Noninvasive method to estimate anaerobic threshold in individuals with type 2 diabetes

The maximal oxygen consumption (VO_2max) and anaerobic threshold (AT) are parameters that have been widely considered as hallmarks of aerobic fitness both for athletes [1], physically active individuals [2] and, in a minor scale, for special populations such as individuals with type 2 diabetes (T2D) [3, 4].

Regular exercise provides many physiologic benefits, reduces risk of disease outcomes, and triggers important psychological gains in healthy and pathological conditions. Moreover, few studies have evaluated the effect of different exercise intensities on T2D [5]. The AT can be used in this specific evaluation and has been considered a gold standard parameter for exercise prescription including cardiovascular risks groups [3, 6].

While several studies have identified the AT through the responses of blood lactate [2, 7, 8] ventilation [9] and blood glucose[7] others have suggested the response of the heart rate variability (HRV) as a method to identify the AT in young healthy individuals [10]. The HRV is a non-invasive measure of the oscillation between consecutive cardiac cycles (as measured between each R-R complex). This technique has been used for the noninvasive evaluation of the sympathetic (SNS) and parasympathetic nervous system (PNS) activity both during resting and low intensity exercise[11].

During low to moderate exercise intensity the heart rate (HR) increase is mainly controlled by the PNS withdrawal [12]. On the other hand, at higher intensities (e.g. above AT), there is a reduction in the parasympathetic modulation concomitant to an increase in sympathetic activity [13, 14] and thus a decrease in HRV is observed [15]. Such transition between the increase in the SNS activity and the vagal tone withdrawal occurs close to 60% VO_2max, being close to the intensity at which the lactate (LT) and ventilatory (VT) thresholds have been observed for healthy young individuals [10, 15]. However, the validity of the HRV in estimating the LT and VT for individuals who may have autonomic dysfunction associated with T2D [16‐18] has not been previously examined.

The HRV as a measure to estimate the AT is noninvasive and inexpensive, providing information regarding the autonomic regulation during exercise. In spite of presenting a higher sympathetic activity during resting and light exercise [19, 20] the HRV responses of T2D is thought to be sensitive to predict the LT and VT noninvasively by considering the intensity at which the vagal withdrawal occurs. However, the validity of HRV in estimating the LT and VT for individuals with T2D has not been investigated yet.

Thus, the purposes of present study were 1) to analyze the possibility of identifying the heart rate variability threshold (HRVT) by considering the responses of parasympathetic indicators during incremental exercise test, and 2) to compare and verify the relationships between the HRVT with the VT and LT for individuals with (TD2) and without type-2 diabetes (ND).

The results presented in table 2 show no significant differences among the workloads (W) corresponding to AT identified by different methods both groups. For both groups the HR, [Lac], VO₂ and RPE corresponding to the AT identified by different techniques did not differ from each other (Table 2). Further, for the T2D group significant correlations (p < 0.01) between LT and peak power (W_peak) (r = 0.89), VT and W_peak (r = 0.75), HRVT-RMSSD and W_peak (r = 0.80) and between HRVT-SD1 and W_peak (r = 0.80) were observed. High correlations (p < 0.001) were also found in the ND group for the LT and W_peak (r = 0.95), VT and W_peak (r = 0.92) and HRVT-RMSSD and W_peak (r = 0.89) and between HRVT-SD1 (r = 0.90). The relationship among exercise intensities (W) corresponding to the studied thresholds are presented in table 3 for both groups.

The Bland and Altman plots showed good agreement for the VO₂ corresponding to LT, VT, HRVT-RMSSD and HRVT-SD1 based on the low bias and relatively narrow limits of agreement [Bias (± 95% of confidence interval)] presented in both groups (T2D and ND) (Table 4 and Figure 2). In addition, the Kappa index also evidenced a good to excellent agreement among studied parameters for both T2D (K = 0.66 to K = 0.92) and ND (K = 0.62 to K = 0.92).

The ability of the HRV responses in estimating LT and VT was investigated in individuals with T2D and ND. The findings confirmed our hypothesis that AT could be identified in T2D through the parasympathetic markers of the HRV responses similarly to ND, being not different and highly correlated to the LT and VT for both groups (Tables 2 and 3, Figure 2). The HRV prediction of AT through autonomic variables such as RMSSD and SD1 has been studied in different populations [10, 15, 28, 29]. However, this is the first study in which the AT was identified through HRV responses in T2D, in spite of the fact that this disease is associated with autonomic dysfunction [30, 31].

The technique of HRV, as used in the present study, considered a progressive reduction of RMSSD and SD1 to intervals lower than 3 ms as markers of AT intensity as proposed by others [10, 26]. Lima and Kiss [26] found a high relationship between workloads (Watts) corresponding to HRVT and LT for healthy subjects (r = 0.76). In this study, even higher correlations were observed between LT and RMSSD threshold (r = 0.85 and r = 0.94) and between LT and SD1 threshold (r = 0.87 and r = 0.96) for T2D and ND groups, respectively.

