Article Text

Intrauterine growth restriction predicts lower lung function at school age in children born very preterm
  1. Eveliina Ronkainen1,2,
  2. Teija Dunder3,
  3. Tuula Kaukola4,
  4. Riitta Marttila4,
  5. Mikko Hallman1,4
  1. 1PEDEGO Research Center, and Medical Research Center Oulu, University of Oulu, Oulu, Finland
  2. 2Department of Children and Adolescents, Oulu University Hospital, Oulu, Finland
  3. 3Division of Allergology and Pulmonology, Department of Children and Adolescents, Oulu University Hospital, Oulu, Finland
  4. 4Division of Neonatal Medicine, Department of Children and Adolescents, Oulu University Hospital, Oulu, Finland
  1. Correspondence to Dr Eveliina Ronkainen, Oulu University Hospital, Department of Children and Adolescents, P.O. Box 23, Oulu FIN-90029 OYS, Finland; eveliina.ronkainen{at}


Objective Children born preterm have lower lung function compared with term-born children. Intrauterine growth restriction (IUGR) may predispose individuals to chronic obstructive pulmonary disease. The purpose of this observational study was to investigate the role of IUGR as predictor of lung function at school age in children born very preterm. We further studied the difference in lung function between term-born and preterm-born children with distinct morbidities.

Design Preterm-born children and age-matched and sex-matched term-born comparison groups (88 of each) were studied at the mean age of 11 years. Spirometry and diffusing capacity of the lung for carbon monoxide (DLCO) were recorded. All preterm-born subjects with IUGR (n=23), defined as birth weight less than −2 SD, were compared with preterm-born subjects without IUGR (n=65).

Results In the preterm-born children exposed to IUGR, the forced expiratory volume in 1 s (FEV1) was 5.7 (95% CI −10.2 to −1.3) and DLCO 9.2 percentage points lower (95% CI −15.7 to −2.7) than in the preterm-born children with appropriate in utero growth (expressed as percentage of predicted values). The effect of IUGR in decreasing FEV1 and DLCO remained significant after adjustment for bronchopulmonary dysplasia (BPD). Further study indicated that after adjustment with IUGR and BPD, prematurity explained reduction in FEV1 but not in DLCO.

Conclusions In children born very preterm, IUGR is an independent risk factor for a lower lung function in school age. We propose that IUGR and BPD are the major early factors predisposing the children born very preterm to lower lung function.

  • premature birth
  • bronchopulmonary dysplasia
  • intrauterine growth restriction
  • lung function test
  • pulmonary outcome

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

What is already known on this topic

  • According to the fetal origins hypothesis, an adverse intrauterine environment resulting in intrauterine growth restriction may predispose individuals to pulmonary disease in adult life.

  • Low birth weight appears to be associated with decreased lung function in later life in term-born or near-term-born children.

  • It is unclear whether poor in utero growth impairs pulmonary health in children born very preterm.

What this study adds

  • Intrauterine growth restriction (IUGR) is associated with impaired lung function in schoolchildren who were born very preterm.

  • The effect of IUGR on lung function remained significant after adjustment for bronchopulmonary dysplasia and prematurity.

  • IUGR in children born very preterm may require a special consideration for pulmonary health.


Low birth weight (BW) is associated with decreased lung function in children1–4 and adults.5 ,6 According to the fetal origins hypothesis, an adverse intrauterine environment resulting in low BW or intrauterine growth restriction (IUGR) affects airway growth and peripheral lung development, predisposing individuals to chronic obstructive pulmonary disease (COPD) in adult life.7 ,8 The distinction between prematurity-associated low BW and the effect of IUGR is difficult to reproduce in historical cohorts.9 The association between BW and later lung function was significant after adjusting for length of pregnancy in cohorts containing mostly term-born or near-term-born children.1 ,5 Few studies have addressed this question among exclusively prematurely born children. Greenough et al10 found that small for gestational age (SGA) children born preterm had significantly lower lung function at infancy compared with those born with appropriate weight for gestation. According to Morsing et al,11 there was no difference between preterm-born schoolchildren with or without IUGR, though preterm-born children with IUGR differed from term-born controls. In children born very preterm, bronchopulmonary dysplasia (BPD) adversely influences respiratory outcome.12–14

The primary aim of this study was to investigate whether IUGR is an independent risk factor associating with defects in lung function among schoolchildren born preterm. We further studied whether the risk factors associating with premature birth may account for the lower respiratory function of preterm-born compared with term-born children.


