Development of a Wearable Device for Osteoporosis Prevention and Fracture Risk Reduction in Women

CURRENT STATUS AND JUSTIFICATION OF THE PROPOSAL Osteoporosis is an important public health problem and it is characterized by the deterioration of the microarchitecture of bone tissue and a low bone mineral density (BMD) that leads to a reduction in bone strength and the consequent increased risk of skeletal fractures 1. One in three women will suffer an osteoporotic fracture in their lifetime 2. In fact, it is estimated that more than 200 million people suffer from osteoporosis and these figures are increasing due to the aging of the population and the change in lifestyles 3. A recent systematic review gathered epidemiological information on selected musculoskeletal injuries and provided pooled injury-specific incidence rates. Authors reported that the most common fractures in the whole adult population based on the pooled incidence rates were distal radius fractures (212 per 100,000 person-years)4. Distal radius fractures account for one in five bony injuries in both primary and secondary care5. According to data from the International Osteoporosis Foundation (IOF), in Spain there would be more than 2.8 million women over the age of 50 with osteoporosis, with more than 330,000 fragility fractures per year and projections suggest that by 2030 these figures will increase 28.8%1. Consequently, there is a significant economic burden on the healthcare system for patients who experience an osteoporosis-related fracture. Direct costs for osteoporosis, such as medications, occupational therapies, and inpatient and outpatient visits, are expected to increase to $25 billion by 2025 as the population ages 6. Together with the indirect costs, such as loss of productivity, are expected to rise to more than $95 billion in 2040 6. Osteoporosis, and the 4.3 million fragility fractures that it causes, cost the health care systems of Europe over €56 billion each year based on data for 2019 7. Only in our country it is estimated that the cost associated with sick days due to fragility fractures amounts to €355,000 each year. Only the direct and indirect costs associated with these fractures would exceed €4.2 billion, with a projected increase of 30.6% by 2030 1. It should not surprise us, therefore, that in postmenopausal women, osteoporotic fracture costs exceed the annual costs for breast cancer, myocardial infarction (MI), and stroke 6. The loss of the protective effect of estrogens on the skeletal structure after menopause intensifies the decrease in bone density as a result of a sudden imbalance between bone formation and bone resorption. This decrease in estrogens (e.g., estradiol) leads to an increase in the rate of bone remodeling due to an increase in the number of osteoblasts and osteoclasts; although osteoclast-mediated removal of bone matrix is much greater than osteoblast bone formation8. Therefore, the loss of BMD is very evident, especially in the 3-5 years following the last menstrual period 9, in a 2% loss rate per year 10. Parallel to this decrease in BMD, it is common to find a decrease in muscle mass and strength, probably derived from the decrease during menopause of certain precursors (e.g., insulin-like growth factor type 1, -IGF-1-) 11. It is important to note that bone strength is determined not only by bone mass, but also by the size, shape, structure, or properties of collagen 12. Since densitometry (DEXA) captures only bone mass, there is growing interest in the use of other three-dimensional imaging modalities, such as quantitative computed tomography (QCT) to assess bone strength. QCT can help differentiate between the cortical and trabecular compartments of bone and their relative contribution to bone strength in vivo. Similarly, various biomarkers have been identified to detect the dynamics of bone remodeling 13,14. The National Bone Health Alliance and the American Association for Clinical Chemistry have suggested that the most widely accepted biomarkers for the evaluation of interventions in osteoporosis are the C-terminal telopeptide of collagen type 1 -CTX-1- (bone resorption) and the amino-terminal propeptide of procollagen type 1 -P1NP- (bone formation) 13,14. There are well-known modifiable and non-modifiable risk factors for osteoporosis. Non-modifiable items would include age, family history of fragility fractures, type 1 diabetes, rheumatoid arthritis, or cancer treatment (e.g., aromatase inhibitors or chemotherapy). In the first group, smoking, excessive alcohol consumption, the use of corticosteroids, insufficient calcium or vitamin D, low body mass index (BMI), sedentary lifestyle15, or the lack of physical activity, which would be another of the main risk factors can be included. However, bone loss can remain clinically silent. A recent retrospective healthcare study found that 79.4% of patients did not have a diagnosis or treatment on-board for osteoporosis 12 months before the fracture 6. It surprises that less than half of women at high risk of fracture are treated despite the high