The relationship between creamatocrit and cumulative percentage of total milk volume: a cross-sectional study in mothers of very preterm infants in Bangkok, Thailand

Table 3 Random-coefficient mixed model analysis results

Model	-2 log likelihood	AIC	BIC	Regression equation Y = b₀ + b₁ × X + b₂ × X²	p-value
Model	-2 log likelihood	AIC	BIC	Regression equation Y = b₀ + b₁ × X + b₂ × X²	b₀	b₁	b₂
1	699.7	703.7	706.0	Y = 8.3047 + 0.09863 × Xc	< 0.001	< 0.001	-
2	670.4	678.4	682.9	Y = 8.2525 + 0.09874 × Xc	< 0.001	< 0.001	-
3	706.7	710.7	713.0	Y = 7.8581 + 0.09019 × Xc + 0.00058 × Xc²	< 0.001	< 0.001	0.004
4	673.6	681.6	686.2	Y = 7.8134 + 0.08553 × Xc + 0.00066 × Xc²	< 0.001	< 0.001	0.002
2.1	670.4	678.4	682.9	Y = 3.3157 + 0.09874 × X	< 0.001	< 0.001	-
				SE(b₁) = 0.008011 95% CI of β₁ = (0.08212, 0.1154)

The mean cumulative percentage of total milk volume (CP-MV, represented by X) from 155 bottles was 57%. A value of 50% was used to center X (represented by Xc). The relationship between X and Xc can be expressed as Xc = X – 50. Creamatocrit was represented by Y. Abbreviations: AIC, Akaike information criteria; BIC, Bayesian information criteria; CI, confidence interval; SE, standard error; Xc, centered X

ISSN: 1746-4358