vertebral Compression
Fracture
The short-term aims of treating acute VCF are pain relief and mobilisation in order to prevent further bone loss and fracture. Pain relief can usually be accomplished with oral analgesics, such as non-steroidal anti-inflammatory drugs. Short term use of oral narcotics such as paracetamol with codeine or methadone is needed in some patients, although the latter drugs may cause constipation. The use of calcitonin, discussed elsewhere in this volume (Reginster, this volume), may have an
Clinical consequences
of vertebral compresion
fracture
analgesic effect and encourage early mobihsation. A period of bedrest is often necessary until the acute pain begins to subside, usually after l-2 weeks. A firm mattress and a pillow under the knees are often helpful. Deep breathing helps to avoid atelectasis. Ileus develops in a few patients, who may then need to be hospitalised for parenteral fluid therapy. Paravertebral muscle spasm is often a prominent feature and muscle relaxants combined with moist heat may help to reduce this. Patients may require assistance from family, friends or social services, and may require hospitalisation if they have no immediate support. Mobilisation should begin as soon as the acute pain begins to subside (l-2 weeks following fracture). Patients should be encouraged to sit or stand for short periods, several times each day, and then slowly begin ambulation. It is our experience that patients with severe pain may benefit from a rigid thoracolumbar hyperextension orthosis or a three-point semirigid thoracolumbar extension orthosis as tolerated. Some patients may also benefit from a cane to provide extra support. Once the acute pain has subsided, the patient should begin a programme of spine extension exercises to strengthen the paraspinal musculature. Strong spinal extensors decrease the flexion forces on the spine. Patients with thoracic compression fractures may also benefit from pectoral stretching and deep breathing. All patients will benefit from instruction on how to avoid spinal flexion when lifting or bending. Within two to three months, most patients are pain free and can resume activity. Some patients, however, may still require a corset for support. Risk of Fracture Perhaps the greatest long-term consequence of vertebral compression fracture is the increased risk of further vertebral Both low bone mass and previous fracture fracture. independently predict subsequent fracture (Ross et al, 1991). Women with low bone mass have a seven-fold increase in risk of fracture, compared with a 25-fold increase in risk in women with low bone mass and a single fracture (Ross et al, 1991). Patients with vertebral compression fracture are also at risk of concomitant peripheral fracture. Vega et al reported that one third of patients with osteoporotic vertebral compression fracture were also found to have one or more peripheral fractures (Vega et al, 1990). The most common sites were wrist (14.1%), humerus (6.2%), proximal femur (4.7%) and tibia (2.8%). Patients with vertebral compression fracture who had femoral fractures had significantly lower bone mass in the lumbar spine, femoral neck or radial shaft than non-fracture patients. Vertebral Compression
Fracture and Mortality
There are no data to suggest that vertebral compression fractures are associated with excess mortality, however, recent data suggest that patients with low bone mass have increased non-trauma mortality (Browner et al, 1991). Conclusion Vertebral compression fracture is often defined as a radiographic event; a method of definition that has proven confusing since not all radiographic fractures are clinically diagnosed or clinically symptomatic. It may prove useful to define vertebral compression fracture as a clinical event characterised by acute pain and height loss. Height loss is a
S.L. Silverman:
Clinical consequences
of vertebral compresion
fracture
strong predictor of disability, as well as adverse psychosocial outcomes. Height loss may be considered as a surrogate measure of vertebral fractures in clinical trials. The short-term goals in the management of vertebral osteoporosis are to reduce pain and increase mobilisation, while in the long-term we should aim to prevent further bone loss and fracture.
