Search tips
Search criteria 


Logo of archdischArchives of Disease in ChildhoodVisit this articleSubmit a manuscriptReceive email alertsContact usBMJ
Arch Dis Child. 2007 September; 92(9): 776–780.
PMCID: PMC2084021

Very prematurely born infants wheezing at follow‐up: lung function and risk factors



To determine whether abnormalities of lung volume and/or airway function were associated with wheeze at follow‐up in infants born very prematurely and to identify risk factors for wheeze.


Lung function data obtained at 1 year of age were collated from two cohorts of infants recruited into the UKOS and an RSV study, respectively.


Infant pulmonary function laboratory.


111 infants (mean gestational age 26.3 (SD 1.6) weeks).


Lung function measurements at 1 year of age corrected for gestational age at birth. Diary cards and respiratory questionnaires were completed to document wheeze.

Main outcome measures

Functional residual capacity (FRCpleth and FRCHe), airways resistance (Raw), FRCHe:FRCpleth and tidal breathing parameters (TPTEF:TE).


The 60 infants who wheezed at follow‐up had significantly lower mean FRCHe, FRCHe:FRCpleth and TPTEF:TE, but higher mean Raw than the 51 without wheeze. Regression analysis demonstrated that gestational age, length at assessment, family history of atopy and a low FRCHe:FRCpleth were significantly associated with wheeze.


Wheeze at follow‐up in very prematurely born infants is associated with gas trapping, suggesting abnormalities of the small airways.

Keywords: lung function, prematurity, wheeze

In the past, prematurely born children were reported to suffer frequent respiratory symptoms at follow‐up.1,2,3 Such infants, particularly if they went on to develop bronchopulmonary dysplasia (BPD), had often been exposed to high peak inflating pressures and inspired oxygen concentrations and suffered severe lung injury. Nowadays, very prematurely born infants (that is, those born at <29 weeks of gestational age) may have mild or even no respiratory distress in the immediate neonatal period.4 Nevertheless, they are frequently wheezy at follow‐up.5 The limited post‐mortem data available from very prematurely born infants demonstrate that even those who had had BPD had minimal lung inflammation and fibrosis, but had evidence of arrest in alveolar development.6,7 Therefore, it cannot be assumed that such infants will have abnormalities of airway function at follow‐up. Unfortunately, there is only one study8 exclusively reporting lung function at follow‐up of very prematurely born infants (<29 weeks of gestational age) and as the infants were born approximately 15 years ago, they were exposed to different perinatal care practices. The aim, therefore, of this study was to characterise the lung function abnormalities of very prematurely born infants with wheeze at follow‐up, as this will inform their optimum management. In addition, we wished to determine risk factors for wheeze at follow‐up.


Data from infants who had been recruited into either a randomised trial of ventilation strategies (UKOS)9 or a prospective follow‐up study examining the impact of respiratory syncytial virus (RSV) infection in prematurely born infants (RSV study)10 and had lung function assessment at 1 year of age corrected for prematurity were analysed in this study. We combined the two datasets so that we were able to report lung function results from a large number (n = 111) of very prematurely born infants. All the infants entered into UKOS9 and assessed at 1 year corrected11 were born prior to 29 weeks of gestation. Only data from infants born prior to 29 weeks of gestation entered into the RSV study10 were analysed in this study.

