Table 2.

New spirometric prediction equations obtained from the study sample in comparison to the South African equations.

Outcome (Sex Specific)	South African (Black) Population	Mozambique (Local) Population
FVC (Males)	−3.08 − 0.024 × Age + 4.8 × Height; RSS = 0.54	−2.271 − 0.019 × Age + 3.989 × Height; RSS = 0.43; adj Rsquare = 0.61
FVC (Females)	−3.04 − 0.023 × Age + 4.5 × Height; RSS = 0.41	−2.761 − 0.019 × Age + 3.989 × Height; RSS = 0.43; adj Rsquare = 0.61
FEV1 (Males)	−0.54 − 0.027 × Age + 2.9 × Height; RSS = 0.46	−3.504 − 0.023 × Age + 4.426 × Height; RSS = 0.37; adj Rsquare = 0.65
FEV1 (Females)	−1.87 − 0.028 × Age + 3.4 × Height; RSS = 0.39	−0.170 − 0.023 × Age + 2.150 × Height; RSS = 0.37; adj Rsquare = 0.65
Ratio FEV1/FVC (Not sex specific)	-	0.921 − 0.0027 × Age; RSS = 0.06; adj Rsquare = 0.22

Legend: The regression estimates are smaller in magnitude for Mozambican compared to South African equations, however, the direction of association is the same. We modelled the ratio of forced expiratory volume in 1 s/forced vital capacity (FEV1/FVC). The adjusted, comparably low value for R square (=0.22) indicates that age is only explaining 22% of the observed variability in the ratio of FEV1 and FVC. For the individual outcomes for FEV1 and FVC, both values for R square were higher than 0.6 and hence more than 60% of the variation observed in FEV1 and FVC are explained by the covariates in the regression equation. The Global Lung Initiative (GLI) equations used in this article are based on generalized additive models for location scale and shape (GAMLSS) models and, hence, the regression estimates are not directly comparable and therefore not included in Table 2.