References Browner WS, Seeley DG, Vogt TM, Cummings S. Non-trauma mortality in elderly women with low bone mineral density. Lancer 1991;338:355-358. Cooper C, Atkinson ES, O’Fallon WM, Melton LJ. The incidence of clinically diagnosed vertebral fractures: a population based study in Rochester, Minnesota, 1985-1989. J Bone Miff Res 1992;7:221-228. Cooper C, Shah S. Hand DJ et al Screening for osteoporosis using individual risk factors. Osreoporosis Int 1991; 2:48-53. Ettinger B, Black D, Cummings SR. Only severe vertebral deformities are associated with lower bone density, more back pain and back-related disability. Presented at the Eighth Annual Meeting of the International Workshop on Bone Densitometry, Bad Reichenhall, Germany, April 1991. Ettinger B, Block JE, Smith R, Cummings SR. Harris ST, Genant HK. An examination of the association between vertebral deformities, physical disabilities and psychosocial problems. Moturitas 1988;10:283-296. Filliponi P. Analgesic effect of eel calcitonin 0%CT) and salmon calcitonin (SCT) in osteoporotic pain: a multicentre trial. In: Osteoporosis 1990. Eds: Christiansen C, Overgaard K. Osteopress, Copenhagen, 1990: pp 1444-1448. Franceschini R, Bottaro P, Paoopoulos C, Messina V. Long-term treatment with salmon calcitonin in postmenopausal osteoporosis Curr Ther Res 1983;34:795800. Gennti C, Agnusdei D, Camporeale A. Use of calcitonin in the treatment of bone pain associated with osteoporosis. C&if Tissue Id 1991:49 (Suppl 2):S9s13. Grazier KL, Holbrook TL, Kelsey J. Stauffer RN. The frequency of occurrence, impact and cost of musculoskeletal conditions in the United States. American Academy of Orthopedic Surgeons, Chicago, 1984. Holbrook TL, Grazier K, Kelsey JL, Stauffer RN. The frequency of occurrence, impact and cost of musculoskeletal conditions in the United States. American Academy of Orthopedic Surgeons. Chicago, 1985.
s31 Kaplan FS. Osteoporosis. CIBA Clinical Symposia, CIBA Pharmaceutical Company, Summit, NJ, 1983;35(3). Leidig G. Minne HW, Sauer P et al. A study of complaints and their relation to vertebral destruction in patients with osteoporosis. Bone Miner 1990;8:217219. Linnel PW, Hermansen SE, Elias MF et al. Quality of life in osteoporotic women. J Bone Mitt Res 1991;6 (Suppl 1) abs 96, S106. Lyritis GP, Magiasis B, Eliopoulos A, Tsekoura M, Ioakimidis D. Analgesic effect of salmon calcitonin in cases of osteoporotic vertebral fractures. In: Osteoporosis 1990. Eds: Christiansen C, Overgaard K. Osteopress. Copenhagen, 199Ozpp 1392.1395. Lyritis GP, Mayasis B, Tsakalakos N et al. The natural history of the osteoporotic vertebral fracture. C[in Rheum 1989;8 (Suppl 2):66-69. Melton LJ, Kan SH, Frye MA, Wahjner HW, O’Fallon WM. Riggs BL. Epidemiology of vertebral fractures in women. Am J Epidemiol 1989;129:1000-11. Montagnani M, Gonnelli S, Fraocini G, Piolini M, Gennari C. Analgesic effect of salmon calcitonin nasal spray in bone pain. In: Calcitonin ‘88 New Therapeutic Perspectives: The Nasal Spray. Ed: Mazzuoli GF. Saodoz AG, Basl, 1989: pp 126-133. Nelson DA, Kleerekoper M, Havsta S, Peterson E. Vertebral deformities in osteoporosis may be transient events: the “rebound” phenomenon. J Bone Min Res 1991;6 (abs 766):s275. Phillips S, Fox N, Jacobs J. Wright WE. The direct medical costs of osteoporosis for American women aged 45 and older. Bone 1988;9:271-90 Ringe JD. Clinical evaluation of salmon calcitonin in bone pain. In: Osteoporosis 1987. Eds: Christiansen C, Johansen JS, Riis BJ. Osteopress, Copenhagen 1987:1262-1264. Ross PD, Davis SW. Epstein RS, Wasnich RD. Preexisting fractures and bone mass predict vertebral fracture incidence in women. Ann Intern Med 1991;114:919-923. Ross PD. Ettinger B, Davis JW, Melton LJ, Wasnich RD. Evaluation of adverse health outcomes associated with vertebral fractures. Osteoporosis Id 1991;1:134-140. Silverman SL, Wallace D. Unpublished observations. Vega E, Mautalen C, Ghiringhelli G, Fromm G. Other skeletal fractures in osteoporotic females with vertebral fractures. In: Osteoporosis 1990. Eds: Christiansen C, Overgaard K. Osteopress, Copenhagen, 1990:~~ 504-6. Watts NB, Genant HK, Harris ST et al. Clinical correlates of radiographically apparent vertebral fractures. CalcifTiss 1991;42 (Suppl):A39.
Copyright
The Clinical Consequences
8756-3282/92 $5.00 + .OO o 1992 Pergamon Press Ltd.
of Vertebral Compression
Fracture
S.L. SILVERMAN.