The infants from both studies were examined in the Amanda Smith infant pulmonary function laboratory at King's College Hospital, London using the same equipment and techniques. No infant was tested within at least 2 weeks of a respiratory tract infection. Length was measured using a stadiometer (R511A Rollametre cm.100; Raven Equipment, Dunmow, Essex, UK) and weight using electronic scales (MPBT35 digital baby/toddler scale; Marsden Weighing Machine Group, Henley‐on‐Thames, Oxfordshire, UK) prior to lung function measurements. The Institute of Child Health Growth charts were used.12 The infants were sedated with 80–120 mg/kg of chloral hydrate and monitored by pulse oximetry (Datex‐Ohmeda 3800; Datex‐Ohmeda, Hatfield, UK) throughout the pulmonary function testing and afterwards until they were awake. Once asleep, the infant was laid supine in the plethysmograph (Department of Medical Engineering, Hammersmith Hospital, London, UK), which had a total volume of 90 litres and included a heated, humidified rebreathing system. The infant breathed through an appropriately sized Rendell‐Baker face mask, sealed around the nose and mouth with silicone putty. The mask deadspace was calculated as 50% of the water‐displacement measurement with the silicone putty in situ.13 Pressure at the airway opening (Pao) was measured using a differential pressure transducer (range ±5 kPa; MP45, Validyne Engineering, Northridge, CA, USA) connected to a port in the mask support. The mask support also incorporated a thermistor measuring airway temperature and was connected to a heated pneumotachograph (Fleisch, Lausanne, Switzerland) to measure airflow. The pneumotachograph was attached to a differential pressure transducer (range ±0.2 kPa; MP45, Validyne). The total deadspace of the breathing apparatus, including the mask and all tubing connecting the transducers, measured by water displacement,13 was 31 ml. Pressure changes within the plethysmograph were measured using a differential pressure transducer (range ±0.2 kPa; MP45, Validyne). All signals were amplified (CD18 carrier amplifiers; Validyne) and the flow signal integrated electrically to give tidal volume (FV156 integrator; Validyne). The resultant four channels of data were acquired, analysed and displayed in real time on a personal computer (Gateway GP7‐500; Gateway, Dublin, Ireland) running a computer program custom‐designed using LabWindows software (National Instruments, Austin, TX, USA) with analogue‐to‐digital sampling at 200 Hz (PC‐LPM‐16PnP; National Instruments). All channels were calibrated prior to each patient test, as previously described.11

Following application of the face mask, time was allowed for the infant to settle and tidal breathing was then recorded for 1 min, from which a minimum of 20 breaths were recorded for analysis of tidal breathing, including calculation of the time taken to achieve peak expiratory flow, expressed as a proportion of expiratory time (TPTEF:TE), and respiratory rate. At least 20 acceptable breaths were analysed (median 28, range 24–36). Plethysmographic estimation of lung volume (plethysmographic functional residual capacity, FRCpleth) was then undertaken. FRCpleth was calculated from a minimum of three end‐inspiratory occlusions. A time‐based trace of all four data channels and an x/y plot of Vpleth/Pao during each occlusion were displayed by the computer. Occlusions were considered acceptable if Vpleth and Pao were in phase and no airflow was evident. The infant was then switched to the rebreathing bag. Individual breaths acquired during periods of rebreathing were displayed as x/y plots of Vpleth/flow by the computer. Only technically acceptable breaths, that is, the loop was closed or nearly closed at points of zero flow, were used in the analysis. Airways resistance (Raw) was calculated electronically using an established formula by applying a regression line to the selected portion of the loop. Raw was calculated during initial inspiration between 0 and 50% maximal inspiratory flow, and during expiration between 0 and 50% maximal expiratory flow.14 During all Raw measurements, the computer calculated the apparatus resistance of the selected portion of the individual breath by relating ΔPao to Δflow and then subtracted this value from the total measured resistance.

On completion of the plethysmographic measurements, lung volume was estimated by a helium gas dilution technique (FRCHe). FRCHe was measured whilst the infant lay undisturbed on the base of the plethysmograph, using the same mask with silicone putty. During the initial stages of the study, FRCHe was determined using a water‐sealed spirometer (Pulmonet III; Gould, Bilthoven, The Netherlands), as described previously.11 The deadspace of the breathing apparatus, including the mask and filter, was 66 ml. Most infants in the UKOS respiratory follow‐up study11 and all in the RSV study10 were tested using the EBS 2615 system (Equilibrated Bio Systems, New York, USA), which consisted of a 500‐ml rebreathing bag in a closed heliox circuit. The deadspace of this system, including mask and filter, was 22 ml. The system was modified to produce a time‐based display of flow and tidal volume, allowing accurate switching into the circuit at end expiration. An online display of the helium dilution curve allowed precise determination of gas equilibration. For both FRCHe techniques, the mean of two recordings that were within 10% of each other was taken. The FRCHe of 12 infants was measured using both devices in order to assess comparability. The agreement between the two devices was assessed using the Bland‐Altman method.15 The mean (SD) difference in measured FRCHe was 3.4 (12.5) ml (1.76% of the mean measured FRCHe) with 95% limits of agreement of 21.2 to 27.9 ml.11 The FRCHe results from the EBS 2615 from these 12 infants were included in the study.

