 |
 |
 |
|
| home | CME | help | feedback | subscriptions | archive |
|
© 1997 The Journal of Bone and Joint Surgery, Inc.
Current Concepts Review - Injury of an Extremity as a Risk Factor for the Development of Osteoporosis*  , TAMPERE, FINLANDInvestigation performed at the Section of Orthopaedics, Department of Surgery, Tampere University Hospital, and the Accident and Trauma Research Center, Urho Kaleva Kekkonen Institute for Health Promotion Research, Tampere
Low bone mass occurs with a relatively high prevalence in the elderly population and is thought to be a natural part of aging. When low bone mass is associated with architectural deterioration of the bone, decreased bone strength, and increased susceptibility to fractures, the condition is called osteoporosis77. Osteoporotic fractures are a major public-health problem not only because of the associated morbidity and mortality but also because of the health-care costs that they generate. The number and rate of fractures of the hip have increased in developed countries worldwide during recent decades, and they are expected to rise dramatically as the populations age12,34,67. The two most important risk factors for osteoporosis are an insufficient bone mass by the time of skeletal maturity (low peak bone mass) and the rapid loss of bone that occurs in some women after menopause6,9-11. The average lifetime bone loss after the acquisition of peak bone mass is estimated to be 40 per cent (less than 1 per cent per year)16,68. A decrease in bone-mineral density of one standard deviation from an age-matched norm (an amount equivalent to approximately 10 to 15 per cent of an individual's bone mass) is associated with a 50 to 100 per cent increase in the risk of fracture31-33. A positive family history, increased age, white race, female gender, delayed puberty, various hormonal disturbances and nutritional deficiencies, low body weight, many diseases, many drugs, and smoking are well recognized risk factors for generalized osteoporosis. However, there are some other forms of osteoporosis that, in contrast to generalized osteoporosis, result in a loss of bone in only certain regions of the skeleton and are usually specific to the event causing the bone loss. This kind of local bone loss can, for example, be seen in transient osteoporosis associated with pregnancy and after radiation therapy, physical inactivity, or musculoskeletal injury14,20,31,43,47,64,71,76. Injury with subsequent immobilization and disuse of the injured extremity leads to a rapid loss of bone21,22,74,84. Post-traumatic bone loss is best described as a high-turnover condition—that is, both bone formation and resorption are increased but the latter overcomes the former, resulting in net bone loss56,63. During the healing phase of an injury, three factors are thought to be involved in the process of bone loss. First, the injury itself may produce a catabolic response that leads to a negative mineral balance. Second, with severe injury, the additional stress induced by operative intervention may cause an additional catabolic effect on the skeleton. Finally, post-traumatic immobilization, long-term disuse of the injured extremity, diminished weight-bearing, and loss of normal muscle function are known to result in rapid initial bone loss, the persistence of which presumably depends on several factors, such as the severity of the injury, the treatment, and the functional recovery of the patient1,21,22,46,55,56,74,83-85. The long-term effects of different types of musculoskeletal injuries on bone-mineral density at various skeletal sites have not been extensively reviewed, as far as we know. In this article, we summarize the current knowledge on bone loss after various injuries of the extremities and the role of this post-traumatic bone loss in the development of generalized and localized osteoporosis and associated fractures. This article briefly describes the terminology and methodology that are used in the determination of bone-mineral density and content, reviews the studies of post-traumatic osteoporosis, summarizes the current knowledge on post-traumatic osteoporosis and its risk factors, and provides guidelines for the prevention and treatment of the condition.
In clinical practice, the measurement of bone density is used both to identify osteoporosis and to monitor the response of the skeleton to different bone-targeted therapies. Absorptiometric methods are used to measure the bone-mineral content, which is expressed in grams. This value can be converted into an areal bone-mineral measurement (bone density) by dividing the bone-mineral content by the area scanned: for example, grams per square centimeter7,8. Although, when it is determined in this way, bone density is an areal density and not a true volume density, it is commonly reported along with bone-mineral content because the correction of bone-mineral content for area removes some, but not all, of its dependency on the size of the bone. Measurements of changes in bone mass are commonly used as a surrogate for the assessment of the risk of fracture. Prospective studies have demonstrated an indirect relationship between bone density and the occurrence of fractures, but because of the overlap of measurements obtained from osteoporotic and control subjects the predictive power of these absorptiometric variables is somewhat limited13,15,17,25,31,86. This also emphasizes the fact that, although bone mass contributes a great deal to the structural strength of bone, it is not the only determinant of strength. The contribution of other properties, such as the size (geometry) or quality of the bone, are also important to the strength of the bone19,72,75. The bone density in the calcaneus and that of the radius were reported to be similar with regard to the ability to predict a fracture of the hip15, whereas the areal bone density in the femoral neck was found to be a better predictor of fracture of the hip than the areal bone density of the spine, radius, or calcaneus13. The association between the risk of fracture and low bone mass is well established. An inverse relationship between the size of the bone and the risk of fracture also seems intuitively likely as, given the same mechanical stresses, a small bone is more likely to break than a large one of similar volumetric density, although the geometry of the bone also clearly influences this relationship19. Studies of differential changes in bone density should provide more information about the relative contributions of bone size to skeletal strength in adults. In the growing skeleton, changes in bone size seem to be more important than changes in bone density7.
