METHODS

Study population

Details of the VLBW cohort and the comparison group of children have been described previously.⁶ Briefly, of the 399 infants with a birth weight ≤1500 g born in the County of Merseyside during 1980–81, 219 survived to the age of 15 years. Ten children have either refused assessment, gone abroad, or could not be traced. Thirty four children with clinical disability, predominantly cerebral palsy and severe learning disability, were excluded because the disability would have affected the lung function tests. Forty seven children from twin pregnancies were also excluded as measures of IUGR would differ from the singletons. The remaining 128 children comprised the index cohort. When the index population was followed up at 8 years of age, a comparison group individually matched for age, sex, and school was obtained. Subsequently, both the index cohort and the comparison group were assessed for lung function and respiratory health at 15 years of age.

Lung function tests

The standard lung function test used to measure airway obstruction is the forced expiratory spirogram. This allows an assessment of the rate of change in volume that occurs as a function of time. The lung function tests were performed using a portable spirometer, Vitalograph-Alpha-II, with an attached printout. The output provided numerical as well as graphical data. The graphical data enabled the detection of any error in subject technique. The Vitalograph was calibrated daily before use.

All assessments were carried out by one examiner (CJS) either at school or at home. The examiner demonstrated to each subject the technique for performing the forced vital capacity (FVC). The subjects were given an opportunity to perform a few practice efforts. Later each subject was required to perform a minimum of three reproducible FVC measures (within 5% of maximal FVC). The spirogram that produced the highest sum of FVC and FEV₁ was used in the analysis.⁷ Forced expiratory volume in 1 second (FEV₁), the ratio of FEV₁/FVC, and forced expiratory flow when 25–75% (FEF_25–75%) of FVC was expired were used as markers of airway obstruction. The lung function indices are calculated using standard reference values as percentage of predicted for FVC, FEV₁, and FEF_25–75%.⁸ To avoid the potential pitfalls of correction for size, the FEV1/FVC ratio was also calculated as a method of size compensation.

Lung function was compared in those who did and did not receive respiratory support. Respiratory support was defined as any form of supplementary oxygen, continuous positive airway pressure, or intermittent positive pressure ventilation.

Physical measures and IUGR

Height was measured, using the Leicester height measure, to the nearest 0.5 cm; weight was determined to the nearest 0.5 kg with the child lightly clothed using SECA patient scales.

For the index cohort, the birth weight ratio (observed birth weight/expected birth weight for gestational age) was determined to assess the degree of IUGR. The expected birth weight for the gestation was based on Scottish national data.⁹

Questionnaires

The Medical Research Council’s questionnaire on respiratory symptoms was used to obtain information on respiratory morbidity.¹⁰ Details of respiratory management during the neonatal period were extracted from hospital neonatal notes.

Information on smoking habits was obtained by two questionnaires. One questionnaire completed by the mother provided information on her smoking during the pregnancy and on her current smoking habits. The other questionnaire, completed by the children, provided data on their current smoking habit. The children were classified as smokers if they had been smoking for six months or more.

Statistical methods

The data were analysed using the statistical package SPSS, PC version 10.07. Means were compared using the paired t test, and adjustment of the respiratory function tests for maternal smoking (current and during pregnancy) and teenage smoking was made by analysis of co-variance. Categorical variables were analysed by the χ² test with Yates’s correction for small 2×2 tables (that is, any expected frequency <5).

The effects of age, sex, and school (as a proxy for social class) were excluded by the matched study design.

The study was approved by the local ethics committee, and informed written consent was obtained from the parents of the children in the study.

RESULTS

Table 1 shows the characteristics of the index cohort and the comparison group. Mean gestational age of the comparison group is not known, but from their mean birth weight it is reasonable to assume that it is near to term. Only four children had a birth weight <2400 g, the lowest being 2098 g; six weighed 2400–2499 g. The remaining 118 all weighed ≥2500 g. Height and weight were significantly different between the index cohort and the comparison group.

Table 2 shows the lung function indices, after adjustment for the smoking habits of the mother and the teenager. FVC and FEV₁ in the index cohort and comparison group were not significantly different, either before or after adjusting for smoking. However, FEV₁/FVC and FEF_25–75% differed significantly, even after adjusting for smoking, indicating that there is a persisting medium and small airway obstruction among VLBW adolescents compared with the comparison group.

To determine whether this difference is associated with preterm delivery or IUGR, the birth weight ratio was calculated for the index cohort and plotted against the lung function parameters. The birth weight ratio ranged from 0.4 to 1.2, mean 0.8 (SD 0.2). No significant correlation was observed for any of the lung function parameters.

