How ‘healthy’ do children really need to be? Going beyond the limits

Abstract: 
The establishment of healthy reference ranges and the upper and lower limits of normal for pulmonary function tests are critical for understanding the difference between health and disease in the clinical and research settings. Despite American Thoracic Society guidelines1recommending that reference values for normal lung function be derived from never‐smokers who are free of respiratory diagnoses, the interpretation of a healthy population has been widely debated and subsequently differed markedly between studies. The Global Lung Function Initiative has demonstrated that high‐quality spirometry2 and gas transfer3data, obtained from ‘healthy’, lifelong non‐smokers with modern equipment and good quality control, allow for the collation of large data sets of healthy individuals and the accurate derivation of predicted and upper and lower limits of normal.

While epidemiological studies exploring specific developmental hypotheses might validly exclude a range of adverse exposures such as preterm birth, maternal smoking during pregnancy and early life wheeze or seek to adjust for socio‐economic status of the individual4; arguably, a clinical test should be matched to the broader community in which it is being used and remains sensitive to the identification of clinically relevant disease. Indeed, the limited number of studies in this area (all using spirometry) suggests that the inclusion of children and adults with such adverse exposures often results in only minor deviations from the predicted norms of highly selected healthy populations.5, 6

In this issue of the Journal, Shackleton et al.7 present an elegant approach to exploring this issue and its impact on respiratory mechanics of young children. The authors assessed the impact of including preschool‐aged children with a history of preterm birth, early life wheeze, asthma diagnoses and/or recent respiratory symptoms in healthy reference ranges for respiratory impedance using the forced oscillation technique (FOT). Importantly, the study contains a large sample of carefully phenotyped children and concludes that preterm children and early life wheezers can be included for the generation of normative data without impacting the overall prediction equations.

The addition of children with early life wheeze in normative data appears warranted, with only small differences (<0.1 z‐scores) between the stringently healthy group and the early life wheezers. This finding is bound to be of importance when collecting normative data as approximately 20% of children with viral wheeze in early life have no signs or symptoms of asthma by school age.8 The inclusion of this population will ease the burden of recruitment for the large‐scale studies required to determine normative values and potentially as healthy participants in research studies. In addition, (as noted) parent recall about early life wheeze is a known issue, and these research findings strengthen the confidence that the lack of recall about early life wheeze is not of critical importance in research studies.

While the preterm group presented in this study were not statistically different from the stringently healthy group, the effect was actually larger than for the asthma group in some cases (e.g. −0.37 z‐scores in reactance at 6 Hz (X6)). As an a priori z‐score difference was not defined, the relevance of this finding remains unknown. Indeed, the true impact of prematurity on lung growth may have been masked by the assignment of preterm children to the asthma or current symptoms groups if they presented with these known sequelae of preterm birth. As such, we would caution against the blanket inclusion of preterm children in healthy reference ranges, particularly given the small sample size presented and relatively late median gestational age (36.1 weeks) at delivery. Indeed, previous studies have shown marked alterations in respiratory impedance in the years after preterm birth, particularly in children born less than 32 weeks’ gestation.9, 10 Recent studies also suggest increasing airway obstruction throughout childhood and adolescence in those born very preterm.11Diminished lung function is also evident after late preterm birth,12 which must be considered in the context of ‘healthy’ reference ranges.

The results of this study should not be extrapolated beyond (predominantly) Caucasian preschool children, nor beyond the FOT outcome measures examined (e.g. area under the reactance curve (AX)). Regardless, this work provides a sound starting point for further work toward more robust healthy reference ranges, particularly in that section of the population that can be difficult to recruit and obtain acceptable and repeatable lung function measures, such as young children.