Statistical model estimates
Parameters | Linear mixed model | Interpreted estimates | |||
Estimate# | 95% CI | p-value | Estimate¶ | 95% CI | |
Mean parameters | |||||
Below 2 weeks of age | |||||
Newborn scheduled measurement (intercept) | 3.679 | 3.457–3.900 | <0.0001 | 39.6 ppb | 31.7–49.4 ppb |
Above 2 weeks of age | |||||
Infant scheduled measurement (intercept) | 4.276 | 4.052–4.500 | <0.0001 | 71.9 ppb | 57.5–90.0 ppb |
Infant RTI measurement | −1.556 | −1.786 to −1.326 | <0.0001 | −78.9% | −83.2 to −73.4% |
Infant age months | 0.052 | 0.037–0.068 | <0.0001 | 5.4% | 3.7–7.0% |
Variance parameters | |||||
Between-subject variance (random intercept) | 0.068 | 0.029–0.308 | 0.04 | ||
Within-subject variance (residual error) | 0.49 | 0.400–0.607 | <0.0001 |
RTI: respiratory tract infection. #: estimated parameters from the linear mixed model used for analysis of log-transformed nasal nitric oxide (NO). Newborn babies only had scheduled measurements. For healthy infants (HI) >2 weeks old only, the model included RTI (yes/no) and age (months). A random intercept for each child was used in the random part of the model. ¶: estimates transformed to the nasal NO scale (a ratio scale, determined as the antilog of the estimated mean parameters, exp(estimate)). Intercepts are to be interpreted as median values whereas differences between groups are reported as percentages.