Table 2.
The patients were stratified by sex and placed into a logistic regression model predicting group. The model for females showed that no variables were significant. The model for males, however, had a significant AHI variable (P value = .026) and a significant two-way interaction between age and AHI (P value = .023). Older ages with higher AHI values are associated with a higher chance of being compliant. There were sixty-six observations with 32 compliers that were used in the analysis for males and 29 observations with 12 compliers that were used for the analysis of females.
| Sex | Variable | Coefficient | P-value |
|---|---|---|---|
| Male | Age/10 | −0.83 | .148 |
| AHI/10 | −1.58 | .026 | |
| Inter | 0.3 | .023 | |
|
| |||
| Female | Age/10 | −0.57 | .612 |
| AHI/10 | 0.23 | .877 | |
| Inter | −0.01 | .975 | |