Are spontaneous fractures possible? An example of clinical application for personalised, multiscale neuro-musculo-skeletal modelling
Introduction
It is estimated that in Europe 179,000 men and 611,000 women, mostly elderly, suffer a hip fracture each year (Melton et al., 2003). Complications of such fractures lead 20–25% to death within the following 12 months (Trombetti et al., 2002, Leibson et al., 2002). All the others will suffer minor or major impairment. Some of the most catastrophic projections suggest these figures could double by 2050 (Gullberg et al., 1997), due to the progressive ageing of the population. A considerable problem is that the aetiology of these fractures is frequently unclear, jeopardising any attempt to reduce their incidence. While most of those fractures cannot be associated with a high-energy trauma (low-energy fracture) (Rockwood et al., 1991, Rüedi and Murphy, 2001), an unknown fraction is reported to occur in complete absence of trauma (spontaneous fractures).
The primary explanation for these fractures is osteoporosis, a metabolic disease that decreases the total mass of mineralised tissue in the body, by increasing the degree of porosity of bone tissue. Osteoporosis significantly reduces the ability of bones to withstand biomechanical loading without damage (AJM, 1991). However, some biomechanical studies seem to suggest that only extremely severe osteoporosis can weaken bones to the point where they fracture under physiological loading conditions (Taddei et al., 2008). On the other hand some clinical studies show that nearly half of the fractures observed occur in patients whose bone mass is reduced only to a point that is considered physiological in association to ageing (osteopenia) (Siris et al., 2004, Wainwright et al., 2005, Schuit et al., 2004). Because of this, some authors suggest that, especially for the hip, truly spontaneous fractures (that occur during normal daily activity) are nearly impossible, and that low-energy hip fractures are almost always due to a combination of osteopenia/osteoporosis and a fall (Cotton et al., 1994, Jeffery, 1974). Indeed, ageing usually involves also a weakening of the muscles (sarcopenia) and a multifactorial degradation of the neuromotor control, which tends to increase significantly the probability of falling (Roos and Dingwell, 2010).
In the biomechanical literature in order to quantify the forces acting on the hip joint during normal physiological activities it is almost always implicitly postulated that the neuromotor control operates under optimal conditions (Praagman et al., 2006, Thelen et al., 2003, Zajac et al., 2002). Joint forces generated by a normal neuromotor control are not sufficient to fracture a bone, even if severely osteoporotic. Thus, one could conclude that spontaneous fractures are not possible. But it has been suggested that when neuromotor control is sub-optimal, the forces transferred to the skeleton during normal daily activities can considerably increase (Martelli et al., 2011), thus inducing overloading. Moreover, (Yang et al., 1996) developed an in vitro simulation to determine a possible biomechanical background for spontaneous hip fractures, concluding that abnormal muscle contraction of the rotator muscles could induce hip fracture. If these conjectures are true, then truly spontaneous fracture could be possible, at least in principle.
This dichotomy cannot be easily resolved with observation. In theory, if spontaneous fractures were not possible, all low-energy fractures should be a consequence of falling. On the contrary, if spontaneous fractures were possible, then some patients would first fracture, and then fall as a consequence. As hip fractures are always associated with sudden and intense pain, in principle one could ask patients if they first perceived pain and then fell, or vice-versa. Some clinical literature reports that between 6% and 10% of the hip fractures in the elder population cannot be associated to a major or minor fall (Parker et al., 1997, Tinetti et al., 2003). There are two arguments however, that cast serious doubts on these observational studies. The first, frequently reported in the medical literature, is that many elderly after a femoral neck fracture experience mild delirium, including memory loss, which makes their reconstruction of the sequence of the events unreliable (Björkelund et al., 2010, Milisen et al., 1998). The second is that the femoral neck fracture typically occurs in less than one millisecond (Juszczyk et al., 2011), while the time required to perceive the pain (Treede et al., 1988, Hauck et al., 2007) and that to perceive the proprioceptive stimuli (Behm et al., 2004) are comparable. Thus, whether the fracture is spontaneous or caused by the fall, the perception of the pain and the loss of stability occur simultaneously.
If observational clinical studies cannot conclusively confirm/exclude the existence of spontaneous fractures, at least biomechanical modelling can provide a tentative answer to a more fundamental question: are spontaneous fractures physically and physiologically possible?
The aim of this work is to use a validated stochastic multiscale body-organ model to estimate:
- (1)
if theoretically spontaneous fractures of the femoral neck during level-walking are physically and physiologically possible,
- (2)
if spontaneous fractures are possible in principle, which level of negative synergy of neuromotor degradation and bone tissue osteoporosis is needed to produce a spontaneous fracture.
Section snippets
Overview
The multiscale model (Viceconti and Kohl, 2010) used in this study was generated starting from an extensive data collection of the anatomical, physiological, and physical properties of the musculoskeletal system of an individual subject. These data were used to generate an inverse dynamics model of the lower body that takes the kinematics and the ground reactions during level walking as an input, and predicts the joint net moments at the hip, knee, and ankle. This inverse dynamics model was
The musculoskeletal model
The musculoskeletal model of the lower-limb was defined as a 7-segment, 10 degrees-of-freedom (DOFs) articulated system actuated by 82 muscle-tendon units. Three idealised joints articulated each lower-limb, resulting in three DOFs at the hip joint (i.e. a ball-and-socket joint), and one DOF (i.e. a hinge joint) at the knee and the ankle. Joint parameters were identified accordingly to the ISB standards (Wu et al., 2002). Inertial properties were derived from the CT scan assuming homogeneous
Definition of stochastic variables (Table 1)
To account for the intra-subject neuromotor repeatability all seven repetitions of the gait cycle were included in the model, assigning to each of them the same probability.
