American Journal of Respiratory and Critical Care Medicine

Rationale: After recent standardization of forced expiratory maneuvers for both infants and preschool children, longitudinal measurements are now possible from birth.

Objectives: The aim of this study was to investigate the evolution of lung function during the first 6 years of life after a clinical diagnosis of cystic fibrosis (CF) in infancy in children with CF and in healthy control subjects.

Methods: The raised volume technique was used during infancy and incentive spirometry during the preschool years.

Measurements and Main Results: Forty-eight children with CF and 33 healthy control subjects had up to seven (median, 3) measurements. Over these early years, the diagnosis of CF itself accounted for a significant mean reduction of 7.5% (95% confidence interval, 0.9 – 13.6%) in FEV0.75 and 15.1% (95% confidence interval, 3.6 – 25.3%) in FEF25–75. Wheeze on auscultation, recent cough, and Pseudomonas aeruginosa (PsA) infection (even if apparently effectively treated) were all independently associated with further reductions in lung function. Premorbid lung function did not predict infection with PsA.

Conclusions: This is the first study to describe physiologic measurements from infancy through the preschool years in subjects with CF and healthy control subjects, the understanding of which is critical for future intervention trials. Airflow obstruction in uncomplicated CF persists through the preschool years despite treatment, with PsA acquisition being associated with further deterioration in lung function, even when apparently eradicated. This suggests that new therapies are needed to treat the airflow obstruction of uncomplicated CF, and rigorous strategies to prevent PsA acquisition.

Scientific Knowledge on the Subject

Lung function has been found to be reduced in infants with cystic fibrosis. Spirometry, now standardized for preschool children, allows longitudinal testing from infancy to school age.

What This Study Adds to the Field

Lung function was persistently reduced in patients with cystic fibrosis, particularly in patients with either current or previous Pseudomonas infection.

Airway inflammation and infection in patients with cystic fibrosis (CF) cause considerable morbidity, and respiratory failure accounts for more than 90% of deaths. It has been shown that the infection–inflammation cycle starts in early childhood (13). The London CF Collaboration (LCFC) has previously shown that infants with CF have evidence of airflow obstruction shortly after diagnosis, independent of previous respiratory illness, positive bacterial airway cultures, administration of antibiotics, or the presence of respiratory symptoms or signs at time of testing (4, 5). This obstructive defect persisted over the next 6 months, despite treatment in specialist CF centers (6). The raised volume technique (RVT), which produces forced expiratory curves similar to the expiratory flow–volume curve of standard spirometry, has been standardized (7) and has been shown to be a sensitive method of identifying reduced lung function in infants with CF (5, 6, 8, 9). Recent modifications of spirometric techniques to facilitate their application in the preschool age group mean that it is now possible to undertake continuous measurements of forced expiratory flows and volumes (FEFV) from infancy through to school age (1013). However, little is known about the evolution of lung function in young children with CF. One group has reported a correlation between FEFV variables in infancy and preschool age in 14 children with CF, but serial measurements were not available in healthy control subjects (14). Others have shown a persistent elevation of specific airway resistance from the preschool years into early school age, but did not obtain such measurements in infancy (15).

Changes of lung function over time are clearly important; if the CF-related airway defect cannot be reversed or further deterioration prevented with current treatment, new therapeutic strategies must be sought. We hypothesized that airflow obstruction, determined by FEFV shortly after clinical diagnosis in children with CF, would persist into the preschool years, despite treatment. Furthermore, we hypothesized that complications of CF, such as Pseudomonas aeruginosa (PsA) infection and other clinical factors predictive of poor lung function would additively impact on airflow obstruction.

Some of the results of these studies have been previously reported in abstract form (1618).

Study Population
Infants with CF.

Children diagnosed with CF by sweat test or CFTR mutation analysis before their second birthday were recruited into the LCFC study between January 1999 and December 2002 (46). The LCFC comprises five pediatric CF centers (Great Ormond Street Hospital for Children, Kings College Hospital, Lewisham University Hospital, The Royal Brompton Hospital, and Barts and The London Children's Hospital). Children were excluded if they had congenital cardiorespiratory or neurological abnormalities. Newborn screening did not become available in London and the southeast of England until July 2007. All children were treated according to national and European standards of care with regard to Staphylococcus prophylaxis, PsA infection treatment, and clinical follow-up (1921). Staphylococcal prophylaxis consisted of an oral antistaphylococcal antibiotic that was discontinued by the age of 5 years unless the child had become chronically infected with this organism. PsA infection treatment was a combination of oral ciprofloxacin and a nebulized antibiotic, usually colistin, for 3 months, unless the child was clinically unwell and so received intravenous ceftazidime and an aminoglycoside for 2 weeks, followed by 3 months of nebulized colistin.

Healthy control subjects.

Healthy infants born at Homerton or University College London Hospitals were recruited as part of an epidemiologic study (22, 23). Children were ineligible if they had required hospitalization for respiratory illness, had a history of wheeze before recruitment, had any congenital abnormalities, required ventilation at birth or were born at less than 36 weeks' gestation.

Follow-up at Preschool Age

Families of children recruited as above were invited to return for follow-up between September 2002 and April 2005, when the child was between 3 and 5 years of age. Exclusion criteria for children with CF and healthy control subjects included the following: neuromuscular disease, obstructive sleep apnea, or cardiac disease (recently operated on or requiring surgery). Healthy control subjects were also excluded if they had been hospitalized for respiratory illness (e.g., croup, pneumonia, bronchiolitis), had doctor-diagnosed asthma, were currently using inhaled bronchodilators, or had ever used inhaled steroids. On the day of testing, a detailed family history, including whether the mother currently smoked or had done so during pregnancy, was obtained and the family asked if the child had coughed or wheezed in the preceding week. All children were examined for any respiratory signs (crackles or wheeze), and a cough swab for microbiological culture was obtained from the children with CF.

