The cohort consisted of 93 ELBW children and 87 controls. Baseline characteristics can be found in Table 1. Investigated perinatal characteristics of the study population can be found in Table S1. Blood samples were successfully taken from 59 out of 93 cases (failure to collect blood was due to refusal, too low sample volume or collection failure after 1 or 2 attempts).
Blood pressures were significantly higher amongst ELBW children than controls (Figure 1, Table 2). Average SBP were in the 75th percentile for cases compared to the 47th percentile for controls (p<0.001). Average DBP were in the 68th percentile for cases versus the 54th percentile for controls (p<0.001). Elevated SBP and systolic hypertension occurred significantly more frequently amongst ELBW children than controls (OR=6.15, p<0.001 and OR=13.6, p=0.002). Elevated DBP was far less prevalent than elevated SBP and found to not significantly differ between ELBW children and controls. While 38.6% of ELBW children had elevated blood pressures (either elevated SBP or DBP), this was only 8% for the controls (OR=5.5, p<0.001).
On average, ELBW children had an eGFR of 94.1 ml/min/1.73m2, while healthy controls had an eGFR of 106.5 ml/min/1.73m2 (p<0.001, Figure 1). Amongst ELBW children, 34.5% had reached a kidney function decline (defined as eGFR <90 ml/min/1.73m2) and 12.7% in the control group (Table 1).
Perinatal risk factors in relation to outcomes in adolescence
None of the investigated perinatal factors showed an association with elevated SBP. Former ELBW children with an elevated DBP were born at a higher gestational age (29.5 vs. 27.2 weeks, p=0.001). These children were subsequently discharged at a significantly lower weight (1786 vs. 2240 g, p=0.01). Sex, being small for gestational age (SGA), birthweight, intra-ventricular hemorrhage (IVH), broncho-pulmonary dysplasia (BPD), retinopathy of prematurity (ROP), postnatal ibuprofen use, or steroids did not associate with elevated SBP or DBP. No direct correlation was observed between SBP or DBP percentiles and eGFR for the population of ELBW children.
Male ELBW children had a tendency of developing eGFR <90 ml/min/1.73m2 in comparison to female ELBW children (OR=3.33, p=0.055). ELBW children who developed an eGFR <90 ml/min/1.73m2 received significantly longer ventilation therapy (17 vs. 9 days, p=0.006). They additionally tended to suffer more frequently from IVH (40% vs. 15.8%, p=0.056) (Supplementary Table S2). Birthweight, being SGA and gestational age played no role. Ibuprofen and steroids had no effect on the outcome of ELBW children (Supplementary Table S2). None of the ELBW children had developed microalbuminuria by age 11 (data not shown).
The most significant perinatal determinants for an unfavorable renal outcome in ELBW children were used to create a model that could illuminate their risk of a lower eGFR by age 11, defined in the study as a single measurement of eGFR <90 ml/min/1.73m2. A point was assigned for sex (male), ventilation therapy (>10 days) and intraventricular hemorrhage IVH (any), representing vulnerability through sex, exposure to intensive treatment and the occurrence of neonatal comorbidities.
As a proof of concept, we modelled the risk for kidney function decline (defined as eGFR <90 ml/min/1.73m2), using the criteria sex (male), ventilation (>10 days) and IVH (any). Amongst ELBW children 34.5% progressed to kidney function decline. This risk progresses linearly between 0% and 80% varying on the points scored in the model (Figure 2).