Methods Used to Measure Bone Density Several techniques are now available for the measurement of bone-mineral content and bone-mineral density, including single-photon absorptiometry, single and dual-energy x-ray absorptiometry, and quantitative computed tomography53,54,69,73,75. With each of these methods, bone density is determined by measurement of the attenuation of a beam of energy as it passes through the skeleton. As a screening test, single-photon absorptiometry provides a simple, rapid, and inexpensive method with which to assess bone density at sites within the appendicular skeleton (for example, the distal aspect of the radius, the middle of the forearm, and the calcaneus). Because the cortical area of the wrist is closely associated with total body calcium and because epidemiological studies have indicated that bone mass of the forearm can predict osteoporotic fracture at other skeletal sites, single-photon absorptiometry is a useful method for the measurement of menopausal osteoporosis and the effects of different bone-targeted therapies6. Dual-photon absorptiometry with transmission scanning from a radioisotope source was developed to provide a method with which soft-tissue density could be subtracted from bone density, allowing measurements to be made at sites within the axial skeleton (for example, the hip, pelvis, and spine) with increased accuracy. Dual-photon absorptiometry was subsequently replaced by dual-energy x-ray absorptiometry, which uses x-rays as an energy source. Currently, dual-energy x-ray absorptiometry is the preferred method for the assessment of bone-mineral status in clinical practice because of its accuracy, precision, stability, and low dose of radiation as well as the speed and ease of scanning7,8,75. Quantitative computed tomography provides an accurate measurement of the three-dimensional geometry of bone and its trabecular bone compartment, but it does so at the expense of increased exposure of the patient to radiation. The technique is now available for the evaluation of peripheral bones as well, with clearly lower doses of radiation. With radiographic absorptiometry, standard radiographs of the hands are subjected to computer-controlled analysis. Also, radiographic measurements of cortical width have been used in studies of bone loss. Ultrasound techniques, which provide information on bone mass and structure without the use of ionizing radiation, are relatively inexpensive and reliable, but their ability to predict future fractures requires additional investigation. The latest in the line of new techniques is quantitative magnetic resonance imaging, which has been shown to reveal the trabecular structure and architecture of bone. However, because of its cost and the time required for measurement, it is presently limited to use as a research tool27. As dual-energy x-ray absorptiometry is currently considered the method of choice for the assessment of bone density, it is important to recognize some practical points that should be kept in mind when such measurements are performed or when the results of such measurements are interpreted. First, the sites of measurement (regions of interest) should be constant between subjects, as bone-mineral values vary greatly between different bones of the skeleton and between different regions of an individual bone. Second, the regions of interest should be far enough from the surface of the joint (for example, at least fifteen millimeters from the femorotibial joint line in the knee area) to avoid the disturbing effects of the possible secondary changes caused by osteoarthrosis. Third, the use of a large enough analysis box (region of interest) has been shown to improve the precision of measurement, thus ensuring that changes found in the bone measurements are real. (We have used a forty-five-millimeter-high box for the measurement of bone-mineral density of the proximal aspect of the tibia and the distal aspect of the femur.) Finally, it was recently shown that the efficacy of a study may be enhanced further by use of anatomical regions of interest instead of regions of interest of fixed length75. The primary difference between these two regions is that the length of the anatomical region of interest is not fixed but rather is determined as a certain percentage of the total length of the bone in question. The height of the subject is used as a basic determinant of the individual bone length. In this way, exactly the same anatomical region (and, thus, the same average composition of trabecular and cortical bone) can be obtained from different subjects75.
