Premium
Prediction of underwater residual lung volume in healthy men and women
Author(s) -
Cicone Zackary S.,
Nickerson Brett S.,
Esco Michael R.
Publication year - 2021
Publication title -
clinical physiology and functional imaging
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.608
H-Index - 67
eISSN - 1475-097X
pISSN - 1475-0961
DOI - 10.1111/cpf.12719
Subject(s) - hydrostatic weighing , underwater , medicine , limits of agreement , cohort , statistics , regression analysis , residual , nuclear medicine , mathematics , body weight , algorithm , oceanography , geology
Regression equations are commonly used to predict residual lung volume (RV) during underwater weighing when measurement is not practical. However, the equations currently available were derived from on‐land measures of RV and may account for changes in lung capacity during submersion, thus leading to inaccuracies in assessment of percent body fat (%BF). The purpose of this study was to (1) develop a new equation (RV NEW ) for the prediction of underwater RV, (2) cross‐validate RV NEW and compare it to existing RV equations, and (3) compare the effects of RV NEW and existing equations on underwater %BF. One‐hundred seventy‐five healthy adults were recruited to complete simultaneous hydrostatic weighing and RV measurements. The sample was randomly divided into development ( n = 131) and cross‐validation ( n = 44) cohorts. Regression analysis in the development cohort resulted in the following equation: underwater RV = −3·419 + 0·026 × height (cm) + 0·019 × age (y) ( p < 0·001; R 2 = 0·53; SEE = 0·26). In the cross‐validation cohort, Bland‐Altman analysis revealed that the new equation provided the best overall agreement with underwater RV (bias ± 1·96 SD , 0·07 ± 0·5 L), while existing equations produced significantly different values from measured RV and wider limits of agreement. When used to calculate %BF, the new RV equation produced the strongest agreement with underwater %BF (−0·5% ± 3·8%), although all equations produced strong correlations (all r > 0·95) and limits of agreement ≤4·7%. The results of this study suggest that RV NEW may be more appropriate for RV estimation during hydrostatic weighing than existing equations. However, its applicability to populations outside the current study needs to be examined.