Exercise in the Prevention of Age-Related Fragility Fractures (Narrative Review)

Abstract

Background

Loss of bone integrity and the age-associated decrease of the neuromuscular function make elderly subjects prone to fragility fractures.

Summary

Exercise is a strategy to counteract these age-associated changes and impairments. Because of the tight relationship between muscle and bone – anatomically, functionally, and biochemically – physical activities and targeted exercises, which induce muscle contraction and sufficient mechanical stress, influence bone metabolism. Exercise proved to have a positive effect on bone mineral density. The young skeleton is especially susceptible to impact and strenuous stimuli. This also applies to the neuromuscular system in the case of balance training. Therefore, the best time to start preventing fragility fractures is at young age. Despite the lower responsiveness in older age, targeted training is also very important at an advanced age. Lowering the modeling threshold, osteoanabolic treatment seems to increase the responsiveness to mechanical loading. In case of antiresorptive treatment, a more intensive training may be necessary.

Key Message

A multiple component exercise intervention reduces the risk of age-associated fragility fractures. Depending on the mode of exercise, it mainly affects bone integrity or the neuromuscular system. The effect of exercising also depends on age and bone-specific medications influencing the sensitivity of these structures. However, despite a lower sensitivity to exercise at higher age, targeted training is especially important when getting older to decrease the risk of fragility fractures. In case of prevalent fragility fractures, patients should exercise as well; the training stimulus simply needs to be adapted.

Introduction

According to the algorithm for the management of patients at risk of osteoporotic fractures, independent of fracture risk, all patients – those at low, high, and very high risk of fragility fractures – should exercise regularly [1]. Besides pharmacological management, risk appropriate exercise and fall prevention are essential components of primary and secondary fracture prevention. Exercise interventions have been shown to increase bone mineral density (BMD) [2] and reduce the risk of falls in the elderly [3], but primary prevention begins much earlier. The most important time for osteoporosis prevention is adolescence since the young organism seems to be very susceptible to physical stimuli that increase bone density [4, 5]. This narrative review covers the effect of different exercise regimens on the musculoskeletal system, bone integrity, and fracture risk at different ages. Additionally, the combined effect of bone-specific medication and exercise is highlighted.

Exercise Regimens

Since a significant relationship between muscle mass and BMD has been established [5, 6], in regard to exercising, bone health has primarily been associated with resistance and strength training. Several systematic reviews and meta-analyses have shown that resistance training has a positive effect on BMD [7, 8]. BMD adaptation is load dependent and high-intensity resistance exercising led to significantly higher BMD increases than low-intensity resistance exercising [9]. Kitagawa et al. [10] concluded that high-intensity resistance exercising is superior to moderate resistance training when combined with impact training exercises.

Impact exercises have a special position in the field of exercise interventions as in some publications, they are also called weight-bearing exercises and listed together with resistance exercises [11, 12]. While it is true that resistance exercises can be “weight bearing,” when it comes to the core understanding of weight-bearing exercises, any physical activity that involves standing on the feet and applying additional force or a controlled impact to the bones must be considered a weight-bearing exercise or, in another terminology, an impact exercise [13]. Weight-bearing or impact exercises range from brisk walking to running or dancing to competitive gymnastics and similar high-impact sports. Resistance training, on the other hand, is an independent exercise modality and not all resistance exercises meet the weight-bearing criterion. This distinction is necessary and important in order to avoid an inadmissible combination of these different exercise modalities. Weight-bearing exercises have been showing significant effects on bone formation in prepubertal, early pubertal, and pubertal girls [11, 14]. Nevertheless, studies, which evaluate the isolated effect of impact exercises only, are scarce. Often, the effects of impact exercises are evaluated together with resistance training. Watson et al. [15], Manaye et al. [16], and Kistler-Fischbacher et al. [17], for example, showed optimal results regarding BMD increases and physical function via combined high-impact and resistance training.