Clinical Assessment of Children with CF

Information about mode of presentation, age at presentation, genotype, number of courses of intravenous antibiotics for respiratory exacerbations, results of all cough swab cultures, and date of first growth of PsA was obtained from the child's clinical records. Each visit was coded according to whether PsA had been grown from cough swabs before that visit. Retrospectively, because the classification was not available at the commencement of the study, children were subdivided into PsA subgroups: (1) never grown, (2) previously grown but now “free” or (3) “still” grown, either intermittently or chronically (24). A child was defined as growing PsA intermittently when less than 50% of the upper airway cultures taken in the child's final year within the study were positive for PsA and as growing PsA chronically if more than 50% of the cultures were PsA positive during this period.

Written, informed consent was obtained from the parents of all children. Ethical approval was obtained from the North Thames Multicentre Research Ethics Committee and the local research ethics committees of participating hospitals.

Measurement of Lung Function

All lung function measurements were performed in the pediatric respiratory laboratory at the UCL Institute of Child Health, London. Testing was postponed for 3 weeks after any respiratory exacerbations in children with CF or upper respiratory tract infections in either group.

After sedation with chloral hydrate (50–100 mg/kg), measures of FEFV were obtained using the RVT, according to international guidelines, in sleeping, young children less than 2 years of age (hereafter referred to as “infant lung function tests”). In this technique, the lungs are passively inflated toward total lung capacity, before inflating a thoracoabdominal jacket to force expiration (5, 7).

Preschool children performed multiple-breath inert gas washout, specific resistance measurement by plethysmography, and incentive spirometry (Jaeger MasterScope spirometer; VIASYS Healthcare, Hoechburg, Germany), in that order, according to internationally accepted guidelines (10, 12, 13). Longitudinal data from infancy were, however, limited to those from FEFV maneuvers and these form the basis of this report.

FVC, forced expired volume in 0.5 second (FEV0.5), and FEF between 25 and 75% of FVC (FEF25–75) were reported if at least two technically acceptable curves were obtained (10). FEV0.75 and FEV1 were also calculated when possible (5, 10).

Repeated Visits

The minimum inclusion requirement into this longitudinal study was technically satisfactory FEFV results on at least one occasion during the first 2 years of life and on at least one occasion during the preschool years. Valid results from all visits were included in the multivariable analysis.

Statistical Analysis

Height, weight, and body mass index (BMI) were converted to z scores and unpaired t tests (SPSS for Windows, version 11.0; SPSS, Inc., Chicago, IL) used to assess changes in somatic growth between children with CF and healthy control subjects during the study period (25). FVC, FEV0.5, and FEF25–75 results from the first infant visit were converted to z scores to allow comparison of baseline lung function (unpaired t tests) between children who were and were not followed up into preschool age (26).

Multilevel multivariable linear regression modeling (MLwiN, version 2.12; Institute of Education, Bristol, UK) was used to compare changes in lung function in the first 6 years of life between healthy control subjects and children with CF while accounting for factors known or suspected to influence lung function, including sex, maternal smoking status (during pregnancy and current), birth weight, gestational age, height, weight, BMI, and age (27, 28). These highly flexible models adjust for the correlated nature of repeated measurements in individuals and allow inclusion of variable numbers of measurements per child to provide the most precise characterization of changes over time (2729). FEFV variables, height, and weight were logged before modeling. Initially, the univariable relationship between each FEFV measure (FVC, FEV0.5, FEV0.75, FEV1, and FEF25–75) and potential explanatory variables was examined. In addition to age as a continuous variable, a factor was also included to denote whether infant (sedated) or preschool (awake) lung function tests were used. This variable quantified the extent to which differences in technique, equipment, or measurement conditions could account for differences over and above any general age trend.

A multivariable model was used to quantify the extent to which the effects attributable to CF were independently associated after accounting for other factors. Each FEFV outcome measure was modeled separately. CF disease and prior growth of PsA were both included in the model, whereas other variables were only retained if they made a significant contribution (P < 0.05) to the multivariable model for each specific FEFV outcome measure. Potentially relevant, although previously nonsignificant, factors were also added to ensure these did not make a significant contribution once other factors had been adjusted for (30).

A further model for each of the five FEFV outcome measures was created in an identical manner, after substituting PsA subgroups for CF disease and prior growth of PsA (24).

This study was powered to identify clinically significant group differences between those with CF and healthy control subjects. Assuming factors known to influence lung function in health account for at least 40% of the variation in lung function (31), then 35 per group would be required to determine whether CF per se or PsA status independently account for a further 10% of variation in lung function with 90% power at the 5% significance level (32).

Study Participants

During the study period, infant lung function was measured in 70 infants with CF, 52 (74%) of whom returned for at least one preschool visit (Figure 1). Four children could not produce technically acceptable FEFV loops at any preschool visit, leaving 48 children with CF with at least paired infant and preschool results. Forty-two healthy control subjects with technically acceptable infant data were invited for follow-up. Of the 35 (83%) who attended, 2 were unable to produce technically acceptable data, leaving 33 healthy control subjects with paired infant and preschool lung function data (Figure 1). Each group had a median of three visits (range, 2–7 children with CF; 2–6, healthy control subjects). FEFV data from 129 infant lung function visits (85 CF, 44 healthy control subjects) and 123 visits during the preschool years (79 children with CF, 44 healthy control subjects) were available for longitudinal analysis using multivariable modeling.

