Spectrochemical analysis of blood combined with chemometric techniques for detecting osteosarcopenia.

Among several complications related to physiotherapy, osteosarcopenia is one of the most frequent in elderly patients. This condition is limiting and quite harmful to the patient's health by disabling several basic musculoskeletal activities. Currently, the test to identify this health condition is complex. In this study, we use mid-infrared spectroscopy combined with chemometric techniques to identify osteosarcopenia based on blood serum samples. The purpose of this study was to evaluate the mid-infrared spectroscopy power to detect osteosarcopenia in community-dwelling older women (n = 62, 30 from patients with osteosarcopenia and 32 healthy controls). Feature reduction and selection techniques were employed in conjunction with discriminant analysis, where a principal component analysis with support vector machines (PCA-SVM) model achieved 89% accuracy to distinguish the samples from patients with osteosarcopenia. This study shows the potential of using infrared spectroscopy of blood samples to identify osteosarcopenia in a simple, fast and objective way.

Relationship between phase angle and physical performance measures in community-dwelling older adults.

INTRODUCTION: The phase angle (PhA) is a measure of great clinical relevance provided through the Bioelectric Impedance Analysis (BIA). PhA is related to health status. Physical performance measures are also similarly associated to the health status of older individuals, however, studies which asses the relationship between these two measures are scarce. OBJECTIVE: To identify the relationship between PhA and physical performance measures in community-dwelling older adults in a Brazilian sample. METHODOLOGY: This was a cross-sectional study in which 200 community-dwelling older adults up to 65 years of age of both genders were recruited. Physical performance was evaluated by walking speed and handgrip strength, and the PhA was derived from BIA. Linear regression models were used to estimate the associations between PhA and physical performance measures. Two models were built: the first model was adjusted by handgrip and walking speed; and the second model additionally included the number of chronic diseases, gender, age and body mass index (BMI). RESULTS: A total of 200 subjects were evaluated through BIA. Men showed a mean age of 72.13 ± 3.42 years and women 71.94 ± 3.35 years. Mean PhA among men was 5.99 ± 0.67, while the mean obtained for women was 5.43 ± 0.70. Linear regression showed that handgrip strength (ß: 0.036; p-value < 0.001; ß: 0.024; p-value: 0.005) and walking speed (ß: 0.495; p-value: 0.044; ß: 0.619, p-value: 0.009) were correlated with the PhA in both models. CONCLUSION: The results of our study revealed that PhA is a good marker of physical performance for the Brazilian community-dwelling older adults studied.