Table 2 Tabulated comparison of reference SaO2 linear model coefficients (top row of coefficients; adapted from reference3) with linear model coefficients obtained from equivalent fits to daytime mean SpO2 (middle row of coefficients) and overnight SpO2 (bottom row of coefficients) using our data.

From: Pulse oximetry values from 33,080 participants in the Apple Heart & Movement Study

Linear regression results for blood oxygen saturation: fit coefficients for reference model (MRef)

Model

R2

SEE

Constant

Age (years)

Baro. Press. (mm Hg)

Weight (kg)

Sex (m = 1, f = 0)

Height (cm)

Reference SaO2 (Crapo et al.3)

0.56

0.85

89.41

−0.0362

0.0128

−0.0159

n.s.

n.s.

Daytime SpO2

0.23

1.13

89.25 (88.68, 89.83) [p < 1.0e−10]

−0.0309 (−0.0322, −0.0296) [p < 1.0e−10]

0.0100 (0.0094, 0.0105) [p < 1.0e−10]

−0.0149 (−0.0157, −0.0140) [p < 1.0e−10]

−0.0168 (−0.0686, 0.0351) [p = 0.36]

0.0123 (0.0098, 0.0149) [p < 1.0e−10]

Nocturnal SpO2

0.32

1.21

87.12 (86.50, 87.74) [p < 1.0e−10]

−0.0431 (−0.0445, −0.0417) [p < 1.0e−10]

0.0140 (0.0134, 0.0146) [p < 1.0e−10]

−0.0182 (−0.0192, −0.0173) [p < 1.0e−10]

0.1629 (0.1071, 0.2187) [p < 1.0e−10]

0.0069 (0.0042, 0.0097) [p < 1.0e−10]

  1. Values in parentheses indicate 99.5% confidence intervals for fit coefficients. Values in brackets are regression coefficient p values, corresponding to two-sided t tests under the null hypothesis that the coefficient is equal to zero.
  2. R2 coefficient of determination, SEE standard error of the estimate, n.s. not significant.
Back to article page