|Home | About | Journals | Submit | Contact Us | Français|
Premorbid infant lung function predicts childhood wheeze, but it is unclear whether lower infant lung function is most closely associated with atopic or non-atopic preschool wheeze.
To examine the association between premorbid infant lung function and preschool wheeze according to atopic or non-atopic wheeze phenotype. Additionally, to explore the relations of ADAM33 polymorphism with lung function during infancy, preschool wheeze and atopy.
Infant lung function was measured in147 healthy term infants aged 5-14 weeks. Raised volume rapid thoracoabdominal compression was performed to determine FEV0.4. Atopic status was determined by skin prick testing at 3 years and wheeze ascertained from parental questionnaires (1 and 3 years). ADAM33 polymorphisms were examined using haplotype analysis.
Early infancy V’maxFRC and FEV0.4 were lower in those who wheezed in the first year (p=0.002 and p=0.03), and lower V’maxFRC was associated with wheeze in the third year (p=0.006). Non-atopic children who wheezed in their third year of life had lower FEV0.4, compared to non-atopic children who did not wheeze (p=0.02), whilst atopic children with wheeze did not (p=0.4). No ADAM33 haplotype was associated with infant lung function, preschool wheeze or atopy after correction for multiple testing.
Lower premorbid infant lung function was present in infants who subsequently wheezed during the first and third years of life. Lower FEV0.4 was associated with non-atopic wheeze but not atopic wheeze at 3 years of age. The relation between ADAM33 polymorphism, infant lung function and preschool wheeze requires examination in larger studies.
Lower infant lung function is a risk factor for wheeze in infancy and early childhood1-4. Previous studies have demonstrated abnormalities of infant lung function are present in those children who subsequently experience wheeze and these abnormalities can be detected prior to the first wheezing illness1-4.
Lower levels of respiratory function soon after birth have been shown to precede and predict the development of wheezing illnesses during the first year of life5 and also the first three years6. The results of longer-term follow up studies are less consistent. Although tracking of lung function from infancy, through childhood7 and into adult life8 is well described, the relationship between early lung function and phenotypes of wheezing illness is less clear. The Tucson study studied 125 infants and found that reduced V’maxFRC in the first 6 months of life was associated with transient wheeze but not with wheezing after 3 years of age9. In contrast, other studies suggest early impairment of lung function is a risk factor for wheeze which persists into later childhood7;10.
The original ‘wheeze phenotypes’ derived from the the Tuscon study were defined retrospectively based on age of onset and persistence of symptoms9. In a cohort prospectively phenotyped using clinical features11 thought to be associated with later asthma, high risk infants have been found to have lower forced expiratory volumes at 8-20 months than wheezy infants at lower risk of subsequent asthma12. However, this study of high risk infants did not measure premorbid infant lung function nor establish the relationship between lung function and clinical wheeze phenotype in an unselected population. It is not clear how premorbid infant lung function relates to the atopic and non-atopic wheeze phenotypes.
Common genetic polymorphism is known to impact neonatal disease and outcome13 and reduced early life lung function may be, in part, genetically determined. ADAM33 is a positionally cloned asthma susceptibility gene that shows strongest linkage with a combined asthma and bronchial hyperresponsiveness phenotype14. ADAM33 is expressed in lung fibroblasts and bronchial smooth muscle and is expressed during embryonic lung development15. Previously, ADAM33 polymorphisms have been associated with impaired lung function at 3 years of age, specifically deficits in FEV1 and carriers homozygous for the A allele of F+1 SNP had double the risk of transient early wheeze16.
The Southampton Women’s Survey was the first to measure premorbid FEV0.4 in a longitudinal pregnancy cohort study17. Using this outcome measure, we aimed to explore the associations between premorbid infant lung function and preschool wheeze according to atopic phenotype. As a secondary aim, we explored the relationship between ADAM33 polymorphism, lung function in infancy and preschool wheeze.
The Southampton and South West Hampshire Local Research Ethics Committee approved the protocol and written consent was obtained from the children’s mothers. Detailed information regarding the participants and infant lung function testing have been described17. In brief, caucasian infants, born at at least 37 weeks gestation without major congenital anomalies and prior to any respiratory infection were recruited from the Southampton Women’s Survey (SWS), a population-based birth cohort18. Lung function was measured, between May 1999 and October 2002, when the infants were 5-14 weeks old.
