Lung function at 16–19 years in males and females born very prematurely

To determine if there were differences in lung function at 16–19 years of age between males and females born very prematurely.


| INTRODUCTION
Male compared with female infants have structurally different lungs at birth. 1,2 Androgen receptors have been demonstrated throughout the fetal lung 1 and murine studies have shown that exposure of lung tissue to dihydrotestosterone changes the amount of terminal lung buds. 2 Amniotic fluid sampling in humans have highlighted that male fetuses reach surfactant maturity approximately 1.5 weeks after females. 3 Males have greater respiratory morbidity in infancy 4 and childhood. 5 Furthermore, at 11-14 years of age, mean airway function forced expiratory volume in the first one second (FEV 1 ) in males was poorer than in females 6 and males were more likely to have a clinically abnormal FEV 1 (z score below the 5th centile). 6 Studies in adults, however, have suggested that females may be more adversely affected by premature birth. Nineteen-year-old women born before 32 weeks of gestation with a birth weight below 1.5 kg, had a higher incidence of shortness of breath during exercise compared with their male peers. 7 In addition, females born at less than 35 weeks of gestation between 1925 and 1949 and studied between 1987 and 2006 had a higher lifetime incidence of chronic obstructive airways disease (COPD) and asthma. 8 During puberty, the last positive effector on lung function, 9 there is rapid growth of the parenchyma and airways. 10 This growth results in flow and volume differences between boys and girls, particularly in late adolescence 11 and may explain the superior outcomes of male compared with female adults born prematurely. 7,8 Both studies, however, were undertaken in populations that were not routinely exposed to antenatal corticosteroids or postnatal surfactant. Chronic respiratory problems such as COPD can be predicted by poor lung function postpuberty. 12 It is, therefore, important to determine if the differences in lung function between males and females routinely exposed to antenatal corticosteroids and postnatal surfactant persist postpuberty and are associated with reduced exercise capacity and the aim of this study.

Young people who had been recruited into the United Kingdom
Oscillation Study (UKOS) 13 [15][16][17][18][19][20] Spirometry, plethysmography, impulse oscillometry, DLCO, and FRC by helium dilution were measured using Vyair lung function equipment. Lung clearance index was calculated using the SF6 (Innocor). Lung function results were converted to z scores using global lung initiative (GLI) reference equations or appropriate reference equations for those lung measurements not included in the GLI reference equations. [19][20][21][22][23] Abnormal lung function was defined as lung function below the 5th centile for normal as reported previously. 24 A shuttle sprint 25 test was used to assess exercise capacity in meters.
Participants completed a questionnaire regarding their respiratory symptoms, a current diagnosis of asthma, and exercise undertaken on a weekly basis. Smoking status of the participants was set at "yes" if either they were a self-reported smoker or had a salivary cotinine level greater than 15 ng/mL. Puberty was previously assessed when the participants were 11-14 years of age. 26 More than 90% of children were found to be of Tanner stage 2. 27,28

| Analysis
FEF 75 was the primary outcome of this study to maintain consistency as the primary outcome of other studies of this cohort. 14,26 Demographic factors, lung function, exercise and respiratory symptoms were compared by sex with analyses adjusted for the nonindependence of multiple births using mixed effects multiple linear regression where the participant was the random effect (continuous outcomes) 29 or using logistic regression with robust standard errors (categorical outcomes). 30 The effects of sex on lung function used the same models as described above with adjustment for the following confounders: antenatal steroids (yes/no), birthweight, gestational age at birth, oxygen dependency at 36 weeks corrected (yes/no), administration of postnatal corticosteroids (dexamethasone, yes/no), maternal smoking in pregnancy (yes/no), and the participant's age at assessment.