The Osteoporosis Medical Center, Beverly Hills, California; Wadsworth VA Medical Center and the Rheumatology Division, UCLA School of Medicine, Los Angeles, California, USA. Address for correspondence and reprints: Dr S. Silverman, California, 90210, USA.
The Osteoporosis
Abstract
consequences of vertebral compression fracture. This paper reviews the relationship between VCF as a radiographic and clinical event and looks at the clinical consequences of VCF as it is known.
Vertebral compression fractures (VCFs) may be defined radiographically or as a clinical event. The prevalence of these fractures in women aged 50 and over has been estimated at 26% when defined as a reduction in vertebral height >15%. Retrospective reviews of case records have shown a clinical detection rate of VCF in white women of 153/100,000 person years. Of these clinically detected VCFs, 84% were associated with pain. VCF may be defined as a clinical event characterised by loss of height and acute pain. The pain of acute fracture usually lasts 4 to 6 weeks with intense pain at the site of fracture. Chronic pain may also occur in patients with multiple compression fractures, height loss and low bone density but is probably due to structural changes or osteoarthritis. Radiographic VCF may not be symptomatic. The greater the deformity, the greater the likelihood of pain and disability. As height is lost, patients experience discomfort from the rib cage pressing downward on the pelvis. Patients develop a thoracic kyphosis, a lumbar lordosis, and a protuberant abdomen with prominent horizontal skinfold creases. The reduced thoracic space may result in decreased exercise tolerance and reduced abdominal space may give rise to early satiety and weight loss. Sleep disorders may also occur. Patients lose self esteem. Self care may become difficult. They are often depressed. They become fearful of further fracture. They have distorted body image and poor health perception. Patients with one vertebral fracture are at increased risk of peripheral fracture and further vertebral fracture. The aims of acute management are to reduce symptoms and mobilise the patient as quickly as possible. Long-term goals are to maintain or increase skeletal mass and improve mobility and function. Key Words: Osteoporosis fracture - Disability
- Compression
Medical Center, 450 N Bedford Drive, Suite 111, Beverly Hills,
The Incidence Fracture
and Prevalence
of Vertebral
Compression
Vertebral compression fracture may be defined either as a radiographic event or as a symptomatic clinical event. Radiographic VCFs have been defined according to the presence of a crush or wedge fracture and, recently, according to objective criteria relating to percentage loss of vertebral height as assessed on X-ray (eg. reduction in vertebral height >15%). Vertebral deformity scores were then developed to summarise radiological observations in multiple vertebrae. Recently, population standards for specific vertebral heights have been collected and individual vertebral heights can be compared to these norms. The prevalence of radiographic VCF is high. Using a reduction in vertebral height of ~15% as a criterion for VCF, Melton et al reported a prevalence of VCF in Rochester, Minnesota in 1981-1983, as high as 26% in women aged 50 and over (Melton et al, 1989). The corresponding radiographic incidence of vertebral fracture rose from 5/1,000 person years (py) in women aged 50-54 years to 29.6/1,000 py in women aged 85 years and over. Most of these radiographic changes are, however, asymptomatic and are therefore not clinically diagnosed. Cooper et al, while studying Rochester men and women of all ages, found an age- and sex-adjusted incidence rate for clinically diagnosed vertebral fracture of 123/100,000 py for the years 1985-1989; an incidence that is approximately 35% of that estimated from the prevalence data (Cooper et al, 1992). The age-adjusted incidence in women (153/100,000 py) was almost twice that in men (81/100,000 py). The incidence in women is comparable to that for hip fracture, which has an overall ageand sex-adjusted rate of 114/100,000 py. Not all patients with clinically diagnosed vertebral fracture are symptomatic. In the study by Cooper et al, 16% of diagnoses were made incidentally in the course of radiographic investigations into unrelated disorders or as part of a workup in a high risk patient, on corticosteroids, for example. Of all the clinically diagnosed VCFs, 84% were associated with pain. When broken down further, it was shown that VCFs associated
fracture - Vertebral
Introduction The incidence of symptomatic vertebral compression fracture (VCF) is high and is probably similar to that of hip fracture. However, there is very little detailed knowledge of the clinical s21
S.L. Silverman: Clinical consequences of vertebral compresion fracture
S28
with severe trauma (traffic accidents or falls from greater than standing height) were always symptomatic, while only 81% of the moderate to minimal traumatic VCFs were associated with back pain. This included spontaneous VCFs that accounted for 46% of the latter total. Not all patients with symptomatic VCF are clinically diagnosed. The pain may be of short duration and responsive to analgesics, so it is likely that many vertebral compression fractures are not clinically diagnosed by physicians. Indeed some may be seen by chiropractors and masseuses, thus bypassing the attention of the physician (Ross et al, 1991). Fig. 1 shows the incidence of VCF.
