TABLE 4.
Unadjusted modela |
Adjusted modelb |
|||||
β | P-value | Q-valuec | β | P-value | Q-valuec | |
Active standing | ||||||
Δ LFn (n.u)†¶ | −0.531 | 0.001 | 0.003 | −0.528 | 0.019 | 0.034 |
Δ LF/HF†¶ | −0.639 | < 0.001 | < 0.001 | −0.643 | 0.001 | 0.003 |
Paced breathing | ||||||
Δ LFn (n.u)‡ | −0.658 | < 0.001 | < 0.001 | −0.716 | < 0.001 | < 0.001 |
Δ LF/HF‡ | −0.620 | < 0.001 | < 0.001 | −0.935 | < 0.001 | < 0.001 |
aSimple linear regression with the Δ HRV index as independent variable and the annual change in the cognitive Z-score as dependent variable, bMultiple linear regression adjusted for age, sex, education, physical activity, morbidity (additive index), resting HRV and baseline cognitive Z-score, cP-value corrected for multiple testing by means of the Benjamini–Hochberg procedure with a 5% False Discovery Rate (FDR), †Dependent variable: annual change in episodic memory Z-score (model 1), ‡Dependent variable: annual change in executive functioning Z-score (model 2), ¶One outlier removed from the analyses. Significant results are shown in bold typeface. MCI, Mild Cognitive Impairment; β, standardized regression coefficient; LFn, normalized low frequency power; n.u, normalized units; LF/HF, low frequency power to high frequency power ratio; Δ index, index during challenge – index at rest.