Table 2.
Summary of the SLCMA results for the 62 CpGs with R2 > 3%
| Adversity | Number of R2 > 3% CpGs | Range of R2 | Range of (P-values | Number of FDR < 0.05 CpGs |
|---|---|---|---|---|
| Neighborhood disadvantage | 17 | 3.0–4.2% | 1.3 × 10−7—7.1 × 10−6 | 4a |
| Job loss | 15 | 3.1–3.7% | 5.8 × 10−7—8.8 × 10−6 | - |
| Low family income | 13 | 3.0–3.8% | 1.7 × 10−6– 2.5 × 10−5 | - |
| Financial hardship | 9 | 3.0–3.7% | 5.9 × 10−7– 8.5 × 10−6 | - |
| Major financial problem | 5 | 3.0–3.8% | 2.6 × 10−7– 4.7 × 10−6 | - |
| Income reduction | 3 | 3.0–3.3% | 1.5 × 10−6– 4.5 × 10−6 | - |
The R2 values reflect the increase in the variance of DNA methylation explained by the first hypothesis chosen after accounting for covariates. P-values were calculated using selective inference, which assesses the significance of the increase in R2 explained. See Supplementary Material, Table S2 for the full list of the 62 CpGs. SLCMA = structured life-course modeling approach.
aFour CpGs for neighborhood disadvantage passed an FDR < 0.05 significance threshold: cg20102336, cg08638097, cg23405172 and cg14212190.