Radiographic
Clinically
VCF
diagnosed
1531100,000
VCF
py
\
/
(Nelson et al, 1991). They found that only 35% of new VCFs
(15% reduction in vertebral height) were painful, while only one in five were associated with a positive bone scan (positive bone scans were associated with ~25% reduction in vertebral height). Nelson et al. therefore, suggested that more stringent criteria be used to diagnose VCF and that this type of fracture be defined only as a clinical event. Further complicating the radiographic diagnosis of VCF in this study was the observation that vertebral height loss may be transient in patients in whom the percentage loss was less than 20-25%. Using a morphometric criterion of a 210% change in anterior, middle or posterior height, a ‘rebound’ phenomenon was observed in 54% of vertebrae with reduction in middle height, 28% with reduction in anterior height and 42% with reduction in posterior height. Lyritis et al studied the natural history of VCF in 210 patients (Lyritis et al, 1989), and reported two types of VCF. In the first, pain was acute and severe, improving gradually, with obvious radiographic deformity at onset. In the second, the pain was less severe and of shorter duration; however, a new attack of pain occurred within 6 to 16 weeks and often recurred for a period of 6-18 months. With the second type of pain, radiographic deformity was not evident during the first attack but wedging gradually developed over the next few months. This study suggests heterogeneity in the clinical manifestation of vertebral fractures. The diagnosis of symptomatic VCF is often delayed. Cooper et al, while studying the Rochester population, found that the median duration of symptoms before fracture diagnosis was 7 days, although in 25% of patients, symptoms had persisted for one month or longer before the diagnosis was made (Cooper et al, 1992). Fractures following severe trauma (a traffic accident or a fall from greater than a standing height) were diagnosed more rapidly. The Pain of Acute Vertebral Compression
Fig. 1. Incidence of vertebral compression
fracture.
The Relationship Between Vertebral Compression as a Radiographic and a Clinical Event
Fracture
Recent data from two large prospective studies have provided insight into the clinical correlates of new VCF as defined radiographically. Watts et al compared the radiographic and clinical manifestations of new VCFs in a US multi-centre study into the effects of etidronate (Watts et al, 1991). Four out of six patients (67%) with two or more new VCFs had height changes (>I cm) and five out of six patients (83%) had acute pain. All three of the patients with one new VCFs and greater than 40% reduction in vertebral height had height changes, while two out of the three patients (67%) had pain. Of the 10 patients with one new VCF and 25-40% reduction in vertebral height, seven (70%) had height loss and pain. Of the six pateints with one new VCF and 20-25% reduction in vertebral height, three (50%) had height loss and one (16%) had pain. This study suggests that the clinical measures of height loss and, to a lesser extent, acute pain are sensitive indicators of new VCF with moderate to severe deformity. In a large study investigating the effect of fluoride on VCF, Nelson et al reviewed the clinical significance of new VCF
Fracture
Acute VCF pain usually lasts from 2 weeks to 3 months and is associated with intense, deep pain at the site of fracture, with palpatory tenderness over the site. Leidig et al studied pain in patients hospitalised due to VCF (Leidig et al, 1990). The pain was described as intermittent or chronic, deep, worse with prolonged posture, bending, walking or standing. Pain was moderate to severe and sometimes intolerable. Pain was relieved by lying down, heat, diversion and pain relievers in that order. Pain was associated with physical limitation in bending, arising and dressing. Acute VCF pain is felt deeply and diffusely in the back, is worse with motion and reduced by rest. The patient is in discomfort while sitting or standing, and an acute episode may be so painful that the patient may lie on the floor unable to move. The spine is tender to deep palpation and to percussion, and paraspinal vertebral muscle spasm is often present. Radiculopathy is not uncommon with compression fractures and may cause either unilateral or bilateral pain, radiating anteriorly along the rib of the affected nerve root. Acute pain often resolves after two weeks of bedrest, but may persist if there is continued pressure on surrounding structures or if the episode is repeated at another site. Subacute pain, on the other hand, can last for several months. This pain appears to be mechanical in origin, with patients describing the pain as a mild to moderate backache. There is also evidence of muscle spasm and ligament tenderness on examination. Chronic pain may occur in patients with multiple compression fractures, height loss and low bone density.