Concerning fragility fracture prevention, one more exercise modality is very important: balance training. There are factors that increase fall risk, which cannot be influenced by exercising like neurological disorders, impaired vision and hearing, medications inducing an increased risk of falling, and environmental factors (poor lighting, stairs, slippery floors, obstacles, etc.). However, there are fall risk factors that can be positively influenced by physical activity, such as age-related loss of fitness [18]. Sherrington et al. [19] reported that balance training interventions showed the greatest effects in reducing fall risk compared to other exercise interventions. However, resistance training focusing on lower extremity strength as a critical requirement for safe locomotion [20] and core strength training can mitigate age-related deficits in balance [21]. It seems that not a single exercise modality but exercise interventions with multiple components may be the best real-life strategy to reduce fall risk [22, 23]. Moreover, besides the classical exercise routines, more athletic training methods like high-velocity, respectively, power training at maximal speed as well as agility exercises bear considerable potential to counteract the risk of fall [24–26].

Nonimpact endurance training such as bicycling or swimming showed to have no or in comparison to resistance and impact exercises inferior effects on bone health [27, 28]. The following sections will explain in more detail the reason for the efficacy of targeted exercise regimes on fragility fracture prevention, the muscle-bone connection.

Muscle-Bone Connection: Mechanical Interaction

Muscle and bone are tightly connected. Both the myoblast and osteoblast have their origin in the mesenchymal stem cell. Alongside all other organs, muscle and bone change with increasing age. Bone and muscle mass are built up in youth and gradually decrease thereafter; these changes run parallel [29]. Muscle and bone form a single unit, having a number of regulatory factors in common. Maximum muscle and bone mass as well as their loss associated with aging is genetically determined. Besides other factors like hormones, nutrition, endocrine disorders, and potentially drugs (e.g., glucocorticoids), mechanical factors affect both muscle and bone metabolism [30, 31].

Muscle contraction contributes to bone loading via compressive, tensile, and bending forces. Because of its elasticity, bone is a little bit deformable. Deformation is given in μStrain – 1,000 μStrain equaling a change in length of 0.1%. Very low strains less than 10 μStrain prevail in everyday life, whereas large strains above 1,000 μStrain occur only occasionally [32]. In humans, strain is maintained below 2,000 μStrain – even during strenuous activities [33]. The Mechanostat thesis proposed by Harold Frost many years ago illustrates the control loop system. If forces produce a deformation of bone exceeding a certain threshold (above the modeling threshold), bone is added leading to a reduced deformation. In contrast, in case of reduced bone loading (below the remodeling threshold) bone is removed. The aim of this control-loop system ws continuous optimization of bone strength with a minimum of material [34].

There is evidence that most cells in the body are able to sense their mechanical environment. The most important targets of skeletal loading seem to be the osteocytes [35]. They are distributed three-dimensionally throughout the bone and connected to each other via their cytoplasmatic processes within canaliculi in trabecular and cortical bone. Osteocytes comprise the majority of bone cells (90–95%) with a life span of up to 25 years. Hence, they are the predestinated mechanosensors. Within one bone remodeling unit, there is an intensive crosstalk between the different cells; multiple functional interconnections of osteocytes with each other, with osteoblasts, osteoclasts, and with the vasculature exist. Osteoclast and osteoblast activities are carried out in concert [36]. The process of converting mechanical stimulus into a biological response is called mechanotransduction.

Depending on mechanical stimuli (e.g., muscle contraction), bone can adapt its mass, shape, and microarchitectural properties [37]. A hindlimb suspension model showed that a lack of mechanical loading leads to an increase of apoptosis of osteocytes and microarchitectural changes – a decrease of trabecular as well as cortical thickness and cortical porosity – leading to a reduction of bone strength already within 2 weeks [38]. The close relationship between muscle and bone is also emphasized by evidence indicating the negative association between muscle strength and physical function on one side and the risk of fragility fracture on the other side [39–41]. From a preventive point of view, it must be taken into consideration that the human body reacts sensitively to different exercise stimuli at different ages (Fig. 1).

Fig. 1.

a–d Age-dependent reactivity on different training modalities.