Baseline characteristics and infant lung function were compared in children with CF who were and who were not followed up (Table 1). All the children with CF who were followed up were pancreatic insufficient. Age at diagnosis and the proportion of boys were lower among those not attending for follow-up, but these differences were nonsignificant. There were no significant differences in genotype, mode of presentation, age, body size, or lung function at the time of the first infant lung function tests between the two groups. Similarly, there were no differences in either infant lung function or background characteristics for the healthy control subjects who were and were not followed up (data not shown).


With Paired Results

No Paired Results
Cystic Fibrosis
Total number4822
Age diagnosis, wk*9.4(0–87.4)6.1(0–64.6)
 Homozygous ΔF508/ΔF5083266.71150.0
 Heterozygous ΔF508/other1020.8940.9
Mode presentation
 Meconium ileus and/or antenatal bowel pathology1837.51150.0
 Failure to thrive816.7418.2
 Chest infections612.529.1
 Chest infections and failure to thrive1429.2418.2
 Family history and asymptomatic24.214.5
Data from first infant visit
 Age, y0.7 (0.4)0.7 (0.4)
 Height z score−0.4 (1.4)−0.1 (1.3)
 Weight z score−1.3 (1.5)−0.9 (1.1)
 FVC z score−1.8 (1.4)−1.7 (1.0)
 FEV0.5 z score−2.0 (1.8)−2.1 (1.3)
 FEF25–75 z score
−1.4 (1.9)
−1.8 (1.5)

*Median (range).

Mean (SD).

z scores for forced expiratory flows and volumes were calculated using reference equations in Reference 26.

Children with CF were similar to healthy control subjects with respect to gestational age, birth weight centiles, parental atopic status, prenatal smoke exposure, and parental occupation (Table E1 in the online data supplement). Among the mothers who smoked during pregnancy, five (4 healthy control subjects, 1 mother of a child with CF) had stopped by the time the child attended for preschool tests, whereas one mother of a child with CF who had not smoked in pregnancy was smoking by the time of follow-up. Four (8.3%) of the children with CF and three (9.0%) of the healthy control subjects were of non-Caucasian or mixed ethnic origin. At the time of the first visit, despite being slightly older, children with CF were shorter and lighter than the healthy control subjects (Table 2). The height and age at time of each visit for those with and without CF is shown in Figure E1. By the end of the study, there had been a significant improvement in weight and BMI z scores in the children with CF, but they remained shorter than the healthy control subjects (Table 2).


First Visit

Change Over Study Period (Last Visit–First Visit)

95% CI (CF–HC)
95% CI (CF–HC)
Boys, n (%)21 (44)16 (48)
Age, yr0.6 (0.4–1.0)*0.2 (0.1–0.6)*0.1 to 0.53.9 (1.0)4.3 (0.8)−0.8 to 0.1
Age range, yr0.14 to 1.770.09 to 1.92
Weight z score−1.3 (1.5)−0.2 (1.0)−1.6 to −0.51.1 (1.3)0.5 (1.0)0.1 to 1.1
Height z score−0.4 (1.4)0.2 (0.9)−1.2 to −0.10.0 (1.3)0.0 (0.9)−0.5 to 0.5
BMI z score
−1.6 (1.5)
−0.6 (1.2)
−1.6 to −0.4
1.7 (1.4)
0.8 (1.2)
0.4 to 1.5

Definition of abbreviations: BMI = body mass index; CI = confidence interval; CF = children with cystic fibrosis; HC = healthy control subjects.

*Median (interquartile range).

P < 0.01.

P < 0.05.

Twenty of the children with CF presented with respiratory symptoms. These children were similar in terms of height, weight, and FEFV z scores at the time of the first lung function test, but were older than the 28 children who did not present with respiratory symptoms (Table E2).

Of the children with CF, 22 (46%) had had intravenous antibiotics for a respiratory exacerbation. The median number of courses was one (range, 1–9) per child. These 22 children had a total of 78 visits to the lung function laboratory, with a maximum of five courses of intravenous antibiotics for respiratory exacerbations between visits.

Among the children with CF, six had wheeze on auscultation on a total of 11 (8 infant) lung function test visits. No parent reported wheezing in the week preceding testing, because under these circumstances testing would have been deferred. Six children had crackles on auscultation on a total of seven (2 infant) lung function test visits. Cough in the week before testing, ranging from an occasional dry cough with chest physiotherapy to a chronic wet cough, was reported by parents of 37 children on 79 test visits.

By the end of the study, 32 (67%) children with CF had grown PsA on cough swabs, 3 of whom grew mucoid strains. The median age of the first PsA growth was 1.4 (range, 0.7–4.8) years. Eight (17%) children first grew PsA before their first lung function test. Twenty children (42%) had grown Staphylococcus aureus (1 sample of which was methicillin-resistant S. aureus) by the end of the study (median age of 1.5 [range, 0.2–4.6] yr at first isolation). Nineteen (39%) children had grown Haemophilus influenzae by the end of the study (median age when first isolated, 1.9 [range, 0.4–4.6] yr).

Lung Function

Logged height, logged weight, age, BMI z score, gestation, maternal smoking status, type of lung function test (infant or preschool), a diagnosis of CF, genotype, mode of presentation, presence of wheeze on auscultation, presence of crackles on auscultation, cough in the preceding week, and previous PsA infection were univariably associated with all five FEFV outcome measures.