S.L. Silverman: Clinical consequences of vertebral compresion fracture The Natural History Fracture Pain
of Acute
Vertebral
Compression
The natural history of acute vertebral compression fracture pain has only recently been studied in clinical trials. Filliponi found that spontaneous VCF pain at rest was more severe than pain on movement or with pressure (Filliponi, 1990). The pain of acute VCF may not decrease significantly during the first 7 to 10 days. Montagnini et al, in a IO-day study, found no significant change in the reduction of pain, as assessed by the patient or physician, and no significant change in analgesic consumption (Montagnini et al, 1989). In a study assessing pain in 56 hospitalised women with acute VCF, pain (as assessed by the physician) had decreased by only 22% at day 7, and by only 33% at day 14 (Lyritis et al, 1990). Analgesic consumption fell by 16% at day 5 and by 33% at day 14. Gennari et al found that spontaneous pain, measured with a visual analogue scale, was only significantly decreased by day 15, but at day 30 it had decreased by approximately 40% (Gennari et al, 1991). In the same study pressure-provoked pain only began to decrease by day 20 and had decreased by 50% at day 30. There are few long-term studies of the pain of acute VCF. Ringe found, when studying patients with corticosteroidinduced osteoporosis and back pain, that pain in the control patients persisted throughout the 6-month study period (Ringe, 1987). The aim of analgesic use in VCF is not only to reduce pain but also to maximise function to a level that will allow patients to mobilise. In one study by Filliponi, NSAID consumption was variable over the first four weeks following fracture (Filliponi, 1990). Although NSAID consumption tended to decrease by the second week, more NSAID consumption was needed as activity increased. Long-term Fracture
Consequences
of Vertebral
Compression
As height is lost, there is a reduction in the size of the thoracic and abdominal cavities with resulting long-term consequences (Kaplan, 1983). Patients develop thoracic kyphosis and lumbar lordosis, and may have reduced exercise tolerance as a result of restricted lung volume due to the loss of support for the rib cage and stooped posture. Patients develop a protuberant abdomen with horizontal skinfold creases near the pelvic rim. The abdomen may be flat on awakening but becomes more distended as the day progresses. Early satiety is often noted, with secondary weight loss. Patients notice increasing waist size, and may feel bloated after eating. As the rib cage approaches the pelvic rim, patients may experience discomfort and difficulty with sleeping. Vertebra1 compression fracture may be associated with chronic back pain, which may be caused by structural changes or unrelated degenerative osteoarthritis of the spine. Patients may experience chronic back pain and muscle fatigue when sitting or standing for long periods and often find they are most comfortable in bed. They have been found to have reduced spinal extensor muscle strength. They may have difficulty getting to sleep or have fitful, non-restorative sleep. The pain and subsequent sleep disorder may result in the development of fibromyalgia (Silverman & Wallace, unpublished observations). Patients experience difficulties when travelling in cars. Car doors are often heavy, and height loss may result in driving difficulties. Exercise is difficult unless the back is supported, and lifting and bending is avoided. Patients lose self esteem and have distorted body image
S29 (Linnel et al, 1991). They are dissatisfied with their current health and are pessimistic about their future health. Since self care and dressing often become difficult, patients become dependent on family members for help. They are often depressed. Patients are often afraid of being pushed over in crowds or climbing stairs. Some patients may become dependent on analgesics including narcotics. The Mechanism and Causes Compression Fracture
of Bone Pain in Vertebral
The mechanism of bone pain in vertebral compression fracture is complex and has been reviewed by (Gennari et al, 1991). The sensation of bone pain primarily depends on the action of nociceptors in the periosteum and of the joint capsule; the cortex and bone marrow are insensitive to pain. Bone pain in VCF arises from two mechanisms: the direct action on bone nociceptors; or a secondary mechanical effect. Vertebral compression fracture can stimulate nociceptors directly by structural damage or by direct compression of tissue, or indirectly by the release of chemical mediators associated with tissue injury. Mechanical pressure on insensitive areas of bone may also trigger pain by altering the shape of a nearby joint, thereby stimulating distant nociceptors. Using