Assessment of autonomic neuropathy in type 1 diabetics by cardiovascular reflex testing and heart rate variability: A case series

Patients with diabetes mellitus suffer from pathological involvement of the nervous system. Autonomic neuropathy is a dysfunction of the nerves that has the function of regulating functions such as perspiration, blood pressure, and heart rate. Cardiovascular autonomic neuropathy (CAN), despite being considered a common chronic complication in patients with diabetes mellitus (DM) and associated with increased cardiovascular morbidity and mortality, still remains underdiagnosed [1],[2]. The signs and symptoms associated with the presence of CAN appear late in the course of the disease, making it necessary to use cardiovascular lex tests for early identification of CAN [3],[4]. Considering the scarcity of data in the literature on the prevalence of CAN in the Brazilian population with type 1 DM (DM1), this case series has a relevant role in the early identification of this complication as well as the determination of possible clinical and laboratory factors related to its presence.

The prevalence ranges from 2.6% to 90% among patients with DM [5], depending on the diagnostic method used and increases with age, duration of diabetes, and presence of inadequate glycemic control [6]. It is usually associated with peripheral sensory-motor neuropathy of the lower limbs (in up to 62.5% of cases) as well as with the presence of other cardiovascular risk factors (CVR), such as hypertension, dyslipidemia, diabetic nephropathy and retinopathy, and peripheral vascular disease (PVD).

Cardiovascular autonomic neuropathy has been pointed out as one of the possible causes of sudden death and represents a strong predictor of CVR in both DM1 and DM2. Diabetic patients with coronary artery disease (CAD) have mortality rates of 16–53% in five years, depending on the severity of the CAD. In the DIAD study (detection of silent myocardial ischemia in asymptomatic diabetic subjects), carried out in patients with DM2, the presence of CAN, determined by the reduced Valsalva's ratio, was predictive of silent myocardial ischemia, regardless of the presence of traditional RCV factors such as high blood pressure (HBP), age, gender, and smoking. In the EURODIAB (European Epidemiology and Prevention of Diabetes) study, CAN was identified in 1/3 of patients with DM1 and was associated with other CVR after adjustment for age, glycated hemoglobin (HbA1c) levels, and duration of diabetes [3]. Data from the ACCORD (Action to Control Cardiovascular Risk in Diabetes) study demonstrated that patients with CAN had a mortality rate about 1.55–2.14 times higher than patients without CAN.

Considering the scarcity of data in the literature on the prevalence of CAN in the Brazilian population with DM1, this case series have a relevant role in the early demonstration of this complication.

This case series will report two clinical cases of patients diagnosed with diabetes mellitus type 1 (DM1) for over five years, where patient 1 presented alterations in the Ewing test, for evaluation of diabetic autonomic neuropathy and heart rate variability (HRV). And patient 2 was presented within the normality standard.

Patient 1 was a 25-year-old Caucasian female, who diagnosed with DM1 14 years ago, was overweight [body mass index (BMI) 27.0], abdominal circumference 108 cm, normotensive, normocardic, normoglycemic, denied hypertension, alcoholism, and smoking, and practiced physical activity regularly. She reported symptoms such as dyspnea, dizziness, visual disturbances, pre-syncope when standing, constipation, urinary incontinence, polyuria, urinary urgency, repeated urinary tract infection, and hyperhidrosis. She has no other comorbidities.

Patient 2 was a 20-year-old Caucasian male, who diagnosed with DM1 seven years ago, presented with ideal weight (BMI 21.1), abdominal circumference 84 cm, normotensive, normocardic, normoglycemic, denied hypertension, alcoholism, and smoking, and practiced physical activity regularly. As for the symptoms questionnaire, he was asymptomatic. He denied any associated comorbidities.

Both the patients were evaluated on the same day, and were previously instructed not to ingest caffeine, alcoholic beverages, and oral medications such as antidepressants, anxiolytics, and cardiovascular drugs, for a period of 8 hours before the tests and the suspension of physical activity for 24 hours.

