2.4. Statistical Analysis

Data were analyzed using SPSS (version 22, Chicago, IL, USA). Using G*power calculations for power analysis, a TTP effect size = 0.14 was observed between controls and MetS using a sample size of 200. The margin of error = 0.05, and the actual power achieved = 86.1%. Continuous data were presented as mean ± standard deviation (SD) for normal variables, and non-Gaussian variables were presented in median (1st and 3rd) percentiles. All continuous variables were checked for normality using the Kolmogorov–Smirnov test. Non-Gaussian variables were log-transformed prior to parametric analysis. An independent t-test and a Mann–Whitney U were performed to compare the mean and median differences in Gaussian and non-Gaussian variables. A linear regression analysis was performed to determine the explained variation and correlation between TTP and MetS components after adjusting for age and gender. Correlations between variables were done using a Pearson’s and a spearman correlation analysis. A sensitivity and specificity analysis was performed for TTP with MetS. p value <0.05 was considered statistically significant. All figures were plotted in MS Excel except for the receiver operating characteristics (ROC) curve. The ROC curve was plotted in MedCalc^® Statistical Software version 19.5.2 (MedCalc Software Ltd., Ostend, Belgium; https://www.medcalc.org; accessed on 20 May 2021). The curve was performed to optimize cutoff points for TTP based on high sensitivity and specificity, PPV and ± LR, using independent group design (condition variable MetS) and the criterion variable TTP, with a higher value indicating a positive condition. The area under the curve (AUC) was also calculated to provide a comparison of the performance of TTP as a biomarker. An AUC of 0.9 to 1 is considered excellent, 0.8 to 0.9 is considered good, 0.7 to 0.8 is considered fair, 0.6 to 0.7 is considered poor, and 0.5 to 0.6 is considered very poor.