Study population and setting
We performed a retrospective review of the medical records of BPD infants admitted to the Neonatal Intensive Care Unit (NICU) at Zhangzhou Hospital between Jun 2016 and May 2020. The study was approved by the Institutional Review Board of the Zhangzhou Hospital, with waiver of informed consent. Case with MBD were identified as case group. Inclusion criteria consisted of the following: (1) gestational age<32 weeks, (2) survival up to 36 weeks post-menstrual age (PMA), (3) diagnosis of BPD according to National Institutes of Health (NIH) that defined as requirement of oxygen support (>21%) for at least 28 days and a subsequent assessment at 36 weeks PMA or discharge, whichever comes first. At the time of assessment, infants with no oxygen requirement were classified as having mild BPD, infants requiring <30% oxygen were classified as having moderate BPD and infants with a need for positive pressure ventilation/continuous positive pressure (PPV/CPAP) and/or oxygen requirement ≥30% were classified as having severe BPD [8], and (4) diagnosis of MBD that defined as peak serum ALP higher than 900 U/L and serum phosphate level lower than 1.8 mmol/L, which yielded a sensitivity of 100% at a specificity of 70%, with or without radiographic changes of long bones [9]. Exclusion criteria included the following: (1) gestational age≥32 weeks, (2) presence of significant congenital anomalies, (3) death before 36 weeks PMA. For each MBD case, two contemporaneous with BPD but without MBD matched at equivalent gestational age and gender (male: female=1:1) were randomly selected.
Clinical data on demographics and putative risk factors for MBD were collected, including birth weight, gender, use of antenatal steroids, histologic chorioamnionitis (HCA), fetal growth restriction (FGR), maternal hypertensive disorders without FGR, prolonged rupture of membranes (PROM), type of feeding, initiation of oral vitamin D supplementation, feeding volume at the end of the 4th week after birth, postnatal dexamethasone application, patent ductus arteriosus (PDA), cholestasis, use of mechanical ventilation, late onset sepsis (LOS), thyroid function tests and platelet (PLT) count.
FGR was defined as antenatal diagnosis by any 2 of the following: estimated fetal weight below the 10th percentile according to the care provider reference curve, abnormal fetal Doppler findings, growth arrest, or maternal hypertensive disorders. PROM was defined as rupture of the membranes occurring earlier than 24 h before delivery [10]. HCA was defined as the infiltration of neutrophils in any of the following structures: placental disc, the chorioamniotic membranes, and the umbilical cord [11]. Cholestasis was defined as direct bilirubin >2 mg/dL [12]. Maternal hypertensive disorder was defined as systolic blood pressure ≥ 140 mmHg and (or) diastolic blood pressure ≥ 90 mmHg [13]. PDA was diagnosed with echocardiography confirmation. Thyroid-stimulating hormone (TSH) = 10 mU/L was used as the cutoff level for hypothyroidism [14] and PLT<100×10^5/L was considered to be thrombocytopenia [15]. LOS was diagnosed with clinical suspicion and evaluated by microbial culture or two non-specific blood indicators occurring after 72 h of birth [16, 17].
In our NICU, we started parenteral nutrition in the first hours after birth and introduced enteral feeding with breast milk as soon as the infant reached a tolerated respiratory rate without hemodynamic instability. The amount increased gradually with a rate of 20–30 mL/kg/day without signs of feeding intolerance. Oral vitamin D (800–1000 IU) supplementation was started when the amount of breast milk feed reaches 50 mL/kg/day, and liquid human milk fortification was added when the amount of breast milk feed reaches 80 mL/kg/day. Premature formula was also used if the amount of breast milk is insufficient.
Statistical Analyses
Data were analyzed using SPSS version 26. Dichotomous or categorical variables were compared between cases and controls with the use of χ2 analysis or the Fisher exact test or z-test. Continuous variables were compared between groups via the Mann-Whitney U test. A risk model was derived via logistic regression, with 95% CIs calculated for ORs using the forward LR method. A P-value of ≤ 0.2 was used in univariate analysis for inclusion of putative risk factors into the multivariate (adjusted) model. A P-value of < 0.05 was considered statistically significant [16].