Low socioeconomic status and lung function.

The study by TAYLOR-ROBINSON et al. [1] is unique because of the large number of subjects with cystic fibrosis, the careful follow-up and detailed statistical analysis upon which the authors base their conclusions. This remarkable study was performed in a population with reduced life expectancy. Evaluating the impact of socioeconomic status (SES) on lung function over many decades is particularly daunting in subjects with respiratory disease. Ever since the pioneering work of VILLERMÉ [2, 3], the founding father of social epidemiology, numerous studies have documented an association of SES with growth as well as with mortality in adulthood. It is also generally accepted that adverse conditions during childhood lead to stunted growth. For example, in a study of Indian children [4], stature was 4% smaller in the group with the lowest, compared with the highest, SES, and associated with up to 16.7% differences in forced expiratory volume in 1 s (FEV1) and forced vital capacity (FVC); adjusting for age and differences in height did not explain these differences. It is therefore attractive to conclude that the lower (“worse”, “poorer” or “compromised”) FEV1 and FVC represent damage associated with poor SES. Whereas this may well be the case, association is no proof of causality. The conclusion hinges entirely on the notion that stature is a good proxy for the dimensions of the chest cavity. However, poor SES or nutritional and health problems (lumped together as SES for short) during early childhood are associated with smaller leg length relative to the height of the upper body segment [5], which biases prediction equations of lung function based on height. There is no evidence that growth stunting due to poor SES is limited to height, so it is logical to assume that chest dimensions will also be affected. If poor SES leads to a 1%, 2% or 3% lower width, depth and height of the thoracic cage, this leads to a 3%, 6% and 9% lower volume of the chest cavity, respectively. Taking into account measurement error and individual variability, it will be difficult to assess such small differences, but they cannot be discarded and might (in part) explain the observed difference in growth of FEV1. Measurements of chest circumference in a large number of subjects might throw some light on the role of differences in body frame.