Skip to main content
. Author manuscript; available in PMC: 2019 May 2.
Published in final edited form as: Brain Behav Immun. 2019 Jan 9;77:168–177. doi: 10.1016/j.bbi.2019.01.001

Fig. 3.

Fig. 3.

Mediation analyses or the relationship between ELA, CMV seropositivity, and the CD4/CD8 ratio, CD4+ CD57+ cells, and CD8+ CD57+ cells as outcomes with standardized coefficients. (A) ELA considered as an independent variable, CD4/CD8 ratio as the dependent variable, and CMV titer as mediator variable. (B) ELA as independent variable, arcsine-transformed percentages of CD4+ CD57+ cells as dependent variable, and CMV titer as mediator variable. (C) ELA as independent variable, arcsine-transformed percentage of CD8+ CD57+ as dependent variable, and CMV titer as mediator variable. Statistic modeling: ELA was treated as continuous variable (age at adoption); values for control youth were set to 0. In each path analysis, the dependent outcome controlled for sex and age at blood draw. Significance was set at *p < 0.05, **p < 0.01, ***p < 0.001.