Blood gas measurements were collected on healthy lifetime nonsmokers at sea
level (n = 96) and at an altitude of 1,400 meters (n = 243) to establish r
eference equations. At each study site, arterial blood samples were analyze
d in duplicate on two separate blood gas analyzers and CO-oximeters. Arteri
al blood gas variables included Pao(2), Paco(2) pH, and calculated alveolar
-arterial Po-2 difference (AaPo(2),). CO-oximeter variables were Hb, COHb,
MetHb, and Sao(2). Subjects were 18 to 81 yr of age with 166 male and 173 f
emale. Outlier data were excluded from multiple regression analysis, and re
ference equations were fitted to the data in two ways: (1) best fit using l
inear, squared, and cross-product terms; (2) simple equations, including on
ly the variables that explained at least 3% of the variance. Two sets of eq
uations were created: (1) using only the sea level data and (2) using the c
ombined data with barometric pressure as an independent variable. Compariso
ns with earlier studies revealed small but significant differences; the dec
line in Pao, with age at each altitude was consistent with most previous st
udies. At sea level, the equation that included barometric pressure predict
ed Pao(2) slightly better than the sea level specific equation. The inclusi
on of barometric pressure in the equations allows better prediction of bloo
d gas reference values at sea level and at altitudes as high as 1,400 meter
s.