Cross-validation Group Research Articles

Estimation of Lower Body Muscle Power from Vertical Jump in Youth

ABSTRACT Incorporation of the vertical jump (VJ) test in health-related fitness batteries is predicated on the availability of prediction equations to accurately estimate VJ power. The purpose was to develop and validate a generalizable equation to estimate VJ power from VJ height, body mass, age, and sex in 10- to 18-year-olds. The sample (N = 529) performed a VJ test and had VJ power assessed via mechanography. Validation and cross-validation groups were randomly formed to assess accuracy. Results indicated that estimates of VJ power from VJ height and body mass were accurate (R = .93, SEE = 275 Watts). Age and sex did not add substantially to the model. Upon cross-validation, accuracy was maintained. The newly developed model was more accurate than previously published equations. The following equation provides accurate and feasible estimates of VJ power (Watts) = −1354.820 + (35.455*VJ[cm]) + (43.942*body mass[kg]). This equation provides a feasible way to estimate power in field-based settings.

ESTIMATING BODY FAT PERCENTAGE THROUGH BODY MASS INDEX AND HANDGRIP STRENGTH IN MIDDLE AND OLDER-AGED ASIAN ADULTS

Objectives: To determine the ability of handgrip strength combined with body mass index (BMI, kg/m2) to estimate body fat percentage (BF%) in middle-aged and older Asian adults. Methods: Middle-aged and older Asian adults (n=459, males=197) were randomly divided into a validation and model development group (n=303) and cross-validation group (n=156). A whole-body scan using dual energy x-ray absorptiometry measured BF%. Bland-Altman plots, standard error of the estimates, total errors and mean absolute errors were used to compare prediction equations. Stepwise regression analysis was used to determine a new prediction equation for middle-aged and older Asian adults. Right and left handgrip strength, age, sex and BMI were included in the analysis. Results: A previously developed prediction equation that included handgrip strength poorly predicted BF% in our current sample with the mean difference being -6.0 ± 4.2%. Predicted BF% values were significantly lower than measured BF% values (22.7% vs. 28.7%, p<0.05). A new prediction equation was developed that included sex, BMI, left handgrip strength and age. Validation of the new equation revealed a constant error of 0.2 ± 3.9% with there being no significant difference between measured and predicted BF% (28.2% vs. 28.5%, p=0.467). Previously developed BF% equations using BMI, but not handgrip strength, had similar constant errors and mean absolute errors compared to the new prediction equation. Conclusion: Handgrip strength does not appear to improve the estimation of body fat percentage from BMI prediction equations in middle and older-aged Asian adults.

Panoramic prediction equations to estimate implant- to-mandibular canal dimensions in the mandibular posterior region: implications for dental implant treatment

BackgroundTo develop and cross-validate site-specific panoramic radiography (PAN) analysis prediction equations of implant-to-mandibular canal dimensions (IMCD) in mandibular regions posterior to the mental foramen, and to help determine in which instances CBCT technology will be a justified adjunct in clinical practice.MethodsIMCD by PAN (Pan-D) from implant site-specific regions (first premolar, second premolar, first molar, and second molar sites) were collected from 40- to 70-year-old adolescents. They were randomly assigned to validation (n = 144) and cross-validation (n = 148) groups. The cone-beam computed tomography (CBCT) technique was used as the criterion method for the estimation of IMCD (CBCT-D). The PAN analysis equations were developed using stepwise multiple regression analysis and cross-validated using the Bland–Altman approach.ResultsThere was a significant relationship between PAN-D and CBCT-D for both validation (R2 = 57.8 %; p < .001) and cross-validation groups (R2 = 52.5 %; p < .001). Root means-squared error (RMSE) and pure error (PE) were highest for the first molar (RMSE = 1.116 mm, PE = 1.01 mm) and the second molar region (RMSE = 1.162 mm, PE = 1.11 mm).ConclusionsPAN-D has the potential to be developed as an indirect measure of IMCD. However, the findings suggest to exclude scoring of the first and second molars when assessing IMCD via PAN. Use of CBCT may be justified for all IMCD estimations in the first and second molars regions.Trial registrationThis study has been registered and approved by the Ethics Committee of the Martin-Luther University, Halle, Germany (2020-034).

