Muscle Strength, Bone Mass, and Age-Related Bone Loss

THE EFFECTS OF DISUSE, or conversely, of elevated physical activity on muscle strength and bone mass are well known. Disuse causes muscle wasting and bone loss; physical activity increases muscle strength and bone mass. The association between muscle strength and bone mass is therefore clearly established, although neither the dominant effect of muscle on bone mass, nor the cause and effect relationship is established by these simple correlations alone. Frost1 proposes that “voluntary muscle forces … dominate a bone's postnatal structural adaptations to mechanical usage, modified … by body weight and one's voluntary physical activity.” Neither body weight nor physical activity is independent of muscle mass, but it is true that muscle forces place greater loads on bones than do gravitational forces associated with weight. This was shown theoretically nearly 50 years ago by Pauwels,2 and is easily deduced from a simple analysis of muscle pull and levers such as the one described by Frost. More recently, this has been demonstrated using experimental data from implanted hip prostheses.3 This analysis shows that >70% of the bending moments on a bone are transmitted by muscle force rather than body weight. Computational models employing finite elements together with the application of computer-aided optimization to predict femoral geometry during growth are unable to simulate the morphology of the femoral shaft without the addition of muscle forces.4 Application of various combinations of axial, bending, and torsional loading to a cylinder were not able to generate femoral diaphyseal morphology when compared with the cross-sectional geometry of real bones, but the introduction of adductor muscle forces in various combinations with these loadings did produce geometries comparable to observed femoral geometry. Therefore, based on the predictions of theoretical analyses and computational models, and based on supporting experimental data, it is well established that forces applied to bone are primarily the result of muscular contraction. To demonstrate that muscle force dominates skeletal adaptation requires that greater muscle mass/strength is associated with greater bone mass, independent of body size contributions. A multitude of studies have shown such correlations,5-9 but have not always demonstrated their independence from other body size measures.10 Schonau et al.,8 for example, show that as much as 76% of the variation in bone strength index in the distal radius can be explained by grip strength alone. However, high correlations between muscle strength/mass at one site and bone strength/mass at another unrelated site suggest that the relationship may be related to other confounding factors,11 or to stimuli that have independent effects on both systems.12 Myburgh et al.5 showed that 67% of the bending rigidity of the ulna can be explained by biceps strength, but the biceps has no attachment to the ulna and would not directly generate loads on the ulna. Likewise, muscle strength of the quadriceps and hamstring muscles independently predicts bone mineral density (BMD) of the humerus and spine.13 Inferences about the direct effect of muscle strength on bone adaptation based on such correlations are spurious. Still, there are studies that demonstrate that muscle strength has effects on bone mass or BMD that are independent of age, weight, height, or years of estrogen use.6,14,15 The effect of muscle strength on bone mass has been complicated somewhat by the failure to find correlations between muscle strength and bone mass in subjects with normal or even elevated bone density. In the hormonally replete adult, mechanostat theory16 predicts that once skeletal loading by muscle activity brings bone density to within the normal range, remodeling will be suppressed. Because modeling is inactive or slow in adults, higher bone masses will not be achieved. Consequently, correlations between muscle strength and BMD in active or athletic individuals are generally not high13,17,18 but are not expected to be. The absence of a correlation between muscle strength and bone mass in active subjects is not a sufficient argument against the idea that muscle force exerts a dominant effect over bone's structural adaptation. To demonstrate rigorously that voluntary muscle forces dominate the adaptation of bone it is necessary to show that a decline in muscle mass precedes a decline in bone strength under conditions of disuse or hormonal insufficiency, and conversely that muscle strength increases before bone mass under conditions of hypervigorous mechanical usage or the re-establishment of hormonal balance. The time course for the development of muscle weakness relative to bone loss following injury, and reversal of these trends subsequent to rehabilitation and recovery, recently was shown nicely during a “fortuitous” experiment in which a knee ligament rupture occurred during a strength training program.7 In this study, a healthy, physically active 26-year-old woman participated in a lower limb strength training program for 46 weeks. At week 57, she ruptured her left anterior cruciate ligament (ACL) in an unrelated activity. A ligament reconstruction was performed arthroscopically 10 days after the injury. Following ACL injury, the isometric extension strength of the injured limb immediately fell to zero, eliciting an immediate and total reduction in strain applied to the patella. Bone mineral apparent density (BMAD) subsequently declined by ∼25% over the 14 weeks following the injury. About 18 weeks following injury, she began rehabilitation. Extension strength increased, and strain on the patella was nearly triple that prior to injury (presumably because of the effects of bone loss), but there was no noticeable increase in BMAD (Fig. 