RESUMO
O propósito deste estudo foi analisar o comportamento da frequência cardíaca (FC) versus a carga de trabalho crescente (CTC) em teste de esteira, utilizando três modelos matemáticos (linear, linear com dois segmentos de reta e sigmóide) e verificar qual o melhor modelo que possibilita a identificação de um limiar de FC que pudesse servir de preditor para os limiares ventilatórios (LV1 e LV2). Vinte e dois homens realizaram um teste incremental (re-teste: n=12), com velocidade inicial de 5,5 km.h-1 e incrementos de 0,5 km.h-1 a cada minuto, até a exaustão. Medidas contínuas de FC e trocas gasosas foram convertidas para médias de 5 e 20 segundos. Somatória dos resíduos quadrados e quadrado médio do erro foram usados para verificar o melhor ajuste. A relação FC/CTC foi melhor representada pelo modelo Lin2 no grupo teste e re-teste (p<0,05). Foi possível identificar um ponto de deflexão de FC, utilizando o modelo Lin2 (limiar de FC) em todos os indivíduos no teste (164 ± 16,6 bpm; 83,6 por cento FC MÁX) e no re-teste (162 ± 20,0 bpm; 83,9 por cento FC MÁX). O limiar de FC (Lin2PDFC) ocorreu a 9,2 ± 1,3 km.h-1 (67,9 por cento VelMÁX) e foi menor que o LV2 (LV2= 10,6 ± 1,5 km.h-1; 77,3 por cento VelMÁX; p< 0,05), mas não diferente de LV1 (8,4 ± 1,2 km.h-1; 61,6 por cento VelMÁX; p> 0,05). Durante teste incremental em esteira, a relação FC/CTC parece ser bem descrita por uma função linear com 2 segmentos de reta, a qual permite a determinação de um limiar de FC que se aproxima do LV1.
The objective of this study was to analyze the heart rate (HR) profile plotted against incremental workloads (IWL) during a treadmill test using three mathematical models [linear, linear with 2 segments (Lin2), and sigmoidal], and to determine the best model for the identification of the HR threshold that could be used as a predictor of ventilatory thresholds (VT1 and VT2). Twenty-two men underwent a treadmill incremental test (retest group: n=12) at an initial speed of 5.5 km.h-1, with increments of 0.5 km.h-1 at 1-min intervals until exhaustion. HR and gas exchange were continuously measured and subsequently converted to 5-s and 20-s averages, respectively. The best model was chosen based on residual sum of squares and mean square error. The HR/IWL ratio was better fitted with the Lin2 model in the test and retest groups (p<0.05). The Lin2 model permitted the identification of the HR threshold (Lin2HRDP) in all subjects of the test (164 ± 16.6 bpm; 83.6 percent HR MAX) and retest groups (162 ± 20.0 bpm; 83.9 percent HR MAX). Lin2HRDP (9.2 ± 1.3 km.h-1; 67.9 percent speedMAX) was lower than VT2 (10.6 ± 1.5 km.h-1, 77.3 percent speedMAX; p<0.05), but did not differ from VT1 (8.4 ± 1.2 km.h-1, 61.6 percent speedMAX; p>0.05). During a treadmill incremental test, the HR/IWL ratio seems to be better fitted with a Lin2 model, which permits to determine the HR threshold that coincides with VT1.
RESUMO
BACKGROUND: Volatile anesthetics have been shown to decrease baroreflex. In humans, cardiovagal baroreflex curve is best described to be sigmoidal and analysis of sigmoid model can provide more information, such as threshold (THR), saturation (SAT), maximal gain (Gmax), operating range (OR, difference between THR and SAT), operational point (OP, relative position of resting set point within the OR) in addition to the linear gain. We compared the changes in the sigmoid model of the systolic blood pressure (SBP)-RR interval relation between awake state and general anesthesia. METHODS: Sigmoid curve was obtained using the 'modified Oxford technique' (bolus sodium nitroprusside followed by bolus phenylephrine) to perturb SBP in awake group and 2 vol% sevoflurane anesthesia group. Radial arterial beat-by-beat SBP and raw RR interval data points were plotted and then binned in 2 mmHg SBP increments. Curve fittings were performed using symmetric four-parameter sigmoid model and asymmetric five-parameter sigmoid model. Gmax was calculated from the first derivative of the logistic function. RESULTS: Compared with awake control, general anesthesia caused decreases of linear gain, Gmax and OR. RR interval THR and SBP THR were unaffected, whereas SBP SAT increased and RR interval SAT decreased, resulting from reduced linear gain and curvature parameter. Resting set point fell and OP shifted significantly to THR region. CONCLUSIONS: Sigmoidal baroreflex curve revealed diminished gains, OR and OP during general anesthesia. Moreover, our results demonstrate that the analysis of sigmoid baroreflex arc across the entire baroreflex range provide more unique information beyond that of linear gain.