Lung function was measured, with infants lying supine, in quiet sleep augmented with chloral hydrate. An inflatable jacket connected to a rapid inflation system was placed around the infant’s chest and abdomen and a leak-free facemask with Fleisch pneumotachograph (Dynasciences, Blue Bell, CA) was held over the nose and mouth. Data were collected using RASP software (Physiologic Ltd, Newbury, Berks, UK) and analysed in SQUEEZE (Paul Dixon, London).
Respiratory rate was measured during tidal breathing. As previously described17, partial expiratory flow-volume loops were obtained by performing a rapid thoracoabdominal compression (RTC) at end tidal inspiration and V’maxFRC calculated. Compliance of the respiratory system (Crs) was calculated from flow-volume curves performed at raised volumes (RV) performed by passive inflation at 30cm water. At least two acceptable, reproducible (within 10%)17 raised volume and partial expiratory curves were obtained. FEV0.4 and FVC were obtained from flow-volume curves recorded at raised volume RTC. For each infant, FEV0.4 and Crs were calculated from the best raised volume loop (FEV0.4 plus FVC) and the best V’maxFRC was used. For each outcome there were approximately 100 infants with acceptable lung function data as recording each measure of lung function was not possible for all infants.
Mothers completed contemporaneous respiratory symptom diaries during their child’s first year of life, and at ages one and three years they completed a questionnaire which included the question ‘has your child experienced any episodes of chestiness associated with wheezing or whistling in his/her chest since they were last seen?’. Skin reactivity to cata, doga, house dust mitea (Dermatophagoides pteronyssinus), milka, grass pollensa, and eggb was assessed at age 3 years (Hollister-Stier, Spokane, WAa, Alyostal, Antony, Franceb). Atopy was defined as a wheal to any allergen at least 3mm in diameter in the presence of appropriate control results.
DNA was isolated from cord blood using a salting-out procedure19. The following five SNPs were chosen for genotyping as they span the ADAM33 gene region (based on the LDU map from Simpson et al16). ADAM33 SNPs Bp1 (rs487377), Fp1 (rs511898), STp7 (rs574174), Vm3 (rs628977), V4 (rs2787094) were genotyped by TaqMan SNP genotyping assays (Table E1).
The infant lung function data were logarithmically transformed to achieve normality and adjusted for age and sex where necessary. t-tests were used to assess differences between geometric mean group values of infant lung function measurements according to wheeze status at 1 and 3 years, and after subdividing the 3-year wheeze group according to atopic status. Agreement between the parent-completed symptom diaries and the nurse-administered questionnaires was assessed by weighted Kappa (κ) analysis.
Assuming 50 children in each outcome group, there is 80% power at the 5% level of significance to detect a difference of 0.57 SDs in any infant lung function measurement between the two outcome groups. Matched genotype and lung function data were available for 103 children for respiratory rate, 82 for Crs, 105 for V’maxFRC and for 76 children for FEV0.4.
147 infants had lung function measurements, of which 146 provided questionnaire data at age one year, 141 provided questionnaire data at 3 years of age and 113 had skin prick testing. Symptom diaries were returned by 102 mothers. There was strong concordance between the symptom diaries and questionnaires for the reported incidence of wheeze, (κ =0.71, p<0.001). Given this, all subsequent analyses are based upon questionnaire data. A total of 114 DNA samples were available for ADAM33 SNP genotyping; genotype assignment approached 99% (6 failed PCR reactions out of a total 570) with a maximum of two unassigned genotypes for each SNP and all genotype frequencies were in Hardy-Weinberg Equilibrium. (See online supplement).
During the first year of life, 74 infants had at least one episode of wheezing. Thirty three children experienced wheeze in the third year of life, of these 23 had also experienced wheeze in the first year of life whilst 10 had not. Of the 113 children who were skin prick tested for allergy, 20 (17.7%) had at least one positive reaction and were classed as atopic. Of the children who experienced wheeze 75% were atopic and 25% were not.
