Intraindividual variability of infant whole-body plethysmographic measurements: effects of age and disease

The intraindividual variability of whole-body plethysmographic measurements was studied in a large series of consecutive infants (N = 144), divided into two groups: a group of infants born very prematurely (PM, N = 63), with (N = 28) or without (N = 35) a history of bronchopulmonary dysplasia (BPD), and a group of infants with persistent respiratory symptoms (PRS, N = 81), i.e., wheezing (N = 53) or cough (N = 28). The intraindividual variability was determined within each test and between tests, separated by a 10-min interval. In both study groups, the between-test variability was significantly larger than that within tests. Expressed as the median coefficient of variation (CV), the between-test repeatabilities in the PRS group were 8.0% for thoracic gas volume (TGV), 17.5% for airway resistance (Raw), and 18.4% for specific airway conductance (sGaw), and in the PM group, 8.9% for TGV, 20.4% for Raw, and 20.7% for sGaw. However, the individual range of CVs was large, ranging from 3 to 19% for TGV and from 5 to 55% for sGaw. With respect to TGV, the difference between the groups was statistically significant (P = 0.03). In infants with a history of BPD, there was also a significant negative age dependency in CVs of sGaw (r = -0.50, P = 0. 009), showing larger variation among younger individuals. The presenting symptom (wheezing or cough) in the PRS group did not influence the measurement variability significantly, and neither did the degree of bronchial obstruction. We conclude that on a group basis, the repeatability of infant body plethysmographic measurements may be satisfactory for scientific studies demonstrating pharmacodynamic effects; however, the intraindividual measurement variability should be reported for each test conditions and for infant groups in each study. Due to the large range in individual variation and the influence of age and disease processes on the variation, for an individual child there is only questionable benefit from a given measurement, unless the intrasubject, between-test variability is assessed individually before interventions, such as a bronchodilation test.