Background: A non-invasive method, that used a general transfer function (TF), has been shown to accurately predict central aortic pressure from peripheral tonometry measurements in the adult population. However applying the same TF to estimate central aortic systolic pressure (aSP) in children has not yet been fully evaluated. This study aimed to assess the usage of adult TF to estimate aSP in children (aSP_{est_ch}) by establishing and testing different linear single/multivariate regression models between the adult TF estimated aSP (aSP_{TF_ad}) and the measured aSP (aSP_{meas_ch}).

Methods: 218 healthy, pre-pubescent children aged 8 years (113 male) had tonometer measured radial and carotid arterial pressure waveform recorded using the SphygmoCor device. Central aSP_{TF_ad} was estimated from the radial pressure using the TF (SphygmoCor, AtCor Medical), and the carotid systolic pressure taken as a surrogate for central pressure (aSP_{meas_ch}). The study group was divided into two groups: one to estimate the models (n=50, 19 male); another to test the models (n=168, 94 male). Models 1 and 2 were two simple linear regression models, whilst models 3 and 4 were two multivariate regression models.

Results: In the tested group, the aSP_{est_ch} from all models showed high correlations and low average differences with aSPmeas_ch (model 1 R²=0.88, difference=1.6+2.6 mmHg; model 2 R²=0. 88 difference=1.8+3.4 mmHg; model 3 R²=0.89 difference=−0.7+2.5 mmHg; model 4 R²=0.89 difference=−1.2+2.7 mmHg, all p<0.0001).

Conclusion: Central aSP in children can be estimated accurately using the adult TF from the radial pulse by incorporating the now defined linear relationship between aSP_{TF_ad} and aSP_{meas_ch}.