No abstract

A positive association between fat and bone mass is maintained through a network of signaling molecules. Clinical studies found that the circulating levels of adiponectin, a peptide secreted from adipocytes, are inversely related to visceral fat mass and bone mineral density, and it has been suggested that adiponectin contributes to the coupling between fat and bone. Our study tested the hypothesis that adiponectin affects bone tissue by comparing the bone phenotype of wild-type and adiponectin-knockout (APN-KO) female mice between the ages of 8-37 weeks. Using a longitudinal study design, we determined body composition and bone density using dual energy x-ray absorptiometry. In parallel, groups of animals were killed at different ages and bone properties were analyzed by microcomputed tomography, dynamic histomorphometry, 3-point bending test, nanoindentation, and computational modelling. APN-KO mice had reduced body fat and decreased whole-skeleton bone mineral density. Microcomputed tomography analysis identified reduced cortical area fraction and average cortical thickness in APN-KO mice in all the age groups and reduced trabecular bone volume fraction only in young APN-KO mice. There were no major differences in bone strength and material properties between the 2 groups. Taken together, our results demonstrate a positive effect of adiponectin on bone geometry and density in our mouse model. Assuming adiponectin has similar effects in humans, the low circulating levels of adiponectin associated with increased fat mass are unlikely to contribute to the parallel increase in bone mass. Therefore, adiponectin does not appear to play a role in the coupling between fat and bone tissue.

No abstract

With the emergence of wearable technology and machine learning approaches, gait monitoring in real-time is attracting interest from the sports biomechanics community. This study presents a systematic review of machine learning approaches in running biomechanics using wearable sensors. Electronic databases were retrieved in PubMed, Web of Science, SPORTDiscus, Scopus, IEEE Xplore, and ScienceDirect. A total of 4,068 articles were identified via electronic databases. Twenty-four articles that met the eligibility criteria after article screening were included in this systematic review. The range of quality scores of the included studies is from 0.78 to 1.00, with 40% of articles recruiting participant numbers between 20 and 50. The number of inertial measurement unit (IMU) placed on the lower limbs varied from 1 to 5, mainly in the pelvis, thigh, distal tibia, and foot. Deep learning algorithms occupied 57% of total machine learning approaches. Convolutional neural networks (CNN) were the most frequently used deep learning algorithm. However, the validation process for machine learning models was lacking in some studies and should be given more attention in future research. The deep learning model combining multiple CNN and recurrent neural networks (RNN) was observed to extract different running features from the wearable sensors and presents a growing trend in running biomechanics.