The study population consisted of 88 schoolchildren born before 32 weeks of gestation between 1997 and 2003 who were participating in prospective cohorts.15–17 The comparison group comprised 88 age-matched and sex-matched term-born controls recruited from the population register. We approached families with altogether 198 preterm-born and 632 term-born schoolchildren. Children were studied at the mean age of 11 years. Preterm-born children were younger, because we recruited the control child after the preterm-born child had visited (10.9, SD 1.4 years vs 11.6, SD 1.7 years, p<0.001). Infants with serious congenital defects were excluded. The study was approved by the Ethics Committee of Oulu University Hospital. All parents and children provided written informed consent.

The follow-up included measurements of spirometry and diffusing capacity of the lung for carbon monoxide (DLCO). Spirometry was measured using USB spirometry (SpiroStar USB, Medikro Oy, Kuopio, Finland), and DLCO using a single-breath technique (Jaeger MasterScreen PFT; Viasys Healthcare GmbH, Hoechberg, Germany), according to the guidelines. Reproducibility was considered acceptable if the variation between measurements did not exceed 5%. The best result was approved.18–20 Lung function data were standardised for height, age and gender. For spirometry we report results using Finnish reference (percentages)21 and multiethnic Global Lungs Initiative 2012 regression equations (Z-score).22 For DLCO we used reference values by Polgar and Promadhat.23 Current morbidities and exposures were determined by parental responses to the questionnaire.24

The length of gestation was always measured by ultrasonographic examination at <18 weeks of gestation. IUGR was defined as being born SGA; that is, BW >2 SD below the mean for gestation according to the reference values of the Finnish growth curves.25 All children with IUGR had at least one of the following associated perinatal factors: (A) placental perfusion defect upon pathological examination; (B) severe defect in placenta or in attachment of the umbilical vessels to fetal membranes with abnormal umbilical blood flow (absent or reverse diastolic flow in the umbilical artery); (C) preterm premature rupture of membranes for >7 days and evidence of fetal compression. The examination of the placentas has been previously reported.26 ,27 BPD was defined as lung disease requiring oxygen supplementation for ≥28 days, and its severity was graded by oxygen requirement at 36 postmenstrual weeks.28

The sample size was originally calculated to allow the detection of at least five percentage points difference in forced expiratory volume in 1 s (FEV1) (assumable clinically significant difference) between preterm-born and term-born children (α error 0.05 and power 90%). The minimum sample size required was 74 pairs. In retrospect, the study was 80% powered (with 5% α error) to detect a 7.5% points difference in FEV1 between those with and without IUGR. Statistical procedures were performed with IBM SPSS V.20.0 for Windows software (SPSS, Chicago, Illinois, USA). Comparisons between preterm and term children were performed by paired-samples t tests and between the IUGR and the no-IUGR preterm groups with independent samples t test for continuous variables and by Pearson's χ2 test for categorical variables. General linear model analysis was done to determine which of the risk factors predicted lung function in school-age children. p<0.05 was considered statistically significant.


The participating children were more premature at birth and had longer neonatal respirator treatment than those who did not participate. There was no statistically significant difference in the incidence of IUGR between the groups (26% vs 15%, respectively; p=0.08) (see online supplementary table S1). Twenty-three of the 88 children born preterm were defined as IUGR; only one term-born child had IUGR. The length of pregnancy was very similar between the IUGR and the no-IUGR cases. A higher percentage of the IUGR cases was girls (70% vs 38%, p=0.01) compared with those without IUGR (table 1).

