INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Keywords: infant; small for gestational age; fetal growth retardation; birthweight; function test; respiratory; sex

The clinical importance of fetal growth restriction was first recognized by Lubchenco and coworkers in 1963, who reported that perinatal mortality and morbidity were increased among infants whose birthweight fell at or below the 10th percentile for gestational age (1). More recently, it has been shown that infants who are small for gestational age (SGA) are at increased risk of sudden death in infancy (2) and of wheezing and respiratory infection in early childhood (3, 4). Furthermore, diminished airway function has been reported in adults who were of low birthweight (5, 6), leading to the speculation that this association is mediated by "fetal programming" (5). The physiologic mechanisms underlying these associations remain unclear. If the "fetal programming" hypothesis is correct, this might suggest an association between birthweight and airway function in infancy and early childhood. However, there have been few published studies specifically examining this association (7, 8) and this aspect has received little attention in previous epidemiologic studies of infant respiratory function (9). The associations between maternal smoking in pregnancy and low birthweight (10, 11) and between impaired airway function in infancy (12) and an increased risk of lower respiratory illness in early childhood are well recognized (15), but it is unclear whether there is an association between low birthweight and impaired airway function in infancy that is independent of exposure to maternal smoking. This study was therefore established to examine the hypothesis that airway function is impaired in infants who are of low birthweight for gestational age but who have not been exposed to maternal smoking pre- or postnatally.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Study Population

Infants were recruited from the maternity units at the Homerton and University College Hospitals, London. Healthy, singleton infants (> 35 wk gestation) were eligible for inclusion if born to a white northern European mother who did not smoke in pregnancy or postnatally. Infants with congenital abnormalities or with neuromuscular or cardiorespiratory disorders were ineligible, as were those who required any ventilatory assistance during the neonatal period or who had experienced any lower respiratory illness (LRI) prior to testing.

Infants were classified according to birthweight and gestational age using the sex-specific Child Growth Foundation (CGF) algorithms (16) as well as the Gestation Related Optimal Weight or "GROW" program (17). The latter takes maternal characteristics such as height, booking weight, ethnic group, and parity into account as well as infant birthweight, gestation, and sex. Gestational age was based on ultrasound assessment before 20 wk. Infants with a birthweight  10th percentile according to either the CGF or GROW programs were classified as SGA, whereas those between the 20th and 95th percentile were classified as appropriate for gestational age (AGA). Local Research Ethics Committees approved this study and informed written consent was obtained from parents.

Family history of respiratory illnesses, first-degree family history of asthma, maternal age on leaving full time education, and parental occupational status were obtained from the mother at the time of the lung function test. Additional details were obtained from the obstetric notes. Exposure to maternal smoking pre- and postnatally was assessed from parental report. Current smoking exposure was validated by cotinine assay of infant urine and maternal saliva obtained at the time of lung function testing (18). Five infants whose mother's salivary cotinine concentrations ranged from 20.8 to 434.6 ng ml¹ were excluded from the study as these are consistent with values reported from active smokers (> 15 ng ml¹) (19, 20). Maternal salivary cotinine for the remaining study population was negligible (geometric mean [range]: 0.125 [0.0001-4.096] ng ml¹), with no significant difference between mothers of SGA and AGA infants.

Respiratory function was measured between 4 and 12 wk postnatally, when infants had been well and free from upper respiratory tract infections for at least 3 wk. Body weight and crown-heel length were measured as described previously (21) and expressed as sex-specific z scores (16). All infants were studied supine following sedation with 60 mg kg¹ chloral hydrate, administered orally. Heart rate and oxygen saturation were monitored continuously during the test (CO₂SMO, Model 7100, Novametrics Medical Systems Inc., Wallingford, CT).

Respiratory Function Tests

Airway function was assessed from both partial (22) and raised lung volume (23, 24) forced expiratory maneuvers, using the rapid thoracoabdominal compression (RTC) technique as described previously (13, 25, 26). Measurements were performed in accordance with recent recommendations (22), ensuring that flow limitation, as indicated by reproducible flow-volume (F-V) curves with no further increase in maximal flow at FRC (V'maxFRC), was achieved despite increasing jacket pressures. V'maxFRC, was reported as the mean (SD) of the three highest flows at FRC (22, 27).