cost of fractures and the availability of affordable medications 2. Consequently, according to the IOF, there is a large unmet need as it relates to the prevention of postmenopausal osteoporosis because many women are not undergoing the recommended diagnostic screenings or assessing baseline osteoporosis risk 15. Prevention should be the priority when pursuing the reduction of the burden caused by osteoporosis16. In addition to prevention, the optimization of the management, from first aid to rehabilitation, should be well considered to reduce excess costs. The approach in this population group is usually multimodal and includes both pharmacological and non-pharmacological interventions. In Europe, the most commonly used drugs for osteoporosis are bisphosphonates, raloxifene, parathyroid hormone drugs, and denosumab 17. However, the widespread and long-term use of these drugs is limited due to possible side effects 18. Therefore, diet and physical exercise, together with other environmental and lifestyle factors, would constitute the main non-pharmacological strategies for the management of this pathology 19. However, although physical activity is a modifiable factor that contributes to the improvement of bone mass and strength, our understanding of how to quantify the dimensions of physical activity that are osteogenic (including frequency, intensity, time, and type) is incomplete. There is no doubt that to initiate an osteogenic response, the bone must be subjected to a stimulus that exceeds a threshold determined by the usual range of deformation in the predominant load direction19. Precisely, the mechanical stress to which exercise (impacts and muscle contraction) subjects the bone (e.g. tension, compression, hydrostatic pressure) can promote osteogenic differentiation of mesenchymal stem cells, increasing osteoblastic regulation through different mechanisms 20. To increase effectiveness, the exercise must: (i) be dynamic, not static; (ii) exceeding an intensity threshold; (iii) exceeding a strain frequency threshold; (iv) be relatively short but intermittent; (v) imposing an unusual load pattern on the bones; (vi) include adequate availability of calcium and vitamin D 21. However, the optimum loading patterns of bone that produce clinically significant benefits have not been clearly defined to date, since it has not been possible to easily quantify exercise intensity in terms of bone-loading forces in clinical settings22. Q1 Recent guidelines for the prevention and management of osteoporosis through physical exercise suggest interventions that last longer than 8 months and that combine high impacts and high-intensity strength training. Interventions in which fast walking (e.g., \> 6 km/h) is combined with activities that incorporate a relatively short number of impacts (e.g., 10-50 jumps per day), at intensities exceeding 2-4 times body weight (\~ 3.9 G), at least 4 sessions a week and using multiaxial exercises to generate an unusual load on the hip, can help postmenopausal women to compensate for age-related bone loss 23. These would be effective strategies to preserve BMD in postmenopausal women with a level of evidence 1a24. However, there is no evidence of its effectiveness in preventing distal radius fractures Q2 and also dropout rates in these types of programs are high25 and, despite the evidence, participants are often reluctant to exercise, citing barriers as poor health, lack of time due to family responsibilities, lack of companionship or encouragement, lack of access to adequate facilities or opportunities, or cost. Patients with osteoporosis often report a lack of time and transportation limitations as the main barriers that would limit their participation in supervised training programs 26. For these reasons, it has been suggested that intervention trials should use settings that are easily accessible to participants and provide flexible exercise schedules 25. In the literature, it is suggested that in interventions aimed at bone health in young adult women, it is crucial to minimize the perceived demands of time (e.g., short and flexible intervals) and environmental barriers related to the convenience and accessibility of physical activity 25. Therefore, to increase adherence in this population group, regular contact with study personnel, personalization of progression, and adequate exercise monitoring (avoid safety concerns and lack of confidence) are suggested. In a recent study from our group, evidence showing that non-supervised exercise programs could be effective to improve BMD in adult women was found. However, only five studies were available in postmenopausal women and overall registered a substantial dropout rate (above 40%) due to unwillingness to continue and illness among the main reasons. Therefore, the necessity to investigate the efficacy of remote/assistive technologies for delivering and monitoring non-supervised exercise interventions was highlighted. The question that should be asked is, in what way all these issues can be managed, facilitating the monitoring of the