Data were analysed using the Chi²test to compare the prevalence of respiratory symptoms among both ex-preterms and those of the general population. If numbers were too small to use Chi²test, we made use of Fisher's exact test. We studied the independent effects of birth weight, gestational age, gender, duration of mechanical ventilation, smoking habits of the parents during the youth of the child and family history of atopic diseases and asthma on the presence of wheeze using multiple logistic regression analysis. Likewise we studied the independent effect of these potential risk factors on the presence of asthma, shortness of breath (with or without exercise), hay fever and eczema respectively. All statistical procedures were performed using SPSS 10.0. P-values less than 0.05 were considered to be significant (2 sided tests).

The results of the analyses of the ex-preterms (GA ≤ 32 weeks) are shown in Table 2. The prevalence of doctor-diagnosed asthma and shortness of breath during exercise was significantly higher in the preterm than in the general population, whereas eczema and hay fever were significantly lower. The premature born women reported more symptoms like wheeze than the controls. Women with a birth weight less than 1500 gram (GA >32 weeks) reported more often wheeze and shortness of breath but less allergy and eczema than the female controls. We found no such differences in males.

111 Children developed BPD (8.2%); 28 of them (25%) died. Boys (n = 72) were more prone to develop BPD than girls (n = 39). The response rate among BPD-patients was 69%. Since the number of BPD patients with a GA >32 weeks was only 3, we decided to analyse the results of the patients with a GA = 32 weeks (table 3). Compared to female controls, 19-year-old females with BPD reported a higher prevalence of doctor-diagnosed asthma, wheeze and shortness of breath during exercise. The BPD males reported significant less hay fever and eczema than the male controls.

In regression analyses dyspnea, asthma, wheeze, dyspnea on exertion, hay fever and eczema were assessed as outcome parameters. Dyspnea was significantly related to long term mechanical ventilation and BPD, maternal asthma and current smoking. An inverse relation was found with gestational age. Asthma was significantly related to maternal asthma. Wheeze was significantly related to female gender and current smoking habits and tended to be related to maternal smoking during the youth of the participant. Shortness of breath during exercise was related to female gender and smoking in the past. We found no significant associations of birth weight, gestational age, duration of mechanical ventilation, gender, smoking habits or BPD to hay fever and eczema (see table 4). Young adults with recurrent respiratory infections in infancy reported more asthmatic symptoms than those without respiratory infections (p < 0.001). No significant differences were found between recurrent respiratory infections and hayfever or eczema. Young adults with sepsis during the neonatal period reported less hayfever than those without sepsis (p = 0.03), but no significant differences were found between sepsis and asthma or eczema.

The pathophysiology of neonatal RDS is not completely understood, but it has been demonstrated that factors such as mechanical ventilation and oxygen lead to an inflammatory process, which could result in an early Th1- response. Moreover, in most reported studies the rates of re-hospitalisation of preterm and/or (extremely) low birth weight infants during the first two years of life, approach or exceed 50% 7 . Respiratory illnesses and especially respiratory infections are the most common indication for re-hospitalisation in this patient group. Also in our cohort the re-hospitalisation-rate in early childhood was high (34%) 14 . In contrast re-admission rates for normal birth weight infants are much lower (about 20%) 7 . Asthma is often characterised by symptoms like shortness of breath and wheeze; reversible airway obstruction; airway hyper-responsiveness and airway inflammation. In children and young adults, asthma is associated with atopy through IgE-dependent mechanisms, and airway-inflammation is partly related to helper T type 2 (Th2) lymphocytes and eosinophil mediation 20 . Preterm born adults report asthma-like symptoms, but less allergy compared to controls. Decreased risk of atopy is also found in a Finnish prospective birth cohort study comparing term and preterm adults: high gestational age increased the risk of atopy at the age of 31 21 . The early Th1- response, in combination with serious infections in the first two years of life, could be an explanation for the lower prevalence of atopy, which is in line with the hygiene-hypothesis 22 . Others found that children who were septic in the neonatal period were less likely to have asthma 23 . We could not confirm this. However, young adults who were septic during the neonatal period did report less hayfever. In our study, adults with recurrent respiratory infections in infancy did not report less but more asthma. This is remarkable considering that early exposure to endotoxins or other allergens enhance Th1-type cytokine responses tip the balance away from Th2-type responses that favour the development of allergic diseases including asthma 24 . The high rate of respiratory symptoms might be due to sustained increased vulnerability of the immature airways in a way that mimics asthma, but is not exactly the same.

