ABC | Volume 111, Nº1, July 2018

Original Article Oliveira et al Nonlinear dynamics in young adults with type 1 diabetes Arq Bras Cardiol. 2018; 111(1):94-101 Table 5 – Association of linear and nonlinear indexes of heart rate variability in the group with type 1 diabetes mellitus Index 0V% 1V% 2LV% 2ULV% SE SDNN r -0.494 0.043 0.522 0.357 0.478 ICC -1.204 0.058 0.410 0.295 0.037 95%CI -3.204- -0.156 -0.795-0.506) -0.126-0.690 -0.345-0.630 -0.836-0.495 RMSSD r -0.690 -0.081 0.550 0.627 0.650 ICC -3.337 -0.072 0.539 0.617 0.065 95%CI -7.271- -1.274 -1.045-0.438 0.122-0.759 0.270-0.799 -0.782-0.510 LF ms 2 r -0.692 0.034 0.721 0.495 0.669 ICC -0.054 0.000 0.000 0.021 0.002 95%CI -1.009-0.447 -0.907-0.476 -0.849-0.491 -0.866-0.487 -0.903-0.477 HF ms 2 r -0.365 0.036 0.469 0.202 0.381 ICC -0.032 0.001 0.020 0.010 0.001 95%CI -0.967-0.459 -0.905-0.476 -0.869-0.486 -0.888-0.481 -0.905-0.476 LF n.u. r 0.776 -0.080 -0.638 -0.601 -0.640 ICC 0.871 -0.295 -1.840 -1.575 -0.081 95%CI 0.754-0.932 -1.470-0.321 -4.416- -0.489 -3.911- -0.350 -1.062-0.433 HF n.u. r -0.776 0.080 0.638 0.601 0.640 ICC -6.750 0.228 0.648 0.612 0.075 95%CI -13.779- -3.064 -0.473- 0.595 0.329-0.815 0.260-0.796 -0.764-0.515 LF/HF r 0.776 -0.080 -0.688 -0.601 -0.704 ICC 0.321 -0.712 -0.658 -0.294 -0.702 95%CI -0.294-0.644 -2.264-0.102 -2.161-0.131 -1.468-0.321 -2.246-0.107 Values in bold represent p < 0.05. 0V: pattern without variation; 1V: pattern with a variation; 2LV: pattern with two similar variations; 2ULV: pattern with two different variations; SE: Shannon entropy; SDNN: standard deviation of all normal R-R intervals recorded in a time interval, expressed in milliseconds; ICC: intraclass correlation coefficient; 95% CI: 95% confidence interval; RMSSD: square root of the mean of the squared differences between adjacent normal R-R intervals, in a time interval, expressed in milliseconds; LF: low frequency component; HF: high frequency component; n.u.: normalized units. is five-fold higher than in patients without this complication, 26 being suggested as a diagnostic test by the American Diabetes Association® (ADA). 26 Some limitations should be pointed out, such as the cross‑sectional study design, which made it impossible to follow the individuals for a longer time, not allowing to know whether the influence of DM1 on the ANS would remain or worsen in the long term. Also, the anthropometric characteristics, such as weight and BMI, were different between the groups and higher in the group with DM1; however, the mean values were within the normal range, that is, below the values considered for obesity and overweight (BMI < 25 kg/m 2 ). 14 Despite the abovementioned limitations, some positive points should be emphasized, such as the use of new methods of non-linear HRV analysis, such as the symbolic analysis, which was able to identify autonomic alterations in DM1 and can be used to evaluate and monitor this population. Moreover, we observed that the symbolic analysis showed a moderate and high agreement with some indexes evaluated in this study, indicating that this method should be included in the HRV analysis, associated with the traditional indexes of the time and frequency domains, being crucial in the clinical follow-up of the ANS status in this population. However, considering the small number of studies with this population that used methods such as 99