Predicting optimal continuous positive airway pressure in Japanese patients with obstructive sleep apnoea syndrome

Several algorithms that predict the optimal CPAP have been developed for Caucasian patients with OSA syndrome, but these algorithms do not allow for racial differences in craniofacial anatomy. We investigated whether an equation that included data on craniofacial structure, physique and severity of OSA could more accurately predict the optimal CPAP for Japanese patients with OSA syndrome.

In 170 Japanese patients with OSA syndrome, the optimal CPAP was determined by manual titration during polysomnography. An equation predicting the optimal pressure was derived from anthropometric, polysomnographic and cephalometric data. This equation was validated in another 110 Japanese patients with OSA syndrome.

Stepwise multiple regression analysis identified AHI, BMI, mean SaO(2) and a cephalometric parameter: the angle between a line from point B to the menton (Me) and a line from Me to the hyoid bone (H) (BMeH), as independent predictors of optimal CPAP. The following equation was constructed to predict the optimal CPAP: 27.78 + (0.041 x BMeH) + (0.141 x BMI) + (0.040 x AHI) - (0.312 x mean SaO(2)). This equation accounted for 47% of the variance in optimal pressure (R(2) = 0.47, P < 0.0001). The measured optimal pressure and the pressure calculated using this equation were very similar in the other 110 patients with OSA syndrome (9.5 +/- 3.0 and 9.2 +/- 2.1 cmH(2)O, respectively).

Optimal CPAP was more accurately predicted by combining a cephalometric parameter with BMI and polysomnographic data in Japanese patients with OSA, suggesting that craniofacial structure may be important in the pathogenesis of OSA syndrome among Asians.