program and solving the barriers outlined? Mobile and digital technologies (mHealth) allow complementing exercise programming for the management of osteoporosis by providing information on skeletal mechanics during locomotion23. However, and even though wearable activity trackers offer potential as a multifaceted intervention to help people become more active, no study has examined the usefulness of these wearable devices in postmenopausal women nor in the upper extremities Q3. The importance of reaching a minimum threshold of stimulation to obtain bone adaptations has been justified, since insufficient loads would not have the desired effect, for this reason, precise quantification of the mechanical loads to which the musculoskeletal system is subjected is necessary, what could be achieved with certain biosensors28. However, interventions based on this type of technology aimed at preventing osteoporosis are very scarce29, even though authors agree on their potential, not only for increase the benefits of the interventions, but mainly to improve adherence to them. In summary, the scientific community face a group of women with a high risk of suffering from osteoporosis (potential incidence of fragility fracture), with the subsequent incidence for their functionality and quality of life, on the one hand, and for the health system, on the other. The strategies for managing this group suggest the incorporation of activities that require impact and strength training; however, direct measurement of bone strains requires invasive surgical procedures and indirect measurement of ground reaction forces are limited to a fixed place and time, being thus impracticable in long-term clinical measurements. Assessing the effects of this type of intervention could help to record variables that, after proper analysis (e.g., machine learning), facilitate decision-making and reduce the use of resources in the health system. Therefore, the questions that arise are: ¿can indirect surrogate of bone strains that are now readily measurable with a novel accelerometer-based measurement device be used? ¿Can these wearable devices be used to investigate the effect of exercise-induced impact loading on bone mineral density, geometry and metabolism in posmenopausal women with the intention of assessing the intensity and amount of impact loading required to gain adaptive effects at the wrist level? Starting hypothesis: The lifestyle intervention controlling the intensity of impact and exercises with wearable devices will improve bone composition and structure, physical fitness and quality of life in postmenopausal women. Evaluating the effect of a bone-targeted exercise training programme on extra-skeletal risk factors of osteoporotic fractures and physical performance scientist will be able to design a technological smart cloud platform with digital technologies (artificial intelligence, a mobile-based application and wearable managed in the cloud) that could be implemented (after the proper identification key factors for the transfer of research results obtained in the experimental phase) to the public socio-health system contribute in the prevention of osteoporosis and frailty fractures. Procedures Objective 1 Participants will be scheduled for three days to complete the measurement protocol. On the first day, they will attend the hospital and complete the sociodemographic and clinical information. They will then receive the accelerometer and all the questionnaires to be completed at home. In addition, from the hospital center, the participants will be cited another day for the biochemical analysis, QCT and densitometry. On a second day, the participants will be cited to the university laboratory to return the accelerometers and questionnaires, and they will take the physical fitness tests. Finally, in a third visit to the laboratory, the participants will be familiarized with the use of the recording wearable. Finally, all participants will receive a daily vitamin D supplement of 600 IU/day. If serum levels are deficient, it will be supplied for 4 months: 1 vial of 25,000 IU per month if \<20ng/ml, 1 vial of 25000 IU every 15 days if \<15ng/ml, 1 vial of 25,000 IU every week if \<10 ng/ml. Calcium intake will be calculated using the 3-day recall and supplemented up to 1200 mg per day. Lifestyle intervention A. Program to increase physical activity and impacts For the intervention, a biosensor -wearable device- (Fitbit Versa) will be used. The device is linked to an App (Weapom) that will allow users to quantify the number of steps (including the rate of steps per minute), the impacts and their intensity. Users will be given feedback from the application on the level of achievement of the objectives proposed in the impact program (EV). It will be proposed to achieve 50 vertical and multidirectional jumps per day (milestone 1. The number of jumps\> 3.9 G per day will be recorded 28, divided between 3-5 sets of 10-20 repetitions with a minimum 1-2 minute rest between sets. A frequency between 4-7 days per week will be used. In the same way, it