When potential demographic and clinical predictors were inserted to the multivariable model (Table E3), logged height was the strongest predictor for all five lung function indices. The longitudinal association between each FEFV measure and height according to disease status is shown in Figure E3. Type of lung function test (infant vs. preschool) explained some of the variability in FEV0.5 but not in any other measure (Table E3). CF status and previous PsA infection were included in the model for all the FEFV measures (Table E3).

After adjustment for height, a diagnosis of CF per se (Table 3) accounted for a significant mean reduction of 7.5% in FEV0.75 and 15.1% in FEF25–75, but not in FVC, FEV0.5, or FEV1 when compared with healthy control subjects studied over the 6 years. This can also be seen in Figure E2, where, despite considerable within-subject variability in the rate of increase in lung function with somatic growth, especially with respect to FEF25–75, results from many of the children with CF lie below those from the healthy control subjects.


FVC (n = 249)

FEV0.5 (n = 252)

FEV0.75 (n = 225)

FEV1 (n = 174)

FEF25–75 (n = 240)
CF−2.6 (−9.0 to 4.4)−4.3 (−11.0 to 2.9)−7.5 (−13.6 to −0.9)−7.1 (−14.0 to 0.2)−15.1 (−25.3 to −3.6)
PsA−10.1 (−15.9 to −4.0)−7.1 (−13.1 to −0.7)−9.0 (−14.8 to −2.7)−10.9 (−17.4 to −3.8)−4.1 (−14.4 to 7.4)
Wheeze on auscultation*−19.7 (28.9 to −9.4)−19.2 (−28.0 to −9.3)−19.7 (−29.0 to −9.1)−33.1 (−45.8 to −17.5)
Cough in the preceding 7 d

−6.5 (−11.6 to −1.0)

−14.2 (−22.2 to −5.4)

Definition of abbreviations: CF = cystic fibrosis; PsA = Pseudomonas aeruginosa.

*Six children on 11 test occasions.

Thirty-seven children on 79 test occasions.

All results are expressed as mean (95% confidence interval) percent change compared with the healthy controls. Percent change indicates overall change attributable to the clinical variable (CF, PsA, cough, or wheeze) that was included within the model for each FEFV measure after adjustment for height (Table E3 in the online supplement). Because wheeze on auscultation had no significant effect on FVC and cough did not influence FVC, FEV0.75, or FEV1, they were not retained in the model for these variables and hence appear as gaps within the table. For PsA status, each visit was coded according to whether PsA had been grown from cough swabs before that visit and the figures for PsA indicate the percent change in FEFV if PsA had been grown before any specific set of lung function tests. Significant changes are shown in bold (P < 0.05). In the multilevel multivariable model, variables have a multiplicative effect. For example, if CF was associated with a 7.5% decrease in FEV0.75 and prior PsA with a 9.0% decrease, then FEV0.75 in a child with CF who had previously grown PsA would on average be 100 times (0.925 × 0.91) or 84% of that in a healthy control of similar height (i.e., have a 16% reduction).

Growth of PsA before any lung function test during the study independently accounted for a further reduction in all FEFV parameters, except FEF25–75 (Table 3). Due to the multiplicative effect of variables within the model, children with CF who had grown PsA before any specific test occasion had a total mean reduction of 13% in FVC, 11% in FEV0.5, 16% in FEV0.75, 17% in FEV1, and 19% in FEF25–75, compared with the healthy control subjects of similar height.

Wheeze on auscultation was independently associated with a significant further reduction in all FEFV parameters except FVC after adjustment for the above factors (Table 3). Crackles on auscultation were not independently associated with a reduction in any of the FEFV parameters and were therefore not included in the model. Cough in the preceding week was associated with diminished FEV0.5 and FEF25–75 (Table 3). The total number of courses of intravenous antibiotics before each lung function test was not independently associated with any of the FEFV measures and so not included in the model. Presentation with chest symptoms did not account for any further significant reductions in any of the FEFV measures. Figure 2 illustrates the different predicted values for the children in the different diagnostic subgroups. The lines show predicted values of FEV0.75 and FEF25–75 (from the models given in Table E3) for each subgroup against height. Although the absolute difference increased with height, the percentage differences in FEFV measures remained constant with growth.

PsA Subgroups

Low lung function at the time of first infant test was not associated with subsequent PsA infection as shown by the similar results between the 16 children with CF who never grew PsA and the 24 children who grew it after their first infant visit (Figure E3).

Post hoc analysis showed that, of the 32 children with CF who had ever grown PsA, 19 had grown PsA but had ceased to isolate this organism by the end of the study, whereas 13 continued to grow PsA (9 intermittently and 4 chronically) (24). Five children had grown a mucoid strain of PsA—one chronically, two intermittently, and two were “free” of mucoid PsA at the end of the study.

Substitution of CF disease and PsA status before the lung function test with PsA subgroups (Table E4) showed that the 19 children who had grown PsA but had ceased to isolate this organism by the end of the study had a mean reduction of 9% (95% confidence interval [CI], 1–16%) in FVC, 11% (95% CI, 3–18%) in FEV0.5, 14% (95% CI, 7–21%) in FEV0.75, 16% (95% CI, 8–24%) in FEV1, and 19% (95% CI, 6–30%) in FEF25–75, compared with the healthy control subjects. Although these changes were greater than those seen in children who had never grown PsA, they were similar to those observed in the 13 children who still grew PsA at the end of the study, in whom mean reductions of 9% (95% CI, 1–17%) in FVC, 8% (95% CI, 1–16%) in FEV0.5, 13% (95% CI, 5–20%) in FEV0.75, 13% (95% CI, 4–21%) in FEV1, and 12% (95% CI, 3–25%) in FEF25–75, were observed when compared with the healthy control subjects.