1). Isometric muscle strength reached a peak about 44 weeks following the injury, when less than half of the density loss had been recovered. Two years following injury, and more than 1 year after isometric extension strength of the lower limb had returned to normal, normal BMAD was nearly recovered. Strain index, muscle strength (upper) and bone mineral apparent density (BMAD, lower) over a 3-year training program before and after rupture of the anterior cruciate ligament at week 57 (vertical dashed line). Solid lines represent the limb with ACL rupture; dashed lines represent the uninjured control limb. Wide solid line (upper panel): Patellar strain index, defined as the product of the maximal isometric strength during lower limb extension and the patello-femoral contact area, normalized to bone mineral density3. Narrow solid line (upper panel): Isometric extension strength of the lower limb. Solid line (lower panel): BMAD of the patella. Note that strain index and extension strength dropped immediately after the injury; BMAD dropped to a minimum by 12 weeks following injury. Extension strength (narrow solid line, upper) began to increase prior to recovery of BMAD (solid line, lower). Extension strength peaked more than 1 year before BMAD returned to preinjury values. (With permission from Sievänen et al.(7)) This natural experiment demonstrates several important concepts crucial to Frost's arguments. First, in cases of disuse, the loss of muscle strength precedes the loss of bone mass. Second, the recovery of muscle mass precedes the normalization of BMD. These data, then, prove the direct cause and effect association of muscle strength on bone mass. However, because this was an experiment of nature, it was not possible to include negative controls (the contralateral leg served as an adequate positive control). Ideally, an animal experiment could be conducted in which ligament resection and repair were performed, but the return of muscle strength was prevented. This would prove that the recovery of bone mass was solely the effect of muscle recovery. This experiment demonstrates the lag that Frost predicts between the relatively rapid response of muscle and the sluggish adaptation of bone to a change in mechanical environment. This lag is partly explained by the relative rates of adaptation in each of these tissues, but is also related to bone's “stiffness lag”19 which involves the delay between the formation of new bone and the establishment of fully mineralized and mature bone. This creates a potential problem in rehabilitation, because for a period of time the muscle will be much stronger than the bone. At this time, the muscle, pulling with great force, can cause avulsion fractures. An analogous situation is found in rapidly growing and very active adolescents, who will occasionally avulse their tibial tubercle (Osgood-Schlatter's disease) because the muscle is strong while the strength of the rapidly growing bone is insufficient for the muscle force placed on it. Frost proposes that the effect of decreasing muscle strength after the age of 30 is to place bones adapted to the stronger muscles of youth into a virtual “disuse” situation, subsequently stimulating remodeling and the net loss of bone. The accuracy of this view depends on the natural history of the decline in muscle strength during the lifespan, and its relation to patterns of age and hormone-related bone loss. Marcus11 claims that age-related bone loss cannot be ascribed to loss of muscle because bone loss occurs earlier than the loss of muscle strength. This view is supported by some research,20,21 much of which suggests that muscle strength begins to decline in the latter portion of the fifth or early sixth decades.22-26 Vandervoort and McComas27 showed that maximum voluntary torque of ankle dorsi- and plantarflexors was maintained at least to age 50, although contraction times appear to be prolonged beginning in the fourth decade. Narici et al.28 showed that both grip strength and relaxation times of the adductor pollicis began to decline between 55 and 60 years of age. However, other studies demonstrate that loss of muscle strength begins shortly after the age of 30,29,30 progressing slowly at first but more rapidly around the age of menopause (Fig. 2).31-33 Timing of the strength loss varies by anatomical location, occurring slightly later in the lower extremities, where bone tends to be lost earlier, than in the upper extremities.34 The loss of muscle strength beginning in the fourth decade correlates well with morphological data which show that muscle fiber number reaches a peak at ∼25 years of age and decreases ∼40% by age 80. Maximum isometric strength of three muscle groups plotted as a percentage of the strength measured for a group of 30-year-old men, and running speeds for a 200 m sprint expressed as a function of maximum running speed. These data show that maximum isometric strength of most muscle groups in the back, arm, and leg begins to decline about the age of 30 years. (Used with permission from Brooks and Faulkner.(33)) These studies, however, tend to be cross-sectional and do not consider the possible interaction of secular or cohort effects.29 Kallman et al.29 followed 864 subjects (not divided by gender, unfortunately), 342 of which were followed longitudinally for 5–13 years. These data show a clear decline in grip strength beginning on average between 40–50 years (Fig. 3). People are variable, and not everyone begins to lose strength at the same age or at the same rate. The averaged data show that muscle strength increases, on a populational basis, during the fourth decade (age 30–39). However, more than half (52%) of the subjects lose grip strength when they are younger than 40 years of age. By age 60, 71% of all subjects had lost grip strength. The mean data obscures the fact that most people do lose muscle strength well before the onset of bone loss. Regression showing the relationship between grip strength and age in 847 men and women. Grip strength begins to decline by age 40 in this sample of combined genders. (Used with permission from Kallman et al.