Objective. To determine the risk of chronic lung disease (CLD) in small for gestational age (SGA) preterm infants in comparison to appropriately grown and large for gestational age (LGA) infants.
Methods. Observational study derived from a geographically defined population (Trent Health Region, United Kingdom). All preterm infants of ≤32 completed weeks’ gestation born to Trent resident mothers admitted to neonatal units between 1995 and 1999 (inclusive) were included. Birth weight percentiles were created for the whole population, and infants were classified as SGA infants (if <10th percentile), appropriately grown (if between 25th and 75th percentiles—reference group), and LGA infants (if ≥90th centile). Both mortality and CLD rates (using both 28 days’ and 36 weeks’ postmenstrual age [PMA] definitions) were determined for these groups of infants.
Results. Four thousand fifty-one preterm infants ≤32 weeks’ gestation were identified. SGA infants showed higher mortality before 28 days’ postnatal age and 36 weeks’ PMA as compared with reference group infants (odds ratio [OR]: 2.01, 95% confidence interval [CI]: 1.49–2.72; and OR: 2.00, 95% CI: 1.49–2.69), respectively.
SGA infants showed a significantly greater risk of developing CLD, both at 28 days’ and 36 weeks’ PMA as compared with the reference group infants (OR: 1.34, 95% CI: 1.03–1.74; and OR: 1.87, 95% CI: 1.39–2.51), respectively. LGA infants showed a trend toward a reduced incidence of CLD in comparison to the reference group, which was statistically significant for the 36 weeks’ definition (OR: 0.76–28 weeks, 95% CI: 0.57–1.01; and OR: 0.55–36 weeks, 95% CI: 0.37–0.81).
Conclusions. Fetal growth seems to influence mortality in general and morbidity, attributable to CLD, in particular in preterm infants. SGA preterm infants are at higher risk of death before 28 days’ and 36 weeks’ PMA and CLD by both definitions. LGA infants show reduced risk of CLD.
- chronic lung disease
- intrauterine growth retardation
- small for gestational age
- bronchopulmonary dysplasia
Chronic lung disease (CLD) is one of the most important morbidities associated with modern neonatal intensive care. This condition was first described by Northway et al in 19671 and subsequently modified by Bancalari et al in 1979.2 Two definitions of CLD remain in common use:
Need for supplemental oxygen/ventilation at 36 weeks’ postmenstrual age (PMA).
The latter is said to be a better predictor of long-term pulmonary outcome in very low birth weight (VLBW) infants.6
Trends in the incidence of CLD have been difficult to follow. A number of studies have reported the experience of units or groups of units and hence have potentially been subject to referral bias and inclusion bias.7–11 Our experience, based on a geographically defined population,12 suggests that the incidence of CLD showed a significant increase between 1987 and 1992 (doubled by 28-day definition and tripled by 36-week definition), but between 1992 and 1997, the incidence of CLD remained static despite significant increases in survival.
There is an increasing recognition of changes to the pathophysiology of CLD, with some VLBW infants developing the condition without preceding respiratory distress syndrome.13 This has led investigators to seek other etiologic risk factors, eg, poor fetal growth. A variety of animals studies have explored this effect and provided evidence that reduced fetal growth may predispose to impaired respiratory function after birth as a result of impaired growth of 1 or more components of the lungs and/or chest wall.14–18
However, the findings are confusing with human studies showing both reduced4 and increased risk of CLD19 in infants who grow poorly in utero. Other evidence suggests a differential effect with less risk of CLD at 28 days’ postnatal age, but with those who had CLD at this age more likely to continue having an oxygen requirement beyond 36 weeks’ PMA as compared with larger infants.20
We wished to explore these issues further with a large population-based cohort of infants ≤32 weeks’ gestation. The specific aim of this study was to explore poor fetal growth as a risk factor for CLD (both definitions).
The Trent Neonatal Survey is an ongoing study of neonatal intensive care activity in the Trent Health Region (United Kingdom), which has a population of around 4.6 million with ∼60 000 births a year. Trent is recognized as being representative of England and Wales as a whole.12 All 16 perinatal services in the region contribute to the study, and units in adjacent regions also permit data collection on Trent infants. The survey was established in February 1990; the database holds information relating to all infants of 32 weeks’ gestation or less born to Trent resident mothers and admitted to a neonatal unit since that time.
The infants used for this analysis were as follows:
All live births (singleton, multiple),
Between 24 to 32 weeks’ gestation (inclusive),
Born between 1995 and 1999 (inclusive),
Without any congenital malformation, and
Of European ethnic origin (white, white).
Infants who still required active respiratory support (mechanical ventilation, continuous positive airway pressure) or oxygen at 28 days of life;
Infants who still required active respiratory support or oxygen at 36 weeks’ PMA.