Muscle-Bone Connection: Biochemical Interaction

In the past, the interaction between muscle and bone was thought to be mainly mechanical. However, nowadays we know that both organs produce and release messenger substances; they have endocrine and paracrine functions. Contracting muscles secrete different myokines like myostatin, irisin, and interleukins implicated in muscle-bone crosstalk [42, 43]. Osteocytes release humoral factors called osteokines including besides others osteocalcin, sclerostin, dickkopf 1, and receptor activator nuclear factor ß ligand (RANKL) [44, 45]. Numerous myokines and osteokines proved to link muscle and bone [46]. Altogether physical activity triggers different pathways (anti-inflammatory, immunological, metabolic, and hormonal) and, thus, ameliorates bone metabolism [47, 48].

Changes in bone turnover markers induced by physical activity have been shown. A 3 months lasting exercise program combining resistance training and weight-bearing exercises increased the bone formation marker pro-collagen type 1 N-terminal peptide (P1NP) but did not show a significant increase of the bone resorption marker type 1 collagen cross-linked C-telopeptide (sCTX) in postmenopausal women [49]. Even a 12 weeks lasting aerobic exercise program including step aerobic (three sessions weekly for 60 min each) increased the bone formation marker bone-specific alkaline phosphatase [50].

Effects of Exercise on Balance

Exercise is a strategy to counteract age-associated changes in physical function including balance and mobility [51, 52]. Deteriorations of these physical functions mean an increased risk of falling. Falls can be devastating for older adults; fallers may lose self-confidence – concerns of falling leading to reduction of physical activities, social isolation, and depression. Additionally, falls may cause fragility fractures. The older we are, the higher is the risk of falling [53]. More than one-third of adults aged 65 years and older fall at least once a year [54]. Most of the time, falls are of multifactorial origin. Both exogenous factors as well as endogenous factors, such as deficits in balance, are important risk factors [55].

According to the world guidelines for falls prevention drawn up by the Task Force on Global Guidelines for Falls in Older Adults, physical activities, especially tailored exercises on balance, gait, and strength are a very important part of fall prevention [56]. Exercise as a single intervention reduces the rate of falls by 23% with high-certainty evidence. Different exercise modalities show different effectiveness: balance and functional exercises reduce the rate of falls by 24% (high-certainty evidence); combined with resistance training, it is even more effective (risk reduction of 39%; moderate-certainty evidence). However, the effect of resistance training, dance, or walking programs as single training modalities is uncertain. Tai Chi may reduce the rate of falling by 19% [22]. It is very important to include balance and functional exercises into an exercise program to achieve a significant effect on the reduction of the risk of falling [23]. A UK consensus statement also puts emphasis on spinal extension exercises in order to improve posture and potentially reduce the risk of falls [57]. Programs reducing the risk of falling in elderly subjects living in the community are the Otago Exercise Program [58, 59], the Falls Management Exercise Program [60–62], and gait adaptability training [63]. With some minor adaptations and performed in supervised sessions, the Otago Exercise Program even proved applicability and efficacy in nursing homes residents [64]. According to a meta-analysis [65], whole-body vibration training also reduces the fall rate.

Rodrigues and colleagues [66] prescribed some practical tips: the exercise program should be 3 h per week, including balance and functional training. Reducing the base of support, moving the center of gravity to the limits of stability are good options to increase balance. It is recommended to combine balance and functional exercises with resistance training. Training intensity has to be increased progressively. Concerning strength training, the exercising person should be able to perform up to 12 repetitions with good form before reaching muscular fatigue. Concerning balance training, subjects should feel confident to actively having control over the balance exercise while fighting for it. When they struggle to keep control over the balance exercise, the task is too difficult; on the other hand, when the balance task is no real challenge, the task is too easy. Exercise can be an individually tailored home-based program or performed in classes (Table 1).

Table 1.

Recommendations of exercise for fall prevention in elderly subjects

Exercise program
Type of exercise	Combination of balance, functional exercises, resistance training
Duration	≥3 h/week
Intensity	Moderate – high; progressive
Mode of exercise	Home-based or group
CAVE	High-risk subjects: no walking, no Tai Chi

Training effects on fall risk reduction have been documented several times, but it is important to note that these training effects do not or hardly persist beyond the end of training [67]. This shows the necessity to integrate exercising into everyday life. Clemson and co-authors [68] showed that after appropriate training, an exercise program to improve balance and strength integrated into everyday life reduces the fall rate by over 30% in people at risk of falling.

