Association of DNA methylation signatures with premature ageing and cardiovascular death in patients with end-stage kidney disease: a pilot epigenome-wide association study

Figures & data

Table 1. Baseline patient characteristics by cardiovascular case status.

Characteristics	Cases (n = 30)	Controls^a (n = 30)	P
Age (years)	63.2 ± 11.4	63.0 ± 11.7	0.95
Male sex	18 (60.0)	18 (60.0)	1.00
Race			1.00
White	13 (43.3)	13 (43.3)
African American	16 (53.4)	16 (53.4)
Others	1 (3.3)	1 (3.3)
Dialysis vintage (years)	4.4 ± 3.6	4.5 ± 3.1	0.90
Vascular access type			0.35
Arteriovenous fistula	20 (66.7)	15 (50.0)
Arteriovenous graft	6 (20.0)	7 (23.3)
Catheter	4 (13.3)	8 (26.7)
Cause of ESKD			0.53
Diabetes mellitus	17 (56.7)	15 (50.0)
Hypertension	6 (20.0)	10 (33.3)
Others	7 (23.3)	5 (16.7)
Body mass index (kg/m^2)	30.6 ± 8.0	30.6 ± 7.3	1.00
Systolic BP (mmHg)	145.4 ± 26.1	156.5 ± 25.9	0.11
Diastolic BP (mmHg)	74.7 ± 15.3	79.4 ± 18.1	0.29
Charlson Comorbidity Index	5.8 ± 1.8	6.1 ± 1.6	0.60
Comorbidities
Diabetes mellitus	21 (70.0)	21 (70.0)	1.00
Ischemic heart disease	5 (16.7)	4 (13.3)	0.72
Congestive heart failure	6 (20.0)	3 (10.0)	0.28
Liver disease	1 (3.3)	2 (6.7)	0.55
HIV/AIDS	0 (0)	0 (0)	1.00
Malignancies	0 (0)	0 (0)	1.00
Laboratory parameters
Blood hemoglobin (g/dL)	11.0 ± 1.3	10.9 ± 0.8	0.59
Serum albumin (g/dL)	4.0 ± 0.3	3.9 ± 0.3	0.22
Serum calcium (mg/dL)	9.3 ± 0.7	9.1 ± 0.7	0.27
Serum phosphorus (mg/dL)	5.7 ± 1.5	5.2 ± 1.2	0.68
Serum ALP (U/L)	114.6 ± 51.3	104.6 ± 43.0	0.42
Serum intact PTH (pg/mL)	342 [226, 708]	339 [219, 504]	0.08
Medications
Statins	7 (23.3)	6 (20.0)	0.75
ESAs	21 (70.0)	27 (90.0)	0.053
Phosphate binders	17 (56.7)	20 (66.7)	0.43
Vitamin D analogs	27 (90.0)	28 (93.3)	0.64
Aspirin	6 (20.0)	5 (16.7)	0.74
Opioids	12 (40.0)	11 (36.7)	0.79
Antibiotics	0 (0)	0 (0)	1.00
Infectious hospitalization	0 (0)	0 (0)	1.00
Culture positive bacteremia	0 (0)	0 (0)	1.00

Data are presented as number (percentage), mean ± SD, or median [interquartile interval].

^aMatched by age, sex, race, and dialysis vintage.

Abbreviations: ALP = alkaline phosphatase; BP = blood pressure; ESAs = erythropoiesis-stimulating agents; ESKD = end-stage kidney disease; HIV/AIDS = human immunodeficiency virus/acquired immunodeficiency syndrome; PTH = parathyroid hormone.

Figure 1. Correlation between chronological age and DNA methylation-based biological age according to four established epigenetic clocks in 60 hemodialysis patients; (A) HorvathAge, (B) HannumAge, (C) PhenoAge, and (D) GrimAge.

Note: The solid line represents least-squares linear regression line, while the dashed line represents the perfect correlation between chronological and biological age.

Figure 2. Violin plots comparing epigenetic age accelerations between cardiovascular cases and controls* according to Horvath-, Hannum-, Pheno-, and GrimAge.

*Matched by age, sex, race, and dialysis vintage.

Note: None of the between-group differences were statistically significant.

Figure 3. Manhattan plot of the association between DNA methylation and cardiovascular death in 60 hemodialysis patients.

Note: The upper (red) and lower (blue) horizontal solid lines represent the genome-wide significance threshold (set at p-value of 9.0x10⁻⁸) and suggestive threshold (set at p-value of 5.0x10⁻⁵).

Table 2. Top 10 most significant CpG sites associated with cardiovascular death based on the genome-wide suggestive threshold at p-value of 5.0 × 10⁻⁵.

	Probe no.	Gene symbol	Chr	Position	Status estimate*	p-value
1	cg22305782	FBXL19	16	30937214	−0.135	2.00E–06
2	cg03111498	VSX1	20	25062860	−0.258	4.00E–06
3	cg19558049	MIER2	19	306297	−0.178	4.00E–06
4	cg15711208	MMP24	20	33836573	−0.258	6.00E–06
5	cg08842907	BSCL2	11	62470818	0.193	8.00E–06
6	cg03874948	-	9	113361270	−0.105	1.00E–05
7	cg25129798	-	2	231189493	0.132	2.00E–05
8	cg05431687	MCL1	1	150552130	−0.217	2.00E–05
9	cg17273588	AFF3	2	100760351	−0.107	2.00E–05
10	cg04578023	TGM2	20	36761479	0.140	2.00E–05

*Difference in methylation M-value between cases vs. controls (reference).

The epigenome-wide association study (EWAS) was performed using the following regression model: CpG ~ status + age + sex + self-reported race/ethnicity + top 5 principal components.

Data availability statement

The data used in these analyses were provided by DaVita Clinical Research. Requests for access to data can be made in writing to DaVita Clinical Research.