a visual analogue scale (VAS), Gennari et al examined the relationship between acute pain and VCF in postmenopausal patients with osteopenia (Gennari et al, 1991). They studied 54 patients with no VCF and 53 patients with VCF. No pain was experienced in 3 1% of non-fracture patients, while only 6% of patients with fracture had no pain. Mild to moderate pain, or a score (VAS) of ~10, was reported in 66% of the non-fracture patients and 56% of the patients with fracture. However, only 2% of the notiracture patients had a score of greater than 10 and none had scores greater than 14, while 40% of patients with fracture had scores of greater than 10 and 8% had scores greater than 15. Thus, while postmenopausal osteopenic patients with or without fracture may experience pain, in general, severe pain was experienced only in patients with fracture. Hospitalisation for Vertebral Compression
Fracture
About one third or more of patients with VCF may seek medical attention, usually for pain. Cooper et al studied 1012 women aged between 48 and 81 years (Cooper et al, 1990). Seventynine (7.8%) were found to have one or more vertebral fractures (defined as a 20% reduction in vertebral height, as viewed on lateral spine X-ray). Back pain was reported in 35.4% of the fracture cases. However, 36.1% of the non-fracture cases also reported back pain, thus questioning whether the back pain was due to the vertebral compression fracture alone. In 1984, we studied the incidence of hospitalisation for vertebral compression fracture in California, finding an incidence in white females of 34.6/100,000 person years. Using a correction factor for attribution of diagnosis (Phillips et al, 1988) and the estimates of Cooper et al, 1992) for the incidence of vertebral compression fracture in white females of 153/100,000 person years, one may estimate a pre-DRG implementation hospitalisation rate as high as 16% for clinically diagnosed fractures. The hospitalisation rate for radiographic VCF would be considerably lower (less than 5%). A similar incidence of hospitalisation of 38/100,000 py was obtained from the National Hospital Discharge Data in 1984, which found 83,000 hospitalisations (Holbrook et al, 1985).
S.L. Silverman:
s30
The Cost of Vertebral Compression
Fracture
The costs of vertebral fractures are reviewed elsewhere (Norris, Kanis, this volume) but are difficult to quantitate. Back pain has been estimated to cost over $1.5 billion per year in the United States (Holbrook et al, 1985) but most of this is due to pathologies unrelated to VCF. Using the California Discharge Data, the cost in California, in 1984, for hospitalised VCF, in white females, was $13.2 million. Correcting for ethnic factors, the estimated cost for hospitalised VCF rises to between $119.8 and $143.4 million in the same year. This compares with the total cost of hospitalised hip fracture in California of $151.7 million and between $1.4 billion and $1.65 billion in the USA. The indirect costs, after hospitalisation and of non-hospitalised vertebral fracture, are not known. The Disability of Vertebral Compression
Fracture
There are little data on the disability that results from VCF. Using the United States Health Interview Survey Handbook, Holbrook et al have estimated that a recognised VCF results in, on average, two weeks of bed rest and one month of restricted activity (Holbrook et al, 1985). Ettinger et al found that 8.4% of patients with moderate to severe vertebral deformity required help at home (Ettinger et al, 1988). When populations are screened for evidence of vertebral deformity, Ettinger et al found that it is not uncommon for women with moderate to severe vertebral deformities (>25% reduction in anterior vertebral height) to be asymptomatic (Ettinger et al, 1988). However, in a larger cross-sectional study, Ettinger et al found that when vertebral height loss was more severe, vertebral deformities were associated with substantial pain and disability and/or loss of height (Ettinger et al, 1991). Leidig et al reported that pain intensity, physical limitations and mood significantly correlated with a spinal deformity index (Leidig et al, 1990). Ross et al re-analysed the original Ettinger data, using risk analysis and odds ratio rather than correlation analysis, and found that significant vertebral deformity was associated with back pain, increased number of doctor visits, and additional expense (Ross et al, 1991). Significant vertebral fracture (235% loss of area in any single vertebra) was associated with back pain, activity limitation due to back pain, medical consultation for back pain, and additional expense. Furthermore, loss of height was a strong predictor of disability and adverse psychosocial outcomes (poor health, mastery, embarrassment due to appearance, and fear of falling). There is little information about the effects of VCF on function and quality of life. Linnel et al compared women with VCFs to independent elderly women (Linnel et al, 1991). They found that women with fractures had greater depression and anxiety. In addition, these women had distorted body image, stooped posture and their general appearance was a major concern. In addition, women with fractures consistently scored lower on measures of well-being and general health. Management