Fracture Prevention

Primary Fracture Prevention

Primary fracture prevention actually starts during childhood and youth because in the toddler phase, our balance system is particularly capable of learning and the young skeleton is very receptive to impact and strength training (Fig. 1). Regular physical activity during the early lifespan is essential for reaching an optimal peak bone mass. However, it is important to continue weight-bearing exercises and/or strength training throughout life. Meta-analyses proved exercise programs to positively influence BMD in men 45+ and postmenopausal women [69, 70]. A finite element model showed that the cessation of physical activities like stair climbing has a negative impact on femoral bone structure, which makes up an important part of bone integrity, leading to a reduced failure load in elderly subjects [71].

Although the neuromuscular system is not as sensitive to balance training in old age as during youth (Fig. 1), balance training is particularly important in advanced age. Most fragility fractures occur because of a fall, an “unexpected event in which the subject comes to rest on the ground, floor, or lower level” [72]. A meta-analysis highlighted the association between sarcopenia, falls, and fractures [39]. A previous fall is associated with an increased risk of a clinical fracture in women (HR: 1.42) and men as well (HR: 1.53) [73]. Regardless of fracture occurrence, in case of a recent fall (within the previous 4 months) the risk of fracturing in the following year is supposed to be imminent: the risk of a hip fracture is 2.8-fold greater in women and 2.3-fold greater in men than in women and men without a fall, respectively [74]. Therefore, it is very important to target balance training and functional exercises, particularly to subjects with past falls; exercise efficacy is especially high in subjects with a high risk of falling [75, 76]. The removal of tripping hazards in the home environment is also most effective for people who are particularly at risk of falling [77]. A meta-analysis proved that the positive effects of exercising on BMD and the risk of falling lead to a significant reduction of overall low-trauma fractures (incidence rate 0.67, p = 0.003) and major osteoporotic fractures (incidence rate: 0.69, p = 0.011) [78].

Secondary Fracture Prevention

Changes in statics after vertebral fractures lead to changes in the transmission of biomechanical forces at the vertebral column. In case of a prevalent vertebral fracture, re-fracture risk is four to five times elevated; prevalent non-vertebral fractures double fracture risk [79]. The risk of a further fracture is highest in the immediate post-fracture interval (so-called imminent fracture risk) [1]. This increased risk of future fractures underlines the importance of secondary fracture prevention. Austrian Society for Bone and Mineral Research [80] recommends pharmacological treatment as well as non-pharmacological interventions to prevent secondary fractures, which are best initiated immediately after the fracture, embedded in a fracture liaison service.

In accordance with the international consensus on the management of vertebral fractures [81], an individualized, guided training program should be initiated as soon as the level of pain subsides or after medical clearance (around 4–12 weeks after the vertebral fracture) – initially, if necessary, in a relieved position. Fracture healing (around the 12th week after fracture) is the time to start a multimodal training program (progressive strength training, functional training, balance training). In case of conservative treatment, peripheral fractures are usually stable under load after bony consolidation and in case of surgical treatment usually immediately afterward. Of course, patients with prevalent nonrecent fractures should be physically active as well. A supervised high-intensity resistance and impact training (HiRIT) proved efficacy in postmenopausal women with reduced BMD, who partially had osteoporotic fractures [15].

Safety

Further analyses of this LIFTMOR trial showed that the 8 months lasting HiRIT was not associated with vertebral fracture progression or incident vertebral fractures in postmenopausal women as well as in men [82, 83]. Several meta-analyses also proved safety of different exercise programs [75, 84, 85].

The crucial factor is individual training control. On the one hand, training must be sufficiently challenging; otherwise it will not be beneficial for balance and movement safety in everyday life; on the other hand, we must pay particular attention to safety during training. Some physical activities, especially those including impact exercises, may be modified in order to be done safely – no activities with weighted, pronounced spinal flexion or twisting of the spine [86, 87]. Patients with vertebral fractures and/or multiple fragility fractures are generally advised to avoid high impact loads on the spine (above the level of everyday stress, e.g., higher impact loads than during brisk walking) [57].

There is no contraindication for physical activity or targeted training; it only needs to be adapted to the corresponding needs or health status (e.g., musculoskeletal problems, cardiovascular or cognitive limitations) [88].