Measurements of airway function at raised lung volume were performed using an adaptation of the technique described by Feher and coworkers (23). Briefly, the respiratory muscles were relaxed by administering four or five lung inflations to a pressure of 3 kPa before inflating the jacket to force expiration from raised lung volume. This maneuver was repeated until a minimum of three acceptable and reproducible F-V curves was obtained. Parameters calculated from the raised volume technique, including forced vital capacity from an inflation pressure of 3 kPa (FVC), forced expiratory volume at 0.4 s (FEV_0.4), and forced expired flow at 75% of expired FVC (FEF₇₅) were reported from the "best" raised volume curve. This was defined as the technically acceptable forced expiratory F-V curve with the highest sum of FVC and FEV_0.4.

Sample Size and Statistical Analysis

It was estimated that 40 infants per group would provide 90% power at the 5% significance level to detect a difference of one standard deviation (SD) in estimates of forced expiratory flows and volumes between SGA and AGA infants after adjustment for potential confounding factors. Comparisons of group characteristics and respiratory function between the groups were performed using t tests, chi-square, or exact tests as appropriate (StatXact v 4.01). The extent to which low birthweight for gestational age is associated with residual variance in forced expiratory flows and volumes was examined using multiple linear regression (SPSS for Windows, Release 8.0.2) after adjustment for body size and sex and after examining for the effects of other potential confounding factors.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Lung function measurements were attempted in 123 infants but were unsuccessful in 15, due to poor quality data or inability to complete the study protocol. Of the 108 infants successfully measured, five were excluded subsequently because maternal salivary cotinine concentrations were consistent with values reported from active smokers (see METHODS) (19, 20). Thus, lung function measurements from 103 infants (38 SGA and 65 AGA) form the basis of this report.

The characteristics of these infants are summarized according to birthweight classification in Table 1. Of the 38 SGA infants, 31 were  10th percentile by CGF classification, 34 by GROW, and 28 by both. In six SGA infants, birthweight was between the 10th and 15th percentile according to the CGF algorithm, but  10th percentile according to the GROW algorithm, while in three infants birthweight was between the 10th and 15th percentile according to GROW but  10th percentile by CGF. A further infant whose birthweight was < 2nd percentile on CGF could not be classified by GROW due to missing data for maternal height. All these infants were classified as SGA. Birthweight was above the 20th percentile by CGF in all 65 AGA infants, but in six fell between the 15th and 19th percentile when classified by GROW.

Boys composed a significantly greater proportion of SGA than AGA infants. At birth, SGA infants were of similar gestational age but lower weight and head circumference than AGA infants, reflecting the selection criteria used. A positive family history of asthma was reported in a similar proportion of SGA and AGA infants, and in four (11%) SGA and eight (12%) AGA infants, the mother was one of the affected family members. There were no significant differences between the groups with respect to maternal age, maternal age at completion of full time education, or the percentage of mothers in nonmanual occupations. The proportion of SGA and AGA pregnancies complicated by prolonged ruptured membranes (more than 24 h but less than 1 wk), antepartum hemorrhage, and pregnancy-induced hypertension was similar (data not shown), but more SGA (21%) than AGA (8%) infants exhibited meconium staining of liquor during labor (95% CI, SGA  AGA: 29%, 0.13%; p = 0.07).

Five SGA (13%) and six AGA (9%) infants experienced an upper respiratory illness prior to testing but all infants had been free from respiratory symptoms for at least 3 wk at the time of the test. There were no significant differences in any airway function between those who did or did not have upper respiratory illness prior to testing.

Characteristics at time of test are summarized in Table 2 according to birthweight status. At time of testing, SGA infants weighed less, were shorter, and of smaller head and chest circumference than AGA infants.

