Sex-specific DNA methylation in saliva from the multi-ethnic Future of Families and Child Wellbeing Study

Figures & data

Table 1. Future of Families and Child Wellbeing Study sample characteristics at age 9, by sex (N = 796).

Characteristic	Overall N = 796^1	Male, N = 403^1	Female, N = 393^1	p-value^2
Sample
Child’s age (months)	110 (108,113)	110 (108, 113)	110 (108, 114)	0.30
Child’s BMI (kg/m^2)	18.0 (16.4,21.6)	17.8 (16.4, 20.7)	18.4 (16.4, 22.4)	0.05
Child’s birth city^3				0.30
Detroit/Toledo/Chicago	180 (22.6%)	85 (21.0%)	95 (24.0%)
Other	616 (77.4%)	318 (79.0%)	298 (76.0%)
Poverty index^3	1.4 (0.7, 2.7)	1.5 (0.7, 2.8)	1.4 (0.7, 2.7)	0.70
Mother’s BMI^3 (kg/m^2)	31 (26, 37)	30 (26, 35)	32 (26, 35)	0.20
Missing	88	43	45
Mother’s race/ethnicity^3				0.60
Black, non-Hispanic	438 (55.0%)	217 (53.8%)	221 (56.2%)
Hispanic	170 (21.4%)	87 (21.6%)	83 (21.1%)
White, non-Hispanic	161 (20.2%)	82 (20.3%)	79 (20.1%)
Other	27 (3.4%)	17 (4.2%)	10 (2.5%)
Mother’s education^3				0.50
Less than high school	263 (33.0%)	128 (31.8%)	135 (34.4%)
High school or equivalent	226 (28.4%)	122 (30.3%)	104 (26.5%)
Some college or technical accreditation	219 (27.5%)	105 (26.1%)	114 (29.0%)
College or graduate school	88 (11.1%)	48 (11.9%)	40 (10.2%)
Mother’s overall health^3				0.50
Great	249 (31.3%)	120 (29.8%)	129 (32.8%)
Very good	283 (35.6%)	149 (37.0%)	134 (34.1%)
Good	209 (26.3%)	102 (25.3%)	107 (27.2%)
Fair/poor	55 (6.9%)	32 (7.9%)	23 (5.9%)
Mother’s smoking habits^3				0.90
1+ packs per day	18 (2.3%)	8 (2.0%)	10 (2.5%)
<1 pack per day	144 (18.1%)	74 (18.4%)	70 (17.8%)
None	634 (79.6%)	321 (79.7%)	313 (79.6%)
Father’s race/ethnicity^3				0.90
Black, non-Hispanic	450 (56.5%)	227 (56.3%)	223 (56.7%)
Hispanic	170 (21.4%)	83 (20.6%)	87 (22.1%)
White, non-Hispanic	146 (18.3%)	75 (18.6%)	71 (18.1%)
Other	26 (3.3%)	15 (3.8%)	11 (2.8%)
Missing	4	3	1
Saliva sample
Average global DNAm (%)	49.55 (49.01, 49.90)	49.51 (49.02, 49.90)	49.56 (48.98, 49.90)	0.60^§
Proportion epithelial cells	0.00 (0.00, 0.01)	0.00 (0.00, 0.02)	0.00 (0.00, 0.00)	0.20
Proportion immune cells	1.00 (0.99, 1.00)	1.00 (0.98, 1.00)	1.00 (1.00, 1.00)	0.20

Note: ¹Median (25%,75%); n (%).

²Wilcoxon rank sum test; Pearson’s Chi-squared test; Fisher’s exact test.

³Measured at baseline (birth of the child).

Continuous variables are presented as Median (25th percentile, 75th percentile) and categorical variables are presented as Count (%). Wilcoxon rank sum test is used to test for differences in continuous variables, and Pearson’s chi-squared test is used to test for differences by sex for categorical variables. The number of children with missing values per variable are presented under the Missing category. The absence of a Missing category under a variable indicates there are no missing values.

DNAm: DNA methylation.

Figure 1. characteristics of differentially methylated sites in the Future of Families and Child Wellbeing Study sample, age 9 (N = 796). a: Adjusted difference in autosomal DNA methylation between female and male children for each site (n = 391,980 sites) in the Fragile Families and Child Wellbeing Study sample, age 9 (N = 796). Effect sizes and -log10 p-values are from the main model containing fixed effects for sex, epithelial cell proportion, and mother’s race/ethnicity and a random effect for sample plate. Positive effect sizes indicate sites for which DNA methylation was higher in female children, while negative effect sizes represent sites for which DNA methylation was higher in male children at age 9. 8,430 sites achieved genome-wide significance at P = 2.4× 10^–7 (horizontal dotted line), with n = 6,425 (76.2%) having higher methylation in female participants. b: Distributions of mean DNAM for differentially methylated sites (n = 8,430; dashed line) and remaining sites (n = 383,550). c: Distribution of differentially methylated sites in relation to CpG islands. Fragile Families and Child Wellbeing Study sample, age 9 (N = 796).

Figure 2. comparison of adjusted sex differences in autosomal DNA methylation between ages 9 and 15 in the Future of Families and Child Wellbeing Study (N = 796, n = 390,659 sites). a: Correlation of adjusted difference in autosomal DNA methylation (%) by sex between ages 9 and 15 in the Future of Families and Child Wellbeing Study (n = 390,659 sites). Effect sizes at each time point are from the main model adjusting for sex, epithelial cell proportion, mother’s race/ethnicity at baseline and a random effect for plate. Positive percentages are sites for which female children had higher DNA methylation than males. Spearman’s correlation = 0.53. b: Miami plot of sex-specific DNAm analyses at ages 9 and 15 in the Future of Families and Child Wellbeing Study sample (N = 796). P-values are reported from the main model containing sex, epithelial cell proportion, maternal race/ethnicity at baseline, and a random effect for sample plate. A threshold of P = 2.4 × 10–7 is used for genome-wide significance.

Figure 3. comparison of adjusted differences in DNA methylation between saliva tissue from the Future of Families cohort and cord blood tissue from Solomon et al., 2022 (Spearman correlation = 0.89).

Data availability statement

The data that support the findings of this study are available on request from the co-authors [DN, CM]. The data are not publicly available due to restrictions; however, data will be released in 2023 per the NIH data guidelines. Survey data for the Future of Families and Child Wellbeing study can be accessed at https://ffcws.princeton.edu/.