Relationship between Bone-Specific Medication and Exercise

According to the Utah Paradigm of skeletal physiology proposed by Frost [89], mechanical stress stimulates the direct activation of bone formation, called modeling; it depresses bone resorption. Elderly people often take bone-specific medication. Therefore, the question arises whether there is a benefit of combining osteoporotic drugs with mechanical loading.

Anabolic Medication

The aim of bone-anabolic preparations is bone formation. Thus, they are supposed to reduce the modeling threshold and to increase the remodeling threshold, set point for activation of resorption and subsequent formation. In addition, anabolic preparations directly stimulate osteoblast activity. That way, formation outweighs resorption. In other words, we have a synergistic effect of mechanical stress and osteoanabolic therapy [90].

Johnson [91] first described the positive effects of physical activity combined with an anabolic agent (sodium fluoride) in 1965. In postmenopausal women with osteoporosis, Riggs and colleagues detected the lower effects of sodium fluoride on the radial shaft compared to more heavily loaded sites, the lumbar spine, and the hip region [92]. The first investigations of a potential association between mechanical strain and the anabolic parathyroid hormone (PTH) were performed in rodents and showed the importance of PTH in the responsiveness to mechanical loading [93]. Some experimental works followed.

For 6 weeks, female C57BL/6 mice were given daily subcutaneous injections of human PTH (1–34) at a dose of 20, 40, or 80 μg per kilogram bodyweight or a vehicle. During the last 2 weeks 35 min after each PTH administration, the tibia and ulna were subjected to dynamic axial loading (12 N, 10 Hz, 40 cycles, 3 days per week). Mechanical loading led to an increase in bone formation in trabecular as well as cortical areas. Mechanical loading with simultaneous administration of PTH – even at very high doses –– did not result in any benefit in the trabecular region but did lead to a further benefit in the cortical region (periosteal and endosteal), showing synergism of mechanical stress and intermittent administration of PTH (iPTH) [94]. In a similar model, Sprague-Dawley rats received daily injections of 50 μg iPTH (1–34) per kilogram body weight or vehicle subcutaneously. The right tibia of some animals was mechanically stimulated using a four-point-bending device 35 min after PTH administration. Isolated PTH treatment increased bone formation by 13.5 μm. Isolated mechanical loading led to a difference of 7.3 μm. The combined therapy showed an additive benefit with an estimated treatment difference of 20 μm [95].

Since all experimental studies indicated synergistic effects of PTH therapy and mechanical stress, clinical studies followed. Using a special processing technique of computed tomography scans, Poole and co-authors investigated the effect of 24 months of teriparatide (recombinant human PTH [1–34]) treatment on cortical thickness of the femur in postmenopausal women. They showed a higher increase in cortical thickness at the inferomedial hip region and the head-neck junction of the superior cortex – areas that are particularly heavily loaded during walking – as well as at the insertion sites of the larger muscle groups such as the attachment sites of the gluteus medius, psoas, and quadratus femoris [96]. So far, only one study evaluating the combination of anabolic therapy and training in humans exists. Postmenopausal osteoporotic women with prevalent vertebral fractures who started teriparatide treatment were randomized to teriparatide alone or teriparatide plus whole-body vibration training (3.6 g, 30 Hz, 1 mm, 12 min 3 days per week). Twelve months of regular vibration training led to a significantly higher increase in BMD of the lumbar spine region compared to those participants not exercising. Thus, whole-body vibration has an additive effect to teriparatide treatment [97].

To date, no data on the combination of regular mechanical stress and the sclerostin antibody romosozumab exist. However, since physical activity lowers the serum levels of sclerostin [98] one could assume that a regular workout and romosozumab treatment may have synergistic effects.

In conclusion, pre-clinical and clinical data show that teriparatide increases the responsiveness of bone tissue to mechanical loading. Concerning romosozumab, no data investigating the influence of the sclerostin antibody on loading-related bone formation changes exist.

Antiresorptives

Antiresorptives have a different mode of action than anabolic substances. Lowering the remodeling set point, the antiresorptives’ main task is to decrease the coupled process of bone resorption and bone formation. The reduction of remodeling spaces results in a slight bone gain [90].