(29)) Male/female differences in the decline of muscle voluntary force and fiber area closely mirror the gender-specific periods of most rapid bone loss. In women there is a significant decline in muscle strength around the time of menopause. In men, muscle strength may be preserved until about the age of 60, not reaching the level found in postmenopausal women until the age of 75.32 The decline in muscle strength may be prevented by hormone replacement therapy (HRT) with estrogen,32,35,36 although some investigators have not found a relationship between estrogen use and muscle strength.37,38 What is crucial to the relationship of muscle mass and bone strength is the demonstration that the introduction of HRT causes increased muscle strength that precedes the estrogen-induced reduction in the rate of bone loss. That experiment, to my knowledge, has not been done. The extent to which mechanical loading dominates the process of bone gain and loss over other nonmechanical factors can be determined by studies which combine variations in hormone/cytokine levels with alterations in mechanical loading of the skeleton. The mechanostat proposes that bone gain and loss are determined within ranges of mechanical stimulation that are bounded by hormonally or metabolically determined setpoints.16,39 The setpoints do not themselves determine whether bone will be gained or lost, only when the remodeling system will be activated above baseline levels, or inactivated. Mechanical usage modulates an activated remodeling system and determines bone balance. If mechanical loading can prevent most or all of the bone loss that occurs from estrogen deficiency, then one might consider the effect of mechanical loading dominant. Human studies suggest that mechanical loading will not prevent bone loss in the presence of low estradiol levels,40 but human studies are difficult to control and are equivocal in answering this question.41 A recent study that examined the effects of estrogen deficiency with either skeletal loading or unloading in the rat model suggests that bone turnover rates are determined by estrogen, but the balance between resorption and formation is determined by the mechanical environment. Skeletal regions such as the distal femoral epiphysis where strain energy is relatively high appear resistant to bone loss, even in the face of estrogen deficiency,42 whereas bone loss will occur following ovariectomy (OVX) in regions where strain energy is lower (the distal femoral metaphysis). Bone of the epiphysis may be resistant to estrogen-deficient loss, but it is not resistant to loss from reduced mechanical strain. Lowering mechanical strain by unloading either through exposure to reduced gravity in space or by sciatic neurectomy will cause significant bone loss from the epiphysis, even without OVX. Finally, in the tibial metaphysis, ovariectomy significantly (p < 0.01) reduced trabecular bone area (Tb.Ar), from ∼22.5% to ∼10%. Treadmill exercise following OVX prevented much of that loss; treadmill exercised rats had nearly as much Tb.Ar as sham animals, ∼19%, and were not significantly different than controls. This is consistent with some other studies that show that exercise can prevent bone loss related to estrogen deficiency.43,44 These data support the proposal that loading dominates the control of bone balance, although hormonal deficiency will cause some loss independently through an increased activation frequency. The idea that mechanical influences dominate control of bone gain during growth and bone loss during aging has more value than as simply an academic argument. The idea strikes at the heart of therapeutic measures for controlling bone mass, which have previously all concentrated on attempting to control the hormonal or metabolic influences on the skeleton. Frost's suggestion is that some loss might be prevented if mechanisms can be found to prevent muscle wasting with age. There is evidence that muscle strength can reduce the risk of vertebral fracture and the development of kyphosis in older women with osteoporosis.9 Some data suggest that estrogen might act to prevent bone loss partly by preserving muscle strength, although its direct effects on bone cells are well known and clearly significant in this process as well. Targeting muscle is an entirely new approach. It may or may not be correct, but it deserves experimental testing to determine the limits of bone response to muscle forces and its independence from related, but confounding, factors. It is likely that both mechanical and nonmechanical factors are important and that the effect of mechanical stimuli are highly dependent on the hormonal/metabolic environment, which acts to alter the sensitivity of cells to mechanical stimuli (i.e., alter the setpoint).16,40,45,46 Current drug therapies for osteoporosis act by changing the activation frequency for bone remodeling (even though some are commonly referred to as antiresorptive agents, they in fact depress both resorption and formation and are not true antiresorptive agents), not by altering the balance between resorption and formation. The latter will be necessary to reverse the effects of osteoporotic bone loss (as opposed to preventing loss, which can be accomplished with current therapies). Arguing whether mechanical stimulation or hormonal/metabolic factors are dominant neglects the more important issue: both mechanical and nonmechanical factors are crucial to reverse bone loss, and it is the interaction between the two that needs to be understood and that holds the key to success.