In academic review articles describing morbidity [1] and mortality [2] risk in patients with diabetes mellitus, Vinik et al. include P&S monitoring in their list of “commercially available computer programs … employed to assess autonomic nerve function.” From an autonomic perspective, the cascade of secondary symptoms typically attributed to diabetes as defined by Vinik et al. and the many cardiovascular risk factors as discussed by DePace et al. [3, 4] are effects of progressively accelerating autonomic dysfunction. This has led DePace et al., in another academic review article, to conclude that “It is unequivocally established that CAN is associated with increased cardiac morbidity and mortality. Identifying and addressing CAN early, especially in a subclinical cardiac patient, will further differentiate which asymptomatic patients require more aggressive therapy. The results from P&S testing documenting CAN may be used as a baseline. One should view these test results as a guide towards more individualized treatment. A more specific selection of medications and dosing based on these results is possible. P&S test results represent objective data which are useful in guiding pharmacological and lifestyle changes. In addition to normalization and improvement of CAN independent, simultaneous P&S testing providing unambiguous parasympathetic and sympathetic activity levels, may guide the physician towards the type and dosing of pharmacological agents necessary to achieve an objective clinical target or outcome. The pharmacopeia includes adrenergic (beta-blockers, anti-hypertensives, bronchodilators and vasopressors) and cholinergic, (anti-depressants, anxiolytics, and anti-psychotics) agents. This would eliminate arbitrarily dosing medications without a clear target outside of HR and BP. Also, the threshold for implanting of prophylactic devices such as cardiac defibrillators may be better defined by assessing and following P&S dysfunction. While further studies are indicated, the clinical and epidemiological data is too compelling not to test for, diagnose, and aggressively treat CAN with abnormal SB to guard the patient’s well being, not only in diabetics, but in all patients with risk factors for heart disease.”

A 2008 vasopressor outcome study [5] compares HRV-alone analyses (both time and frequency domain) with P&S monitoring and demonstrated a greater ability in describing patient outcomes than any of the HRV-alone results. In 184 consecutive, arrhythmia-free, insulin-dependent diabetic patients with previously diagnosed orthostatic hypotension, with persistent dizziness, P&S monitoring demonstrated persistent autonomic dysfunction (sympathetic withdrawal upon standing) and more accurately described the known hemodynamic effects pre- and post-dosing with an alpha-1 adrenergic agonist. Arora et al. conclude, “independent, simultaneous assessment of progressive parasympathetic and sympathetic dysfunction, autonomic imbalance, and the responses of the two ANS branches to therapy enables early detection and early intervention. Orthostatic dysfunction, by way of example, illustrates that frequent, sensitive assessments of both ANS branches can improve the negative outcomes associated with CAN.”

A recently concluded study compares P&S monitoring to HRV-alone monitoring in 2,745 patients (41.5 % female) diagnosed with autonomic neuropathy. The results from both monitoring methods are correlated with: known normals, symptoms of advanced autonomic dysfunction (or DAN in diabetics), symptoms of CAN, age, body mass index (BMI), stage of autonomic decline, and level of autonomic risk as defined by P&S monitoring, HRV-alone monitoring, resting HR, resting BP, and clinical trending. All measures are well correlated (p < 0.05) with age, BMI, resting HR, and resting BP. While different HRV-alone parameters are individually well correlated with known normals (E/I ratio, p = 0.040), symptoms of advanced autonomic dysfunction or DAN (LFnu and HFnu, p < 0.050), symptoms of CAN (sdNN, p = 0.050), stage of autonomic decline (sdNN, p = 0.089), and level of risk (TSP, p = 0.079), there is a lack of consistency. All P&S monitoring parameters are well correlated with all factors (p < 0.020). In measures of specificity, P&S monitoring consistently demonstrates significantly greater positive and negative predictive rates with fewer false-positive and false-negative rates.