An increase in the rate of mineral turnover is one of the early responses of bone to trauma, and several investigators have reported substantial reductions in bone mass after fracture of long bones, not only at the site of the fracture but also at adjacent sites proximal and distal to it1,18,84. Both the amount of bone loss and the extent of recovery have varied greatly among different reports. These differences may be due, in part, to the variable accuracy of the techniques used in the measurements of bone mass18 and to inaccuracies in the recording of numerous accompanying factors, such as the type of injury, type of medical treatment (operative intervention, duration of immobilization, and lack of weight-bearing), degree of pain after the injury, and impaired function of the limb, all of which are known to lead to decreased bone-mineral content in the injured extremity. For example, with tibial fractures, the immediate bone loss reported has ranged from 25 to 50 per cent1,8,84. Although acute bone loss after injury of an extremity is a well known phenomenon, the long-term effects of different types of fractures and other serious injuries to bone have been less well documented. We have observed considerable reductions in bone-mineral density in the bones of injured extremities ten years after initial injuries such as rupture of the rotator cuff, rupture of the knee ligaments, and fracture of the tibial or femoral shaft35-38. In one study, for example, the loss in the region of the knee of the injured limb was 4 to 11 per cent and the loss in the spine was more than 10 per cent nine years after a fracture of the tibial shaft that was treated non-operatively35. Because normal age-related bone loss is less than 1 per cent per year after the age of peak bone mass, permanent bone loss due to a previous injury (post-traumatic osteoporosis) may be a risk factor for osteoporotic fractures in later life.
Rupture of the Rotator Cuff After the bone-mineral density was correlated with various variables describing the outcome of treatment, we concluded that the relative decrease in bone density in the injured extremity was not associated with the size of the rupture, delay between the injury and the operation, type of operation or postoperative treatment, duration of postoperative immobilization, duration of follow-up, age of the patient, muscle strength, level of pain, or subjective assessment of function of the shoulder37. However, the bone density in the injured limb was strongly associated with the objective function of the shoulder the better the observed function of the shoulder, the less bone loss following the injury. The importance of disturbed function of the shoulder in the development of osteoporosis was also observed by Lundberg and Nilsson49, who measured changes in the bone-mineral content in the proximal aspect of the humerus in patients who had so-called frozen-shoulder syndrome. Although their method was fairly imprecise and the measured region of interest was small, the results indicated that chronic frozen-shoulder syndrome leads to a considerable (as much as 50 per cent) loss of bone-mineral density in the proximal aspect of the humerus of the injured extremity.
Fractures of the Distal Aspect of the Forearm (Colles Fractures) In a one-year prospective study, Westlin88 quantitated the changes in the bone mass in the radius and ulna of nineteen women who had sustained a Colles fracture at the age of forty-nine to seventy-nine years. When the bone density was compared with that on the contralateral side, the average decrease was 18 per cent in the shaft of the ipsilateral radius and ulna four months after the Colles fracture; no recovery was noted after one year. Nilsson and Westlin61 measured the bone-mineral content in the forearm of seventy-four perimenopausal and postmenopausal women who had sustained a Colles fracture one month to twelve years previously. They found an average side-to-side difference in the bone-mineral content in the proximal part of the forearm of 9 per cent. They also identified a trend for the amount of post-traumatic osteoporosis to decrease as the time after the injury increased. Those authors also studied sixty-nine patients who had sustained a fracture of the surgical neck or shaft of the humerus or a fracture of the shaft of both bones of the forearm62. Neither the fracture of the surgical neck nor that of the shaft of the humerus caused any post-traumatic osteoporosis, whereas the fracture of both bones of the forearm caused an average side-to-side difference of 15 per cent in the bone-mineral content in the distal part of the forearm. Mallmin et al.51 used single-photon absorptiometry to examine seventy-four patients who had sustained a Colles fracture as the result of a low-energy injury (a fall from a standing height or from less of a height) two months earlier. Control subjects who did not have a fracture were matched by age, gender, and years since menopause. The patients had an average 11 per cent reduction in the bone density in the distal end of the forearm compared with that of the controls. Restoration of bone mass in the injured extremity was not noted in any of these studies. This may be because most patients who have a Colles fracture are women who have postmenopausal osteoporosis and thus have a limited capacity for restoration of bone mass51.
A major difference between the bones of the lower extremity and those of the upper extremity is the amount of weight that they bear. With the exception of the patella, the bones of the lower extremity bear much higher loads than those of the upper extremity. Thus, a fracture or other major injury of the lower extremity, and the associated disability, reduction in use, reduction in weight-bearing, and immobilization may lead to a more prominent loss of bone than that found after an injury of the upper extremity.