Diary cards were completed by the parents when the infants were 11 months of age corrected for gestational age at birth. The parents recorded on a daily basis for a 2‐week period in the UKOS study and for a 4‐week period in the RSV study; for the purpose of this analysis only the first 2 weeks of the RSV diary cards were analysed. Questionnaires also were administered to the parents at follow‐up to further document wheeze, family history of atopy (parents or siblings with asthma, eczema or hay‐fever) and smoke exposure at home. Data were obtained from the medical records regarding maternal smoking in pregnancy, gestation and weight at birth, from which birth centile was calculated. Small for gestational age (SGA) was diagnosed if the infant was less than the tenth percentile in weight for gestational age. In addition, the duration of ventilation and whether the infants had had BPD (oxygen dependency beyond 36 weeks post‐menstrual age (PMA)) were determined.


The infants were divided into two groups according to whether they wheezed on at least 1 day in a 2‐week period (see above) (wheeze group) or did not wheeze during that period (no wheeze group). Comparisons were made using t tests (continuous data) and χ2 tests (categorical data). The two data sets were combined for analysis by wheeze status and allowance was made for the two different source data sets in the modelling. Logistic regression was used to model factors associated with wheeze at age 1 year in two stages: (i) using antenatal and postnatal factors that were significant in unifactorial analyses and (ii) using factors measured at age 1 year that were significant in unifactorial analyses. Backward elimination was used to remove non‐significant factors at each stage. Variables which remained significant in each of stages (i) and (ii) were then modelled together to give a final model. p<0.05 was used to define statistical significance. Results are presented as odds ratios (OR) and 95% confidence intervals (95% CI).


All 76 infants from UKOS who had pulmonary function studies at 1 year of corrected age10 and 35 infants from the RSV study11 were included in this study (table 11).). UKOS was approved by the South Thames Multicentre Research Ethics Committee and the King's College Hospital Research Ethics Committee. The RSV study was approved by the King's College Hospital and Guy's and St Thomas' Hospital Research Ethics Committees. Parents gave informed written consent for their infant to take part in the studies.

Table thumbnail
Table 1 Demographics of the study populations


Comparison of the demographics of the two study groups revealed that the only significant differences between them was that the infants in the UKOS study were shorter (p = 0.03) and lighter (p = 0.001) at 12 months corrected age than those in the RSV study (table 11).

Sixty infants wheezed at follow‐up (median number of days of wheeze was 4 (range 1–14)) (fig 11).). The infants who wheezed at follow‐up compared to those who did not were of lower mean birth weight (p = 0.05), more immature at birth (p<0.001), had required a longer duration of ventilation (p = 0.003), more had had BPD (p = 0.003) and more had a family history of atopy (p = 0.004) (table 22).). The infants who wheezed were also shorter at follow‐up (p = 0.001) and had a lower mean FRCHe (ml) (p<0.02), mean FRCHe:FRCpleth (p = 0.001) and mean TPTEF:TE (p = 0.04) but a higher mean Raw (p = 0.04) (table 33).

Table thumbnail
Table 2 Antenatal and postnatal factors according to wheeze status at follow‐up
figure ac112623.f1
Figure 1 Airway resistance results (individual results are displayed according to wheeze status).
Table thumbnail
Table 3 Lung function at follow‐up according to wheeze status

Regression analysis demonstrated that the factors significantly related to wheeze were gestational age (p = 0.005), family history of atopy (p = 0.009), length at assessment (p = 0.016) and FRCHe:FRCpleth (p = 0.005) (table 44).). Days of wheeze were significantly correlated to FRCHe:FRCpleth (p = 0.015).

Table thumbnail
Table 4 Factors related to wheeze according to logistic regression analysis


We have demonstrated that very prematurely born infants who wheeze at follow‐up have greater evidence of gas trapping than infants without wheeze, as they have a significantly lower FRCHe:FRCpleth. These data suggest the very prematurely born infants who wheeze have abnormalities of the small airways.

Infants were entered into the wheeze group if they wheezed on at least 1 day in a 2‐week period of diary card completion. Using such a definition, it is possible that we included some infants who had very mild symptoms, yet there were significant differences between the groups. In addition, we highlight a significant correlation between days of wheeze and FRCHe:FRCpleth in these very prematurely born infants.

