Adjusting Grip Strength to Body Size: Analyses From 6 Countries.

OBJECTIVES: Absolute handgrip strength and adjusted by body mass index are useful to identify age-related conditions. However, these values are not accurate for older adults with extreme body size because of the nonlinear relationship between strength, height, and body mass. The purpose of this study was to determine cut-off points for age-related conditions of older adults using allometric coefficients to normalize grip strength by body size. DESIGN: Cross-sectional study. SETTING AND PARTICIPANTS: Data from 13,235 older adults of Study on Global Aging and Adult Health conducted in 6 low- and middle-income countries were analyzed. METHODS: Country- and sex-specific allometric exponents for body-size variables (mass and height) were computed with log-linear models. Partial correlation verified whether allometric normalization removed the effect of body size on grip strength. Cut-off points were established (<20th percentile) for low allometrically adjusted grip strength. RESULTS: Allometric exponents for normalization of grip strength were provided for body-size variables, ranging from 0.19 to 2.45. Allometric normalization removed the effect of body size on grip strength (r < 0.30). Overall, frequencies of low muscle strength were overestimated with international criteria (absolute grip strength) compared with the cut-off points proposed in this study. CONCLUSIONS AND IMPLICATIONS: The proposed allometric exponents normalized grip strength according to body-size variables. These exponents improved the accuracy in identifying age-related conditions in older adults with extreme body size. The variability between strength reveals the need for developing specific cut-off points for low- and middle-income countries. New cut-off points of low normalized grip strength with automatized applicability were proposed for health care providers use in clinical practice.

Accuracy of Deep Neural Network in Triaging Common Skin Diseases of Primary Care Attention.

Access to dermatological care can be challenging in certain regions of the world. The triage process is usually conducted by primary care physicians; however, they may not be able to diagnose and assign the correct referral and level of priority for different dermatosis. The present research aimed to test different deep neural networks to obtain the highest level of accuracy for the following: (1) diagnosing groups of dermatoses; (2) correct referrals; and (3) the level of priority given to the referral compared to dermatologists. Using 140,446 images from a teledermatology project, previously labeled with the clinical diagnosis, and their respective referrals, namely biopsy, in-person dermatologist visits or monitoring the case via teledermatology along with the general physician, 27 different scenarios of neural networks were derived, and the algorithm accuracies in classifying different dermatosis, according to the group of the diagnosis they belong to, were calculated. The most accurate algorithm was then tested for accuracy in diagnosis, referral, and level of priority given to 6,945 cases. The GoogLeNet architecture, trained with 24,000 images and 1,000 epochs, using weight random initialization and learning rates of 10-3 was found to be the most accurate network, showing an accuracy of 89.72% for diagnosis, 96.03% for referrals and 92.54% for priority level in 6,975 image testing. Our study population, however, was confined to individuals with chronic skin conditions and, therefore, it has limited value as a triage tool because it has not been tested for acute conditions. Deep neural networks are accurate in triaging, correct referral and prioritizing common chronic skin diseases related to primary care attention. They can also help health-care systems optimize patients' access to dermatologists.

Lipodystrophy diagnosis in people living with HIV/AIDS: prediction and validation of sex-specific anthropometric models.

BACKGROUND: Body composition alterations, or lipodystrophy, can lead to serious health problems in people living with HIV/AIDS (PLWHA). The objectives of this study are to predict and validate sex-specific anthropometric predictive models for the diagnosis of lipodystrophy in PLWHA. METHODS: A cross-sectional design was employed to recruit 106 PLWHA (men = 65 and women = 41) in Brazil during 2013-2014. They were evaluated using dual-energy X-ray absorptiometry, and 19 regions of body perimeters and 6 skinfold thicknesses were taken. Sex-specific predictive models for lipodystrophy diagnosis were developed through stepwise linear regression analysis. Cross-validations using predicted residual error sum of squares was performed to validate each predictive model. RESULTS: Results support the use of anthropometry for the diagnosis of lipodystrophy in men and women living with HIV/AIDS. A high power of determination with a small degree of error was observed for lipodystrophy diagnosis for men in model six (r2 = 0.77, SEE = 0.14, r2PRESS = 0.73, SEE PRESS = 0.15), that included ratio of skinfold thickness of subscapular to medial calf, skinfold thickness of thigh, body circumference of waist, formal education years, time of diagnosis to HIV months, and type of combined antiretroviral therapy (cART) (with protease inhibitor "WI/PI = 1" or without protease inhibitor "WO/PI = 0"); and model five for women (r2 = 0.78, SEE = 0.11, r2PRESS = 0.71, SEE PRESS = 0.12), that included skinfold thickness of thigh, skinfold thickness of subscapular, time of exposure to cART months, body circumference of chest, and race (Asian) ("Yes" for Asian race = 1; "No" = 0). CONCLUSIONS: The proposed anthropometric models advance the field of public health by facilitating early diagnosis and better management of lipodystrophy, a serious adverse health effect experienced by PLWHA.