Fractures of the Femur Finsen et al.24 used single-photon absorptiometry to study the bone-mineral density in twenty-six patients an average of 3.5 years after a fracture of the femoral shaft. Fourteen patients had been managed with intramedullary nailing and twelve, with plate fixation; the bone loss was almost identical in the two groups. Compared with the value in the contralateral limb, adult bone density was decreased by an average of 12 per cent in the distal aspect of the femur, 5 per cent in the proximal aspect of the tibia, and 3 per cent in the tibial shaft. The loss of bone-mineral density was significantly associated with the loss of muscle strength (p < 0.05). We36 used dual-energy x-ray absorptiometry to measure the bone density at different sites in the lower extremities of twenty-nine adult men who had sustained a fracture of the femoral shaft ten years earlier. The patients had initially been managed with an intramedullary nail, which had been removed one to four years after the injury. Compared with the value for the uninjured side, the bone density in the injured extremity was significantly lower in the distal aspect of the femur (-6.8 per cent, p < 0.0001), the patella (-5.4 per cent, p < 0.0001), the proximal aspect of the tibia (-4.7 per cent, p < 0.0001), and the calcaneus (-2.2 per cent, p = 0.03) (Fig. 1, D). No significant difference was found in the femoral neck, while the bone density in the trochanteric area was 6.3 per cent higher in the injured limb than in the uninjured limb. The fracture of the femoral shaft also seemed to result in permanently reduced bone density in the lumbar spine (-9.4 per cent) compared with that in age, height, and weight-matched normal men (Fig. 2). The relative bone densities in the injured extremities were not associated with the type or location of the fracture, the age of the patient, muscle strength, or the duration of follow-up or non-weight-bearing, but they were associated with minimum pain and high functional scores36.
Injuries of the Knee We38 used dual-energy x-ray absorptiometry to study the effects of knee-ligament injuries of varying severity on the bone density of the involved extremity. The bone-mineral content and density were measured at various sites in the lower extremities of forty-two patients an average of ten years after the injury. Eleven patients who had sustained a moderate injury (an isolated rupture of the medial collateral ligament) and who were managed operatively with primary repair of the torn ligament and postoperative immobilization in a plaster cast from the ankle to the groin for six to seven weeks did not have any side-to-side differences in bone density. In the thirty-one patients who had sustained a severe injury (a complete rupture of the anterior cruciate ligament alone or in combination with ruptures of other ligaments) and had been managed operatively with primary repair of the ligament and postoperative immobilization in a plaster cast for six to seven weeks, the average bone loss was 6.0 per cent in the distal aspect of the femur, 9.0 per cent in the patella, and 3.3 per cent in the proximal aspect of the tibia of the injured limb (Fig. 1, A). The calcanei and femoral necks showed no significant side-to-side differences. The age of the patient and the knee stability score38 were not significantly associated with the relative bone density around the injured knee, but all of the scores used to assess function of the knee (the score of Lysholm and Gillquist50, the score of Tegner and Lysholm82, and the International Knee Documentation Committee score38) showed a significant positive association with the bone density in the distal aspect of the femur (p < 0.001, p < 0.001, and p < 0.01, for the respective scores), the patella (p < 0.01, p < 0.001, and p < 0.001), and the proximal aspect of the tibia (p < 0.01 for all three scores); the better the functional score of the injured knee, the better the bone density in that knee. In a unique case study, Sievänen et al.74 used dual-energy x-ray absorptiometry to determine the effects of one year of strenuous strength-training and subsequent rupture of the anterior cruciate ligament on one side on the bone-mineral density in the lumbar spine and the lower extremities. Both the deleterious effect of the rupture and the beneficial (although modest) effect of the preceding training were demonstrated. The immediate post-traumatic decrease in the bone density of the injured extremity was rapid, resulting in a decline of approximately 20 per cent. One year after the injury, the function of the knee, the muscle performance, and the use of the injured limb were completely recovered, but the site-specific bone densities were still about one standard deviation (about 10 per cent) below the baseline bone densities of the subject. After additional follow-up (two years after the injury)72, the results indicated that more recovery had occurred, but the pre-injury values had not been reached. In addition to ligament injuries, meniscal tears also typically affect the function of the knee joint. The long-term effects of medial meniscectomy on bone-mineral content and density in the proximal aspect of the tibia were determined by Petersen et al.66. Dual-photon absorptiometry was performed twelve years after a complete meniscectomy in nineteen patients and after a partial medial meniscectomy in fourteen patients. The average difference in the bone-mineral content between the injured and uninjured limbs after total meniscectomy was 2.8 per cent (the lower values were for the injured limb), whereas the average side-to side difference after partial medial meniscectomy was 1.3 per cent. A more detailed analysis of the scans also revealed that extensive bone-remodeling had occurred in the tibial condyles after the medial meniscectomy. This was attributed to altered loading of the knee joint. The lesser amount of bone loss observed in the patients who had had a partial meniscectomy was explained by a less severe injury and better postoperative function of the knee66.