Predictive equation for assessing appendicular lean soft tissue mass using bioelectric impedance analysis in older adults: Effect of body fat distribution.

Low muscle mass is associated with sarcopenia and increased mortality. Muscle mass, especially that of the limbs, is commonly estimated by dual-energy X-ray absorptiometry (DXA) or bioimpedance analysis (BIA). However, BIA-based predictive equations for estimating lean appendicular soft tissue mass (ALST) do not take into account body fat distribution, an important factor influencing DXA and BIA measurements. To develop and cross-validate a BIA-based equation for estimating ALST with DXA as criterion, and to compare our new formula to three previously published models. One-hundred eighty-four older adults (140 women and 44 men) (age 71.5 ± 7.3 years, body mass index 27.9 ± 5.3 kg/m2) were recruited. Participants were randomly split into validation (n = 118) and cross-validation groups (n = 66). Bioelectrical resistance was obtained with a phase-sensitive 50 kHz BIA device. A BIA-based model was developed for appendicular lean soft tissue mass [ALST (kg) = 5.982 + (0.188 × S2 / resistance) + (0.014 × waist circumference) + (0.046 × Wt) + (3.881 × sex) - (0.053 × age), where sex is 0 if female or 1 if male, Wt is weight (kg), and S is stature (cm) (R2 = 0.86, SEE = 1.35 kg)]. Cross validation revealed r2 of 0.91 and no mean bias. Two of three previously published models showed a trend to significantly overestimate ALST in our sample (p < 0.01). The new equation can be considered valid, with no observed bias and trend, thus affording practical means to quantify ALST mass in older adults.

A New Equation to Estimate Energy Expenditure Using Heart Rate in Children

The first aim of this study was to develop equations to predict physical activity energy expenditure (PAEE) for children utilizing heart rate monitors (HRM) and vector magnitudes (VM) from accelerometers. The second aim was to cross-validate the developed PAEE prediction equations and compare the equations to the pre-existing accelerometer-based PAEE equation (i.e., Trost). Seventy-five students in elementary school (from 10 to 13 years old) were classified into an equation calibration group (N = 50, 33 boys and 17 girls) and a cross-validation group (N = 25, 20 boys and 5 girls). Participants simultaneously wore a portable indirect calorimeter (Cosmed’s K4b2), a heart rate monitor on the chest, and an accelerometer on the right side of the waist. Then, the participants performed a series of various intensity activities. The energy expenditure (EE) measured by K4b2 was set as the dependent variable. Multiple regression analysis was performed to derive the heart rate and accelerometer-based equations. The heart-rate-based EE equation had an explanatory power of adj. R2 = 0.814 and the accelerometer-based EE equation had an explanatory power of adj. R2 = 0.802. The VM-based EE indicated high mean absolute percent errors (MAPE) at light, moderate, and vigorous intensity. The heart-rate-based EE was included in the range of equivalence limit in all activities, but the VM and pre-existing equation showed some overestimation beyond the equivalence range. The agreement errors between the criterion EE and the estimated EE were lower in the heart-rate-based equation than the accelerometer-based equations (i.e., VM and Trost). The approach with the heart-rate-based EE equation demonstrated higher accuracy than the accelerometer-based EE equations. The results of the current study indicate that the heart-rate-based PAEE equation can be a potential method for estimating children’s PAEE.

Development of an assessment measure for sexual disinhibition in dementia.

Sexual disinhibition (SD), a neuropsychiatric symptom characterized by sexually inappropriate comments and/or behaviors, remains under identified in dementia, possibly due to a lack of standardized assessment methods. A recent systematic review of measures used to assess SD proposed four behavioral domains and identified the need for a population-specific measure. The present study addressed this by examining the underlying factor structure of SD to create a new caregiver-report measure. Dementia caregivers (n=622) recruited online were randomly assigned to Initial Validation (n=311) or Cross-Validation (n=311) groups. Initial Validation revealed five behavioral domains that served as provisional scales for cross-validation, leading to measure development. The current study is the first to statistically evaluate the underlying factor structure of SD, resulting in a new measure that can help better characterize and identify SD.