of Vertebral Compression
Fracture
The short-term aims of treating acute VCF are pain relief and mobilisation in order to prevent further bone loss and fracture. Pain relief can usually be accomplished with oral analgesics, such as non-steroidal anti-inflammatory drugs. Short term use of oral narcotics such as paracetamol with codeine or methadone is needed in some patients, although the latter drugs may cause constipation. The use of calcitonin, discussed elsewhere in this volume (Reginster, this volume), may have an
Clinical consequences
of vertebral compresion
fracture
analgesic effect and encourage early mobihsation. A period of bedrest is often necessary until the acute pain begins to subside, usually after l-2 weeks. A firm mattress and a pillow under the knees are often helpful. Deep breathing helps to avoid atelectasis. Ileus develops in a few patients, who may then need to be hospitalised for parenteral fluid therapy. Paravertebral muscle spasm is often a prominent feature and muscle relaxants combined with moist heat may help to reduce this. Patients may require assistance from family, friends or social services, and may require hospitalisation if they have no immediate support. Mobilisation should begin as soon as the acute pain begins to subside (l-2 weeks following fracture). Patients should be encouraged to sit or stand for short periods, several times each day, and then slowly begin ambulation. It is our experience that patients with severe pain may benefit from a rigid thoracolumbar hyperextension orthosis or a three-point semirigid thoracolumbar extension orthosis as tolerated. Some patients may also benefit from a cane to provide extra support. Once the acute pain has subsided, the patient should begin a programme of spine extension exercises to strengthen the paraspinal musculature. Strong spinal extensors decrease the flexion forces on the spine. Patients with thoracic compression fractures may also benefit from pectoral stretching and deep breathing. All patients will benefit from instruction on how to avoid spinal flexion when lifting or bending. Within two to three months, most patients are pain free and can resume activity. Some patients, however, may still require a corset for support. Risk of Fracture Perhaps the greatest long-term consequence of vertebral compression fracture is the increased risk of further vertebral Both low bone mass and previous fracture fracture. independently predict subsequent fracture (Ross et al, 1991). Women with low bone mass have a seven-fold increase in risk of fracture, compared with a 25-fold increase in risk in women with low bone mass and a single fracture (Ross et al, 1991). Patients with vertebral compression fracture are also at risk of concomitant peripheral fracture. Vega et al reported that one third of patients with osteoporotic vertebral compression fracture were also found to have one or more peripheral fractures (Vega et al, 1990). The most common sites were wrist (14.1%), humerus (6.2%), proximal femur (4.7%) and tibia (2.8%). Patients with vertebral compression fracture who had femoral fractures had significantly lower bone mass in the lumbar spine, femoral neck or radial shaft than non-fracture patients. Vertebral Compression
Fracture and Mortality
There are no data to suggest that vertebral compression fractures are associated with excess mortality, however, recent data suggest that patients with low bone mass have increased non-trauma mortality (Browner et al, 1991). Conclusion Vertebral compression fracture is often defined as a radiographic event; a method of definition that has proven confusing since not all radiographic fractures are clinically diagnosed or clinically symptomatic. It may prove useful to define vertebral compression fracture as a clinical event characterised by acute pain and height loss. Height loss is a
S.L. Silverman:
Clinical consequences
of vertebral compresion
fracture
strong predictor of disability, as well as adverse psychosocial outcomes. Height loss may be considered as a surrogate measure of vertebral fractures in clinical trials. The short-term goals in the management of vertebral osteoporosis are to reduce pain and increase mobilisation, while in the long-term we should aim to prevent further bone loss and fracture.