Fractures of the Tibial Shaft Andersson and Nilsson1 used single-photon absorptiometry to determine bone density in twenty-seven patients who had had a fracture of the tibial shaft. They found an average 45 per cent decrease in bone-mineral content in the proximal end of the tibia five months after the injury and an average 25 per cent decrease one year after the injury. In another study3, the same authors evaluated twenty-seven patients who had been managed with immobilization of the limb, either in a cast that allowed weight-bearing during the period of fracture-healing or in an above-the-knee cast that did not permit weight-bearing, after a fracture of the tibial shaft. At one year, there was a small but insignificant difference between the two groups with regard to the total bone-mineral content. The patients who had been managed with a weight-bearing cast had an average bone deficit of 24 per cent whereas those who had been managed with an above-the-knee cast had an average deficit of 29 per cent. Interestingly, in a similar study in which single-photon absorptiometry was used to examine twenty patients who had had a fracture of the tibial shaft two and a half years earlier, the average difference in the bone-mineral content between the injured and contralateral limbs was 7 to 8 per cent in the distal femoral metaphysis and the proximal and distal tibial metaphyses, but only 3 per cent in the femoral and tibial diaphyses22. In a prospective study, Ulivieri et al.84 used dual-energy x-ray absorptiometry to study the acute loss of bone in seven patients who had had operative treatment of a fracture of the tibial shaft. At follow-up intervals of ten to 120 days, the authors measured the bone loss in the portion of the tibia that was distal to the site of the fracture. Total bone-mineral content and bone density of the fractured limb progressively decreased, reaching the maximum reduction (approximately 50 per cent of the initial values) 120 days after the injury. The corresponding values for the contralateral limb remained unchanged. Karlsson et al.46 used dual-energy x-ray absorptiometry to study bone density twenty years (thirty-eight patients) and thirty years (twenty-two patients) after a fracture of the tibial shaft. In the first group, they found an average deficit in the bone density in the trochanteric area of the femur and in the femoral condyles of approximately 4 per cent compared with the corresponding sites in the contralateral bones. In the patients who were analyzed thirty years after the fracture, similar bone loss was found in the femoral condyles only. We35 also performed a study with dual-energy x-ray absorptiometry to investigate the effects of two types of treatment of fractures of the tibial shaft on the bone densities in the injured extremity and in the lumbar spine. In the first group, twenty patients were managed with immobilization in a plaster cast from the toes to the groin for an average of sixteen weeks. All of the fractures healed. Approximately nine years after the injury, the average bone densities in the injured extremity were significantly lower than those in the uninjured extremity: they were 2 per cent lower in the femoral neck (p = 0.02), 4 per cent lower in the distal aspect of the femur (p = 0.004), 4 per cent lower in the patella (p = 0.009), and 5 per cent lower in the proximal aspect of the tibia (p = 0.005) (Fig. 1, B). No significant side-to-side differences were found in the distal aspect of the tibia or the calcaneus. The second group consisted of fourteen patients who had a similar fracture that was also treated with a plaster cast. These fractures did not heal, and eventually they were successfully treated with bone-grafting and immobilization in a cast. Altogether, the injured extremity of these patients was immobilized for an average of twenty-seven weeks. Compared with the values for the uninjured side, the average bone density of the injured extremity was significantly lower in the distal aspect of the femur (-10 per cent; p = 0.0001), the patella (-11 per cent; p = 0.0003), the proximal and distal aspects of the tibia (-9 and -8 per cent; p = 0.0005 and 0.002), and the calcaneus (-6 per cent; p = 0.02) (Fig. 1, C). In the femoral neck, the difference (2 per cent) was not significant. The relative deficits in bone density in the injured extremities differed between the groups of united and non-united tibial fractures. With the exception of the femoral neck, the bone loss was always greater in the latter group, significantly so in the distal aspect of the femur (p = 0.008), the patella (p = 0.006), and the distal aspect of the tibia (p = 0.02). We also found that, compared with male control subjects (men who had not had a fracture), the men who had had a fracture of the tibial shaft had significantly lower bone density in the lumbar spine (F = 5.35, p = 0.008) regardless of whether the fracture had united (average decrease in bone density, 10 per cent) or had gone on to non-union (average decrease in bone density, 12 per cent) (Fig. 2). Neither the fracture type (transverse, oblique, or comminuted) or location (middle or distal third of the shaft) nor the age of the patient had an effect on the bone-mineral loss. However, the duration of immobilization, the pain assessment, the muscle strength, and all three functional scores (the score of Lysholm and Gillquist50, the score of Tegner and Lysholm82, and the International Knee Documentation Committee score38) were significantly associated with the relative bone-mineral density; the shorter the duration of immobilization, the less activity-induced pain in the injured extremity, and the better the muscle