The demographic characteristics of those children reporting wheeze at 3 years and those that did not are listed in table 1. Comparing children who did and did not wheeze, there were no differences in gender or the mother’s age, atopy but those who wheezed at 3 years were more likely to have a mother who smoked during pregnancy. However, those who wheezed in the first year of life were more likely to be born to younger mothers and have mothers who had smoked in pregnancy than those who did not wheeze.
Associations between wheeze or wheeze phenotype and lung function were examined after adjusting for factors that were found to influence measures of lung function on univariate analysis. FEV0.4, and V’maxFRC were adjusted for gender and Crs for gender and age at testing.
V’maxFRC, FEV0.4 and Crs were lower in infants who subsequently wheezed in their first year than in those who did not by 22.5% (p=0.002), 8.1% (p=0.03) and 6.8% (p=0.04) respectively (Table 2). The group mean respiratory rate was 6.8% higher in the wheeze group than in the non-wheeze group (p=0.03).
V’maxFRC was 23.2% lower in infants who wheezed in their third year than in those who did not (Table 3) (p=0.006). However, there were only small non-significant differences in FEV0.4, Crs and respiratory rate between the two groups. When non-atopic children were considered separately from atopic individuals, FEV0.4 was 11% lower in those who wheezed (p=0.02) (Table 4). There were no significant differences in any measure of infant lung function when atopic children with wheeze in the third year were compared to atopic children without wheeze (Table 5).
No ADAM33 SNP haplotype was significantly associated with infant lung function measurements after adjusting for multiple testing by Bonferroni correction (Table 6). However, one rare ADAM33 haplotype (GACTG; Hap Freq = 0.014) indicated a possible association with increased Crs measurements (Tables 6 and E4; corrected P value = 0.062). Haplopower analysis (Online Data Supplement) revealed that with the small sample size (n = 82) we were underpowered (Power = 19%) to detect a common haplotype (freq = 18%) association accounting for 5% of the variance in the mean of the Crs measurement.
This study demonstrates an association between impaired premorbid infant lung function and preschool wheezing illness. This confirms the results of previous studies4;6. Our analyses show for the first time that lower lung function in early infancy is a risk factor for non-atopic wheeze, rather than atopic wheeze. Additionally, this study is the first to report reduced FEV0.4, prior to lower respiratory infection, as a risk factor for wheeze in the preschool years.
Premorbid measures of forced expiratory volume in 0.4 seconds (FEV0.4) during infancy as a predictor of wheeze have not previously been studied, though reduced FEV0.5 following symptoms has been reported in wheezy children aged 8 – 20 months12. Lung function measures derived from raised volume expiratory manoeuvres are believed to discriminate better between healthy individuals and those with respiratory disease when compared to tidal breathing measures22;23. FEV0.4 provides a more robust measure of infant lung function than V’maxFRC24. Furthermore, FEV0.4 is well suited to longitudinal study as it is intuitively more comparable to the measures of lung function, such as FEV1, which are used in older children and adults.
The results of this study may be of relevance to the general population. We recruited healthy infants and only excluded those born before 37 weeks gestation to prevent bias due to effects of prematurity upon lung development. Our results demonstrate that variations in measures of expiratory flow measured in unselected populations are associated with preschool wheeze.
Small study numbers limited the exploration of lung function measures according to wheeze phenotype. However, a significant difference in FEV0.4 was noted in non-atopic children who wheezed during the third year of life compared to non-atopic children without wheeze. There was no difference in any measure of infant lung function when atopic children with wheeze in the third year were compared to non-atopic children without wheeze. This may have been partly related to greater power associated with a more balanced distribution of non-atopic than atopic individuals between the wheeze and no wheeze groups. However, others have speculated that wheezing during the first year of life is often a transient condition due to reduced airway caliber and that wheeze beginning or persisting into later childhood is more likely to be due to the wheeze phenotype recognized as asthma4.