Due to its antiresorptive effect, estrogen is partly responsible for the integrity of bone in young adulthood. The drop in estrogen during menopause leads to a negative balance in bone metabolism with an increase of bone resorption. In contrast, bone metabolism of women on menopausal hormone therapy due to their menopausal symptoms as well as bone metabolism of postmenopausal women taking selective estrogen receptor modulators benefit from the antiresorptive effect of estrogen. Concerning the interaction of this antiresorptive agent and mechanical loading, some studies suggested the effects of estrogen and loading to be additive [99, 100]. Besides the positive effect of suppressed endosteal bone resorption, an experimental study detected the reduction of periosteal bone formation following estrogen treatment in ovarectomized rats [101]. Further investigations of rodents proved the opposite effect of exercise and estrogen on periosteal bone formation [102, 103]. Lanyon and Skerry hypothesized that the main point concerning a competent adaptive response to load bearing is the estrogen receptor, not estrogen itself [104]. In a recent meta-analysis, we showed a nonsignificant additional benefit of hormone treatment plus exercise compared to hormone treatment without exercise on BMD in postmenopausal women [105].

Bisphosphonates and denosumab are frequently used antiresorptive drugs in osteoporosis therapy. Experimental protocols evaluating the combined effect of bisphosphonates and exercise came to different conclusions. In ovarectomized rats, etidronate as well as alendronate administration combined with treadmill running led to superior effects on BMD than either intervention alone [106, 107]. The combined effects of zoledronic acid and running exercise, however, did not induce additive effects [108]. Differences in antiresorptive potency of the drugs may be responsible for the absence or presence of blunting of the osteogenic effect of exercise. A subgroup analysis of a larger meta-analysis suggested a nonsignificant benefit for BMD of the femoral neck of adding exercise to antiresorptive treatment in postmenopausal women [17]. Limitations of this analysis are heterogeneity of the antiresorptive regimen (hormone therapy, different bisphosphonates, and denosumab) as well as heterogeneity of the exercise regimens concerning type, intensity, and duration of the training protocols.

A randomized trial compared 8 months of HiRIT (deadlift, back squat, overhead press, jump drop) with low-intensity Pilates in postmenopausal women partially with and partially without antiresorptive therapy, either bisphosphonates or denosumab. The BMD change was significantly greater for the high-intensity training than for the low-intensity Pilates program. Both exercise regimens reached a greater effect on BMD in case of simultaneous drug therapy. In the small number of patients on drug therapy, no intensity-dependent between-group difference was seen. However in contrast to women of the Pilates group, the change in hip BMD exceeded the least significant change in the women performing the HiRIT [109]. Existing literature indicates that in case of antiresorptive treatment a more intensive exercise program may be necessary to overcome the reduction of periosteal bone formation associated with these drugs.

Conclusion

Physical exercising is a potent treatment modality to reduce age-related fragility fractures. It consists of two main strategies, the increase of bone integrity via strength and (high) impact training and the decrease of fall risk via balance and strength training (Fig. 2).

Fig. 2.

Effect of different training modalities on the risk of fragility fractures.

From childhood through early adulthood is the most important time for preventing bone fractures as this is not only the age when bones are particularly receptive to anabolic stimuli but also when the trainability of the neuromuscular system is at its highest. At a younger age, even low-intensity impact stimuli already have positive effects on bone formation, while later in life, stronger physical stimuli, such as high-intensity resistance training, are necessary to induce adaptation. Bone-specific medications also influence the sensitivity of bone to training. However, both bone integrity and the risk of falls can be positively influenced by physical activity at any age. The exercise stimulus simply needs to be adapted to biological age and any additional health risks that may arise later in life.

Acknowledgment

We would like to thank Galateja Jordakieva for her expertise and the design of Figure 1 to illustrate one of our main messages.

Conflict of Interest Statement

The authors have no conflicts of interest to declare.

Funding Sources

There is no funding source.

Author Contributions

Katharina Kerschan-Schindl and Timothy Hasenoehrl have contributed to the conception of the work, writing, review process, and final approval of the manuscript.

Funding Statement

There is no funding source.