Predicting Cardiorespiratory Fitness Using the 20-m Shuttle Run Test: New Insights Using Nonlinear Allometry.

Recently, doubts have been raised concerning the validity of the 20-m shuttle run test (20mSRT) to predict cardiorespiratory fitness (CRF) in youth. Despite these doubts, authors continue to provide powerful evidence that CRF can be predicted reliably using the 20mSRT albeit using contrasting models. Therefore, we aimed to compare a new linear model with an alternative allometric model to predict CRF (peak oxygen uptake, V˙O2peak) using the 20mSRT. The study included 148 adolescents (43% girls) age 13.37 ± 1.84 yr. Adolescents were randomly assigned to validation (n = 91) and cross-validation (n = 57) groups. V˙O2peak was measured using a gas analyzer in both maximal exercise tests in the laboratory and the 20mSRT. Multiple linear regression methods were applied to develop the linear models using the 20mSRT (laps), body mass index, and body fat percentage. Alternative allometric models were also proposed/fitted using the 20mSRT (laps), height, and body mass. The criterion validity values of both the linear and the allomeric models were found to be acceptable, with R2 = 82.5% and 82.7% respectively, providing reassuring evidence that the 20mSRT can be used with confidence to predict CRF. However, the allometric model identified a height-to-mass ratio, not dissimilar to the inverse body mass index (known to be a measure of leanness), to be associated with CRF. The allometric model also revealed that the rise in energy cost (V˙O2peak) with increasing laps was exponential. This will more accurately reflect the nonlinear rise in energy demand of shuttle running as the test progresses to exhaustion. These observations provided powerful evidence that allometric models are more than satisfactory in terms of both criterion and construct validity when predicting CRF (V˙O2peak) using the 20mSRT.

Development of a New Equation for the Prediction of Resting Metabolic Rate in Sri Lankan Adults.

Resting metabolic rate (RMR) is the key determinant of the energy requirement of an individual. Measurement of RMR by indirect calorimetry is not feasible in field settings and therefore equation-based calculations are used. Since a valid equation is not available for Sri Lankans, it is important to develop a new population-specific equation for field use. The study objective was to develop a new equation for the prediction of RMR in healthy Sri Lankans using a reference method, indirect calorimetry. RMR data were collected from fifty-seven (male 27) adults aged 19 to 60 years. They were randomly assigned to validation (n = 28) and cross-validation (n = 19) groups using the statistical package R (version 3.6.3). Height, weight, and RMR were measured. Multivariable fractional polynomials (MFP) were used to determine explanatory variables and their functional forms for the model. A variable shrinkage method was used to find the best fit predictor coefficients of the equation. The developed equation was cross-validated on an independent group. Weight and sex code (male = 1; female = 0) were identified as reliable independent variables. The new equation developed was RMR (kcal/day) = 284.5 + (13.2 x weight) + (133.0 x sex code). Independent variables of the prediction equation were able to predict 88.5% of the variance. Root mean square error (RMSE) of the prediction equation in validation and cross-validation was 88.11 kcal/day and 79.03 kcal/day, respectively. The equation developed in this study is suitable for predicting RMR in Sri Lankan adults.

Predicting cesarean delivery for failure to progress as an outcome of labor induction in term singleton pregnancy