There have been few reports of lung function at follow‐up of infants born very prematurely. Doyle et al8 demonstrated that children born prior to 28 weeks of gestation and/or with a birth weight of <1000 g, when studied at 8–9 years of age, had diminished respiratory function variables reflecting air flow abnormalities. Their cohort, however, received only rescue surfactant and maternal smoking data were not reported. Others reported findings consistent with signs of dysfunction of the terminal respiratory units, that is, impaired gas mixing.16 That study, however, included more mature infants, the gestational age range of the study population being 25–33 weeks.

On univariate analysis, the infants who wheezed had a significantly higher mean Raw and a lower mean TPTEF:TE, suggesting they had large airways abnormalities. Raw, however, is significantly related to length and as the infants who wheezed were significantly shorter at assessment, this may explain the relationship with wheeze. Indeed, regression analysis demonstrated length at assessment was significantly associated with wheeze, whereas Raw and TPTEF:TE were not. FRCHe (ml) was significantly lower in the wheeze group. As FRCpleth did not differ significantly between the two groups, it is likely that the lower FRCHe in the wheeze group indicated the presence of gas trapping. BPD was significantly more common in the wheeze group. Decreased alveolarisation is seen at post‐mortem in infants with “new” BPD6,7 and in ventilator‐ and oxygen‐exposed baboons.17 Those data would suggest that, as the wheeze group had a significantly greater proportion of BPD infants, they would have had lower lung volumes. Indeed, we have previously demonstrated18 that at 33–39 weeks PMA, infants with mild to moderate BPD had lower lung volumes than those without BPD. In the present study population, however, we saw only significantly low FRCHe, but not FRCpleth, in the wheeze group. They were studied at 1 year of corrected age, so it is possible that abnormalities in lung volume may disappear with increasing postnatal age. Serial measurements are required to test such a hypothesis.

Our FRCpleth results are higher than those reported by Hulskamp et al19 who used a Jaeger plethysmograph. They also documented plethysmographically determined FRC results in infants and children from a series of studies and interpreted the results, except from one report,20 as demonstrating a progressive decline in plethysmographic lung volumes in infants. They hypothesised that the diminished values of plethysmographic lung volumes in healthy infants primarily reflected gradual technological advances and refinements in protocols over the last 30 years rather than any physiological changes in lung growth and development in early life or differences in background characteristics of the tested infants.19 The largest study (n = 264) included in their comparative data, however, reported a mean plethysmographic lung volume of 29.7 ml/kg20 and we recently reported that very prematurely born infants examined using a Hammersmith plethysmograph at a corrected age of 1 year had a mean FRCpleth of 26.7 ml/kg.11 In an in vitro study, we compared results given by the Hammersmith and Jaeger plethysmographs to a lung model and found that the Hammersmith, but not the Jaeger, plethysmograph gave results which were similar to those of the lung model.21 The significantly lower mean FRCpleth results obtained using the Jaeger, compared to the Hammersmith, plethysmograph we therefore interpret as representing under‐recording by the Jaeger plethysmograph.21

There are limitations to our study. In particular, the infants were not recruited a priori to determine lung function abnormalities in very prematurely born infants who were wheezy at follow‐up, but rather into two separate studies. The demographics of the two study populations, however, were very similar and only differed significantly with regard to length and weight at follow‐up (table 11).). Also, as the majority of infants were studied using the same equipment and all using the same plethymograph, we feel our results are valid. The analysis undertaken specifically looked to see if differences in lung function between wheezy and non‐wheezy infants differed by study, demonstrated no significant effect by study. A further possible limitation was that we did not have information for infants entered into UKOS regarding whether they had had an RSV infection. Children born at term22 or prematurely11 have an excess of wheeze at follow‐up following RSV infection and RSV infection has been associated with lung function abnormalities in infants born at term.23,24,25,26 The abnormalities reported, however, are hyperinflation,23,24,25,26 elevated resistance23 and reduced TPTEF:TE.24 Regression analysis demonstrated that neither Raw nor TPTEF;TE was significantly associated with wheeze in the very prematurely born infants we assessed. We would also emphasise that our study was not designed to identify the mechanism of wheeze at follow‐up but rather to determine the lung function abnormalities associated with wheeze in very prematurely born infants, as this could have implications for treatment.