| RESULTS
One hundred and fifty UKOS participants attended for lung function assessment at a mean age of 18 years. The mean height and weight were greater in the males than females. There were no other significant differences in the demographics by sex although the administration of postnatal corticosteroids in the neonatal period was more common among males than females ( Table 1). The participants who were assessed at age 16-19 years had higher mean birthweight and gestational age and their mothers were more likely to be White and not to have smoked in pregnancy. They were also more likely to have had a major cranial ultrasound abnormality and a pulmonary hemorrhage (Supporting Information: E- Table 1).
Males had poorer mean FEF 75 , FEF 50 , FEF 25-75 FEV 1 :FVC ratio, DLCO, and D L CO/VA compared with females and the differences remained significant after adjusting for neonatal factors and age at assessment ( Table 2). These differences in mean z scores ranged from 0.19 to 0.71 and translated into substantial differences in the percentage of participants with lung function below the 5th centile (below the limit of normal). For example, for FEF 75 , 39.6% males and 20.1% females had lung function that was below the limit of normal with an adjusted difference of 19.5 percentage points (Table 3).
Males were able to complete a significantly greater distance during the shuttle sprint test and reported doing more exercise each week (Table 4). Sixteen percent of participants reported wheeze in the past 12 months and around 9% reported asthma but neither varied significantly by sex (Table 4).
A post hoc analysis of lung function was conducted including an interaction term for sex × exercise in the model to attempt to explore the finding that boys had poorer mean lung function but had greater mean exercise capacity. This showed that for most lung function measures, the interaction term was not significant (Supporting Information: E- Table 2), but significant interactions were observed for FRC He and resistance at 20 Hz.

| DISCUSSION
We have demonstrated that at 16-19 years of age, males compared with females had a significantly lower mean FEF 75 , indicating they had poorer small airway function. This difference is equivalent to a difference of 20 percentage points in the proportion with lung function below the limit of normal. This difference remained after adjusting for neonatal factors including postnatal corticosteroid exposure and age at assessment. Similar differences were seen for many of the lung function measurements including obstruction (FEV 1 /FVC) and diffusion capacity of the lungs for carbon dioxide.
There were no significant differences seen with regard to restrictive lung disease. The poorer lung function in the males, however, was not associated with poorer exercise capacity, indeed the males had better exercise performance.
Physical fitness within the general population is due to a combination of factors such as muscle strength 33 and cardiovascular fitness. 34 The complex relationships between individual factors and exercise performance have rarely been studied in the context of investigating differences between males and females. 35 In studies that have been undertaken, it has been shown that muscle mass relative to lean mass was higher in males compared with females and this correlated with improved exercise performance. 33 Higher birthweight has been associated with greater hand grip strength in adolescents especially in females, these associations may be explained by fat-free mass. 36 In addition, improved cardiovascular markers such as resting heart rate and blood pressure were also positively associated with improved exercise capacity as measured by VO 2 max. 34 In one study, VO 2 max was 10% higher in elite male athletes compared with elite female athletes. 37 These factors may have benefited male participants and overcome any deficiencies in lung function in males compared with females. It is important to note, however, that it has been shown that those with the poorest lung function have impaired exercise capacity suggesting that, while it is not a linear relationship, lung function is an important contributor to exercise capacity. 24 Airway hyper-responsiveness has been shown to be higher in those born prematurely. 38 There were no significant differences seen between males and females in this cohort with regard to reported respiratory symptoms. This was despite more males having obstructive lung disease defined as an FEV 1 /FVC ratio below the 5% T A B L E 1 Demographics by sex. 2). These differences may be due to differences in populations as in that study where infants were born at a later gestational age and were not routinely exposed to antenatal corticosteroids or postnatal surfactant. In our study, only 15% of the study population had wheeze and 9% had asthma whereas in the previous study 30% had wheeze and 13% reported having asthma. Similar to our results, that study found no significant relationship of sex and a diagnosis of asthma.
Our results do not explain the findings of one study where females born prematurely were more likely to develop COPD in adulthood. 8 It has been shown that lung function may improve in both boys and girls born prematurely 40 yet, while it is recognized that lung function declines throughout adulthood, 21 few data exist to describe the decline in those born prematurely and routinely exposed    25 Furthermore, using this test in a rehabilitation program, a difference of 48 m correlated with symptomatic benefit and a further benefit was seen with a difference of 79 m. 41 The mean difference between males and females in this study was 227 m and therefore clinically significant and higher than the 10% difference in VO 2 max observed within a previous population. 37 We do acknowledge that this is a single measure of physical fitness and more comprehensive testing may be of benefit in the future. 42 Puberty was not assessed at this follow-up, however, as the average length of puberty is between 2 and 5 years 27,28 and almost all participants had entered puberty at an earlier assessment, we are confident that we are presenting postpubertal lung function. In general, the interaction results were inconclusive, probably due to the relatively low power and hence we will be further investigated in a proposed larger study.
In conclusion, among young people born prematurely, males had poorer lung function compared with females, but this was not accompanied by reduced exercise capacity or increased respiratory symptoms at that time. Whether this difference persists into young adulthood needs assessing.