Maximal static respiratory pressures in children and adolescents.

This study was designed to establish reference values of maximal static respiratory pressures in children and adolescents in our community, and compare them with previous studies. Participants were recruited from three schools (randomly chosen from those located in the metropolitan area of the city of Valencia) after appropriate consent. None of the participants had a previous history of pulmonary, cardiac, and/or skeletal abnormalities, and all of them had normal spirometry. Forced spirometry (Spirotrac III, Vitalograph) and maximal inspiratory (P(ImaxRV)) and expiratory (P(EmaxTLC)) pressure values (Sibelmed 163) were obtained by the same investigator, following national guidelines (SEPAR 1990).We studied 392 subjects (185 males, 207 females) whose ages ranged from 8-17 years. The reproducibility of measurements was investigated in a subgroup of 88 participants (randomly selected from the total sample, and stratified for age and gender) by means of the intraclass correlation coefficient (P(EmaxTLC), 0.98; P(ImaxRV), 0.95). P(EmaxTLC) and P(ImaxRV) values were significantly different between males and females (P < 0.0001) and were normally distributed. A stepwise, linear multiple regression model was built in each gender group (male/female) for the prediction of P(ImaxRV) and P(EmaxTLC) values. Independent variables (weight, height, and age) and their potential interactions were forced to enter the model in order to maximize the square of the multiple correlation coefficient of the resultant equation. This model turned out to be applicable (homoscedasticity, independence, and normality requirements) for P(ImaxRV) (in males and females) and for P(EmaxTLC) (in males but not in females). Variables included in the model were age and the product of weight and height. Their predictive power ranged between 0.21-0.51. In conclusion, P(ImaxRV) and P(EmaxTLC) values increase with age from 8 until 17 years. In all age groups, values were higher in males than in females. Weight, height, and age are included in the predictive equations for P(ImaxRV) (in males and females) and P(EmaxTLC) (in males). Their predictive value is similar to that reported by other authors and ranges between 0.21-0.51. This model is not suitable for the prediction of P(EmaxTLC) in females; the observed mean and range should be used instead. Copyright 2003 Wiley-Liss, Inc.