The robustness and appropriateness of the approach were assessed using the mean relative error (MRE).
The diagnostic performance of our predictive model after cross-validation was evaluated in terms of discrimination, using the area under the receiver operating characteristics (ROC) curve corrected for overoptimism, and calibration, assessed using the mean absolute error.
Interrater reliability was assessed using the mean standard deviation and Pearson correlation coefficients.
Strength was assessed using the mean of a 10RM.
Anxiety and depression were assessed using the mean item score on the locally-validated HSCL-25.
Nephropathy was assessed using the mean of two early morning urine albumin creatinine ratio (ACR) measurements.
The accuracy was assessed using the mean square error and permutation was used to obtain the alpha 0.01 cut-off.
Hyperhomocysteinaemia is usually associated with the genome-wide hypomethylation as assessed using the mean long interspersed nucleotide element-1 (LINE-1) methylation in humans.
Dietary adherence was assessed using the mean of two timed urine collections obtained in the 7-day prestudy period, and no subjects were excluded because of dietary nonadherence.
To identify the most individually predictive features, random forest feature importance was assessed using the mean decrease in Gini index (MDG).
For comparison of baseline characteristics, ordinal measures were assessed using the mean score Cochran Mantel Haenszel test, nominal scale measures were assessed using the generalized Cochran Mantel–Haenszel test, and continuous measures were assessed using the t-test.