strength and functional scores, the higher the relative bone density in the injured extremity and the less bone loss due to the injury (r = 0.42 to 0.78, p < 0.01 to 0.001). Tandon et al.81 performed a prospective follow-up study of similar types of stable and unstable fractures. Twenty-two patients who had a stable tibial fracture were managed with a plaster cast, and fifteen patients who had an unstable fracture of the tibial shaft were managed with external fixation. Both groups were encouraged to bear as much weight as possible beginning on the day after application of the cast or fixator. At four and eight weeks, the patients who had external fixation were found to be bearing a greater amount of weight. Although the patients who had an unstable fracture had a more severe injury, plain radiographs suggested that they had less osteoporosis in the injured tibia sixteen to twenty weeks after the injury. This difference between the groups was attributed to the fact that the patients who were managed with external fixation had improved early function and weight-bearing. Unfortunately, the estimation of bone loss was based on the analysis of plain radiographs only, and no quantitation was performed. Van der Wiel et al.85 prospectively evaluated the loss of bone in the proximal part of the ipsilateral and contralateral femora and in the lumbar spine of seven men and nine women who had a fracture of the tibia. All of the fractures were initially unstable. Twelve patients were managed operatively, and four were managed with immobilization in a plaster cast. All patients walked without bearing weight for an average of eight weeks. Dual-energy x-ray absorptiometry was used to follow the changes in bone density. During the non-weight-bearing period, the bone density decreased in the ipsilateral trochanteric region of the femur and in the lumbar spine by an average of approximately 9 per cent and 1 to 2 per cent, respectively. At one year, the average decrease in bone density in the ipsilateral trochanteric region of the femur and femoral neck was about 15 and 6 per cent, respectively. Eyres and Kanis18 performed a prospective evaluation of the changes in bone density in five patients after a unilateral fracture of the tibia that was treated with intramedullary nailing or application of an external fixator. The densities of the right and left tibiae were measured soon after the fracture and were followed for six months. At three months, bone density distal to the site of the tibial fracture had decreased to approximately 50 per cent of the initial value, and it remained at that level for the next three months. In the second (retrospective) part of the study, the authors measured bone density in twenty-one adult patients who had had a fracture of the middle of the tibial shaft five to eleven years earlier. The patients still had approximately 47 per cent lower bone density in the distal aspect of the injured tibia compared with that in the uninjured tibia. Because the follow-up period was long, the bone loss was regarded as permanent. A profound increase in bone-mineral density (172 per cent) was found at the site of the old fracture. Further analysis of the results showed that there was no difference between the bone density in the injured tibia and that in the uninjured tibia in the patients who had sustained the fracture as a child. This finding supports the concept of Nilsson and Westlin57,59 that children have an excellent capacity for restoration of bone that has been lost secondary to fracture. Henderson et al.28 compared the bone density in the proximal part of the femur and the flexion-extension strength of the knee in fifteen children who had had an uncomplicated fracture of the tibia with those variables in twenty-three children who had had a fracture of the femur. The average difference in bone density between the injured and uninjured limbs was 3.5 per cent. There was no side-to-side difference in the bone density if the injured limb had been immobilized for less than four weeks; however, the bone density was an average of about 4 per cent lower at the involved site if the injured limb had been immobilized for more than eight weeks.
Fractures of the Ankle Petersen et al.65 used dual-photon absorptiometry to measure the bone-mineral content and bone density in the medial and lateral condyles of the tibia in twelve patients who had sustained a displaced fracture of the lateral malleolus. The period of non-weight-bearing in a plaster cast after the operation was six weeks. At six months, the bone-mineral content and bone density in the tibial condyles were 15 and 18 per cent lower on the injured side than on the contralateral side. In the medial tibial condyle, the maximum loss in bone density (12 per cent) was reached at three months, and the maximum loss of bone density (18 per cent) was measured on the lateral side six months after the injury. Slow restoration of bone mass was observed (2 per cent per year), but the bone-mineral content of the involved side still remained 11 per cent lower than that in the proximal aspect of the contralateral tibia three years after the injury. Interestingly, a small (5 per cent) loss in bone-mineral content was also detected in the contralateral tibia after one year; this was most likely due to a systemic effect of the malleolar fracture on bone tissue or to decreased physical activity of the patient, or both. From these reports, it can be concluded that fracture of the ankle may be followed by a marked and long-lasting loss of bone. Early weight-bearing and active exercises of the ankle seem to play a minimum role in the prevention of post-traumatic bone loss.
