Figure 2
From: Sex and race differences in J-Tend, J-Tpeak, and Tpeak-Tend intervals

Summary of intra-subject standard deviations of QTc (top panel), JTc (middle panel) and JTpc (bottom panel) intervals obtained with Framingham formula using the average of last 3 RR interval preceding the interval measurement (Fram 3 RR int), Framingham formula using the 10-second average of RR intervals (Fram 10 s RR), subject-specific linear formula using the average of last 3 RR interval preceding the interval measurement (Lin 3 RR int), subject-specific linear formula using the 10-second average of RR intervals (Lin 10 s RR), subject-specific linear formula using universal rate hysteresis correction (Lin hyster), and curvilinear subject-specific formula using subject-specific hysteresis correction (Curvilin hyster). The population mean + standard deviations of the intra-subject standard deviations are shown for each of the sex- and race-defined (African vs Caucasian) sub-population. In each of these sub-populations, the intra-subject standard deviations of all the intervals obtained with curvilinear model + hysteresis correction were statistically significantly smaller compared to all other correction possibilities (p < 0.001 for all).