Induction of labor is one of the most common interventions in modern obstetrics, and its frequency is expected to continue to increase. There is inconsistency as to how failed induction of labor is defined; however, the majority of studies define success as the achievement of vaginal delivery. Induction of labor in nulliparous women poses an additional challenge with a 15% to 20% incidence of failure, ending in emergency operative deliveries. The Bishop score has been traditionally used before decisions for induction of labor. Nonetheless, it is subjective and prone to marked interobserver variation. Several studies have been conducted to find alternative predictors, yet a reliable, objective method still remains to be introduced and validated. Hence, there is still a need for the development of new predictive tools to facilitate informed decision making, optimization of resources, and minimization of potential risks of failure. Furthermore, a peripartum transperineal ultrasound scan has been proven to provide objective, noninvasive assessment of labor. This study aimed to assess the feasibility of developing and validating an objective and reproducible model for the prediction of cesarean delivery for failure to progress as an outcome of labor induction in term singleton pregnancies. This was a prospective observational cohort study conducted in Cairo University Hospitals and University of Bologna Hospitals between November 2018 and November 2019. We recruited 382 primigravidae with singleton term pregnancies in cephalic presentation. All patients had baseline Bishop scoring together with various transabdominal and transperineal ultrasound assessments of the fetus, maternal cervix, and pelvic floor. The managing obstetricians were blinded to the ultrasound scan findings. The method and indication of induction of labor, the total duration of stages of labor, mode of birth, and neonatal outcomes were all recorded. Women who had operative delivery for fetal distress or indications other than failure to progress in labor were excluded from the final analysis, leaving a total of 344 participants who were randomly divided into 243 and 101 pregnancies that constituted the model development and cross-validation groups, respectively. It was possible to perform transabdominal and transperineal scans and assess all the required parameters on all study participants. Univariate and multivariate analyses were used for selection of potential predictors and model fitting. The independent predictive variables for cesarean delivery included maternal age (odds ratio, 1.12; P=.003), cervical length (odds ratio, 1.08; P=.04), angle of progression at rest (odds ratio, 0.9; P=.001), and occiput posterior position (odds ratio, 5.7; P=.006). We tested the performance of the prediction model on our cross-validation group. The calculated areas under the curve for the ability of the model to predict cesarean delivery were 0.7969 (95% confidence interval, 0.71-0.87) and 0.88 (95% confidence interval, 0.79-0.97) for the developed and validated models, respectively. Maternal age and sonographic fetal occiput position, angle of progression at rest, and cervical length before labor induction are very good predictors of induction outcome in nulliparous women at term.

Framework for Machine Learning of CT and PET Radiomics to Predict Local Failure after Radiotherapy in Locally Advanced Head and Neck Cancers.

Context:Cancer Radiomics is an emerging field in medical imaging and refers to the process of converting routine radiological images that are typically qualitatively interpreted to quantifiable descriptions of the tumor phenotypes and when combined with statistical analytics can improve the accuracy of clinical outcome prediction models. However, to understand the radiomic features and their correlation to molecular changes in the tumor, first, there is a need for the development of robust image analysis methods, software tools and statistical prediction models which is often limited in low- and middle-income countries (LMIC).Aims:The aim is to build a framework for machine learning of radiomic features of planning computed tomography (CT) and positron emission tomography (PET) using open source radiomics and data analytics platforms to make it widely accessible to clinical groups. The framework is tested in a small cohort to predict local disease failure following radiation treatment for head-and-neck cancer (HNC). The predictors were also compared with the existing Aerts HNC radiomics signature.Settings and Design:Retrospective analysis of patients with locally advanced HNC between 2017 and 2018 and 31 patients with both pre- and post-radiation CT and evaluation PET were selected.Subjects and Methods:Tumor volumes were delineated on baseline PET using the semi-automatic adaptive-threshold algorithm and propagated to CT; PyRadiomics features (total of 110 under shape/intensity/texture classes) were extracted. Two feature-selection methods were tested for model stability. Models were built based on least absolute shrinkage and selection operator-logistic and Ridge regression of the top pretreatment radiomic features and compared to Aerts' HNC-signature. Average model performance across all internal validation test folds was summarized by the area under the receiver operator curve (ROC).Results:Both feature selection methods selected CT features MCC (GLCM), SumEntropy (GLCM) and Sphericity (Shape) that could predict the binary failure status in the cross-validated group and achieved an AUC >0.7. However, models using Aerts' signature features (Energy, Compactness, GLRLM-GrayLevelNonUniformity and GrayLevelNonUniformity-HLH wavelet) could not achieve a clear separation between outcomes (AUC = 0.51–0.54).Conclusions:Radiomics pipeline included open-source workflows which makes it adoptable in LMIC countries. Additional independent validation of data is crucial for the implementation of radiomic models for clinical risk stratification.

