Background: There have been numerous studies of asthma in twins, but no study has evaluated whether lung function predictive models yield similar results between twin and general populations. We sought to evaluate this in late childhood and adolescent subjects.
Methods: We generated cross-sectional, sex- and age-specific regression models of FEV(1), and FVC, in a community-based cohort of 3140 healthy, non-smoking Chinese twins using generalized estimating equations to adjust for correlations within twin pairs. We applied the model to a healthy non-smoking general population cohort of 2187 subjects from the same region, and compared %predicted FEV(1) and FVC values between the two populations.
Results: Stratified by age and sex, the associations of height with FEV(1) or FVC varied by age group. During the adolescent growth spurt (age 13 for girls and ages 14-16 for boys), the associations of height with FEV(1) or FVC were nonlinear and greater than that seen at other ages. During adolescence, FEV(1) and FVC for a given height increased with age. The percent predicted values of FEV(1) and FVC in the twin population were similar to that of the general population.
Conclusions: Twin and general populations have similar patterns of lung function change over middle childhood and adolescence. Similar equations may be used to estimate percent predicted values. Finally, a single prediction equation cannot completely describe patterns of lung function from childhood through adolescence due to puberty related changes.