Predictive equations for spirometric reference values in a healthy adult suburban population in Tanzania

This study was conducted in Kinondoni district, Dar es Salaam, Tanzania, with the objectives to generate prediction equations for forced expiratory volume in 1 second (FEV1), forced expiratory volume in 6 seconds (FEV6), FEV1/FEV6 ratio and peak expiratory flow (PEF) from a nonsymptomatic sample of the population, and to compare these equations to published reference values. The study included adults aged ≥15 years who were recruited by use of community based, multistage cluster random sampling. Participants performed spirometry and answered questionnaires regarding respiratory symptoms and socioeconomic conditions. Anthropometric data were collected. Selection of subjects for generation of reference values followed American Thoracic Society (ATS) recommendations. Data were analyzed using multiple regression techniques. Fifty two men and 98 women were selected to the reference value group. FEV1, FEV6, FEV1/FEV6 and PEF were regressed against age, height and weight. For men a curvilinear model was chosen when predicting FEV1, FEV6 and PEF, and a linear model predicted FEV1/FEV6. For women a linear model was used in the regression equations. The reference values generated from our study were lower than in several previously published studies. Our study suggests that assessment of respiratory function should be based on reference values generated from the same population as those being assessed.