A Generalized Equation for Prediction of VO2peak from a Step Test.

This study aimed to develop an equation to reduce variability of VO2peak prediction from a step test and compare VO2peak prediction from the new equation to the Queen's College Step Test (QCST). The development group (n=86; 21.7±2 years) was utilized to develop the SDState step test equation to predict relative VO2peak. The cross-validation group (n=99; 21.6±2 years) was used to determine the validity of the SDState step test VO2peak prediction equation. A regression analysis was used to identify the best model to predict VO2peak. Analysis of variance (ANOVA) was further used to determine differences among predicted and measured VO2peak values. Forward stepwise multiple regression identified age, sex, abdominal circumference, and active heart rate at the 3-min mark of the step test to be significant predictors of VO2peak (mL·kg-1·min-1). No differences among measured VO2peak (47.3±7.1 mL·kg-1·min-1) and predicted VO2peak (QCST, 46.9±9.3 mL·kg-1·min-1; SDState 48.3±5.7 mL·kg-1·min-1) were found. Pearson correlations, ICC, SEE, TEE, Bland-Altman plots, and Mountain plots indicate the SDState step test equation provides less variation in the prediction of VO2peak compared to the QCST. The SDState step test equation is effective for predicting VO2peak from the YMCA step test in young, healthy adults.

Blood-based kinase activity profiling: a potential predictor of response to immune checkpoint inhibition in metastatic cancer

BackgroundMany cancer patients do not obtain clinical benefit from immune checkpoint inhibition. Checkpoint blockade targets T cells, suggesting that tyrosine kinase activity profiling of baseline peripheral blood mononuclear cells may...

The Novel Use of Point-of-Care Ultrasound to Predict Resting Energy Expenditure in Critically Ill Patients.

Accurate estimation of a critically ill patient's caloric requirements is essential for a proper nutritional plan. This study aimed to evaluate the use of point-of-care ultrasound (US) to predict the resting energy expenditure (REE) in critically ill patients. In 69 critically ill patients, we measured the REE using indirect calorimetry (REE_IC), muscle layer thicknesses (MLTs), and cardiac output (CO). Muscle thickness was measured at the biceps and the quadriceps muscles. Patients were randomly split into a model development group (n=46) and a cross-validation group (n=23). In the model development group, a multiple regression analysis was applied to generate REE using US (REE_US) values. In the cross-validation group, REE was calculated by the REE_US and the resting energy expenditure using the Harris-Benedict equation (REE_HB), and both were compared to the REE_IC. In the model development group, the REE_US was predicted by the following formula: predicted REE_US (kcal/d)=206 + 173.5 × CO (L/min) + 137 × MLT (cm) - 230 × (women=1; men=0) (R2 =0.8; P < .0001). In the cross-validated group, the REE_IC and REE_US values were comparable (mean difference, -66 [-3.3%] kcal/d; P=.14). However, the difference between the mean REE_IC and the mean REE_HB was 455.8 (26%) kcal/d (P < .001). According to a Bland-Altman analysis, the REE_US agreed well with the REE_IC, whereas the REE_HB did not. Resting energy expenditure could be estimated from US measurements of MLTs and CO. Our point-of-care US model explains 80% of the change in the REE in critically ill patients.