Table 1

Perinatal characteristics and postnatal growth of children born preterm with or without IUGR

Among the perinatal factors, pre-eclampsia (70% vs 17%, p<0.001) and placental perfusion defect (75% vs 24%, p<0.001) were associated with IUGR. Histological chorioamnionitis was less frequent in the IUGR group than in the no-IUGR group (13% vs 52%, p=0.009). Maternal smoking in pregnancy or passive smoke exposure did not differ between groups (table 1).

Of the prematurely born children, 49 (56%) had BPD (30 mild, 11 moderate and 8 severe). BPD was equally frequent in the IUGR group and the no-IUGR group (65% vs 52%, p=0.3). Blood culture-positive septicaemia was equally frequent in those with and without IUGR (26% vs 20%, p=0.6; table 1). However, infants with BPD had significantly more often blood culture-positive septicaemia than those without BPD (33% vs 8%, p=0.005). Of the infants with BPD and sepsis, 5 (31%) had IUGR and 11 (69%) did not. At 2 years of age, children with IUGR were smaller than children without IUGR. In school-age children, there were no longer statistically significant differences in height between children with or without IUGR (table 1).

As previously reported,14 all of the spirometric measurements (forced vital capacity (FVC), FEV1, FEV1/FVC ratio, peak expiratory flow, maximum expiratory flow at 50% of vital capacity (MEF50) and maximum midexpiratory flow (MMEF)) and DLCO were significantly lower in children born preterm compared with term-born controls. Among children born preterm, FVC, FEV1, MEF50, MMEF, DLCO and DLCO/VA (alveolar volume) were lower in children with IUGR compared with children with appropriate in utero growth (table 2). In our preterm-born population FEV1 did not differ between boys and girls (data not shown).

Table 2

Means of spirometry and DLCO values for preterm-born and term-born children and by IUGR diagnosis

To study risk factors predicting the lung function among children born preterm, we established a general linear model. IUGR, male gender, gestational age (GA) and BPD were considered. Evaluation of the risk factor interdependence revealed correlation between GA as continuous variable and BPD (Pearson's correlation −0.7, p<0.001), and when entered into the model simultaneously, both failed to reach statistical significance due to collinearity. Therefore GA was left out from the model of children born preterm. IUGR did not correlate with GA. Using FEV1 (expressed as % predicted value) as the dependent variable and gender, BPD and IUGR as predictors, the model was statistically significant and accounted for 9.8% of the variance in FEV1 (p=0.008). Using DLCO as the dependent variable, IUGR and BPD accounted 11.7% of the variance in DLCO (p=0.004; table 3).

Table 3

Predicted influence of male gender, BPD and IUGR on FEV1 and DLCO among children born preterm

In order to further analyse the difference in lung function between term-born and preterm-born children with different morbidities we included term-born controls in the general linear model. BPD and IUGR associated with reduced FEV1, MEF50 and DLCO. Prematurity, adjusted with BPD and IUGR, was significant in FEV1 and MEF50, but not in DLCO (table 4).

Table 4

Predicted influence of BPD, IUGR and prematurity on lung function and DLCO among all participating children


In the present study, we investigated whether IUGR may be responsible for the lower spirometry values and diffusing capacity of schoolchildren with a history of premature birth. According to previous studies, BPD was associated with decreased values of spirometry12–14 and diffusing capacity.14 ,29 We detected lower lung function parameters among preterm-born children with IUGR compared with gestation controls, supporting the hypothesis that poor in utero growth is an additive burden to pulmonary health. Although the catch-up in somatic growth was remarkable, the effect of IUGR on lung function in school-age children remained statistically significant after adjustment for BPD and gender.