While it is clear that the neuromotor control does not always work optimally, especially in the elderly, the probability associated to each sub-optimal control pattern is unknown. However, it is reasonable to assume the optimal control solution (HRmin) as the most probable, the worse solution (HRmax) as the least probable,
Simulations
Preliminarily, the multiscale model was used deterministically to explore the risk of fracture associated with extreme values of osteoporosis and neuromotor control degradation. Three conditions were modelled: (i) no osteoporosis but extreme neuromotor degradation, (T-score=0, ND=1); (ii) severe osteoporosis but optimal neuromotor control (T-score=−5; ND=0); (iii) combined severe osteoporosis and extreme neuromotor degradation (T-score=−5; ND=1).
Then, two probabilistic schemes of the multiscale
Results
The explorative deterministic analyses showed that neither extreme osteoporosis, nor extreme neuromotor degradation could alone predict the possibility of spontaneous fractures during level walking. However, the combination of the two factors made this possible. When running the probabilistic scheme representing a population of 80-years old women with variable degrees of osteoporosis, neuromotor control and muscle sarcopenia, the predicted ASFR was 0.4%. When narrowing the population to only
Discussion
In the present work a validated stochastic multiscale body-organ model was used to investigate whether spontaneous fractures of the femoral neck are physically and physiologically possible in subjects affected by osteoporosis.
In the explorative deterministic analyses, the model predicted that in non-osteoporotic subjects even the neuromotor control solution that maximises the hip joint force would not be sufficient to fracture the femoral neck during level-walking. Similarly, even the most
Conflict of interest statement
The authors declare that they do not have any financial or personal relationships with other people or organisations that could have inappropriately influenced this study.
Acknowledgements
The authors would like to thank Luigi Lena for the illustrations and Mauro Ansaloni, Paolo Erani and Mateusz Juszczyk for the support during the experiments. This study was supported by the EU funded projects NMS-Physiome (FP7, grant #24818965) and VPHOP (FP7, grant #223865).
References (49)
- et al.
Hip contact forces and gait patterns from routine activities
Journal of Biomechanics
(2001) Evidence that maximum muscle stress is not a constant: differences in specific tension in elbow flexors and extensors
Medical Engineering & Physics
(1995)- et al.
Are hip fractures caused by falling and breaking or breaking and falling? Photoelastic stress analysis
Forensic Science International
(1994) - et al.
In vitro replication of spontaneous fractures of the proximal human femur
Journal of Biomechanics
(2007) - et al.
Evaluation of the generality and accuracy of a new mesh morphing procedure for the human femur
Medical Engineering & Physics
(2011) - et al.
Musculo-skeletal loading conditions at the hip during walking and stair climbing
Journal of Biomechanics
(2001) Spontaneous fractures of the femoral neck
Orthopedic Clinics of North America
(1974)- et al.
The human proximal femur behaves linearly elastic up to failure under physiological loading conditions
Journal of Biochemistry
(2011) - et al.
Effect of sub-optimal neuromotor control on the hip joint load during level walking
Journal of Biomechanics
(2011) - et al.
The relationship between two different mechanical cost functions and muscle oxygen consumption
Journal of Biomechanics
(2006)
Influence of simulated neuromuscular noise on movement variability and fall risk in a 3D dynamic walking model
Journal of Biomechanics
An accurate estimation of bone density improves the accuracy of subject-specific finite element models
Journal of Biomechanics
Fracture incidence and association with bone mineral density in elderly men and women: the rotterdam study
Bone
The material mapping strategy influences the accuracy of Ct-based finite element models of bones: an evaluation against experimental measurements
Medical Engineering & Physics
Generating dynamic simulations of movement using computed muscle control
Journal of Biomechanics
Multimodal fusion of biomedical data at different temporal and dimensional scales
Computer Methods and Programs in Biomedicine
Multimod data manager: a tool for data fusion
Computer Methods and Programs in Biomedicine
Multiscale modelling of the skeleton for the prediction of the risk of fracture
Clinical biomechanics (Bristol, Avon)
ISB Recommendation on definitions of joint coordinate system of various joints for the reporting of human joint motion—part I: ankle, hip, and spine. International society of biomechanics
Journal of Biomechanics
Biomechanics and muscle coordination of human walking: part I: introduction to concepts, power transfer, dynamics and simulations
Gait & Posture
Ten-year absolute risk of osteoporotic fractures according to BMD t score at menopause: the danish osteoporosis prevention study
Journal of Bone and Minerals Research
Effect of acute static stretching on force, balance, reaction time, and movement time
Medicine and Science in Sports and Exercise
Sarcopenia: prevalence, mechanisms, and functional consequences
Interdisciplinary Topics in Gerontology
Cited by (48)
Predicting bone strength from CT data: Clinical applications
2019, MorphologieAssessment of finite element models for prediction of osteoporotic fracture
2019, Journal of the Mechanical Behavior of Biomedical MaterialsCitation Excerpt :Load can be as high as four times the BW when the heel impacts the ground and during the double support stage of gait (Lopez et al., 2012). With such a significant impact on the femoral head, the effect of soft tissues such as the abductor muscles should also be modelled for more realistic in vivo conditions (Lotz et al., 1995; Viceconti et al., 2012; Tsouknidas et al., 2012; Cadarette et al., 2008). Mostly, a fixed constraint or a vertical displacement constraint was applied on the greater trochanter during sideway fall simulations, while during stance, either no loading or reaction forces due to abductor muscles were included as described above.
Medicine and the virtual physiological human
2019, DHM and PosturographyFemoral fracture load and fracture pattern is accurately predicted using a gradient-enhanced quasi-brittle finite element model
2018, Medical Engineering and PhysicsIntegrative Medicine for Geriatric and Palliative Care
2017, Medical Clinics of North America