What is already known on this topic

  • There are no data documenting the lung function of very prematurely born infants wheezy at follow‐up.
  • Very prematurely born infants with BPD are frequently wheezy at follow‐up; limited post‐mortem data suggest that they have minimal lung inflammation and fibrosis but have evidence of arrest of alveolar development.

What this study adds

  • We have demonstrated, in infants born prior to 29 weeks of gestational age, that gestational age, length at assessment, family history of atopy and a low functional to total lung volume were significantly associated with wheeze at follow‐up.
  • These data highlight that wheeze at follow‐up in very prematurely born infants is associated with gas trapping, suggesting abnormalities of the small airways.

In this population, wheezing at follow‐up was not associated with intrauterine growth retardation or antenatal smoking. Previous studies have demonstrated that wheezing is increased in children who were SGA.27 The present population, however, was born at a very early gestation and it is possible that growth retardation may have less impact on the lung function of such infants. In infants born at term,28 exposure to antenatal smoking has been significantly associated with wheeze and abnormal lung function. The incidence of maternal smoking in that study,28 however, was 44%, whereas only 11% of those infants presently studied were exposed in utero to maternal smoking. Although univariate analysis demonstrated that the infants who wheezed differed significantly from those who did not wheeze with regard to their gestational age, duration of ventilation, BPD status and a family history of atopy, only gestational age and a family history of atopy remained significant factors for wheeze following regression analysis. Duration of ventilation and BPD status are highly negatively correlated with gestational age at birth.10 Their elimination from the multifactorial model suggests that the adverse effect on respiratory health is primarily due to low gestational age rather than its consequences.

In conclusion, we have demonstrated that very prematurely born infants who have wheeze at follow‐up have evidence of gas trapping, suggesting they have abnormalities of small airway function. As it is not certain that such infants would respond to bronchodilator therapy, appropriately designed studies are required to determine whether bronchodilator and/or prophylactic anti‐asthma therapy have benefits in wheezy, very prematurely born infants.


We thank Mrs Deirdre Gibbons for secretarial assistance.


95% CI - 95% confidence intervals

BPD - bronchopulmonary dysplasia

FRC - functional residual capacity

OR - odds ratio

PMA - post‐menstrual age

RSV - respiratory syncytial virus

SGA - small for gestational age


Dr Mark Thomas and Mrs Louise Marston were supported by a grant from the MRC who supported the UKOS trial. Dr Broughton was supported by a peer reviewed grant from the WellChild Trust and Abbott Laboratories.

Competing interests: None.