Another recently concluded study compares P&S monitoring to HRV-alone monitoring in 2,745 patients (41.5 % female) diagnosed with autonomic neuropathy, including treatment for orthostatic dysfunction. The results from both monitoring methods are correlated with the prevalence of orthostatic dysfunction, including orthostatic hypotension (OH, an abnormal drop in BP upon standing), postural orthostatic tachycardia syndrome (POTS, an excessive increase in HR upon standing), postural orthostatic bradycardia syndrome (POTS, an abnormal decrease in HR upon standing), and postural orthostatic hypertension syndrome (POHS, an excessive increase in BP upon standing). For comparison, orthostatic dysfunction is defined both in terms of hemodynamics alone and hemodynamics with autonomic function (sympathetic withdrawal or SW). Total orthostatic dysfunction patients (2,408) are more than the total SW patients (1,127). The discrepancy is because not all patients are neurogenic; some have only vascular dysfunction. Assuming any decrease in BP with upright posture is abnormal, 2,175 patients demonstrate a decrease in BP of more than 5/0 mmHg (clinical and preclinical OH). The P&S method identifies 71.9 % with a false-negative and false-positive rate of 12.3 and 10.1 %, respectively (p = 0.0076). The HRV-alone method identifies 22.3 % with a false-negative and false-positive rate of 26.8 and 21.4 %, respectively (p = 0.0144). POTS is diagnosed in 0.9 % of the patients. The P&S method identifies 48.0 % with a false-negative and false-positive rate of 1.5 and 0.3 %, respectively (p = 0.0187). The HRV-alone method identifies 28.0 % with a false-negative and false-positive rate of 2.6 and 0.4 %, respectively (p = 0.0838). POBS is diagnosed in 3.5% of the patients. The P&S method identifies 45.4 % with a false-negative and false-positive rate of 6.3 and 0.04 %, respectively (p = 0.0251). The HRV-alone method does not identify any POBS. POHS is diagnosed in 4.2 % of the patients. The P&S method identifies 70.0 % with a false-negative and false-positive rate of 21.6 and 0.5 %, respectively (p = 0.0036). The HRV-alone method identifies 11.8 % with a false-negative and false-positive rate of 64.1 and 2.8 %, respectively (p = 0.0853).

Akselrod and the original researchers from MIT and Harvard validated P&S monitoring and its applications throughout many fields of medicine [6–38]. These articles demonstrate specificity, sensitivity, reliability, and repeatability of the method. The HRV standard articles [39, 40] reference the MIT/Harvard work, recognizing its validity. The CPT code 95943 is written based on the MIT/Harvard method.

Differences between the P&S monitoring and HRV-alone monitoring are highlighted in the patient example below (adapted from Aysin and Aysin [41]). A slice of the time-frequency analysis of DB is presented in Fig. 4.1. The upper left graph in Fig. 4.1 is the cardiogram (instantaneous HR plot) in response to six slow, relaxed, deep breaths over 60 s. The upper right plot is the corresponding wavelet analysis results with time on the abscissa, frequency on the ordinate, and frequency (or spectral) amplitude in gray scale (black is zero and white is maximum). The second row presents similar plots for the RA and corresponding wavelet analysis. At each time slice in the RA time-frequency spectrum (a sample is presented in the third row of Fig. 4.1), the FRF is determined and transferred to the HR spectrum (the corresponding sample is presented in the bottom row of Fig. 4.1). Given the FRF, the RFa area is determined (again, the area under the HR spectral curve within the parasympathetic activity window [42], represented in blue, centered on the FRF). The remaining portion of the LF region then represents sympathetic activity. For comparison, the LF and HF regions [39, 40] are also represented on the HR spectrum plot in this example. The apparent significant discrepancies are tabulated; see Table 4.2 [41 (Table 1)]. The region corresponding to parasympathetic modulation (RFa) is missed in HRV-alone analysis. Of course, the isolation of P from S in this example is clear and easy for illustrative purposes. However, in typical, clinical situations, the discrepancies are often more subtle and are the reason for much of the lack of general clinical acceptance.

The results from patients (Table 4.2) show that the HRV-alone method (tHRV) is neither sufficient nor accurate for assessing P&S activity. Table 4.2 data show that during normal breathing (NB), the HF measure is poorly associated (p > 0.050). The data also show that during deep breathing (DB), the HRV-alone LF term increases significantly, suggesting that the LF term captures the expected parasympathetic response. Typically, this would be interpreted as a sympathetic response, whereas the RFa term (from P&S monitoring) captures the parasympathetic activity. Aysin and Aysin [41] conclude, “it has been well established that HRV analysis with and without respiratory [activity] information provides different results” (citing Triedman, Perrott, Cohen, and Saul [14]). Aysin and Aysin continue by stating “Although the differences were more obvious with low breathing rates, using [the P&S method] in general would provide more accurate results, due to accurate localization of parasympathetic activity in the HRV signal” [41].

Aysin et al. also compare the HRV-alone and P&S monitoring methods during orthostatic challenge [43]. In this study, Aysin et al. included in the comparison both spectral analysis and time domain ratio measures from HRV alone (see Table 4.1). One hundred healthy volunteers participated in this study. The methods included 5 min of rest, in a relaxed seated position with sturdy back support, and then a quick PC to standing, followed by 5 min of quiet standing. As documented in Bloomfield et al. [44], this active standing maneuver results in P&S responses that are “nearly identical to the changes in autonomic balance that occurs during the first 5 min of quiet standing.” Their results suggest “nearly identical” is approximately 96.3 %. These results were repeated, using hemodynamic measures as biomarkers [45].

