Table 2.
Outcome | AgeAccel | Time Point | Multiplier | Est. | SE | p | q |
---|---|---|---|---|---|---|---|
Body mass index (kg/m2) | GrimAge | Baseline | — | 0.036 | 0.085 | .674 | .809 |
Follow-up | — | 0.022 | 0.090 | .804 | .889 | ||
Horvath | Baseline | — | 0.101 | 0.049 | .040 | .093 | |
Follow-up | — | 0.085 | 0.053 | .108 | .174 | ||
Timed up and go (s) | GrimAge | Baseline | 10 | 0.004 | 0.001 | .001 | .003 |
Follow-up | 10 | 0.005 | 0.002 | .003 | .009 | ||
Horvath | Baseline | 10 | 0.000 | 0.001 | .734 | .856 | |
Follow-up | 10 | 0.000 | 0.001 | .995 | .995 | ||
10-m walking test (s) | GrimAge | Baseline | — | 0.001 | 0.001 | .054 | .108 |
Follow-up | — | 0.003 | 0.001 | <.001 | .002 | ||
Horvath | Baseline | — | 0.000 | 0.000 | .935 | .982 | |
Follow-up | — | 0.001 | 0.000 | .234 | .339 | ||
6-min walking test (m) | GrimAge | Baseline | 1/1 000 | −0.004 | 0.002 | .005 | .014 |
Follow-up | 1/1 000 | −0.006 | 0.002 | .001 | .003 | ||
Horvath | Baseline | 1/1 000 | 0.000 | 0.001 | .934 | .982 | |
Follow-up | 1/1 000 | −0.001 | 0.001 | .355 | .481 | ||
Grip strength (N) | GrimAge | Baseline | 1/100 | −0.010 | 0.008 | .208 | .312 |
Follow-up | 1/100 | −0.015 | 0.010 | .137 | .213 | ||
Horvath | Baseline | 1/100 | −0.002 | 0.006 | .760 | .863 | |
Follow-up | 1/100 | 0.013 | 0.007 | .054 | .108 | ||
Ankle plantar flexion strength (N) | GrimAge | Baseline | — | −0.116 | 0.066 | .079 | .133 |
Follow-up | — | −0.217 | 0.082 | .008 | .021 | ||
Horvath | Baseline | — | −0.044 | 0.056 | .433 | .568 | |
Follow-up | — | 0.035 | 0.059 | .557 | .709 | ||
Knee extension strength (N) | GrimAge | Baseline | 1/100 | −0.032 | 0.017 | .059 | .113 |
Follow-up | 1/100 | −0.053 | 0.016 | .001 | .003 | ||
Horvath | Baseline | 1/100 | 0.024 | 0.013 | .077 | .133 | |
Follow-up | 1/100 | 0.012 | 0.012 | .317 | .444 |
Notes: AgeAccel = DNA methylation age acceleration; Est. = unadjusted regression coefficient estimate; p = expected false positive rate for a single test; q = expected false positive discovery rate among all outcomes; SE = standard error based on 10 000 bootstrap draws. Outcome values were scaled with the multiplier value. A nonrandom missing data mechanism was utilized in longitudinal models.