The main findings of present study evidenced the HRV as an effective parameter for exercise evaluation in individuals with T2D, due to its noninvasive characteristic as well as because it evaluates variables associated with autonomic function. According to Yeckel et al. [30] the autonomic dysfunction usually precedes the occurrence of T2D. In addition, an elevated sympathetic activity is associated to a higher risk for cardiovascular events such as infarction and stroke [32, 33]. The HRVT seems to identify an exercise intensity that represents the transition from a lower to a higher sympathetic activity and thus to an increased probability for the occurrence of cardiovascular events T2D individuals. Therefore, the identification of the HRVT intensity may be a safe and useful procedure for exercise evaluation and prescription.

Another benefit of HRVT identification is the avoidance of elevated financial costs usually associated with invasive methods. Exercise tests involving blood collection and biochemical measures, and/or the use of expensive apparatus for expired gases analysis, as well as trained professionals dealing with such protocols are expensive and thus of difficult access to a major part of the population.

It has been suggested that vagal withdraw progressively occurs during IT, and is almost completely inhibited at intensities around 50-60% VO_2max[16, 29]. For the T2D participants of the present investigation, the vagal withdrawal, as observed by the RMSSD and SD1 stabilization, occurred at intensities corresponding to 64.8% and 66.6% of VO_2peak, respectively. Previous studies with T2D also verified similar results regarding LT occurring at intensities around 66% VO_2peak[4, 34, 35] what, in turn, may be of practical application.

The fact that participants of the present study had their HRVT being reached at 64.2 to 66.6% VO_2peak may be at least partially explained by their low physical fitness (VO_2peak of 20.6 ± 4.3 mL.kg^-1.min^-1 for T2D and 29.4 ± 6.8 mL.kg^-1.min^-1 for ND). This finding was also observed in a previous study comparing sedentary T2D individuals to physically active T2D and ND counter partners [35]. Moreira et al. also observed a trend towards lower [Lac] during the IT for T2D group, suggesting lower motor units recruitment and thus lower anaerobic glycolytic activity, lower power output, blood lactate and VO₂ at the exhaustion moment as compared to ND, similar to what happened in present study [34].

The RPE may also be an interesting noninvasive tool for exercise evaluation and prescription for both T2D and ND. The RPE integrates cardiovascular, respiratory, neuromuscular and metabolic inputs as a result of increasing exercise intensity [36]. Some studies have demonstrated that during incremental exercise test, RPE scores around 12-13, as measured through the 15-point Borg scale, is related to attainment of the AT [2, 9]. It was observed that RPE scores corresponding to AT in present study were around 12.9 ± 1.8 to 13.6 ± 1.4 for T2D, and 13.7 ± 1.2 to 14.6 ± 1.2 for ND (Table 2). The use of RPE for controlling exercise intensities may be an useful tool to ensure exercise prescription for T2D, once the RPE of 12-14 may help to delimitate the transition of moderate exercise to exercise intensities which eliciting blood glucose to increase [2, 8, 37, 38].

One of the limitations of present study was that the maximal lactate steady state (MLSS) intensity was not determined for comparison to the studied protocols. The MLSS determination is considered the gold standard method among those of exercise evaluation through blood lactate response to exercise. Denadai et al. [39] and Van Schuylenbergh et al. [40] have shown this parameter to be highly correlated and not different to the LT and VT intensities. So we may suggest that the methods adopted on present study (LT and VT determinations), which did not differ from HRVT, would also represent the MLSS intensity.

Several investigations have shown a relationship between the exercise intensity at which the vagal withdrawn is observed with those corresponding to VT [15, 41] and LT [10, 26]. In the present study the HRVT, as identified through the quantitative analyses of the Poincaré plots (SD1), occurred at exercise intensities not different to the LT and VT. On the other hand, in spite of non-detectable statistical differences, the exercise intensities associated to the occurrence of the HRVT-SD1 trend to be slightly lower than LT and VT for both T2D and ND, but without statistical difference (p > 0.05). Similarly to our findings, Shibata et al. [42] observed that HR corresponding to the HRVT was lower than VT, and the authors suggested that vagal withdrawal would precede the VT occurrence during IT. It is expected that mechanisms responsible for homeostasis during exercise may be related to the occurrence of both VT and LT, and is also dependent to the sympathetic/parasympathetic balance. Thus, it is reasonable that the HRVT occurs at exercise intensities closely related to LT and VT intensities. Thus, once the HRVT identified by the RMSSD seemed to better predicts the LT and VT as evidenced by the Bland and Altman analysis, we may suggest that RMSSD may be a more accurate variable and thus should be preferred in relation to SD1.

In conclusion it was possible to identify the HRVT from autonomic indices (RMSSD and SD1) for both T2D and ND, with no differences in relation to LT and VT. Furthermore, high correlations and good agreement between the examined methods were observed, suggesting the validity of HRVT even for T2D. Additional studies are necessary to investigate the reproducibility and sensitivity of AT as identified by HRV measures to various interventions, including physical training.