Skip to main content
. Author manuscript; available in PMC: 2014 Jan 1.
Published in final edited form as: Child Dev. 2011 Sep 2;84(1):58–75. doi: 10.1111/j.1467-8624.2011.01641.x

Figure 3. Representative examples of CpG (cytosine-guanine dinucleotide) sites associated with maternal stress in infancy dependent on gender.

Figure 3

A. A specific CpG site in the promoter of the PKN1 gene (encoding protein kinase N1) was associated with maternal stress in the full group (rho = .40, 6.5% DNA methylation change, slope = .09 (95% CI = .05 – .14), FDR = 0%) and in boys only (rho = .50, 7.5% methylation change, slope = .13 (95% CI = .06 – .20), FDR=0%) but not in girls (rho = .33, 4.5% DNA methylation change, slope = .07 (95% CI = .01 – .13), FDR = 52%).

B. In contrast, while a CpG site in the promoter of the C5orf21 locus was weakly associated with maternal stress in the full group (rho = .23, 5.5% DNA methylation change, slope = .08 (95% CI = .01 – .15), FDR = 0%) it became much stronger in girls only (rho = .50, 10.7% DNA methylation change, slope = .15 (95% CI = .06 – .24), FDR = 0%). In boys, this CpG site tended to be negatively associated with maternal stress, although this was not statistically significant by our criteria (rho = − .17, −3.2% DNA methylation change, slope = −.06 (95% CI = −.17 – .06), FDR = 81%) The average beta scale was restricted to show values between 0.5–1.0 rather than the full range of 0–1.0. Girls are represented as triangles and boys as crosses. Regression lines for the full group are solid, wide-dashed for girls only, and narrow-dashed for boys only.