predicted values

Predicted values are the yardstick that turns a raw lung-function number into a meaningful one. Because a small healthy adult and a tall healthy athlete have very different normal lung sizes, every result is compared against what is expected for a person of the same age, sex, height and — historically — ethnic background.

These expected figures come from reference equations built by testing thousands of healthy non-smokers; the best modern sets, such as the Global Lung Function Initiative equations, span all ages and many populations. A result is then reported as a percent of predicted and, increasingly, as a z-score showing how many standard deviations it sits from the average.

The honest caveat is that no equation fits everyone, and the use of ethnicity-specific norms is being actively rethought, since it can bake historical assumptions and inequities into clinical thresholds. The modern lower limit of normal — the fifth percentile of healthy people — is preferred over a fixed percentage, because what counts as normal naturally narrows with age.