will be recommended to reach 10,000 steps per day (milestone 2. The number of daily steps will be recorded) and, as far as possible, at a rapid pace (milestone 3. The number of daily minutes walking at a pace over 110 steps/minute will be recorded, as this is considered the threshold of moderate-vigorous intensity in this age range 40. Therefore, according to the new WHO guidelines 41 on physical activity and sedentary behavior, it will be quantified whether 150 minutes a week of moderate-vigorous physical activity are achieved. In addition, and in order to specifically focused on radius strength, 50 (3 to 5 cm) drops with the arms completely straight will be register. In parallel, the participants in both groups will receive recipes rich in calcium through the App and will have a calculator to calculate the intake of calcium and other micronutrients through the diet. They will also receive information on osteoporosis and strategies for its prevention (e.g., advice on sun exposure). B. Control group Participants randomly assigned to CON will receive general advice from medical staff on the positive effects of physical activity and nutritional aspects for the prevention of osteoporosis. Participants in this group will carry the device, but will not have access to the Muvone App. Objective 2 Upon placement of the accelerometers, participants completed a standardized exercise protocol during a laboratory-based activity session. The exercise protocol consisted of five countermovement jumps (CMJ) with a 30 s rest between each jump. The jump height was measured with a light barrier system (OptoGait, Microgate, Bolzano, Italy). Participants then performed four treadmill activities (HP-Cosmos®, Traunstein, Germany), walking/running at 4.8, 6.4, 9.7 and 12 km·h-1.The duration of each treadmill bout was 2 min with a 1 min rest between each condition. Participants stood still during rest intervals and completed every exercise protocol in a randomized order. To evaluate spatio-temporal gait parameters during the protocol, a floor-based photocell system (OptoGait, Microgate, Bolzano, Italy) was used. Finally, participants will be instructed to fix their shins on a soft pillow and extend their knees to lean forward. The thigh will be set at 30º from the horizontal. Participants will be raised to reach a proper height from the ground (3 cm or 5 cm between the palm of the hand and the force plates). The subjects will be instructed to keep their elbows fully extended (i.e. outstretched hands) and the arm angle will be set at 15° to the vertical. Each participant will complete ten trials, which involved five trials at h=3 cm, 15 minutes of rest, and five experiments at h=5 cm. To assess the intensity and amount of impact loading, the vertical acceleration peak values will be recorded with an accelerometer based human body movement monitor (Fitbit). The monitor was worn on a wrist 2 cm above the medial epicondyle. For the individual quantification of their daily physical activity, all participants will be asked to carry the monitor during all waking hours for 36 weeks. The number of daily acceleration peaks will be recorded at 33 acceleration levels from 0.3 to 9.9 g. The individual daily average number of peaks at each acceleration level will be calculated for the analysis. The number at different acceleration levels will be analysed to describe the intensity of exercise. Previously, the reliability of the accelerometer-based method using a three-dimensional prototype was tested. However, in this project, the impacts at the wrist level will be measured. The peak values were predominantly obtained immediately after heel contact in different activities; furthermore, the peak acceleration values will be compared with the values obtained simultaneously using a standard optical motion analysis system. Moreover, the acceleration values will also be compared with the GRF measured with a force plate (Kistler 9287A, Kistler Instrument, Switzerland), with the acceleration values multiplied by body weight. Objective 3 In order to identify the key factors for the transfer of research results obtained in the experimental phase to the public socio-health system, the results will be discussed in 3 meetings by a group of experts in the health system (e.g., a district director of Primary Care centers or a politician from the health area of the Junta de Andalucía, two directors of primary care centers, an expert in health promotion, a nurse in care of patients with osteoporosis, etc.). Different models will be designed using biochemical, image, clinical and physical condition data that guide medical personnel on the most relevant factors that determine the response to this type of treatment. Out the whole set of variables included on the electronic healthcare records, a new medical model including only the selected high impact variables (sub-dataset) calculated with machine learning feature selection techniques will be designed. The health gain (QUALYs) and costs will be determined, and the cost-effectiveness ratio (ICUR) will also be calculated.