Association between Airway Function and Birthweight Status

In univariate analyses, FVC and FEV_0.4 were significantly lower in SGA compared with AGA infants (Table 3). These parameters were also associated with body length (Figures 1 and 2) and weight at test but not with sex. FVC was on average (95% CI) 8 ml (6.5, 9.9) and FEV_0.4 8.3 ml (5.3, 10.7) greater for each centimeter increase in body length. After allowing for body length, birthweight status was no longer significantly associated with FVC and FEV_0.4 (Table 3). Thus, the apparent association with birthweight status appears to be mediated primarily through body size rather than a specific effect on lung and airway size, as infants of low birthweight for gestational age were of shorter body length at testing than those whose birthweight was appropriate for gestational age.

View larger version (16K): [in this window] [in a new window]   Figure 1.   Lung volume (FVC) plotted against body length according to birthweight status. Solid triangles, SGA infants; open triangles, AGA infants.

View larger version (17K): [in this window] [in a new window]   Figure 2.   Forced expired volume in 0.4 s plotted against body length according to birthweight status. Solid triangles, SGA infants; open triangles, AGA infants.

FEF₇₅ was also significantly lower in SGA than in AGA infants (Figure 3). However, after allowing for sex (Table 3) birthweight status was no longer significant, reflecting in part the relative excess of boys in the SGA group. In univariate analyses, V'maxFRC was significantly associated with sex but not birthweight status.

View larger version (17K): [in this window] [in a new window]   Figure 3.   Forced expired flow at 75% of vital capacity plotted against body length according to birthweight status. Solid triangles, SGA infants; open triangles, AGA infants.

Because there was a higher proportion of boys in the SGA group, anthropometric and lung function data were also compared by sex (Table 4). There were no significant sex differences in FVC or FEV_0.4, but boys had significantly lower values of FEF₇₅ and V'maxFRC when compared with girls (Figures 4 and 5). After adjustment for body size, these relationships remained significant with FEF₇₅ being on average 35 ml s¹ (95% CI, boys  girls: 61, 8) and V'maxFRC being on average 34 ml s¹ (62, 6) lower in boys than girls.

View larger version (16K): [in this window] [in a new window]   Figure 4.   Forced expired flow at 75% of vital capacity plotted against body length according to sex. Solid triangles, boys; open triangles, girls.

View larger version (15K): [in this window] [in a new window]   Figure 5.   Maximal expired flow at functional residual capacity plotted against body length according to sex. Solid triangles, boys; open triangles, girls.

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

The results of this study suggest that both lung capacity and airway function, as reflected by FVC, FEV_0.4, and FEF₇₅, are diminished during the first few months of life in infants who are born small for gestational age, but this association appears to be primarily mediated through impaired somatic growth rather than through a specific effect on lung and airway growth. In addition, we found that measures that reflect peripheral but not central airway function were significantly diminished in boys compared with girls.

This is the first study to explicitly test the hypothesis that low birthweight for gestational age is associated with impaired airway function in infancy. By limiting analysis to infants who had not been exposed to maternal smoking, the influence of low birthweight relative to duration of gestation (and by inference impaired fetal growth) on airway function in infancy can be examined.

There are several strengths in the current study design. First, lung function tests were performed prior to any lower respiratory infections (LRI), thus enabling respiratory function in SGA and AGA infants to be compared without potential confounding by the effects of LRI on airway function. Second, gestational age was determined from sonographic assessments before 20 wk gestation, which are currently considered to be the most accurate means of estimating gestation and hence birthweight percentiles. Third, we were able to validate maternal reports of postnatal smoking by measuring cotinine, a breakdown product of nicotine, in maternal saliva and infant urine. This allowed the association of low birthweight and airway function to be examined independently of the known associations between maternal smoking, low birthweight, and impaired airway function. Finally, the use of the raised volume technique to measure airway function allowed measures of forced expiration to be compared between infants over an extended volume range (24).

How Representative Is This Population?

Because prematurity, respiratory disease, and ventilatory assistance during the neonatal period are all likely to have a negative impact on airway function, such infants were excluded from this study. As this potentially excludes those infants with more severe fetal growth retardation born by elective premature delivery or with severe respiratory disease, the SGA population studied may be potentially biased toward those with less severe growth restriction. A similar bias could result from excluding those exposed to maternal smoking during pregnancy.

