Statistical analyses
All data were encoded and statistically analyzed using
SPSS Statistics 17.0 for Windows. Between-group changes in
outcomes were compared at baseline and at 2 and 4 weeks.
Effects were evaluated on an intention-to-treat basis. The
statistical analysis was performed with the independent
student’s t test or Mann-Whitney U test for comparing two
different sets of data. In addition, the intrasubject variation at
baseline and at 2 and 4 weeks was determined with repeated-measures
analysis of variance or the Friedman test. If the
repeated-measures analysis or Friedman test revealed significant
differences between individual time points, the
Bonferroni method or the Wilcoxon signed-rank test (correction
by Bonferroni inequality) was performed.
One-way analysis of variance was used for the number of
yoga practice sessions (control versus poor adherence versus
good adherence) and the post hoc test was estimated using
the Bonferroni method. A two-sided p-value < 0.05 indicated
statistical significance. For correction by Bonferroni inequality,
a two-sided p-value < 0.016 indicated statistical signifi-
cance. Basically, intention-to-treat analysis has been done,
but 19 intervention participants were excluded who did not
report how many days they practiced or who withdrew at
only 2 weeks, as indicated in Table 4