1. Rona R J, Gulliford M C, Chinn S. Effects of prematurity and intrauterine growth on respiratory health and lung function in childhood. BMJ 1993. 306817–820.820 [PMC free article] [PubMed]
2. Chan K N, Elliman A, Bryan E. et al Respiratory symptoms in children of low birthweight. Arch Dis Child 1989. 641294–1304.1304 [PMC free article] [PubMed]
3. McLeod A, Ross P, Mitchell S. et al Respiratory health in a total very low birthweight cohort and their classroom controls. Arch Dis Child 1996. 74188–194.194 [PMC free article] [PubMed]
4. Bancalari E, Claure N, Sosenko I R. Bronchopulmonary dysplasia: changes in pathogenesis, epidemiology and definition. Semin Neonatol 2003. 863–71.71 [PubMed]
5. Greenough A, Limb E, Marston L. et al Risk factors for respiratory morbidity in infancy after very premature birth. Arch Dis Child Fetal Neonatal Ed 2005. 90F320–F323.F323 [PMC free article] [PubMed]
6. Erickson A M, delaMonte S M, Moore G W. et al The prognosis of morphological changes in bronchopulmonary dysplasia. Am J Pathol 1987. 127474–478.478 [PubMed]
7. Hussain A N, Siddiqui N H, Stocker J T. Pathology of arrested acinar development in post surfactant bronchopulmonary dysplasia. Hum Pathol 1998. 29710–717.717 [PubMed]
8. Doyle L W. et al Respiratory function at 8–9 years in extremely low birthweight/very preterm children born in Victoria in 1991–1992. Pediatr Pulmonol 2006. 41570–576.576 [PubMed]
9. Johnson A H, Peacock J L, Greenough A. et al United Kingdom Oscillation Study Group. High frequency oscillatory ventilation for the prevention of chronic lung disease of prematurity. N Engl J Med 2002. 347633–642.642 [PubMed]
10. Broughton S, Roberts A, Fox G. et al Prospective study of healthcare utilisation and respiratory morbidity due to RSV infection in prematurely born infants. Thorax 2005. 601039–1044.1044 [PMC free article] [PubMed]
11. Thomas M R, Rafferty G F, Limb E S. et al Pulmonary function at follow‐up of very preterm infants from the United Kingdom Oscillation Study. Am J Respir Crit Care Med 2004. 169868–872.872 [PubMed]
12. Cole T J, Freeman J V, Preece M A. British 1990 growth reference centiles for weight, height, body mass index and head circumference fitted by maximum penalised likelihood. Stat Med 1998. 17407–409.409 [PubMed]
13. Frey U, Stocks J, Coates A. et al Specifications for equipment used for infant pulmonary function testing. ERS/ATS Task Force on Standards for Infant Respiratory Function Testing. Eur Respir J 2000. 16(4)731–740.740 [PubMed]
14. Thomas M, Greenough A, Blowes R. et al Airway resistance estimation by best fit analysis in very premature infants. Physiol Meas 2002. 23279–285.285 [PubMed]
15. Bland J M, Altman D G. Statistical methods for assessing agreement between two methods of clinical measurement. Lancet 1986. 1307–310.310 [PubMed]
16. Hjalmarson O, Sandberg K. Abnormal lung function in healthy preterm infants. Am J Respir Crit Care Med 2002. 16583–87.87 [PubMed]
17. Coalson J J, Winter V, de Lemos R R. Decreased alveolarisation in baboon survivors with bronchopulmonary dysplasia. Am J Respir Crit Care Med 1995. 152640–646.646 [PubMed]
18. Greenough A, Dimitriou G, Bhat R Y. et al Lung volumes in infants who had mild to moderate bronchopulmonary dysplasia. Eur J Pediatr 2005. 164583–586.586 [PubMed]
19. Hulskamp G, Hoo A F, Ljungberg H. et al Progressive decline in plethysmographic lung volumes in infants: physiology or technology? Am J Respir Crit Care Med 2003. 1681003–1009.1009 [PubMed]
20. Milner A D, Marsh M J, Ingram D M. et al Effects of smoking in pregnancy on neonatal lung function. Arch Dis Child Fetal Neontal Ed 1999. 80F8–14.14
21. Broughton S J, Sylvester K P, Page C M. et al Problems in the measurement of functional residual capacity. Physiol Meas 2006. 2799–107.107 [PubMed]
22. Sigurs N, Bjarnason R, Sigurbergsson F. et al Respiratory syncytial virus bronchiolitis in infancy is an important risk factor for asthma and allergy at age 7. Am J Respir Crit Care Med 2000. 1611501–1507.1507 [PubMed]
23. Stokes G M, Milner A D, Hodges I G. et al Lung function abnormalities after acute bronchiolitis. J Pediatr 1981. 98871–874.874 [PubMed]
24. Seidenberg J, Masters I B, Hudson I. et al Disturbance in respiratory mechanics in infants with bronchiolitis. Thorax 1989. 44660–667.667 [PMC free article] [PubMed]
25. Henry R L, Milner A D, Stokes G M. et al Lung function after bronchiolitis. Arch Dis Child 1983. 5860–63.63 [PMC free article] [PubMed]
26. Dezateux C, Fletcher M E, Dundas I. et al Infant respiratory function after RSV proven bronchiolitis. Am J Respir Crit Care Med 1997. 1551349–1355.1355 [PubMed]
27. Rusconi F, Galassi C, Corbo G M. et al Risk factors for early, persistent, and late‐onset wheezing in young children. SIDRIA Collaborative Group. Am J Respir Crit Care Med 1999. 1601617–1622.1622 [PubMed]
28. Dezateux C, Stocks J, Dundas I. et al Impaired airway function and wheezing in infancy: the influence of maternal smoking and a genetic predisposition to asthma. Am J Respir Crit Care Med 1999. 159403–410.410 [PubMed]

Articles from Archives of Disease in Childhood are provided here courtesy of BMJ Publishing Group