The social and demographic characteristics of the mothers of both SGA and AGA infants were similar (Table 1), but mothers in our study were older, of higher social class, and better educated than women in a similar study of preterm infants carried out in this maternity unit (13) or when compared to the national average for age at first delivery (28), suggesting that the SGA infants recruited to this study may have come from more affluent or better educated families. Unfortunately, such details are not available for those who were eligible but not recruited. However, any such potential biases in recruitment would tend to attenuate the associations observed and lead to conservative estimates of the association between airway function and low birthweight for gestational age. A further issue when interpreting our results arises from the relatively low proportion of SGA girls in our study population who were born to nonsmoking mothers. The reason for this unexpected discrepancy is unclear. We were able to confirm, through an audit, that equal numbers of SGA boys and girls were born in the study hospitals during the period of recruitment (data not shown). However, a higher proportion of SGA girls recruited to this study were born to mothers who smoked, thus SGA girls are relatively underrepresented in this analysis based on infants of nonsmoking mothers. Hence, adjustment was made in the regression analyses to investigate differences in outcomes among SGA infants after accounting for the resulting sex imbalance.

Because the optimal method of identifying SGA infants remains unclear, two methods to classify infants' size at birth were used, namely the CGF (16) and GROW (17) algorithms. There was generally good concordance between these two methods (29), with any discrepancies falling between the 11th and 15th percentiles on one or the other chart. By including infants who were identified as SGA by either method, we hoped to avoid misclassification. Furthermore, to maintain a clear dichotomy between the SGA and AGA groups, infants between the 15th and 20th percentile according to CGF charts were not recruited into either group. However, it is recognized that the relationship between morbidity and birthweight is not a dichotomy but a graded risk, as it is dependent on the exposure to risk factors such as smoking, nutrition, and poor socioeconomic status (30).

Although maternal smoking remains one of the strongest associated factors for fetal growth restriction in developed countries (30), maternal nutrition in pregnancy has received increasing attention during recent years (31, 32). In this study, maternal characteristics such as weight at booking and other socioeconomic factors were similar in the SGA and AGA groups (Table 1). In addition, although recognizing that birthweight is influenced by birth order, notably being lower in a first pregnancy, a similar number of SGA and AGA infants studied were firstborn. Furthermore, the incidence of obstetric complications that may predispose to a growth-restricted fetus (such as pregnancy-induced hypertension) was low within the study population and was similarly distributed between mothers of SGA and AGA infants (data not shown). Although it was interesting to note that the incidence of meconium liquor was higher during SGA labors, none of the infants studied had any clinical evidence of meconium aspiration or required any ventilatory assistance during the neonatal period.

Lung Function Parameters

Although FEV₁ is the most frequently used measure of airway function in adults and older children, young infants have a rapid respiratory rate and short expiratory time, which usually preclude its measurement at this age (33). Within the current population, the time of forced expiration (TFE) ranged from 0.36 to 1.77 s (mean 0.82 s). TFE was less than 0.5 s in nine infants, but greater than 0.4 s in all but two infants who had to be excluded when calculating FEV_0.4. In those infants in whom both parameters of timed FEV could be calculated (n = 94) the difference observed between the groups was similar (data available from authors on request).

Interestingly, although both FEF₇₅ and V'maxFRC are considered to be measures of peripheral airway caliber, there was no significant difference in V'maxFRC during early infancy between the SGA and AGA infants. One possible reason for this discrepancy is that these two parameters may reflect the mechanical properties of different generations of peripheral airways. Alternatively, the relatively higher intersubject variability of V'maxFRC (Table 3), which in part reflects the variable extent to which dynamic elevation of FRC occurs during early infancy (34), may mean that it discriminates less well between groups than FEF₇₅. However, as sex differences in expiratory airflow were detected equally well by either technique, this does not appear to be the case (Table 4). The fact that observed sex differences in peripheral airway function were larger both in absolute terms and in relation to the intersubject variability for V'maxFRC than for FEF₇₅ (Table 4) again suggests that these two parameters may be reflecting different aspects of airway function. It also suggests that any decrements of airway function associated with low birthweight may be operating at a site slightly different from those associated with being male.

