TABLE 4.

Paths suggested by the SEM modeling process

Responsea	Predictorb	Estimatec	SEc	Std estimated	P valuee
Cervical length	d-Lactic acid	131.94	87.39	0.0786	0.1321
	TIMP-1	−0.89	0.30	−0.1616	0.0035**
	p62	−70.35	38.70	−0.0938	0.0706!
	Age	0.23	0.06	0.1961	0.0003***
	Mixed race	0.36	0.93	0.0205	0.6990
	Black race	−5.08	1.49	−0.1806	0.0007***
TIMP-1	CST II	0.96	0.50	0.0953	0.0549!
	CST III	0.57	0.17	0.1737	0.0012**
	CST IV	1.53	0.21	0.3826	0.0000***
	CST V	0.58	0.35	0.0830	0.0982!
	Age	−0.05	0.01	−0.2583	0.0000***
	Mixed race	−0.12	0.16	−0.0371	0.4645
	Black race	0.38	0.26	0.0735	0.1486
d-Lactic acid	CST II	−0.004	0.002	−0.1242	0.0135*
	CST III	−0.005	0.001	−0.4183	0.0000***
	CST IV	−0.005	0.001	−0.3866	0.0000***
	CST V	−0.003	0.001	−0.1170	0.0214*
	Age	0.000	0.000	0.0288	0.5647
	Mixed race	−0.0004	0.001	−0.0409	0.4362
	Black race	−0.0002	0.001	−0.0133	0.7956
TIMP-1	∼∼d-lactic acid	−0.08	−0.0830	0.0636
	∼∼p62	0.05	0.0515	0.1729
d-Lactic acid	∼∼p62	−0.02	−0.0245	0.3267

^aThe response column shows the variable affected by the corresponding entry in the predictor columns.

^bEntries marked with two tildes indicate bidirectional unexplained correlations between variables.

^cThe estimate and its standard error (SE) come from the component linear regression models.

^dThe standardized estimate (Std estimate) is given to show the relative effect (path) strengths.

^eThe P values are given. Symbols: !, P < 0.1; *, P < 0.05; **, P < 0.01; ***, P < 0.001.