Appendix 3 Statistics
All data were tested for normal distribution using the residuals for the model used in the specific analysis (Residual Q-Q plots, histogram, scatter plot)
Independent samples T-test was used to test for differences between treatments at baseline.
Paired sample T-Test was used to test for changes over time.
To answer the question “Are there differences between groups at the end of the study, controlling for baseline pre-test values?” and “did the groups change differently?” an analysis of covariance (ANCOVA) was used. The post test results were used as the dependent variable. The pre-test scores obtained before the treatment started were used as covariates. Group was used as a between subject variable. The ANCOVA using pre-test as covariate is more powerful than the ANOVA using pre-test as a linear component to the dependent variable. When participants have been randomized and a pre-test has been collected prior to treatment, the ANCOVA will have higher statistical power [224]
Pearson’s product moment was used to determine correlations and multiple linear regression was used to assess associations.
All data are means with standard error (mean (SEM)) unless otherwise stated.
For the analysis of the completer’s data SPSS 18 for windows (IBM) was used.
Data for Blood pressure and LDL cholesterol was reanalyzed excluding those participants who decreased or ceased medication during the study using ANCOVA as in the original analysis.
For the intention to treat (ITT) analysis with last known value carried forward was used. All randomized volunteers who completed the baseline visit and had all measurements at this visit done were included in the analysis. This is a modified ITT analysis because the volunteers who dropped out of the study were not contacted at the end of the 12mo intervention. Significance level was set at p