To our knowledge, this is the first prospective longitudinal study using objective physiologic outcome measures from forced expiratory maneuvers through the so-called silent period of infancy and the preschool years in both healthy control subjects and children with CF. Over the first 6 years of life, CF per se accounts for a significant reduction in FEV0.75 and FEF25–75. Growth of PsA, wheeze on auscultation, and recent cough are all independently associated with further reductions in lung function. Children who isolated PsA had similar lung function before isolation to those who did not, so any PsA isolation was associated with deterioration, rather than just being a marker of prior poor respiratory status. The negative impact of PsA on lung function was evident whether or not it was still isolated on upper airway cultures before the last preschool lung function test.

Strengths and Limitations of the Study

We successfully retested 48 of 61 (79%) subjects of our CF cohort who were old enough to participate in the preschool studies and these were representative of the entire cohort. These children were recruited from collaborating CF centers in one geographical area, applying similar protocols for management after a clinical diagnosis. Newborn screening in this area was only introduced in July 2007 (33). Some attrition is inevitable in any longitudinal study, as it is difficult to obtain repeated infant and preschool lung function measurements in healthy control subjects due to the mobility of the local population and time-consuming nature of the tests. Although 28% of healthy control subjects whom we contacted were unwilling or unable to reattend, there were no differences in background characteristics and initial lung function results between those who were and were not included in this follow-up. When studying children with CF diagnosed clinically, it is virtually impossible to recruit a prospective control group that is matched for both age and body size, particularly when undertaking repeated measurements. Nevertheless, by studying a prospective control group and using longitudinal multivariable modeling, it is possible to analyze repeated measures and adjust for numerous factors other than CF disease, including age and body size, that influence lung growth and development during early childhood. Another advantage of using multivariable modeling is that the variability of longitudinal changes in outcome and growth is accounted for within the comparisons of the CF group with serially measured healthy control subjects. This is important because improvement in weight from below the 10th centile at age 3 years to above the 10th centile at age 6 years in children with CF has been shown to be associated with better FEV1 and FEF25–75 at age 6 (34). Although weight and BMI were significantly associated with FEFV parameters in univariable analysis, height was by far the strongest predictor and the only measure of body size that remained significant in multivariable analysis. By adjusting for height in all the models, any differences in lung function attributable to differing growth rates within or between the groups was accounted for.

Considerable care was taken to ensure that all children with CF were measured during a stable period. There were no parental reports of current wheezing at the time of testing, although six children were found to be wheezy on examination on 11 occasions. Despite this finding, the children were tested, because they were otherwise well and parents had taken time off work to travel into central London and consequently appointments were difficult to rearrange. Although the use of bronchodilators in this population is controversial, there is evidence that some infants and young children with CF may experience acutely improved lung function after bronchodilator use (3537). It is therefore possible that the observed decrease in flows could be at least partially attributed to alterations in airway tone. This was not assessed in this study due to time constraints.

Although the children were recruited from the five different pediatric CF centers in London, all lung function measurements were undertaken in the same center, by the same personnel, using identical equipment and protocols. Sample size restricted statistical comparisons between centers, but tabulation and graphical display did not reveal any apparent bias in results according to referral center.

Many of these children were followed up on a shared care basis between the LCFC centers and peripheral hospitals, and all centers followed standard published guidelines for staphylococcal prophylaxis and PsA eradication treatment (1921). A potential weakness of this study was the lack of regular, protocol-driven bacteriology studies. This meant that cough swabs were not necessarily taken at regular intervals and laboratory processing may not have always used the most appropriate selective media. More frequent cultures may have increased identification of PsA infection among asymptomatic children. In addition, there was no policy of routine, timed bronchoscopy and lavage at the time this study was conducted, which could have resulted in increased identification of organisms. The effect of either scenario would, however, have been to allocate children incorrectly to the noninfected group, making it more difficult to detect any adverse effects of PsA. The fact that such effects were nonetheless detected makes it more, not less, likely that the findings are indeed genuine. We cannot exclude the possibility that bronchoscopy would have shown that PsA was still present in some of those in whom it was not isolated from upper airway cultures, but upper airway cultures are currently the standard clinical tool, and a negative culture for PsA in this age group has a good predictive value for a negative bronchoalveolar lavage culture (38).

Subdivision into PsA subgroups was performed post hoc because the criteria (based on cough swab and sputum cultures) were not published until after this study had been designed (24). Significant reductions in lung function were seen both in children who had previously isolated PsA but had ceased to do so, and those who continued to isolate PsA when compared with the healthy control subjects. Due to the small number who were chronically infected with PsA (n = 4) and the fact that some children had more visits for lung function than others (range, 2–7), the two subgroups infected with PsA (intermittently and chronic) were combined for the purposes of this analysis.

The pattern of changes observed according to clinical history is intriguing. Whereas there was no change in FVC in those with uncomplicated CF, it was decreased in those who had ever grown PsA (Table 3). There was also an additional reduction in all the FEV parameters in this group, suggestive of more severe airway disease. It is therefore likely that the observed reduction in FVC is secondary to such changes and represents an elevation of residual volume (RV) due to airway closure at low lung volumes. The fact that there was no additional reduction in flows in the subgroup with prior PsA could reflect the fact that, in the presence of an elevated RV, flows will be measured at a relatively higher lung volume. By contrast, in the presence of recent symptoms (cough and wheeze), the additional changes in forced flows and volumes probably reflect more acute changes in airway caliber due to inflammation.