Due to the small numbers involved it was not possible to analyse our data according to age of onset. However, the broader confidence intervals and divergence of results between atopic and non-atopic individuals at age three years may be evidence of comparative heterogeneity of wheeze phenotypes later in childhood. The strong association between reduced measures of forced expiratory flow and wheeze at one year may reflect a common underlying mechanism, similar to that proposed by Martinez, whereby initial airway diameter, length or other characteristics might predispose certain infants to wheezing with viral respiratory infections3. With growth this may resolve in some infants as airway size increases but for those with the smallest airway dimensions the risk of non-atopic wheeze may persist. In contrast, airway dimensions at birth may be less significant in the aetiology of atopic wheeze compared to other factors acquired in association with an atopic phenotype.
As described earlier, ADAM33 is expressed in the embryonic lung15 and SNPs in the gene encoding ADAM33 were previously found to predict impaired lung function at age 3 and 5 years16. However, it is unclear whether ADAM33 polymorphism is associated with abnormal lung function at birth (reflecting altered in utero lung development) or whether post-natal gene-environment interaction alters lung function and increases risk of asthma. Recent observations from the PIAMA birth cohort show that in utero, but not post natal, cigarette smoke exposure interacts with ADAM33 polymorphism to determine childhood lung function and BHR would suggest that ADAM33 polymorphism may alter in utero lung development25. In the current study, no ADAM33 haplotype was associated at the 5% level with lung function or symptoms in the first or third years of life after correction for multiple testing, however one rare haplotype was significantly associated with Crs uncorrected P value = 0.0048). Simpson et al. described that the F+1 polymorphism explained 3% of the variance in lung function (sRaw) at 3 years of age16; power calculations show that we only had 19% power to detect an association that predicts 5% variance in Crs measurements. Given the observation of interaction with in utero smoke exposure and the trend towards association seen in the current study, the association between ADAM33 haplotypes and infant lung function needs to be examined in a meta-analysis of data from all available infant lung function cohorts.
In summary this study has shown a significant relationship between early life lung function including FEV0.4 and respiratory symptoms in the first and third years of life. This is further evidence suggesting that pre-natal factors contribute to the development of wheeze in early life.
Reduced infant lung function is a risk factor for childhood wheeze, but it is not known whether lower premorbid lung function is most closely associated with atopic or non-atopic wheeze. Polymorphisms in ADAM33 have been shown to predict lung function at 3 years of age, but the relationship between such polymorphisms and infant lung function has not previously been explored.
This study demonstrates that children who wheeze at 1 or 3 years have lower premorbid infant lung function than children who do not wheeze, particularly if they are non-atopic. Within this study, ADAM33 polymorphisms were not related to lower lung function in first few years of life.
The authors acknowledge the help of the parents and infants who participated in this study. They are grateful to Claire Foreman and the staff of the Wellcome Trust Clinical Research Facility for collection of the infant lung function measurements and the staff of the Southampton Women’s Survey for collecting and processing the questionnaire data.
Page one KCP is supported by a grant from the British Lung Foundation
JWH is supported by The Medical Research Council UK, and the Asthma, Allergy & Inflammation Research Charity.
MJR-Z was supported by the Asthma, Allergy & Inflammation Research Charity and a research studentship from the Infection, Inflammation and Immunity Division, School of Medicine, University of Southampton
Follow-up of children in the Southampton Women’s Survey has been funded by the Medical Research Council, University of Southampton, British Heart Foundation, and the Food Standards Agency. The measurements of infant lung function were funded by British Lung Foundation, SPARKS (Sport Aiding medical Research for Kids) and Hope.
KCP does not have a financial relationship with a commercial entity that has an interest in the subject of this manuscript. MJR-Z does not have a financial relationship with a commercial entity that has an interest in the subject of this manuscript. ECO does not have a financial relationship with a commercial entity that has an interest in the subject of this manuscript. HMI does not have a financial relationship with a commercial entity that has an interest in the subject of this manuscript. KMG does not have a financial relationship with a commercial entity that has an interest in the subject of this manuscript. SRC does not have a financial relationship with a commercial entity that has an interest in the subject of this manuscript.
GCR does not have a financial relationship with a commercial entity that has an interest in the subject of this manuscript. JBC does not have a financial relationship with a commercial entity that has an interest in the subject of this manuscript. JWH does not have a financial relationship with a commercial entity that has an interest in the subject of this manuscript. JSL does not have a financial relationship with a commercial entity that has an interest in the subject of this manuscript.