Overview
Because most of the previously mentioned long-term follow-up studies were retrospective and the original bone densities at the various skeletal sites were not determined in the patients before the injury18,35-38,46, two factors need to be considered in the interpretation of the results. First, it can be speculated that there was a natural difference in bone density between the non-dominant and dominant limbs and, as a consequence, a fracture occurred in the limb with the lower bone density. This, however, does not seem probable as it has been shown previously that there is no difference in bone density between the dominant and non-dominant lower extremities of uninjured persons36. Second, it is possible that the observed differences in bone density between the injured and uninjured limbs were due to bone hypertrophy in the uninjured limb rather than to bone loss in the injured limb. Previous follow-up studies, however, have shown that, after an injury, bone loss occurs not only in the injured limb but also in the uninjured limb. This loss occurs to a lesser extent in the uninjured limb than it does in the injured limb and is most likely the result of decreased activity1,22,56,65,82,85. Thus, the observed side-to-side differences may actually have been an underestimation of the true loss of bone in the injured extremity. Some investigators have suggested that the bone losses reported in the long-term studies are probably permanent because the average follow-up interval was ten years or more and because previous, short-term studies have indicated that a large part of the restoration process takes place within the first year after an injury1,21. On the contrary, some recent long-term studies have shown that the restoration process of bone can continue even after the first year, as the losses in bone density recorded in those investigations were smaller than those reported in the short-term studies of similar fractures1,61,70,84,85. For example, Van der Wiel et al.85 found a 6 to 7 per cent decrease in the bone density of the ipsilateral femoral neck one year after a fracture of the tibia. In our cross-sectional studies35 and those of Karlsson et al.46, nine and twenty-one years after a similar fracture the average side-to-side differences were only about 2 and 4 per cent. Additional evidence for restoration of bone mass was provided by Karlsson et al.46, who observed that patients who had had a fracture of the tibial shaft and those who had had an injury of a knee ligament had markedly smaller deficits (side-to-side differences) in the bone density in the injured limb fifteen to thirty-eight years after the injury compared with measurements obtained from the same patients in the initial years after the injury2,56. It must be noted, however, that some of this restoration may be due to differences in the measurement technique, as the initial studies2,56 were done with single-photon absorptiometry and the long-term follow-up study46 was done with dual-energy absorptiometry. Also, during the long-term follow-up periods, there may have been a reduction in the bone mass of the contralateral limb, thus decreasing the measured side-to-side differences even further. In summary, the side-to-side comparisons that have been reported suggest that it might be possible to reverse trauma-induced osteoporotic changes years after the injury. It is presumed that bone can be restored by encouraging patients to increase their level of physical activity. Increased physical activity is known to stimulate new-bone formation in certain settings30,39,41,78,79. However, recent experimental and clinical studies have indicated that, in order to restore lost bone in certain situations, the level of physical activity has to exceed the normal activity level, be maintained on a regular basis, and be continued for a long (perhaps indefinite) period30,40-42.
Immobilization A recent finding by Houde et al.30 suggests that immobilization-induced bone loss continues for weeks after the cessation of immobilization; therefore, immobilization may cause an impairment in the bone-cell function that lasts longer than the time that the limb is actually in the cast30,87.
Impaired Function of the Limb Patients who have a severe injury of a knee ligament or a fracture of the tibial shaft have been shown to have the greatest loss of bone density in the patella35,38. Because the patella is essentially a non-weight-bearing bone, receiving most of its mechanical stimulation from tensile and compressive loads created by flexion and extension of the knee, the effect of an injured knee ligament on patellar bone loss may result from an impairment of normal flexion and extension. There is a strong positive association between the prevalence of impaired function of the knee and the finding of a substantial difference in bone density between the injured and uninjured sides in these patients. This suggests that, in patients who have a knee injury, patellar bone loss may occur as a result of diminished function of the quadriceps, resulting in diminished mechanical stimulation of the patella35,38. Thus, for certain bones, impaired function of the limb may very likely play an important role in loss of density.