Comparison to Published Literature

There have been several studies of lung function in either infants or preschool children with CF, but to our knowledge only Marostica and colleagues have followed up children with CF from infancy to the preschool years (14). These authors found a correlation between FEV0.5 z scores in infancy and FEV1 z scores in the preschool years, as well as between FEF25–75 z scores among the 14 children with CF studied on both occasions, but did not undertake longitudinal studies in healthy children (14). Because work is currently in progress to develop more reliable reference equations and hence z scores for preschool spirometry (39), and factors influencing the transition of such z scores between infant and preschool tests have yet to be determined, for this study we applied robust longitudinal methods of statistical analysis to compare changes in lung function with growth between children with CF and healthy control subjects during the first 6 years of life. Using this approach, we found FEV0.75 and FEF25–75 to be persistently reduced in those with CF, independently of cough, wheeze, and PsA status. After adjusting for height and clinical factors, there was a significant reduction in FEV0.75, but not in FEV0.5 or FEV1, among those with CF.

In contrast to its value during infancy, FEV0.5 appears to discriminate poorly between healthy preschool children and those with CF or asthma (4, 6, 11, 40, 41). This probably reflects developmental changes in the expiratory time constant. During infancy, the airways are relatively large in relation to lung volume and lung emptying occurs rapidly during forced expiration (often < 1 s). Consequently, FEV0.5 tends to be reached at low lung volumes and probably reflects the global output of both peripheral and central airway function in infants (42). Subsequent postnatal growth in lung volume is, however, more rapid than that of airway caliber. With the consequent lengthening of the expiratory time constant and slowing of lung emptying with growth, FEV0.5 occurs at a relatively higher lung volume in preschool children than in infants (43). Such developmental changes, rather than specific differences in measurement conditions or protocol may explain why “type of lung function” explained some of the variability for FEV0.5 in the multivariable model but not other FEFV parameters, and why FEV0.75 was a more sensitive parameter at follow-up. The use of FEV1 as a measure of airway function may be limited by the number of children in whom this can be calculated (Table 3), and by the fact that values in healthy preschool children often approximate FVC (10, 39, 40, 44). Further work is needed to explore the relationship of FEV at different timed intervals with growth. The current data suggest that, when using spirometry, different measurements may be required at different ages or disease stages.

Clinical Implications

We have shown that CF per se, in the absence of complications, is associated with decrements in lung function. The basis of these is not clear, but specialist treatment does not appear to ameliorate this. The implication is that new treatments are needed to improve lung health, although future intervention studies with DNase, antiinflammatory agents, or antibiotics in this age group when children are not acutely ill might show an improvement in the lung function of some of these children. It is also known that infection with PsA is associated with reduced FEV1/FVC in 4-year-olds, lower FEV1 in 7-year-olds, and diminished FEFV, faster deterioration in FEV1, and increased mortality in older children and adults (4447). Aggressive treatment aiming to prevent PsA infection has been shown to reduce lung function decline (48). This is one of the few studies to demonstrate a relationship between infection with PsA and diminished lung function in the early years. Reduced lung function was seen in those who had ceased to grow PsA as well as in those who still grew PsA either intermittently or chronically. This has clinical implications for segregation of infants diagnosed by newborn screening (38, 49, 50). Furthermore, our findings suggest that future policies aiming to prevent, not merely eradicate, PsA infection may be needed if we are to improve long-term outcome in children with CF.

The authors thank all the members of the LCFC, the families that participated in this study, and Ms. C. Oliver, Ms. E. Scrase, Dr. H. Ljungberg, and Dr. G. Hulskamp for help testing the children. They also thank Prof. K. Costeloe, Dr. J. Hawdon, and staff at the Homerton University and University College London Hospitals for help in recruiting the healthy control subjects.

