RESULTS
Demographics and Clinical Characteristics :
Of the 598 subjects in the registry who were older than five years of age, 88 (14.7%) patients performed adequate pulmonary function testing at a clinic visit between the ages of five and seven years old and thus were included in the study population. The study population was 45.5% female and had an average gestational age of 26.8 ± 2.9 weeks. 63.6% of subjects identified as non-white race/ethnicity. The mean median household income was $63,000 ± $180,800, and 62.5% had public insurance. 14.5% of subjects reported second hand smoke exposure, and 64.5% reported a family history of asthma. A total of 38 patients (43.2%) required supplemental oxygen at the time of discharge from the NICU with an average amount of 0.35 ± 0.27 liters per minute. Seventy-eight (88.6%) patients had data available that allowed them to be classified according to the severity of their BPD with 11.5% with chronic lung disease of prematurity (no BPD), 7.7% with mild BPD, 33.3% with moderate BPD, and 47.4% with severe BPD. Of the study population, 34.1% had a gastrostomy tube placed, 22.7% underwent Nissen fundoplication, and 18.2% had pulmonary hypertension. In terms of growth parameters, subjects had an average birthweight of 980 ± 518 grams with a percentile of 41 ± 25 corrected for gestational age. At two years of age, subjects had an average weight percentile for age of 30% ± 31%, an average height percentile for age of 24% ± 26%, and an average BMI percentile for age of 49% ± 31%. The study population had an average FEV1 of 84.5% predicted ± 22.7% and an average FVC of 94.2% predicted ± 26.5%. See Table 1 for additional demographic characteristics of the study population. Additional analysis illustrated that the study population was not significantly different from the entire registry population in terms of demographic characteristics, clinical characteristics, or growth parameters (Supplemental Table 1 ).
Predictors of Forced Expiratory Volume in 1 second (FEV1) :
On average, the study population had a FEV1 % predicted of 84.5% ± 22.7%. Of the demographic characteristics included, in univariate regression analysis, only higher median household income (2.9 +/- 1.4, p=0.044) was significantly associated with an improvement in FEV1% predicted. There was no significant association between sex (p=0.55), gestational age (p=0.54), race/ethnicity (p=0.1), use of public insurance (p=0.36), secondhand smoke exposure (p=0.74), and family history of asthma (p=0.99) and FEV1% predicted at six years of age. The need for supplemental oxygen at the time of NICU discharge (-11 ± 5.1, p=0.035) and a greater amount of oxygen required at discharge (-23.7 ± 9.9, p=0.02), the presence of gastrostomy tube (-14.5 ± 5.2, p=0.007), history of Nissen fundoplication (-19.6 ± 5.6, p=0.001), and higher weight percentile at two years of age (p=0.026) were all associated with significantly lower FEV1% predicted at six years of age. However, there was no significant association between a diagnosis of severe BPD (p=0.051), presence of pulmonary hypertension (p=0.44), personal diagnosis of reactive airway disease or asthma in childhood (p=0.61), or prescription for inhaled corticosteroid (p=0.92) and FEV1% predicted at age 6. Weight percentile at age two was shown to be negatively associated with FEV1% predicted at age six (-0.2 ± 0.1, p=0.026) but no association was seen between birth weight percentile (p=0.63), height percentile for age at age two (p=0.07), or BMI percentile for age at age two (p=0.10) and FEV1% predicted.
Multivariate analysis was performed using the variables found to have association with FEV1% predicted in univariate analysis. After adjusting for severity of BPD, a positive association was seen between median household income (2.6 ± 1.3, p=0.04) with FEV1% predicted whereas a negative association was noted between Nissen fundoplication (-17.9 ± 6.2, p=0.005) and weight percentile for age at age 2 (-0.2 ± 0.1, p=0.011) and FEV1% predicted. Severe BPD was adjusted for in the model but was not found to be significant (-5.3 ± 5.5, p=0.34).
Predictors of Forced Vital Capacity (FVC) :
On average, the study population had a FVC% predicted of 94.2% ± 26.5%. In univariate regression none of the demographic characteristics were shown to have a significant association with FVC% predicted including sex (p=0.55), gestational age (p=0.54), race/ethnicity (p=0.1), median household income (p=0.25), use of public insurance (p=0.36), secondhand smoke exposure (p=0.74), and family history of asthma (p=0.99).
In terms of clinical characteristics, a requirement for supplemental oxygen at the time of discharge from the NICU (-14 ± 5.4, p=0.012) and the amount of oxygen required (-36.7 ± 10.2, p=0.001) were both found to be negatively associated with FVC% predicted at six years of age. Additionally, the presence of a gastrostomy tube (-12.5 ± 5.7, p=0.031) and Nissen fundoplication (-19.7 ± 6.1, p=0.002) were also found to be negatively associated with FVC% predicted.
No significant association was seen between FVC% predicted at six years of age and severity of BPD (p=0.08), pulmonary hypertension (p=0.17), personal diagnosis of reactive airway disease or asthma (p=0.28), or prescription for inhaled corticosteroids (p=0.57). A significant negative association was found between weight percentile for age at two years of age (-0.2 ± 0.1, p=0.016) and FVC% predicted at six years of age but no significant association was seen with the other growth parameters analyzed including birth weight percentile (p=0.86), height percentile for age at age two (p=0.11), and BMI percentile for age at age two (p=0.12) and FVC% predicted.
Multivariate analysis was performed using the variables found in univariate analysis to have an association with FVC% predicted. The amount of oxygen required at the time of discharge from the NICU (-26.7 ± 10.8, p=0.016), the presence of Nissen fundoplication (-14.9 ± 6.8, p=0.031), and weight percentile for age at age 2 (-0.2 ± 0.1, p=0.005) were found to be negatively associated with FVC% predicted at six years of age. Severe BPD was adjusted for in the multivariate model but was not found to have a significant association with FVC% predicted (0.2 ± 6.1, p=0.98).