References Browner WS, Seeley DG, Vogt TM, Cummings S. Non-trauma mortality in elderly women with low bone mineral density. Lancer 1991;338:355-358. Cooper C, Atkinson ES, O’Fallon WM, Melton LJ. The incidence of clinically diagnosed vertebral fractures: a population based study in Rochester, Minnesota, 1985-1989. J Bone Miff Res 1992;7:221-228. Cooper C, Shah S. Hand DJ et al Screening for osteoporosis using individual risk factors. Osreoporosis Int 1991; 2:48-53. Ettinger B, Black D, Cummings SR. Only severe vertebral deformities are associated with lower bone density, more back pain and back-related disability. Presented at the Eighth Annual Meeting of the International Workshop on Bone Densitometry, Bad Reichenhall, Germany, April 1991. Ettinger B, Block JE, Smith R, Cummings SR. Harris ST, Genant HK. An examination of the association between vertebral deformities, physical disabilities and psychosocial problems. Moturitas 1988;10:283-296. Filliponi P. Analgesic effect of eel calcitonin 0%CT) and salmon calcitonin (SCT) in osteoporotic pain: a multicentre trial. In: Osteoporosis 1990. Eds: Christiansen C, Overgaard K. Osteopress, Copenhagen, 1990: pp 1444-1448. Franceschini R, Bottaro P, Paoopoulos C, Messina V. Long-term treatment with salmon calcitonin in postmenopausal osteoporosis Curr Ther Res 1983;34:795800. Gennti C, Agnusdei D, Camporeale A. Use of calcitonin in the treatment of bone pain associated with osteoporosis. C&if Tissue Id 1991:49 (Suppl 2):S9s13. Grazier KL, Holbrook TL, Kelsey J. Stauffer RN. The frequency of occurrence, impact and cost of musculoskeletal conditions in the United States. American Academy of Orthopedic Surgeons, Chicago, 1984. Holbrook TL, Grazier K, Kelsey JL, Stauffer RN. The frequency of occurrence, impact and cost of musculoskeletal conditions in the United States. American Academy of Orthopedic Surgeons. Chicago, 1985.
s31 Kaplan FS. Osteoporosis. CIBA Clinical Symposia, CIBA Pharmaceutical Company, Summit, NJ, 1983;35(3). Leidig G. Minne HW, Sauer P et al. A study of complaints and their relation to vertebral destruction in patients with osteoporosis. Bone Miner 1990;8:217219. Linnel PW, Hermansen SE, Elias MF et al. Quality of life in osteoporotic women. J Bone Mitt Res 1991;6 (Suppl 1) abs 96, S106. Lyritis GP, Magiasis B, Eliopoulos A, Tsekoura M, Ioakimidis D. Analgesic effect of salmon calcitonin in cases of osteoporotic vertebral fractures. In: Osteoporosis 1990. Eds: Christiansen C, Overgaard K. Osteopress. Copenhagen, 199Ozpp 1392.1395. Lyritis GP, Mayasis B, Tsakalakos N et al. The natural history of the osteoporotic vertebral fracture. C[in Rheum 1989;8 (Suppl 2):66-69. Melton LJ, Kan SH, Frye MA, Wahjner HW, O’Fallon WM. Riggs BL. Epidemiology of vertebral fractures in women. Am J Epidemiol 1989;129:1000-11. Montagnani M, Gonnelli S, Fraocini G, Piolini M, Gennari C. Analgesic effect of salmon calcitonin nasal spray in bone pain. In: Calcitonin ‘88 New Therapeutic Perspectives: The Nasal Spray. Ed: Mazzuoli GF. Saodoz AG, Basl, 1989: pp 126-133. Nelson DA, Kleerekoper M, Havsta S, Peterson E. Vertebral deformities in osteoporosis may be transient events: the “rebound” phenomenon. J Bone Min Res 1991;6 (abs 766):s275. Phillips S, Fox N, Jacobs J. Wright WE. The direct medical costs of osteoporosis for American women aged 45 and older. Bone 1988;9:271-90 Ringe JD. Clinical evaluation of salmon calcitonin in bone pain. In: Osteoporosis 1987. Eds: Christiansen C, Johansen JS, Riis BJ. Osteopress, Copenhagen 1987:1262-1264. Ross PD, Davis SW. Epstein RS, Wasnich RD. Preexisting fractures and bone mass predict vertebral fracture incidence in women. Ann Intern Med 1991;114:919-923. Ross PD. Ettinger B, Davis JW, Melton LJ, Wasnich RD. Evaluation of adverse health outcomes associated with vertebral fractures. Osteoporosis Id 1991;1:134-140. Silverman SL, Wallace D. Unpublished observations. Vega E, Mautalen C, Ghiringhelli G, Fromm G. Other skeletal fractures in osteoporotic females with vertebral fractures. In: Osteoporosis 1990. Eds: Christiansen C, Overgaard K. Osteopress, Copenhagen, 1990:~~ 504-6. Watts NB, Genant HK, Harris ST et al. Clinical correlates of radiographically apparent vertebral fractures. CalcifTiss 1991;42 (Suppl):A39.