Grip Strength and Demographic Variables Estimate Appendicular Muscle Mass Better Than Bioelectrical Impedance in Taiwanese Older Persons

ObjectivesThis study aimed to develop an equation model combining physical fitness and anthropometric parameters and compare its results with those of bioelectrical impedance analysis (BIA)-measured lean mass (LM) using dual-energy X-ray absorptiometry (DXA)-measured appendicular muscle mass (AMM) as reference. DesignObservational analysis. Setting and ParticipantsHealthy community-dwelling older subjects. MethodsA total of 1020 participants were randomly allocated to the development group (development group, n = 510) or the cross-validation group (validation group, n = 510). Body composition was measured using both DXA and BIA, and physical fitness parameters, including grip strength, timed stepping test, sit-to-stand test, flexibility, and walking speed were also assessed. A prediction equation model of AMM by stepwise linear regression analysis that included or excluded 1 independent variable at each step, based on the P value of significance (P < .05), was developed. ResultsUsing weight, sex, height, and handgrip strength as independent variables, the equation AMM = −9.833 + 0.397 × weight (kg) + 4.433 × sex + 0.121 × height (cm) + 0.061 × handgrip strength (kg) best predicts DXA-measured AMM (adjusted R2 = 0.914, SEE = 2.062, P < .001). The predicted AMM was more highly correlated with DXA-measured AMM than the commonly used BIA-measured LM (R2= 0.9158 and 0.8427, respectively, both P < .001). Using DXA-measured AMM as reference, the Bland-Altman plot showed mean differences of −0.03 kg and −0.12 kg, with limits of agreement of −3.98 to 3.92 kg and −5.97 to 5.73 kg for the predicted AMM and BIA-measured AMM, respectively. Conclusions and ImplicationsThe proposed equation offers a practical alternative method for estimating AMM that is less facility-dependent and more easy to use and affordable than instrumental studies.

What variables determine sprint performance in young athletes?

Summary Objectives To identify sprint performance determinants in young athletes in the field in both sexes. Equipment and methods Two hundred and fourteen track and field athletes, ages between 13 and 15 years old, were divided into validation group (VG) (n = 163) and cross-validation group (CVG) (n = 51). The participants were submitted to biological maturity evaluation, anthropometry, vertical jump, 60-m sprint, and running anaerobic sprint test (RAST). Results The multiple regression analysis selected the variables leg length, vertical jump, RAST total time (RASTtt) and fatigue index for both groups, determined 95.5% and 76.9% of sprint performance on boys and girls, respectively. To confirm our results, we performed a cross-validation on the linear regression, which showed that the measured and predicted sprint performances in CVG were not significantly different (P > 0.05). In addition, the study demonstrated that biological maturity is associated with all variables of the model (P Conclusion The variables leg length, vertical jump, RASTtt and fatigue index explain in different proportions sprint performance in young boys and girls athletes. In addition, this study add to knowledge of researchers and coachers, indicating which variables should be part of the training program of young athletes sprint runners.

Multitrait genomic prediction of methane emissions in Danish Holstein cattle

In dairy cattle, selecting for lower methane-emitting animals is one of the new challenges of this decade. However, genetic selection requires a large number of animals with records to get accurate estimated breeding values (EBV). Given that CH4 records are scarce, the use of information on routinely recorded and highly correlated traits with CH4 has been suggested to increase the accuracy of genomic EBV (GEBV) through multitrait (genomic) prediction. Therefore, the objective of this study was to evaluate accuracies of prediction of GEBV for CH4 by including or omitting CH4, energy-corrected milk (ECM), and body weight (BW) as well as genotypic information in multitrait analyses across 2 methods: BLUP and single-step genomic BLUP (SSGBLUP). A total of 2,725 cows with CH4 concentration in breath (14,125 records), BW (61,667 records), and ECM (61,610 records) were included in the analyses. Approximately 2,000 of these cows were genotyped or imputed to 50K. Ten cross-validation groups were formed by randomly grouping paternal half-sibs. Five scenarios were performed: (1) base scenario with only CH4 information; (2) without CH4, but with information from BW, ECM, or BW+ECM only in reference population; (3) without CH4, but with information from BW, ECM, or BW+ECM in both validation and reference population; (4) with CH4 information and BW, ECM, or BW+ECM information only in the reference population; and (5) with CH4 information and BW, ECM, or BW+ECM information in both validation and reference population. As a result, for each method (BLUP, SSGBLUP), 13 sub-scenarios were performed, 1 from scenario 1, and 3 for each of the subsequent 4 scenarios. The average accuracy of GEBV for CH4 in the base scenario was 0.32 for BLUP and 0.42 for SSGBLUP, and it ranged from 0.10 in scenario 2 to 0.78 in scenario 5 across methods. In terms of bias, the base scenario 1 was unbiased for SSGBLUP; similar results were achieved with scenario 5. Including information on ECM increased the accuracy of GEBV for CH4 by up to 61%, whereas adding information on both traits (BW and ECM) increased the accuracy by up to 90%. Scenarios that did not include CH4 in the reference population had the lowest correlations (0.17-0.33) with single-trait CH4 GEBV, and scenarios with CH4 in the reference population had the highest correlations (0.41-0.81). Thus, failure to include CH4 in future reference populations results in predicted CH4 GEBV, which cannot be used in practical selection. Therefore, recording CH4 in more animals remains a priority. Finally, multiple-trait genomic prediction using routinely recorded BW and ECM leads to higher prediction accuracies than traditional single-trait genomic prediction for CH4 and is a viable solution for increasing the accuracies of GEBV for scarcely recorded CH4 in practice.