Gestation was defined according to the hierarchy specified by the National Confidential Enquiry into Stillbirth and Deaths in Infancy program; mother certain of her dates (most reliable); early dating scan (<20 weeks’ gestation); late dating scan (>20 weeks); and postnatal examination (least reliable). If the difference between maternal dates and early scan was >7 days, early scan was used to determine the period of gestation.21
The infants were divided into categories according to their birth weight (Fig 1). Birth weight percentiles were estimated using quantile regression22 separately for boys and girls. The births lying below the appropriate 10th percentile were classified as small for gestational age (SGA). Infants with birth weights above the 90th percentile were classed as large for gestational age (LGA), and the births that fell between the 25th and 75th percentiles were classed as appropriately grown. Defining the groups in this way also had the advantage of producing groups with approximately equal proportion of each gestational group.
Quantile regression fits a line that minimizes of the sum of the absolute residuals. This method was chosen as it is robust to non-normality of the residuals and to heteroscedasticity, both of which were evident with these data. In addition, it allows the estimation of a regression line for any chosen percentile. The lines fitted were inspected by eye and appeared to fit the data adequately.
The mortality rates were compared between the groups and the results expressed as odds ratios (ORs) with 95% confidence intervals (CIs). The risks of CLD among the survivors were then estimated. Gestation adjusted ORs were also estimated using logistic regression models. Estimates were also obtained for a combined outcome of mortality or CLD.
SAS v8.0 (SAS Institute Corp, Cary, NC) was used for all analyses.
During the 5-year period, there were a total of 4051 preterm infants that were identified between 24 and 32 weeks’ gestation (inclusive). Of these, 2991 (73.8%) infants were singletons. Four hundred one (9.9%) infants were under the estimated 10th birth weight percentile (SGA), 2019 (49.8%) were between the estimated 25th and 75th birth weight percentiles (reference), and 414 (10.2%) were less than or equal to the estimated 90th percentile (LGA).
Figure 1 represents plots showing the percentiles that were used for male and female infants. Shaded areas represent the SGA, LGA, and reference groups.
Mortality to 28 days’ postdelivery and 36 weeks’ PMA was significantly higher in the SGA group (Table 1). Sixty-eight (17.0%) of 401 SGA infants died before 28 days’ postnatal age as compared with 186 (9.2%) of 2019 in the reference group (OR: 2.01; 95% CI: 1.49–2.72) and 35 (8.5%) of 414 LGA infants (OR of LGA infants in comparison to reference group: 0.91; 95% CI: 0.62–1.33). By 36 weeks’ PMA, 73 (18.2%) of 401 SGA infants died compared with 202 (10%) of 2019 infants in the reference group (OR: 2.00; 95% CI: 1.49–2.69) and 36 (8.7%) of 414 LGA infants (OR of LGA infants in comparison to reference group: 0.86; 95% CI: 0.59–1.25).
Ninety-six (28.8%) of 333 SGA infants developed CLD at 28 days as compared with 426 (23.2%) of 1833 in the reference group (OR: 1.34; 95% CI: 1.03–1.74) and 71 (18.7%) of 379 LGA infants (OR in comparison to reference group: 0.76; 95% CI: 0.57–1.01). The comparison groups were balanced to start with in view of the methodology applied. However, we were concerned that the differential mortality rates observed may have unbalanced the groups, and hence we adjusted for gestation. This resulted in an OR for CLD in the SGA infants compared with the reference group of even greater significance (OR: 2.23; 95% CI: 1.57–3.15). The reversed trend, a significant reduction, was seen in the LGA infants compared with the reference group (OR: 0.55; 95% CI: 0.37–0.81). The same pattern was apparent for the 36 weeks’ PMA (Table 2).
Data for singletons only were analyzed separately but the pattern of results showed no significant difference to those obtained for the group as a whole.
Table 3 shows the proportion of infants in each of the 3 groups who survived to 36 weeks’ PMA and required active respiratory support (mechanical ventilation or nasal continuous positive airway pressure or both) at any stage. No significant differences were observed in the proportions of infants in each group needing this type of care.
Large numbers of studies have investigated the risk factors associated with CLD, eg, infection, patent arterial duct, fluid management, mechanical ventilation strategies, etc.4–6,11,19,20,23 With the changing pattern of CLD, more infants are diagnosed to have this condition without preceding severe respiratory distress syndrome and mechanical ventilation.13,24,25 It is in this group of infants that impaired intrauterine growth has been suggested as having a particular etiologic role. Presumably, reduced fetal growth acts as a marker of abnormal intrauterine lung development, which is then reflected in the clinical course post delivery. Our study aimed to look at the influence of impaired intrauterine growth on the development of CLD in a complete cohort of infants.