1. Armstrong DS, Grimwood K, Carzino R, Carlin JB, Olinsky A, Phelan PD. Lower respiratory infection and inflammation in infants with newly diagnosed cystic fibrosis. BMJ 1995;310:1571–1572.
2. Dakin CJ, Numa AH, Wang H, Morton JR, Vertzyas CC, Henry RL. Inflammation, infection, and pulmonary function in infants and young children with cystic fibrosis. Am J Respir Crit Care Med 2002;165:904–910.
3. Nixon GM, Armstrong DS, Carzino R, Carlin JB, Olinsky A, Robertson CF, Grimwood K. Early airway infection, inflammation, and lung function in cystic fibrosis. Arch Dis Child 2002;87:306–311.
4. Ranganathan S, Dezateux CA, Bush A, Carr SB, Castle R, Madge SL, Price JF, Stroobant J, Wade AM, Wallis CE, et al. Airway function in infants newly diagnosed with cystic fibrosis. Lancet 2001;358:1964–1965.
5. Ranganathan S, Bush A, Dezateux CA, Carr SB, Hoo AF, Lum S, Madge SL, Price J, Stroobant J, Wade A, et al. Relative ability of full and partial forced expiratory maneuvers to identify diminished airway function in infants with cystic fibrosis. Am J Respir Crit Care Med 2002;166:1350–1357.
6. Ranganathan S, Stocks J, Dezateux C, Bush A, Wade A, Carr SB, Castle R, Dinwiddie RD, Hoo AF, Lum S, et al.; London Collaborative Cystic Fibrosis Group. The evolution of airway function in infancy and early childhood following clinical diagnosis of cystic fibrosis. Am J Respir Crit Care Med 2004;169:928–933.
7. S. Lum, J. Stocks, R. Castile, S. Davies, M. Henschen, M. Jones, M. G. Morris, S. Ranganathan, P. Sly, R. Tepper, for the American Thoracic Society/European Respiratory Society. ATS/ERS consensus statement. Raised volume forced expirations in infants: recommendations for current practice. Am J Respir Crit Care Med 2005;172:1463–1471.
8. Lum S, Gustafsson P, Ljungberg H, Hulskamp G, Bush A, Carr SB, Castle R, Hoo AF, Price J, Ranganathan S, et al. Early detection of cystic fibrosis lung disease: multiple-breath washout versus raised volume tests. Thorax 2007;62:341–347.
9. Turner DJ, Lanteri CJ, Le Souëf PN, Sly PD. Improved detection of abnormal respiratory function using forced expiration from raised lung volume in infants with cystic fibrosis. Eur Respir J 1994;7:1995–1999.
10. Aurora P, Stocks J, Oliver C, Saunders C, Castle R, Chaziparasidis G, Bush A. Quality control for spirometry in preschool children with and without lung disease. Am J Respir Crit Care Med 2004;169:1152–1159.
11. Aurora P, Bush A, Gustafsson P, Oliver C, Wallis C, Price J, Stroobant J, Carr S, Stocks J. Multiple-breath washout as a marker of lung disease in preschool children with cystic fibrosis. Am J Respir Crit Care Med 2005;171:249–256.
12. Kozlowska WJ, Aurora P. Spirometry in the pre-school age group. Paediatr Respir Rev 2005;6:267–272.
13. Beydon N, Davis SD, Lombardi E, Allen JL, Arets HG, Aurora P, Bisgaard H, Davis GM, Ducharme FM, Eigen H, et al. An official American Thoracic Society/European Respiratory Society statement: pulmonary function testing in preschool children. Am J Respir Crit Care Med 2007;175:1304–1345.
14. Marostica PJ, Weist AD, Eigen H, Angelicchio C, Christoph K, Savage J, Grant D, Tepper RS. Spirometry in 3- to 6-year-old children with cystic fibrosis. Am J Respir Crit Care Med 2002;166:67–71.
15. Nielsen KG, Pressler T, Klug B, Koch C, Bisgaard H. Serial lung function and responsiveness in cystic fibrosis during early childhood. Am J Respir Crit Care Med 2004;169:1209–1216.
16. Kozlowska WJ, Aurora P, Lum S, Saunders C, Ranganathan S, Castile R, Stocks J. Longitudinal assessment of lung function in infants and pre-school children with cystic fibrosis (CF) [abstract]. Arch Dis Child 2004;89:A38.
17. Kozlowska WJ, Saunders C, Lum S, Aurora P, Ranganathan S, Castle R, Stocks J; London Collaborative Cystic Fibrosis Group. Relationship between airway function in pre-school children with cystic fibrosis with that measured during infancy. Eur Respir J 2004;24:213s–214s.
18. Kozlowska W, Bush A, Wade A, Saunders C, Hoo AF, Stocks J. Serial measures of airway function in infants and preschoolers with cystic fibrosis (CF) according to Pseudomonas aeruginosa (PsA) status. Eur Respir J 2007;31:449s.
19. Kerem E, Conway S, Elborn S, Heijerman H. Standards of care for patients with cystic fibrosis: a European consensus. J Cyst Fibros 2005;4:7–26.
20. The Cystic Fibrosis Trust's Standards and Accreditation Group. Standards for the clinical care of children and adults with cystic fibrosis in the UK. Bromley, Kent, UK: Cystic Fibrosis Trust; 2001.
21. Doring G, Conway SP, Heijerman HG, Hodson ME, Hoiby N, Smyth A, Touw DJ. Antibiotic therapy against Pseudomonas aeruginosa in cystic fibrosis: a European consensus. Eur Respir J 2000;16:749–767.
22. Hoo AF, Stocks J, Lum S, Wade AM, Castle RA, Costeloe KL, Dezateux C. Development of lung function in early life: influence of birth weight in infants of nonsmokers. Am J Respir Crit Care Med 2004;170:527–533.
23. Lum S, Hoo AF, Dezateux CA, Goetz I, Wade AM, DeRooy L, Costeloe K, Stocks J. The association between birthweight, sex and airway function in infants of non-smoking mothers. Am J Respir Crit Care Med 2001;164:2078–2084.
24. Lee TW, Brownlee KG, Conway SP, Denton M, Littlewood JM. Evaluation of a new definition for chronic Pseudomonas aeruginosa infection in cystic fibrosis patients. J Cyst Fibros 2003;2:29–34.