Validation of equations to estimate the peak oxygen uptake in adolescents from 20 metres shuttle run test

ABSTRACT This study aimed to develop new prediction models from directly determined peak of oxygen uptake (VO2peak) in adolescents using 20-metre shuttle run test (20 m-SR) and to compare the new models with previously published equations. This study included 148 adolescents (43% girls), aged 13.37 ± 1.84 years old. Adolescents were randomly assigned to validation (n = 91) and cross-validation (n = 57) groups. VO2peak was measured using a gas analyser in both maximal exercise tests in the laboratory as well as by 20 m-SR. The multiple linear regression method was applied to develop the models using BMI, BMI-z score and body fat percentage (%FM). The proposed models presented better predictive performance (R2 = 75.6% to 78.9%) and a lower absolute percentage error (%error = 10.51 to 11.78%) than the previously published equations (R2 = 38.8% to 69.1%; %error = 13.13% to 21.54%). The Model with BMI-z was the best fit equation in girls, and the model with BMI-z and %FM in boys. Therefore, it is recommended that the equations developed in the present study be used in future research and projects in the school environment to estimate VO2peak in adolescents by the 20 m-SR test.

Development and Cross-Validation of Anthropometric Predictive Equations to Estimate Total Body Fat Percentage in Adult Women in Sri Lanka.

Attempts have been made to estimate body fat using anthropometry, and most of them are country-specific. This study was designed to develop and cross-validate anthropometric predictive equations to estimate the total body fat percentage (TBFP) of Sri Lankan adult women. A cross-sectional study was conducted in Galle, Sri Lanka, with two groups: Group A (group for equation development) and Group B (cross-validation group) (n = 175 each) of randomly selected healthy adult women aged 30–60 years. TBFP (%) was quantified with total body DXA (TBFPDXA). Height (m), weight (kg), and skinfold thickness (SFT, mm) at six sites and circumferences (cm) at five sites were measured. In the first step, four anthropometric equations were developed based on the data obtained from multiple regression analyses (TBFPDXA = dependent variable and anthropometric measurements and age = independent variables) with Group A. They were developed on the basis of circumferences (TBFP1), SFTs (TBFP2), circumferences and SFTs (TBFP3), and highly significant circumferences and SFTs (r ≥ 0.6) (TBFP4). In the second step, the newly developed equations were cross-validated using Group B. Three equations (TBFP1, TBFP2, and TBFP4) showed the agreement with cross-validation criteria. There were no differences between TBFPDXA and TBFP estimated by these equations (p > 0.05). They showed higher measurement concordance with TBFPDXA; correlation between measured TBFP with DXA and estimated with TBFP1, TBFP2, and TBFP4, respectively, was 0.80 (R2 = 0.65, SEE = 3.10), 0.83 (R2 = 0.69, SEE = 2.93), and 0.84 (R2 = 0.72, SEE = 2.78). Three anthropometric measurements based on predictive equations were developed and cross-validated to satisfactorily estimate the TBFP in adult women.