Male gender is a risk factor for neonatal RDS and BPD 9 10 11 . Boys with neonatal RDS seem to have more health problems than girls during the neonatal period and school age 13 23 . However, long-term outcome shows gender differences in e.g. school-performances, but not in respiratory health. We found that particularly women reported symptoms as wheeze and shortness of breath. In the 'general' population both incidence and prevalence of wheeze and asthma is higher in males than in females until the age of 16 year 25 26 . In adulthood, asthma occurs more frequently among women 25 26 . The observed variation between males and females in the general population has partly been explained by dys-synnaptic lung growth: the independent growth of the airways in comparison with the lung parenchyma and air spaces. In girls, growth of the airways is proportional to growth of lung parenchyma, whereas in boys growth of the airways lags behind that of lung parenchyma, causing a discrepancy between airway and lung size 27 . Different pubertal patterns of thoracic growth between the sexes results in an approximately 25% higher lung function in males than in females of identical height at the end of puberty. We speculate that a similar process takes place in the preterm born population, although the underlying mechanism is not understood. Another explanation might be that large individual differences exist in physical symptom reports. Women may require a greater amount of cognitive analysis (and thus more attention) to make judgements about physical symptoms compared to men 28 .

There have been few reports of respiratory health during exercise. Our finding of a high percentage of participants that reported shortness of breath during exercise, is in agreement with a study showing low oxygen consumption in low birth weight children compared to children with a normal birth weight 29 . The authors suggested that extremely low birth weight children have a lower level of fitness than controls.

Airway obstruction and airway hyper-reactivity persisted in children and adolescents with BPD 3 4 30 . Long-term studies in children who had BPD as infants showed persisting lung function abnormalities consisting of airway obstruction, airway hyper-reactivity, and hyperinflation 5 31 . Both BPD and asthma are characterised by increased smooth muscle contraction and symptoms of both diseases are therefore perhaps difficult to distinguish. As stated above, airway inflammation is an important feature in children and adults with asthma. Studies showed that inflammation plays an important role in the pathogenesis of BPD. Contrary to asthma, however, the BAL-fluid reflects a Th1-cell subtype 32 33 . Even more than preterm infants without BPD, infants with BPD are likely to be re-hospitalised early in childhood with a respiratory illness 8 34 . The same mechanism as described above could be an explanation for the low prevalence of hay fever and eczema, despite the asthma-like symptoms.

The regression analysis confirmed the association between dyspnea and respectively long-term mechanical ventilation, BPD and smoking of the participant. We expected to find high risks for respiratory symptoms in the young adults with a (very) low birth weight or born (very) prematurely due to the immaturity of the airways at birth. However, the degree of prematurity or dysmaturity did not increase the risk at all. As a matter of fact, the risk for dyspnea was even lower in the children born very prematurely. In seeking to understand this we speculate that these young adults have a bias toward symptom detection and the feeling of distress because they are used to physical limitations. Future research should investigate the extend to which physical symptoms correlate with lung function abnormalities.

A limitation of our study might be that the age range of the preterms and the general population sample is not exactly the same. However, the prevalence of respiratory symptoms is probably increasing with age. Therefore, the differences might even be more obvious when the results of young adults born prematurely could be compared with peers from the general population. As the complete cohort was inhomogeneous in the sense that it consisted of either preterm or small for gestational age infants, we choose to analyse the data of the preterm children (GA ≤ 32 weeks). The possibility that symptoms will disappear and that ex-preterms will "grow out" of their disease after adolescence is likely to be very small because the lungs stop growing and developing after that age. It might even be possible that symptoms come back or become more severe during adulthood, as has been observed in long-term follow-up of asthma 35 .