In a study published inThorax in 1997 Stein and colleagues1 showed that birth weight was associated with adult lung function in an Indian population. We have carried out a similar analysis in a Chinese cohort of 59 men and 61 women born in Hong Kong in 1967 and followed up in 1997. This Hong Kong study has been described recently and has shown a significant inverse association between size at birth and adult blood pressure.2Spirometric tests were performed according to the American Thoracic Society's criteria to assure the quality.3 The same equipment (Sensor Medics 2200) was used in all subjects. Data were analysed by multiple linear regressions; adult height, smoking status, and sex were included as covariates.

The mean birth weight was 3.1 kg (range 2.3–4.1). Table 1 shows that babies of low birth weight (⩽2.5 kg) had lower mean values of forced expiratory volume in one second (FEV₁), forced vital capacity (FVC), and FEV₁/FVC ratio at the age of 30 years, adjusted for sex, adult height, and smoking. However, there was no significant trend (each p>0.1). It can be seen from the findings presented by the Indian study that the previously reported associations between birth weight and FEV₁ in men and between birth weight and FVC in both sexes were largely attributable to a lower lung function in subjects with a birth weight of less than 5 lb (2.27 kg; table 2 in Steinet al 1). In that study there was no obvious trend among the other subjects.

While we fully acknowledge the limitation of a relatively small sample, we believe that it is important to report statistically negative findings and to compare them with the previous Asian study. That our study has shown a statistically significant relation between size at birth and adult blood pressure also suggests that the numbers were sufficiently large to reveal a clinically important association. Having considered findings from our study and the previous one we suggest that, while low birth weight may be associated with a reduced lung function in adults, variation in birth weight among subjects with a normal birth weight did not appear to be relevant.