Cross-validation Of A Prediction Equation For Energy Expenditure Of An Acute Resistance Exercise Bout.

Previously, our laboratory introduced a regression equation for predicting net kcal consumption of a resistance exercise (RE) bout: Total net kcal = 0.874(height, cm) - 0.596(age, years) - 1.016(fat mass, kg) + 1.638(lean mass, kg) + 2.461(total volume x 10-3) - 110.742 (R2 = 0.773, SEE=28.5 kcal). PURPOSE: The purpose of this study was to validate this regression equation using the same variables as predictors. METHODS: Forty-seven healthy, active subjects (23 men, 24 women, 20-58 yrs, 173.5 ± 10.5 cm, 85.5 ± 19.0 kg, VO2max 36.0 ± 8.4 ml/kg/min) were randomly divided into validation and cross-validation groups (nv = 24, ncv = 23). The validation group’s data was used to develop an equation to predict net kcal consumption, which was applied to the cross-validation group’s data to estimate net kcal consumption. Similarly, a prediction equation was derived from the cross-validation group’s raw data and applied to that of the validation group. The strength of the relationship between each group’s measured and estimated net kcal consumption was assessed via correlational analysis. RESULTS: Multiple linear regression yielded the following estimates of net kcal consumption: validation net kcal = 1.125(height, cm) – 0.662(age, years) – 0.800(fat mass, kg) + 1.344(lean mass, kg) + 2.278(total volume x 10-3) – 144.846 (R2 = 0.751, p < 0.0001, SEE=29.7 kcal); cross-validation net kcal = 0.515(height, cm) - 0.520(age, years) - 1.220(fat mass, kg) + 1.995(lean mass, kg) + 2.620(total volume x 10-3) – 59.988 (R2 = 0.823, p < 0.0001, SEE=29.2 kcal). These equations had a cross-validation coefficient of 0.902 and a double cross-validation coefficient of 0.863. CONCLUSION: The strong relationship between the measured and estimated net kcal consumption of both the cross-validation and validation group lead us to conclude that the regression equation derived by this laboratory is valid for estimating net energy expenditure for a total RE bout.

Appling Decision Tree Technique To Predict Fall Risks Classified By Functional Fitness In Older Adults

PURPOSE: To examine an accuracy of the fall risk classification predicted by senior functional fitness tests and health conditions in Korean older adults utilizing a decision tree technique, which is a machine learning algorithm METHODS: A total of 732 older adults (Males 45.9%, age: 73.51±6.20yrs, BMI: 24.07) participated in Korea Survey of National Physical Fitness (KSNPF). All participants performed the senior physical fitness test (SPFT) including body composition (FAT%), 6min walk, hand grip(HG), timed up-and-go (TUG), and chair sit-and-stand. Also, demographic variables (income, employment, etc.) and health markers were measured, which were including blood pressure & waist and hip circumferences as well as physical activity levels and health conditions (disease, medication, and fall experiences). All utilized measures were validated for the Korean elderly in 2014. To determine a classifier of a fall risk and to set cut-off scores of the fall risk classification, a decision tree technique with CHAID algorithm (Kass, 1980) was applied. To examine the accuracy of the fall risk classification predicted by selected variables, 80% of participants were randomly selected to derive the equation for a training group (GR) and the others were assigned for a cross-validation GR (20% for testing). RESULTS: Only TUG and HG were significant classifiers with 87.6% of accuracy in the testing GR and 86.6% of precision in the cross-validation GR, respectively. The cut-off score of TUG was 7.13sec. (chi-square=21.22, p<.001), in which 21.1% of participants were classified into the fall risk GR. Within the TUG GR (<7.13sec.), the cut-off score of HG were set as 23.6kg (chi-square=14.87, p<.001), and 52.8% (<23.6kg) was classified into the fall risk GR. CONCLUSIONS: Among SPFT tests and health related variables, TUG and HG were relatively important to predict the fall risk for the healthy elderly in Korea. Coordination and strength exercise are critical for fall prevention in older adults. *Corresponding author ([email protected])