Objective To study the influential factors of C-reactive protein (CRP) in neonates.
Methods A total of 220 neonates whose serum CRP exceeded 8 mg/mL were recruited in the department of Neonate Intensive Care Unit (NICU) of BaYi Children's Hospital of General Military Hospital of Beijing from January 2009 to December 2010. They were selected to study group According to the diagnostic criteria for neonatal disease of Practical Neonatology and in accordance with the diagnosis and examination results on admission, they were divided into following sub-groups: sepsis sub-group (n=19), premature low birth weight sub-group (n=50), neonatal pneumonia sub-group (n=151), hypoglycemia sub-group (n=14), hemolytic disease sub-group (n=20), respiratory distress syndrome of newborn (NRDS) sub-group (n=8), pathological jaundice sub-group (n=81), intracranial hemorrhage sub-group (n=13), hypoxic ischemic encephalopathy (HIE) sub-group (n=29), asphyxia sub-group (n=20), amniotic fluid pollution sub-group (n=62) and premature rupture of membranes sub-group (n=44). At the same time, a total of 2129 inpatient neonates without corresponding diseases were chosen as corresponding control group (n=201, 170, 69, 206, 200, 212, 139, 207, 191, 200, 158, 176), respectively. Procedures were performed according to the Helsinki Declaration and approved by institutional Review Board (Institutional Review Board of General Military Hospital of Beijing PLA). The blood sampling of neonates were collected according to patient condition after approval was obtained from their families during hospitalization period. There were no statistically significant difference about gender, age, feeding and delivery methods between two groups (P>0.05). The average rank of serum CRP took two-sample rank test (Mann Whitney U test), correlation analysis took Pearson correlation of multiple linear regression analysis.
Results The preliminary screening results by Mann Whitney U test showed mean rank: Serum CRP of neonates septicemia (Yes: 138.32, No: 107.86; U=1381.0, P=0.046), NRDS (Yes: 66.44, No: 112.16; U=495.5, P=0.046), results of Pearson correlation showed possible correlation among serum CRP and hemoglobin (Hb), hematocrit (Hct), total bilirubin (TBIL), direct bilirubin (DBIL), glucose (Glu) (r=-0.149, -0.143, -0.164, 0.246, 0.316, P<0.05). Stepwise regression analysis was used further, dependent variable: Serum CRP, covariates: Neonatal septicemia, NRDS, Hb, Hct, TBIL, DBIL, Glu, only DBIL and Glu entered regression equation, standardized partial regression coefficient were 0.477, 0.311, respectively. The 95% confidence intervals were 0.608-2.265 and 0.588-9.445 (P=0.001, 0.028), respectively, so DBIL and Glu were positively correlated with serum CRP, the model had statistical significance (F=9.298, P=0.001), goodness of fit was slightly poorer (R2=0.341).
Conclusions Serum CRP are positively correlated with DBIL and Glu, furthermore, DBIL has more positive effects on serum CRP than Glu. Some others important influential factors on serum CRP need to be detected.