Table 5 Associations between epigenetic clocks and mortality among cancer survivors and controls in Sample Ba; HRS (2016–2020)

From: Aging measures and cancer in the Health and Retirement Study (HRS)

Cancer survivors (N = 582)

Aging constructs

No. of deaths

Total person-year

HR (95% CI) per 1 SD increase in aging construct, p-valueb

Model 1c

Model 2c

HannumAccel (SD = 5.48 years)

103

2270

1.42 (1.15, 1.76), p = 0.002

1.37 (1.14, 1.63), p < 0.001

HorvathAccel (SD = 7.01 years)

1.26 (0.98, 1.62), p = 0.071

1.21 (0.94, 1.57), p = 0.136

LevineAccel (SD = 6.88 years)

1.64 (1.33, 2.03), p < 0.001

1.56 (1.23, 1.98), p < 0.001

GrimAgeAccel (SD = 4.68 years)

2.08 (1.60, 2.70), p < 0.001

1.80 (1.37, 2.39), p < 0.001

Zhang Score (SD = 0.46 units)

1.72 (1.32, 2.23), p < 0.001

1.62 (1.27, 2.06), p < 0.001

mPOA (SD = 0.09 years of physiological decline per one chronological year)

1.24 (0.98, 1.58), p = 0.073

1.14 (0.85, 1.54), p = 0.381

Controls (N = 2805)

Aging constructs

No. of deaths

Total person-year

HR (95% CI) per 1 SD increase in aging construct, p-valueb

Model 1c

Model 2c

HannumAccel (SD = 5.14 years)

213

11,439

1.17 (1.00, 1.37), p = 0.050

1.13 (0.96, 1.31), p = 0.112

HorvathAccel (SD = 6.25 years)

1.07 (0.87, 1.30), p = 0.551

1.06 (0.87, 1.30), p = 0.555

LevineAccel (SD = 6.74 years)

1.08 (0.91, 1.29), p = 0.376

1.04 (0.86, 1.23), p = 0.647

GrimAgeAccel (SD = 4.63 years)

1.54 (1.32, 1.79), p < 0.001

1.18 (0.97, 1.44), p = 0.092

Zhang Score (SD = 0.43 units)

1.64 (1.39, 1.93), p < 0.001

1.45 (1.21, 1.75), p < 0.001

mPOA (SD = 0.09 years of physiological decline per one chronological year)

1.37 (1.16, 1.61), p < 0.001

1.15 (0.96, 1.36), p = 0.133

  1. Accel age acceleration, mPOA Dunedin methylation-pace of aging, SD standard deviation, HR hazard ratio, CI confidence interval, BMI body mass index, CMV cytomegalovirus.
  2. aSample B included participants who had data on ECs.
  3. bP-values were calculated from multivariable Cox proportional hazards regression.
  4. cModel 1 was adjusted for chronological age, sex, race/ethnicity, as well as the percentage of monocyte, neutrophil, and lymphocyte. Model 2 was additionally adjusted for BMI, smoking status, ever drinking, physical activity, comorbidity index, as well as CMV infection. All analyses were accounted for survey weights.
Back to article page