Greenough et al10 compared lung function of 31 SGA infants (BW <10th centile) with 88 appropriately in utero grown children born preterm using whole body plethysmograph. The main finding was higher airway resistance in the SGA group. Morsing et al11 performed spirometry to children born with IUGR with absent or reversed end-diastolic blood flow in the umbilical artery. All 31 children born preterm in the IUGR group were SGA (BW >2 SD below the mean for gestation). Controls were 31 preterm-born infants with BW appropriate for GA and 31 term-born children with normal in utero growth. The IUGR group had lower FEV1, FEV1/FVC and MMEF compared with the term-born control group, but two preterm-born groups did not differ in school age.

Greenough et al's10 and our study were not originally planned to investigate IUGR. In our study, the sample size was originally calculated for the difference between preterm-born and term-born children making subgroup analysis underpowered. The IUGR and the no-IUGR groups had similar GAs whereas in the IUGR group the girls were over-represented. We found no gender difference in the lung function of children born preterm, though male sex tended to associate with lower FEV1 (p=0.05). This is in line with Fawke et al12 who concluded that male disadvantage in respiratory function decreases with postnatal age. Nevertheless, over-representation of girls with IUGR does not weaken the observed difference between the IUGR and the no-IUGR groups. Present results remain to be further confirmed.

Children born preterm are at high risk for airway obstruction.12–14 The small airways (MEF50, MMEF) and gas exchange (DLCO, DLCO/VA) are particularly affected. Prenatal factors including maternal nutrition and the function of placenta can alter lung development. Placental dysfunction is the most common cause of IUGR.30 In the present study, pre-eclampsia and placental perfusion defect were strongly associated with IUGR. Histological chorioamnionitis associated with normal in utero growth and not with either BPD or lower lung function in school-age children (data not shown). With more advanced treatments, previously observed association between chorioamnionitis and BPD may no longer be evident, as demonstrated by Been and Zimmermann.31 In our cohort, blood culture-positive septicaemia, BPD and lower lung function in school-age children were associated. Others reported that small preterm-born infants had a high risk of BPD when exposed to chorioamnionitis and either mechanical ventilation >7 days or postnatal infection.32

Our prospective population cohort was from a single hospital. However, Oulu University Hospital is a regional centre and about 90% of all very-preterm deliveries in northern Finland take place in Oulu. The fact that participants were born more premature and had been longer in the respirator than the non-participants, potentially adversely influenced the respiratory outcome. Only 14% of the invited term-born controls participated. The rate of family history of atopy was fairly high among the controls suggesting a selection bias in our control population since families with atopy or suspicion of allergies may have been more motivated to participate than others.

In our study, IUGR was defined as being born SGA. Granted, some infants with IUGR have normal BWs and may not be correctly classified.30 For IUGR/SGA we selected a cut-off point of 2 SD corresponding to the 2.3th centile of the mean BW.25 Although segregation of SGA from normal is arbitrary, the 2 SD cut-off was selected because it encompasses higher percentage of fetuses with abnormal fetal growth than the definition of BW below the 10th centile.33

Our definition of IUGR/SGA implied that less than 2.3% of children born in the given gestation are included. However, in our cohort 26% of the children born preterm were SGA while 1.1% of term-born controls were classified as SGA. Statistically, there ought to be the same proportion of SGA children in preterm-born and term-born populations. This suggests that the BW reference for children born very preterm is biased towards normal intrauterine growth.34 We used the Finnish population as reference since all children were of Finnish origin. The imbalance in SGA incidence between term-born and preterm-born children emphasises the need to interpret the results with caution.