We show in images the graphs from the Polar Flow system acquired during the tests. In Figure 1 we present the HRV analysis of patient 1 and in Figure 2 the same analysis of patient 2. We excluded the first 5 minutes and the final 5 minutes.

The HRV analysis involved time and frequency domains. For the time domain, the following indices were selected.

• SDNN: RR intervals between each normal heartbeat, expressing sympathetic nervous system interference in the heart.
• pNN50: percentage of adjacent iRR (interval between two waves r) values greater than 50 ms. Represents parasympathetic influences on iRR, because actions controlled by the parasympathetic nervous system are faster than those modulated by the sympathetic nervous system; when greater than 50 ms, they may signify greater vagal interference in heart function.
• RMSSD: square root of the sum of the square of the difference between iRR. Similar to pNN50, RMSSD expresses parasympathetic nervous system interference in the heart, and the higher its value, the greater the vagal action in the heart.

For the frequency domain we use the variables described below:

LF: low frequency component (0.04–0.15 Hz), whose values express sympathetic cardiac tonus, although some authors report a certain vagal influence on these values.

HF: high frequency component (0.15–0.4 Hz), whose values express parasympathetic cardiac tonus.

In the analysis of the data described in Table 1, it shows that patient 1 presented results regarding the time analysis within the parameters of normality, which indicates that the cardiac automatism is under influence of the sympathetic nervous system that is superposing the parasympathetic system. Patient 2, on the other hand, presents with altered results, i.e., higher than the normality level, mainly in pNN50 with a value of 0.56 heart. Remembering that values higher than 0.50 represent the parasympathetic influences in the iRR, because the actions controlled by the parasympathetic nervous system are faster than those modulated by the sympathetic nervous system; when greater than 50 ms, they can mean greater vagal interference in the functioning of the heart.

Cardiac activity during the orthostatism test, for patient 1 is shown in Figure 3 and the same result for patient 2 is shown in Figure 4. This test is altered for the two patients analyzed, showing a bradycardia in the first minute that represents a higher than expected vagal activity, which is demonstrated in Table 2, should be between 1.17 for both the patients. Patient 1 presented the index of 1.34 already whereas patient 2 with the value of 1.42, as already reported higher than normal indices.

The third test to which they were submitted was the respiratory test, where Figure 5 demonstrates the outcome of patient 1 and Figure 6 of patient 2. Table 3 shows the analysis of the graphics and calculation of the respiratory coefficient. It is noted that only patient 2 achieves the expected coefficient for age.

Figure 7 and Figure 8 show the Valsalva test result of patients 1 and 2, respectively. Table 4 shows the result obtained and the expected for normality according to age.

We used results from two patients where one was symptomatic and the other was asymptomatic for autonomic neuropathy. They were approximately the same age, which may have interfered in the symptomatology that is the duration of the DM1.

The HRV test showed that in the time domain, patient 1 presented a RMSSD slightly higher than the SDNN value, which means more vagal control over cardiac function; however, the PNN50% value did not indicate this parasympathetic activity. Patient 2, on the other hand, presented an SDNN higher than the RMSSD, indicating prevalence of sympathetic stimulation over the heart. When we look at the frequency domain, we can confirm the high influence of the parasympathetic system with a variable of HF higher than LF, which represents sympathetic activity. In contrast the asymptomatic patient showed prevalence of sympathetic activity with LF extremely higher than HF [1],[2].

In the orthostatic and postural hypotension tests, the patients showed indices within the normal range [7].

In the breath test and the waltz test, we observed an alteration in the results of the symptomatic patient, reaching an index lower than expected for his age. On the other hand, the asymptomatic patient was above the expected index. These tests evaluate the parasympathetic component of the autonomic nervous system. The result of two altered tests confirms the diagnosis of CAN [4],[5].

Symptomatology is a striking feature in the diagnosis of diabetic autonomic neuropathy. Ewing’s tests are essential to close the diagnosis, and also to evaluate and diagnose the asymptomatic cases that may be present.