Table 2.

Associations between cytokine levels of NE infants and perinatal inflammation and infection.

	Neonatal bacteremia (Na = 14/159)		Raised neonatal CRPb (Na = 12/158)		Histological funisitis (Na = 12/45)		Raised maternal CRPc (Na = 17/156)
	Adjusted geometric mean ratio (95% CI)	p value	Adjusted geometric mean ratio (95% CI)	p value	Adjusted geometric mean ratio (95% CI)	p value	Adjusted geometric mean ratio (95% CI)	p value
TNFα	0.98 (0.55, 1.77)	0.951	1.27 (0.68, 2.37)	0.446	1.79 (0.85, 3.81)	0.125	1.10 (0.77, 1.57)	0.597
IL-8	0.96 (0.18, 5.00)	0.959	0.89 (0.15, 5.15)	0.895	1.81 (0.32, 10.28)	0.493	0.54 (0.20, 1.47)	0.223
VEGF	0.58 (0.20, 1.71)	0.325	2.19 (0.70, 6.82)	0.176	1.35 (0.38, 4.78)	0.632	1.23 (0.63, 2.40)	0.540
IL-1α	1.63 (0.80, 3.31)	0.180	1.00 (0.47, 2.14)	0.999	2.88 (1.27, 6.58)	0.013	1.19 (0.77, 1.83)	0.435
IL-6	0.58 (0.20, 1.67)	0.308	1.36 (0.44, 4.24)	0.591	2.54 (0.82, 7.87)	0.103	0.68 (0.35, 1.31)	0.249
IL-10	0.67 (0.26, 1.72)	0.399	0.89 (0.33, 2.44)	0.824	0.78 (0.20, 3.07)	0.719	0.99 (0.55, 1.78)	0.961

Multivariable linear regression analysis; differences are shown as the geometric mean ratio in cytokine levels (pg/ml) with 95% confidence intervals (95% CIs), adjusted for infant sex and time of cytokine blood sampling.

Bold values indicate statistical significance p < 0.05.

^aN denotes the number of positive results out of the total numbers of test results available. Within this nested cohort of NE infants with cytokine data, blood culture and qPCR data were available for all infants (n = 159), neonatal CRP was available for 158 infants, placental histology data were available for 45 NE cases, and maternal CRP data were available in 156 NE cases.

^bDefined as neonatal CRP > 97th centile among ABAaNA study non-NE infants (CRP > 31.7 mg/l).

^cDefined as maternal CRP > 90th centile among mothers of ABAaNA study non-NE infants (CRP > 86.6 mg/l) as previously described.¹⁷