• H W Glindmeyer, J J Lefante, C McColloster, R N Jones, and H Weill.
• Department of Medicine, Tulane University School of Medicine, New Orleans, Louisiana 70112.
• Am. J. Respir. Crit. Care Med. 1995 Feb 1; 151 (2 Pt 1): 412422412-22.

AbstractNormative spirometric values were derived from 5,042 white (of mainly European ancestry) and black (of mainly African ancestry) men and women paper plant workers who are never-smokers, with no respiratory symptoms or diagnoses and no history of occupational exposure to fibrogenic dusts or irritant chemicals. This cohort was selected from a much larger population under long-term respiratory surveillance (n > 50,000 at 50 plants). Standardized equipment, procedures, and data reduction methods complied with ATS recommendations. Data were collected by the medical departments of the participating companies as part of their routine health surveillance, and the graphic and numeric test results were transmitted to the Tulane University Section of Environmental Medicine for centralized quality assurance, interpretation, and archiving. The large numbers allow derivation of gender- and race-specific reference values. Lower limits of normal were derived and depend upon residual variation and any changes in variation with age. The results indicate that polynomial regression equations provide a significantly better fit than linear regressions with breakpoints. In addition to being more biologically plausible, the polynomial model more closely matches observed longitudinal changes in lung function with age. The age range of the cohort, 18 to 65, provides a regression that more closely matches the observed values in this range, because it does not include "supernormal" elderly survivors, which can lessen the slope of the regression and artifactually increase the predicted values of 50 to 65 yr olds. The regression equations derived for black men and women do not support the use of a single race adjustment (0.85 or 0.88) for all age, sex, height, and spirometric test parameter combinations. These race- and gender-specific regression equations, with their respective lower limits of normal, should improve the detection and quantification of adverse health effects in working individuals and populations.

17 keywords...

hide…