Table 3.

Associations between adverse events in childhood and CRP (mg/L) at 15 years.^a

Adverse events z-score	1: Age 15 log CRP			2: Change in log CRP e
	β	SE	p-value	β	SE	p-value
Baseline models
1.5 years	0.06	(0.02)	0.001	0.06	(0.02)	0.01
2.5 years	0.03	(0.02)	0.23	0.02	(0.03)	0.42
3.5 years	0.03	(0.02)	0.17	0.02	(0.02)	0.46
4.5 years	0.02	(0.02)	0.26	0.01	(0.03)	0.59
6 years	−0.03	(0.02)	0.17	−0.06	(0.02)	0.02
7 years	0.00	(0.02)	0.87	−0.02	(0.02)	0.45
8 years	0.05	(0.02)	0.02	0.06	(0.02)	0.02
Sum scoreb	0.05	(0.02)	0.04	0.04	(0.03)	0.18
Significant baseline models with potential mediatorsc
Events at 1.5 years	0.05	(0.02)	0.01	0.04	(0.02)	0.07
BMI	0.10	(0.01)	<.0001	0.08	(0.01)	<.0001
Depression	0.11	(0.06)	0.08	0.14	(0.07)	0.048
Smoking	0.10	(0.09)	0.26	0.16	(0.11)	0.13
Events at 8 years	0.06	(0.02)	0.01	0.06	(0.02)	0.02
BMI	0.10	(0.01)	<.0001	0.08	(0.01)	<.0001
Depression	0.09	(0.06)	0.13	0.12	(0.07)	0.11
Smoking	0.06	(0.09)	0.54	0.14	(0.11)	0.20
Sum score (all events)	0.03	(0.02)	0.20	0.01	(0.03)	0.66
BMI	0.10	(0.01)	<.0001d	0.07	(0.01)	<.0001d
Depression	0.12	(0.07)	0.08	0.15	(0.08)	0.06
Smoking	0.09	(0.11)	0.41	0.12	(0.13)	0.36

Note: In the models not adjusted for age 10 CRP, the sample sizes in the baseline models range from N = 2802 (for the model of adverse events at 1.5 years) to N = 2047 (for the model using the summary score). In models adjusted for age 10 CRP, the samples sizes in the baseline models range from N = 1888 (model of adverse events at 1.5 years) to N = 1427 (model using the summary score).

^aLinear regression coefficients; models adjusted for gender, age at blood measure, ethnicity, family income at age 8 (5 categories + missing category), mother’s education at gestation (4-categories + missing category), and medication. Bolded values indicate associations examined in mediation models.

^bSum of adverse events, across all measures between 1.5 and 8 years.

^cBMI = body mass index; depression measured at age 14 using the Moods and Feelings Questionnaire. BMI, depression, and smoking were simultaneously included in each mediation model; therefore, coefficients reflect the main effects of each potential mediator holding constant the other variables in the model.

^dSignificant indirect effect detected (i.e., the indirect effect is the effect of the adverse event score on the inflammatory outcome that is mediated through the potential mediator); significant indirect effect coefficients are provided in the Results section of the text (i.e., not presented in the table); the coefficient for direct effect of the parameter is presented in the table.

^eAge 10 CRP is included in the model.