As expected, Aysin et al. [43] find a statistically significant decrease in the parasympathetic activity as measured by P&S monitoring (RFa, p < 0.005) when the patient assumes a head-up posture. A similar result is also found with the HRV-alone measures that are considered to represent parasympathetic activity (i.e., HF, rmsSD, pNN50; p < 0.005). The P&S monitoring measure, SB (=LFa/RFa), and the HRV-alone measure, LF/HF, significantly increase from baseline to standing (p < 0.005), as does the P&S monitoring sympathetic activity measure (LFa, p = 0.023). However, the changes in the HRV-alone measures, which are assumed to represent sympathetic activity (sdNN and LF), were not statistically significant (p > 0.050). Aysin et al. observe that this “result is important in that significant changes in sympathetic activity is properly identified only by using HRV Spectral analysis with RA analysis [the P&S method]. Although sympathetic activity is not directly related to respiration, parasympathetic activity is, as its measure is computed based on respiration. Therefore, using RA analysis in HRV analysis should make a difference. The reason that LF was not significantly different between baseline and stand is explained by the fact that the LF area includes both P and S modulations resulting from varying respiratory rates. If an individual has low respiratory rates, Vagal (parasympathetic) activity will contribute to the LF area and power in LF area will be mixed. Therefore, change in sympathetic activity from sitting (baseline) to stand will not be reflected properly. In clinical settings, detection of significant changes is very important in determining treatment [43]. Aysin et al. [43] conclude that “even under normal breathing rates, [P&S Monitoring] seems to provide the most accurate estimates of both ANS branches, and therefore would be a better clinical tool in diagnosing and treating medical conditions resulting from autonomic dysfunction.”

Aysin and Aysin [41] also presented a second example with a breathing rate that is >0.15 Hz and satisfies the HRV-alone spectral analysis criterion for respiratory frequency [39, 40]. In this example (see Fig. 4.2), the respiratory frequency while the subject is standing is 0.26 Hz (15.6 breaths per minute). The top two plots in Fig. 4.2 present the independently and simultaneously captured HBI (cardiogram) and respiratory activity (“RA”) signals, respectively. A higher respiratory rate is obvious and reflected in the RSA in the cardiogram (HR plot, top). The next two plots are the HR and RA time-frequency results from the two independent and simultaneous wavelet analyses, respectively. The higher frequency content in both signals is apparent. The bottom two plots are single slices from the RA and HR spectra, respectively. As depicted, the FRF is detected from the RA spectrum and transferred to the HR spectrum. Then from the HR spectrum, parasympathetic activity is computed from the area under the HR spectrum within the blue-shaded area (RFa). Notice this RFa is completely contained within the HF. The more narrow RFa is a more accurate parasympathetic measure than the broader HF measure. From Figs. 4.1 and 4.2, both RFa and LFa computations result in an increased specificity over the HRV-alone methodology.

P&S assessment defines independent, simultaneous measure of P and S activity. So what?!? Examples will be demonstrated throughout the rest of this book. In general, independent, simultaneous assessment of P&S activity enables physicians to quantify and document patients’ responses to therapy, disease, lifestyle, and history. It also simultaneously measures parasympathetic activity (RFa) and sympathetic activity (LFa) without extraneous signal content (noise). Specific, independent, simultaneous measures of P&S activity enable a true measure of autonomic balance (SB = LFa/RFa). Wavelets enable P&S measures for arrhythmia patients through the ability to analyze the transient nature of arrhythmia [46]. Wavelet spectral analysis resolves the shortcomings of the Fourier transform (FFT), enabling customized and individualized interventions. The results are reduced morbidity and mortality, reduced medication load and hospitalizations and reduced healthcare costs, and improved patient outcomes.

In HRV-alone, the Parasympathetic response is often lost in the Low Frequency measure, inflating the assumed Sympathetic Response, masking orthostatic dysfunction, and causing false positive indications of Sympathetic Excess.

The next section is composed of excerpts from a manuscript first published as an abstract entitled “Enhanced frequency domain analysis replaces older heart rate variability methods.” The abstract was accepted at the Fourth Annual Diabetes Technology Meeting, Philadelphia, PA, 28–30 October 2004, and presented as a poster [47]. This manuscript details the differences between P&S monitoring and HRV alone for DB and Valsalva over the adult age range. The original data (as presented below) are based on a cohort of 389 patients (161 females, 354 type 2 diabetics and 35 type 1 diabetics). This is a subpopulation from a patient cohort of 511 patients diagnosed with type 2 diabetes (248 females, age = 63.4 ± 13.1, range 25–96) [48].