Among the 128 index cohort, 83 had received respiratory support. Table 3 shows no difference in lung function indices between those who received respiratory support compared with those who did not. The consequence of bronchopulmonary dysplasia (BPD) for subsequent respiratory function is of interest. However, oxygen dependency for more than 28 days as a marker of BPD occurred in only eight of our cases, which did not allow a meaningful examination of the problem.

Table 4 shows the prevalence of respiratory symptoms or diseases in the index cohort and comparison group. Cough, wheezing, and asthma were significantly more common among the index cohort compared with the comparison group. Although bronchitis and pneumonia were more prevalent among the index cohort, the difference was not significant.

DISCUSSION

The aim of the study was to determine whether being of low birth weight had an adverse effect on respiratory function measured in adolescence and, if so, whether this could be attributed to IUGR or preterm delivery. Very low birth weight was associated with medium and small rather than large airways obstruction, and this effect was mediated through preterm delivery rather than IUGR.

An advantage of this study is that it reports on respiratory function in adolescence of a cohort of VLBW infants and a matched comparison group from a geographically defined population. Furthermore it deals with respiratory function in adolescence, and not preadolescence, as in the majority of other studies. A similar study of adolescent respiratory function from a tertiary referral centre may have been biased because referral may have been because of severe respiratory problems.⁵

A significant reduction in both FEV₁/FVC ratio and FEF_25–75%, reflecting the medium and small airway obstruction in adolescents, was observed in the VLBW index group. Smoking is potentially an important variable when making comparisons of respiratory function as it is known to impair lung growth.¹¹ Our data on smoking were limited as they were not obtained with the intention of assessing its effect on respiratory function. However, adjusting for smoking as a binomial variable (smoking/non-smoking) had a negligible effect on the index and comparison group differences in FEV₁/FVC and FEF_25–75%. Similar medium and small airway obstruction was observed in a cohort of 14 years old children.⁵ However contrary to our observations, a reduction in FEV₁, a marker of upper airways obstruction, was also noted. Several studies of preadolescent subjects who were born preterm, reported a reduction in expiratory flow rates.^3,4,12 Though these studies were in preadolescent subjects and are not strictly comparable to our current study, it suggests that the reduction in expiratory flow noted in VLBW infants in childhood persists into adolescence.

Whether the effect on respiratory function is attributable to IUGR or preterm delivery is clinically important. It has been proposed that intrauterine factors that impair fetal growth may increase the risk of developing chronic obstructive airways disease in adulthood.¹³ These subjects were born at a time when methods for gestational assessment were imprecise, and it is uncertain whether they were growth retarded or premature. In a later cohort, among preterm infants, the birth weight for gestational age was significantly associated with FVC and FEV₁ but not with FEF_25–75% and FEF_75–85%.¹⁴ However, in the total sample of 2036, the proportion who were preterm (<37 weeks) was only 13%, of whom 79 (3.8% of the total sample) were <2500 g. In contrast, we found no correlation between birth weight ratio and any of the airflow measures, indicating that prematurity rather than IUGR is important. Another study of 300 VLBW infants compared those who were IUGR with those who were not and found no difference in airflow variables.¹⁵ Similarly, in a study of term infants, no association was found between low birth weight (as a measure of growth restriction) and lung function.¹⁶

Respiratory support in the neonatal period appeared not to be of importance in influencing lung function in adolescence. In contrast, a hospital based study found significant differences between those who received ventilatory support and those who did not.⁴ A possible explanation of these contradictory observations may result from the bias that can occur when hospital populations are assessed. A regional hospital centre is likely to include infants transferred from other neonatal units because of severe respiratory problems.

The association of long term respiratory morbidity with low birth weight is less certain. In our study the reduction in airflow indices was associated with a significant increase in the prevalence of chronic cough, wheezing, and asthma among the VLBW subjects. An increased prevalence of asthma and recurrent bronchitis was also found in a cohort of children at age 8 years.¹⁷ However, subsequent assessment of the cohort at 14 years found similar rates of asthma and pneumonia when compared with normal subjects.⁵ In another study the increased prevalence of asthma persisted into adolescence.¹⁸

In conclusion, the reduction in adolescent lung function indices, reflecting medium and small airway obstruction, is associated with preterm delivery rather than IUGR but is not associated with respiratory support in the neonatal period. The possible confounding effect of smoking could not be discounted entirely because of the limitation in the data on smoking. Compared with normal birth weight infants, VLBW infants have an increased prevalence of chronic cough, wheezing, and asthma in adolescence.