Statistical Analysis Methods for Comparing Untreated and Treated Conditions in a Study

Categorical variables were presented as percentages, and quantitative variables as medians with ranges or mean and SD, according to their distribution. The Shapiro-Wilk test was used to test the normality of continuous variables.

Regarding independent sample analyses, logistic regression models were used to compare demographic, clinical, PSG, 24-hour ABPM, and PAT variables according to untreated and treated conditions. Covariables displaying differences between conditions in univariate analysis at p < 0.10 were included in logistic models to analyze the adjusted association between the ABPM and PAT and the 2 conditions. The assumption of linearity to the logit required in logistic regression for independent continuous variables was verified.

Regarding dependent sample analyses, Student t test or Wilcoxon signed-rank test was used to compare paired differences between continuous variables in untreated and treated conditions. The McNemar test or exact McNemar test was applied for paired data at 2 categories and the Bowker test for paired data at 3 categories or more. Given the exploratory nature of our study, multiple-testing correction was not used.

In the independent sample, mixed model regression was performed to study the evolution of (1) SBP, (2) DBP, and (3) HR over 24 hours, taking into account the repeated measures (time periods 7:00 AM -1:00PM; 1:00 PM -7:00PM; 7:00 PM -1:00 AM; 1:00 AM -7:00 AM) and the treatment group (treated or untreated). Patients were considered as random effects. Time, treatment group and its interaction with time periods, age, and hypertension were considered as fixed effects. The same methodology was used to study BP and HR changes in the dependent sample where the same patients were seen in 2 conditions. Finally, logistic regression models were implemented to study ABPM between patients treated with stimulants alone or combined with anticataplectic drugs. Statistical significance was set at p < 0.05. Statistical analyses were performed with SAS, version 9.4 (SAS Institute, Cary, NC).