The Barker hypothesis states that adverse intrauterine growth predisposes individuals to chronic disease in adult life.35 The mean FEV1, adjusted for height and age, rose by 0.06 L (95% CI 0.02 to 0.09) for each 450 g increase in BW, independent of smoking habits and social class in Englishmen born in 1920–1930.7 Rona et al reported that BW adjusted for GA, parental smoking and social factors was associated with FEV1 in school-age children.1 IUGR associates with poorer lung function at 8–9 years in term-born children.3 Among a Scottish cohort born in 1997–1999, for each millimetre increase in first trimester size, FEV1 was higher by an average of 6 mL at 10 years of age.4

Very preterm birth is associated with lower lung function at school age.12–14 Severe complications of prematurity were intensified by the effect of suboptimal fetal growth, causing so-called double jeopardy.36 According to our data, the sequential insults, IUGR and BPD together with prematurity, associated with deterioration of lung function in school-age children. In contrast, the deficit in diffusing capacity of children born preterm was explained by BPD and IUGR without the added impact of prematurity. We propose that the preterm pulmonary insults, IUGR and BPD independently predict lower lung function in school-age children.

The risk for mild-to-severe BPD was up to 3.8-fold greater in infants with SGA compared with infants of appropriate size for GA.36 The association between lower-than-expected BW and moderate-to-severe BPD was still evident in large contemporary cohorts.37 ,38 In our follow-up study, there was no detectable difference in the risk of BPD between those with IUGR and those with appropriate size for GA. This might be explained by the limited number of cases in our cohort. However, we found no significant association between BPD and IUGR when the whole cohort of 198 invited preterm-born children was considered.

The lack of a precise definition of BPD is a limitation in the present study. Currently the need for oxygen is determined on the basis of a physiological assessment rather than on the basis of individual assessment. However, since our study population was born before the introduction of the standardised oxygen reduction test this remains a disadvantage in our study.39 The rate of surfactant replacement treatment was 64% in our preterm-born population. This is equivalent to a large North American registry data, where surfactant replacement rate varied from 64% to 66% during the years 2000–2009.40 Nevertheless, the neonatal treatment practices that were used for the present study population are somewhat outdated in comparison to the current treatment protocols.

In conclusion, we found preliminary evidence that IUGR and BPD are independent risk factors that predict limitations in the lung function of preterm-born school-age children. Our result is in concordance with the fetal origins hypothesis. School-age children with IUGR had a lower lung function in spirometry and particularly in diffusing capacity, despite significant catch-up of somatic growth. The observed deficits in lung function were clinically insignificant in most cases, since all mean values were above the lower limit of normal (% predicted >70 or Z-score >−1.64). However, due to a possible predisposition to COPD, these individuals should avoid any excess burden on pulmonary health, such as smoking. The causes of fetal growth restriction need to be studied prospectively in order to better understand the mechanisms of fetal programming and eventually to aim to decrease the risk of IUGR and its adverse consequences.


The authors thank Riitta Vikeväinen, research nurse, Tytti Pokka, MSc, biostatistician, and personnel at the outpatient clinic of the Department of Children and Adolescents and at the Pulmonary Function Laboratory of Internal Medicine, Oulu University Hospital. We would also like to thank the children and the families who participated in this study.


Supplementary materials

  • Supplementary Data

    This web only file has been produced by the BMJ Publishing Group from an electronic file supplied by the author(s) and has not been edited for content.


  • Contributors All authors have made a substantial contribution to prepare the manuscript (planning, conduct and reporting). ER wrote the first draft of the manuscript. Each author listed on the manuscript has seen and approved the submission of this version of the manuscript and takes full responsibility for it. ER, TD and MH are responsible for the overall content as guarantors.

  • Funding This work was funded by the National Graduate School of Clinical Investigation/University of Oulu Research Fund, the Alma and K. A. Snellman Foundation (Oulu, Finland), the Foundation for Pediatric Research (Helsinki, Finland), Väinö ja Laina Kiven säätiö (Helsinki, Finland), and the Sigrid Jusélius Foundation (Helsinki, Finland).

  • Competing interests None declared.

  • Patient consent Obtained.

  • Ethics approval The Ethics Committee of Oulu University Hospital.

  • Provenance and peer review Not commissioned; externally peer reviewed.