Association between Low Birthweight and Airway Function

In a study based on nine SGA infants, whose gestational age ranged from 33 to 41 wk, Dahms and coworkers (7) reported elevated compliance and crying vital capacity but normal functional residual capacity when compared with appropriately grown infants of similar birthweight and concluded that intrauterine stress leads to increased pulmonary maturity. However, this study was small and gestational age was determined from physical appearance and neurologic characteristics, which are less accurate than sonographic assessments. We are unaware of other attempts to assess lung function in SGA infants. A number of studies have been published reporting an association of low birthweight with diminished airway function in older children (8) and adults (5, 6, 35). With the exception of one study of British school children by Rona and coworkers (8), which included maternal report of birthweight and gestational age, other published studies (5, 6) have not taken account of information on gestational age when assessing associations between airway function and birthweight in their study population. Thus, it is unclear whether the associations reported between low birthweight and diminished airway function reflect prematurity or low birthweight for gestational age.

Although the "fetal programming" hypothesis has been suggested as a possible explanation for an association between birthweight and adult airway function, other factors such as social class at birth and various intervening social and biologic events merit consideration (36). Recent evidence from the 1958 British Birth Cohort suggests that health in later life is strongly associated with social class at birth for factors such as birthweight, childhood material circumstances, height, educational attainment, and smoking behavior (36). Thus, although Rona and colleagues (8) reported a significant association between birthweight (adjusted for gestational age) and lung function (FVC and FEV₁) in children aged 5 to 11 yr, it is possible that this association reflects the impact of other intervening exposures, related to low birthweight and impaired airway function but occurring prior to the test occasion. Although the current report focuses on a cross-sectional comparison of airway function between SGA and AGA infants shortly after birth, follow-up studies of this cohort are currently being undertaken to determine whether there is impairment of airway function with subsequent growth and development.

Sex and the Airways

Our findings suggest that peripheral airway function is reduced in boys shortly after birth. This is consistent with most previously published observations during infancy and childhood (13, 25, 37). Preliminary anatomic evidence is consistent with the finding of diminished airway function in boys. Airway structure has been shown to differ in male and female infants, and it has been suggested that the greater amount of smooth muscle and thicker airway wall in boys may provide part of the explanation for sex differences in airway function and susceptibility to respiratory disease in early life (42).

As airway function appears to be reduced in boys when compared with girls, any further decrements in airway function associated with low birthweight for gestational age might increase the risk of respiratory symptoms and the need for assisted ventilation. Given the exclusion criteria used in this study, this could have resulted in only the fittest of SGA boys being eligible for inclusion in this study. There was, however, no evidence from our audit of births that SGA boys were more likely than girls to be admitted to neonatal special or intensive care units during the study period (data not shown).

Interpretation

Fetal and early postnatal life are periods of rapid growth and development of the respiratory system. Bronchial development and airway branching are mainly complete by Week 16 of gestation (43). Thus, any insult occurring in the first few months of pregnancy may cause developmental alterations, resulting in changes to the airway branching system (44). As hypothesized by Martinez (45), minor alterations in lung structural development during fetal life may have marked postnatal consequences, leading to critical disturbances in airway caliber in response to subsequent respiratory infections and resulting in severe and potentially fatal respiratory compromise. In this study, we have found that SGA infants have diminished airway function. Although this appears to be mediated primarily through the reduction in body size, it could contribute to the increased incidence of respiratory morbidity observed among SGA infants in the first year of life (3) as infants with diminished premorbid lung function are known to be at increased risk of subsequent wheezy illnesses (14, 46). Bearing in mind the early formation of airways during prenatal development, it will be particularly important to examine the pattern of subsequent growth and development in these SGA infants to ascertain whether somatic growth is associated with a "catch up" of airway function or continuing impairment.

Conclusions

The findings of the present study suggest that airway function is diminished in infants who are small for gestational age at birth but who have not been exposed to maternal smoking. Although this appears to be mediated primarily through reductions in body size it could contribute to the increased incidence of respiratory morbidity observed in such children during early life. Furthermore, peripheral airway function as reflected by FEF₇₅ and V'maxFRC is significantly lower in boys when compared with girls. Further follow-up of this cohort is required to establish the pattern of growth and development of the airways in relation to sex, birthweight status, and subsequent somatic growth.