Data from existing epidemiologic studies in relation to this topic are somewhat confusing. Hakulinen et al 198819 reported on CLD at 28 days in a preterm cohort from a single center. They found no significant difference in the proportion of SGA infants in the CLD and non-CLD groups of infants (12% vs 14%, respectively). On further analysis, however, they noted that preterm infants with CLD were significantly lighter even when compared with preterm infants of equally low mean gestational age (for each gestation age group, the standardized mean birth weight of non-CLD infants was 0 (standard deviation: 1), whereas this was significantly lower at −0.67 (standard deviation: 1.1; P < .05 for CLD infants). No information on CLD at 36 weeks’ PMA was provided. Palta et al4 reported on traditional risk factors for CLD (oxygen dependence on day 30 of life) in a cohort of VLBW infants from 7 neonatal units in Wisconsin and Iowa. They found that infants who were SGA were at lower risk of CLD than infants whose size was average for gestational age but with the same weight. This comparison seems inappropriate, because the methodology will have produced a comparison group, the gestation of which will have been consistently less than that of the SGA infants. Korhonen et al,20 in a retrospective study, reported on the risk factors associated with CLD (both definitions) in a cohort of VLBW infants from 1 hospital. SGA infants were less common in the CLD group than among infants without CLD (17% vs 35%; P = .0131) at 28 days’ postnatal age. However, numbers were too small to adequately assess the influence of infants born SGA. Another study published only in abstract form26 looked at outcome of SGA infants born between 24 and 26 weeks’ gestation and found that SGA infants were at higher risk of CLD at 36 weeks as compared with appropriate for gestational age infants (65% vs 32%). This was a retrospective review of admissions to a single neonatal unit between January 1983 and December 1992; limited details are available from the abstract.
We believe that the apparent variation in the findings of the above studies is likely to have been produced, predominantly, by 2 factors. A number of the studies are single center and hence referral and inclusion bias may have affected the results. More important is the need to adjust for the fact that increased numbers of low gestation SGA infants die before they can develop CLD. How any such effect was interpreted will have varied between studies. In the data reported here gestation was higher in the SGA groups at both time points at which unadjusted comparisons were made (28 days’ and 36 weeks’ PMA), ie, differential mortality did occur.
Clearly one explanation for our findings is that gestation has been consistently miscalculated. This seems unlikely. The Trent Neonatal Survey has used the same approach to data collection for 10 years. A small, independent team obtains the data using a standardized approach with team members cross checking each other. In addition, for this aspect of data collection, the high number of early dating scans now performed in this population (60%–70%) provides an additional data quality check. The consistent behavior of the groups, in terms of the need for respiratory support, also indicates that the groups were of comparable maturity. We have no information on mode, duration, and maximum peak inspiratory pressure used.
The data indicating that LGA infants have improved survival and a reduced incidence of CLD as compared with the reference group of infants are interesting, and we can find no comparable data in the literature. The finding supports the importance of normal (or supra normal) intrauterine growth in achieving healthy outcomes after delivery. The converse is also true, and the data regarding the poor outcome of SGA infants indicates that greater focus on optimizing intrauterine growth, the factors that affect it,14–18,24,25 and the timing of delivery may offer great improvements for neonatal outcomes.
We wish to acknowledge the continuing help and collaboration of the hospitals delivering perinatal care in both Trent and adjacent regions. This study is one of the Trent Infant Mortality and Morbidity Studies, which are funded by the Trent Regional Health Service. Elizabeth S. Draper, MPhil, is supported by a grant from Leicestershire Health Authority.
- ↵Cooke RWI. Factors associated with chronic lung disease in preterm infants. Arch Dis Child.1991;66 :776– 779
- ↵Shennan AT, Dunn MS, Ohlsson A, Lennox K, Hoskins EM. Abnormal pulmonary outcomes in premature infants: prediction from oxygen requirement in the neonatal period. Pediatrics.1988;82 :527– 532
- ↵Avery ME, Tooley WH, Keller JB, et al. Is chronic lung disease in low birth weight infants preventable? A survey of eight centers. Pediatrics.1987;79 :26– 30
- ↵Manktelow BN, Draper ES, Annamalai S, Field D. Factors affecting the incidence of chronic lung disease of prematurity in 1987, 1992, and 1997. Arch Dis Child Fetal Neonatal Ed.2001;85 :F33– F35
- ↵Charafeddine L, D’Angio CT, Phelps DL. Atypical chronic lung disease patterns in neonates. Pediatrics.1999;103 :759– 765
- ↵Draper ES, Manktelow B, Field DJ, James D. Prediction of survival for preterm births by weight and gestational age: retrospective population based study. BMJ.1999;319 :1093– 1097
- ↵Marshall DD, Kotelchuck M, Young TE, et al. Risk factors for chronic lung disease in the surfactant era: a North Carolina population-based study of very low birth weight infants. Pediatrics.1999;104 :1345– 1350
- ↵Bardin C, Zelkowitz P, Papageorgiou A. Outcome of small-for-gestational age and appropriate-for-gestational age infants born before 27 weeks of gestation. Pediatrics.1997;100(2) . Available at: www.pediatrics.org/cgi/content/full/100/2/e4
- Copyright © 2003 by the American Academy of Pediatrics