25. Cole TJ, Freeman JV, Preece MA. British 1990 growth reference centiles for weight, height, body mass index and head circumference fitted by maximum penalized likelihood. Stat Med 1998;17:407–429.
26. Jones M, Castile R, Davis S, Kisling J, Filbrun D, Flucke R, Goldstein A, Emsley C, Ambrosius W, Tepper RS. Forced expiratory flows and volumes in infants. Am J Respir Crit Care Med 2000;161:353–359.
27. Edwards LJ. Modern statistical techniques for the analysis of longitudinal data in biomedical research. Pediatr Pulmonol 2000;30:330–344.
28. Schluchter MD, Konstan MW, Drumm ML, Yankaskas JR, Knowles MR. Classifying severity of cystic fibrosis lung disease using longitudinal pulmonary function data. Am J Respir Crit Care Med 2006;174:780–786.
29. Brown H, Prescott R. Applied mixed models in medicine. Hoboken, NJ: Wiley; 1999.
30. Draper NR, Smith H. Applied regression analysis, 2nd ed. Hoboken, NJ: Wiley; 1981.
31. Dezateux C, Lum S, Hoo AF, Hawdon J, Costeloe K, Stocks J. Low birth weight for gestation and airway function in infancy: exploring the fetal origins hypothesis. Thorax 2004;59:60–66.
32. Cohen J. Multiple regression and correlation analysis. Statistical power analysis for the behavioral sciences, 2nd ed. London, UK: Lawrence Erlbaum Associates; 1998. p. 407.
33. U.K. Newborn Screening Programme Centre. Cystic Fibrosis Screening Programme UK [Internet] [accessed 2007 Oct 9]. Available from:
34. Konstan MW, Butler SM, Wohl ME, Stoddard M, Matousek R, Wagener JS, Johnson CA, Morgan WJ. Growth and nutritional indexes in early life predict pulmonary function in cystic fibrosis. J Pediatr 2003;142:624–630.
35. Hiatt P, Eigen H, Yu P, Tepper RS. Bronchodilator responsiveness in infants and young children with cystic fibrosis. Am Rev Respir Dis 1988;137:119–122.
36. Nielsen KG, Pressler T, Klug B, Koch C, Bisgaard H. Serial lung function and responsiveness in cystic fibrosis during early childhood. Am J Respir Crit Care Med 2004;169:1209–1216.
37. Thamrin C, Gangell CL, Udomittipong K, Kusel MM, Patterson H, Fukushima T, Schultz A, Hall GL, Stick SM, Sly PD. Assessment of bronchodilator responsiveness in preschool children using forced oscillations. Thorax 2007;62:814–819.
38. Rosenfeld M, Emerson J, Accurso F, Armstrong D, Castile R, Grimwood K, Hiatt P, McCoy K, McNamara S, Ramsey B, et al. Diagnostic accuracy of oropharyngeal cultures in infants and young children with cystic fibrosis. Pediatr Pulmonol 1999;28:321–328.
39. Stanojevic S, Wade A, Lum S, Stocks J. Reference equations for pulmonary function tests in preschool children: a review. Pediatr Pulmonol 2007;42:962–972.
40. Vilozni D, Barak A, Efrati O, Augarten A, Springer C, Yahav Y, Bentur L. The role of computer games in measuring spirometry in healthy and “asthmatic” preschool children. Chest 2005;128:1146–1155.
41. Vilozni D, Bentur L, Efrati O, Minuskin T, Barak A, Szeinberg A, Blau H, Picard E, Kerem E, Yahav Y, et al. Spirometry in early childhood in cystic fibrosis patients. Chest 2007;131:356–361.
42. Ranganathan SC, Hoo AF, Lum SY, Goetz I, Castle RA, Stocks J. Exploring the relationship between forced maximal flow at functional residual capacity and parameters of forced expiration from raised lung volume in healthy infants. Pediatr Pulmonol 2002;33:419–428.
43. Stocks J. Pulmonary function tests in infants and young children. In: Chernick V, Boat TF, Wilmott RW, Bush A, editors. Kendig's disorders of the respiratory tract in children, 7th ed. Philadelphia: Elsevier; 2006. pp. 129–167.
44. Kosorok MR, Zeng L, West SE, Rock MJ, Splaingard ML, Laxova A, Green CG, Collins J, Farrell PM. Acceleration of lung disease in children with cystic fibrosis after Pseudomonas aeruginosa acquisition. Pediatr Pulmonol 2001;32:277–287.
45. Kraemer R, Blum A, Schibler A, Ammann RA, Gallati S. Ventilation inhomogeneities in relation to standard lung function in patients with cystic fibrosis. Am J Respir Crit Care Med 2004;171:371–378.
46. Kraemer R, Delosea N, Ballinari P, Gallati S, Crameri R. Effect of allergic bronchopulmonary aspergillosis on lung function in children with cystic fibrosis. Am J Respir Crit Care Med 2006;174:1211–1220.
47. Wilmott RW, Tyson SL, Matthew DJ. Cystic fibrosis survival rates: the influences of allergy and Pseudomonas aeruginosa. Am J Dis Child 1985;139:669–671.
48. Frederiksen B, Koch C, Hoiby N. Antibiotic treatment of initial colonization with Pseudomonas aeruginosa postpones chronic infection and prevents deterioration of pulmonary function in cystic fibrosis. Pediatr Pulmonol 1997;23:330–335.
49. Armstrong DS, Grimwood K, Carlin JB, Carzino R, Olinsky A, Phelan PD. Bronchoalveolar lavage or oropharyngeal cultures to identify lower respiratory pathogens in infants with cystic fibrosis. Pediatr Pulmonol 1996;21:267–275.
50. Farrell PM, Li Z, Kosorok MR, Laxova A, Green CG, Collins J, Lai HC, Rock MJ, Splaingard ML. Bronchopulmonary disease in children with cystic fibrosis after early or delayed diagnosis. Am J Respir Crit Care Med 2003;168:1100–1108.
Correspondence and requests for reprints should be addressed to Wanda Kozlowska, M.B.B.S., M.R.C.P.C.H., Portex Anaesthesia, Intensive Therapy and Respiratory Medicine Unit, UCL Institute of Child Health, 30 Guilford Street, London WC1N 1EH, UK. E-mail:


No related items
American Journal of Respiratory and Critical Care Medicine

Click to see any corrections or updates and to confirm this is the authentic version of record