Non-Weight-Bearing The importance of weight-bearing in minimizing the amount of post-traumatic osteoporosis was shown in patients who had a non-union of a fracture of the tibial shaft35,81. For those patients, the duration of immobilization and, especially, the period of non-weight-bearing were considerably longer than they were for a group of patients who had a union. This difference was probably the main cause of the prominent bone loss found in the former group, a loss that was still clearly visible ten years after the injury in nearly all bones studied35. Van der Wiel et al.85 showed that, after a tibial fracture, the loss of bone density in the lumbar spine took place during the period of non-weight-bearing only, while the loss of bone in the injured limb continued for months after weight-bearing was reinitiated. This finding supports the previous speculation that non-weight-bearing is the most important single factor in the determination of post-traumatic bone loss in the lumbar spine35,36,48. A substantial loss of bone-mineral density was found in the lumbar spine of patients who had a united or non-united fracture of the tibial shaft or a fracture of the femoral shaft35,36. This bone loss was greatest in the patients with a non-united fracture of the tibial shaft, who also had the longest period of non-weight-bearing, and was least in the patients with a fracture of the femoral shaft (Fig. 2), who had the shortest period of non-weight-bearing. We38 showed that years after a severe injury of the cruciate ligament the bone-mineral density of the lumbar spine was normal. Similar results were obtained by Sievänen et al.74, who, in their case study, found no decrease in the bone density of the lumbar spine either immediately after an injury of the anterior cruciate ligament or during the one-year rehabilitation period, although marked reductions in bone-mineral density (as much as 20 per cent) were found in the bones of the injured extremity. The lack of bone loss in the spine may be explained by the type of treatment used for these patients. Unlike fractures of the femoral or tibial shaft, injuries of a knee ligament are commonly treated with immediate partial weight-bearing and muscle-strengthening exercises, thus avoiding the deleterious effects of immobilization and non-weight-bearing. Therefore, it is likely that, after severe fractures of the lower extremity, the relatively long periods of non-weight-bearing and disuse are the primary reasons for loss of bone in the lumbar spine.
Age and Gender After a fracture of the tibial shaft, acute bone loss in the distal aspect of the femur is generally somewhat more pronounced in male patients than in female patients56. This may be attributable to the higher peak bone mass found in male patients56. However, there may be less permanent post-traumatic osteoporosis in male patients because they may have a greater potential for restoration of the lost bone56. Bone loss may not be restored in elderly patients, as little or no restoration of bone-mineral content was identified in postmenopausal women who had sustained a Colles fracture60,88.
Karlsson et al.43-45 and Finsen et al.23 investigated whether a severe injury of the lower extremity is a risk factor for a subsequent new fracture of a bone in the ipsilateral limb. In an epidemiological study, Finsen et al.23 found that patients who had had a previous fracture of the femoral neck, femoral shaft, or tibia had a greater risk of a subsequent fracture on that side than patients who had not had a previous fracture. They also found that femoral and patellar fractures were more often observed on the same side as the previous fracture of the lower extremity. Those authors concluded that post-traumatic osteoporosis may play a role in the pathogenesis of certain fractures. Karlsson et al.43 studied whether fractures sustained earlier in life can be predictive of osteoporosis-related fractures later in life. They observed that persons who had had a previous fracture of the tibia or ankle had an increased risk of osteoporosis-related fractures during the three to four decades after the initial fracture. More detailed analysis showed that a previous fracture of the ankle was associated with a twofold increase in the occurrence of new fractures and that there was a 24 per cent higher prevalence of new fractures in the injured lower limb than in the uninjured, contralateral limb43. The corresponding values for a fracture of the tibial shaft were almost a 2.5-fold increase in the total number of new fractures and a 35 per cent higher prevalence of new fractures in the previously injured lower limb than in the uninjured, contralateral limb45. However, in this context it should be noted that the increased prevalence of refracture in the patients who had had a previous fracture cannot be attributed solely to the osteoporosis induced in the bones of the injured limb by the injury43-45; these patients may well have been affected by other factors (such as an injury-provoking lifestyle) that rendered them more prone to injuries in general.
Bone loss in an injured extremity seems to evolve in two phases. First, acute bone loss occurs during the early post-traumatic period and is very rapid. The fact that the amount of immediate bone loss seems to depend on the severity of the injury suggests that the injury itself has a catabolic effect on bone. However, it must be kept in mind that severe injuries are usually associated with longer periods of immobilization, more complete loss of weight-bearing, and pain, all of which may also contribute to the increased loss of bone. In the latter, more chronic phase, factors such as functional impairment of the involved extremity and persistent pain may influence the amount of bone lost. These effects are usually long-lasting and may prohibit patients from using the extremity, thus interfering with the normal process of restoration of bone tissue. Over-all, these observations support the concept that the treatment of injuries of the lower extremity should involve short periods of immobilization, early weight-bearing, and a well planned program of rehabilitation4,40. Patients should also be encouraged to increase their level of general physical activity in order to minimize age-related bone loss and the subsequent risk of osteoporotic fractures.
*No benefits in any form have been received or will be received from a commercial party related directly or indirectly to the subject of this article. No funds were received in support of this study.
  CiteULike  Connotea  Del.icio.us  Facebook  Technorati  Twitter What's this?
This article has been cited by other articles:
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Top
Introduction
Bone Density as a...
| home | CME | help | feedback | subscriptions | archive | site index |
|
Stanford University Libraries' HighWire Press (TM) assists in the publication of JBJS Online. |