†: p-values analyzed among groups by linear mixed model statistics.
(all p-values analyzed by linear mixed model statistics >0.05).
PPO markers significantly associated with BI using mixed model statistics were also tagged by linear regression when the probability threshold was increased to P < 0.005 (data not shown).
In a repeated measures design analysed with linear mixed model statistics, participants with missing data can be included in the analysis, the time points of assessment can vary and it is possible to specify the best variance-covariance structure for the data [ 35].
Mixed-model statistics are a recommended statistical method for analyses of longitudinal data with long follow-up periods, uneven time intervals, and unequal numbers at each assessment [ 38].
It is recommended to investigate the possible bias from missing data by comparing subsamples of patients with different attrition patterns using mixed-model statistics [ 40].
Studies with small between-groups differences (i.e. %) is a common point of discussion using mixed models; as statistics are performed on all 3 trials (repeated factor) it can lead to statistical significance.
However, mixed model test statistics are prone to increased type I error rates, particularly if utilized with large amounts of missing data.
This is not too surprising since mixed model test statistics are based on asymptotic approximations, and others [13], [14] have raised concerns about inflated type I error rates when using these tests.
Because of the clustering effects of twin pairs, groups were compared by means of multilevel modeling (Linear mixed models (LMM), SPSS statistics, version 20) [ 36].
A repeated measures analysis was conducted using linear mixed models in IBM SPSS Statistics 21 [ 56], with the random statement to fit a random intercept model.
