Skip to main content
. 2021 Oct 26;10(21):4937. doi: 10.3390/jcm10214937

Table 4.

Adjusted Cox proportional hazards analyses exploring associations of gene expression with outcomes.

Composite Primary Endpoint
HR 95% CI p-Value
FTMT Δct, T2 vs. T1 + T3 2.396 1.043–5.502 0.039
SIRT71 Δct, T2 + T3 vs. T1 5.495 1.784–16.923 0.003
SMIM20 Δct, T2 + T3 vs. T1 9.511 2.698–33.530 0.000459
TFRC Δct, T3 vs. T1 + T2 1.042 0.347–3.126 0.942
ADM Δct, T2 + T3 vs. T1 1.296 0.456–3.683 0.626
ECE1 Δct, T2 + T3 vs. T1 1.489 0.588–3.767 0.401
All-Cause Death
HR 95% CI p-Value
FTMT Δct, T2 vs. T1 + T3 4.448 1.497–13.216 0.007
SIRT71 Δct, T2 + T3 vs. T1 7.122 1.848–27.438 0.004
SMIM20 Δct, T2 + T3 vs. T1 7.44 1.882–27.930 0.03
TFRC Δct, T3 vs. T1 + T2 2.056 0.700–6.036 0.19
ADM Δct, T2 + T3 vs. T1 1.497 0.460–4.869 0.502
ECE1 Δct, T2 + T3 vs. T1 1.083 0.310–3.783 0.9

Adjusted Cox proportional hazards models using backwards methods. Δct (change in change in gene expression cycle threshold) gene expression tertiles according to generalized additive models of mRNA levels: T1 (lower expression), T2 (intermediate expression), T3 (higher expression). ADM (adrenomedullin), ECE1 (endothelin converting enzyme 1), FTMT (mitochondrial ferritin), TFRC (transferrin receptor), SMIM20 (small integral membrane protein 10), SIRT7 (sirtuin-7).