The Impact of Blood Pressure and Baroreflex Sensitivity on Wind-Up

Abstract

BACKGROUND: 

Elevated resting blood pressure (BP) and spontaneous baroreflex sensitivity (BRS) are associated with hypoalgesia to acute pain. These associations are significantly altered in chronic pain. We investigated whether degree of wind-up (marker for central sensitization) is similarly influenced by BP and BRS, and whether these associations are altered by chronic pain.

METHODS: 

BP and BRS (sequence method) were assessed in 30 healthy and 26 chronic back pain subjects who then completed a standardized thermal stimulation protocol to assess wind-up. This protocol was performed under placebo and α-2 adrenergic (ADRA2) blockade with yohimbine in counterbalanced order to test for the influence of ADRA2 mechanisms.

RESULTS: 

1) In healthy controls, higher systolic BP was associated with lower wind-up (P < 0.05) but this was reversed in chronic pain subjects (P < 0.05); 2) higher BRS was associated with lower wind-up in healthy controls (P < 0.05) but not in the chronic pain group; 3) higher systolic BP was associated with lower BRS only in the chronic pain group (P < 0.05); and 4) ADRA2 receptor blockade did not significantly affect wind-up.

CONCLUSIONS: 

These findings suggest that hypoalgesia associated with elevated resting BP and BRS in healthy individuals involves both diminished central sensitization (reflected in wind-up) and enhanced descending inhibition. The presence of chronic pain significantly alters the nature of these interactions. The reversal of normal interactions between overlapping systems modulating cardiovascular systems and pain in chronic pain patients may shift the healthy buffering of BP and heart rate toward instability and eventual higher BP and cardiovascular morbidity.

In Brief

IMPLICATIONS: Elevated resting blood pressure and greater spontaneous baroreflex sensitivity inhibit wind-up (an index of central pain sensitization) in healthy individuals, but not in individuals with chronic pain. Systems modulating cardiovascular function and pain overlap and chronic pain-associated dysfunction in these modulatory systems may be an important risk factor for cardiovascular disease.

Injury to tissues and peripheral nerves can result in enhanced responsiveness to pain transmission, depressed inhibition, and eventual amplification of pain due to peripheral and central sensitization.^1,2 Wind-up is a frequency-dependent increase in the excitability of spinal cord neurons that is evoked by afferent C-fiber stimulation.³ This phenomenon may underlie the continuation of pain sensation with prolonged or repetitive stimuli. Wind-up can be used as an investigative tool for central sensitization.³ Chronic pain patients exhibit greater wind-up than do pain-free controls.⁴

Previous work has documented functional interactions between the cardiovascular and pain regulatory systems.⁵ Studies indicate that elevated resting blood pressure (BP),^6–8 baroreceptor stimulation,⁹ and elevated spontaneous baroreflex sensitivity (BRS)¹⁰ all are associated with diminished responsiveness to acute pain stimuli in healthy individuals. Antinociception related to elevated BP may be mediated by a circuit involving baroreceptors, the nucleus tractus solitarius, the paraventricular hypothalamus, and the rostroventromedial medulla (RVM), which sends projections to the spinal cord via the dorsolateral funiculus.^11,12 This descending pain inhibition reduces pain transmission at the dorsal horn. If the critical frequency of C-fiber activation is not achieved, wind-up may not be induced. One aim of this study was to examine whether elevated BP and BRS are associated with reduced wind-up in healthy individuals.

The extent to which this hypothesized inverse relationship between wind-up and both resting BP and BRS might be affected by chronic pain is unknown. Chronic pain depends on the convergence of central descending modulation, spinal neuronal plasticity, and primary afferent drive.¹³ Persistent afferent input from injured peripheral nerves can produce neuroplastic changes in the RVM which results in descending facilitation of spinal nociceptive input.¹⁴ If sufficient temporal summation and recruitment of C-fibers occur as a consequence of enhanced spinal sensitivity, wind-up can be triggered.

Elevations in BP stimulate the baroreflex: baroreceptor output travels to the nucleus tractus solitarius and stimulates descending nociceptive inhibition.¹² BRS is diminished during periods of stress and pain.^10,15,16 Reduced BRS would reduce this baroreflex-dampening of pain. A net descending pain facilitation arising from the RVM, related in part to altered BRS,¹⁷ would further enhance afferent noxious input to the spinal cord and wind-up.^14,18 Chronic pain might therefore be expected to alter associations between wind-up and both BP and BRS. Prior work indicates that hypoalgesia related to BP and BRS is significantly altered by chronic pain.^19–21

A final aim of this study was to address potential modulators of wind-up. Animal studies suggests that wind-up can be modulated by both spinal and supraspinal sites of opioid and noradrenergic action.^22–26 Impaired α-2 adrenergic inhibitory activity (ADRA2) could increase wind-up in chronic pain.²² A final aim of this study was therefore to examine whether endogenous ADRA2 mechanisms modulate wind-up in humans, and explore the degree to which ADRA2 changes associated with chronic pain contribute to any alterations in the relationships between BP, BRS, and wind-up.

In summary, this study tested: 1) whether higher resting BP and greater spontaneous BRS were associated with lower wind-up in healthy controls, 2) whether these associations are altered in chronic pain patients, and 3) the extent to which ADRA2 mechanisms contribute to these relationships.

METHOD

Design

This study used a randomized, double-blind, placebo-controlled crossover design with administration of yohimbine, a selective ADRA2 receptor antagonist.

Sample

The sample consisted of 30 healthy pain-free controls and 26 individuals with chronic low back pain (LBP) recruited through online advertisements. Inclusion criteria were: age between 18 and 55 yr; current normotensive status; no history of cardiovascular disease, hypertension, liver or kidney disorders, or opiate dependence; no current daily use of opioid analgesics, and no current use of anti-hypertensive, opioid analgesic, or antidepressant medications. All subjects were free of current major depression, panic disorder, or posttraumatic stress disorder based on results of a structured clinical interview (Structured Clinical Interview for DSM-IV Axis I Disorders; 27). Potential subjects who were pregnant were excluded and all participants were asked to refrain from use of analgesics, nonsteroidal anti-inflammatory drugs, or any medications potentially affecting BP for 12 h prior to and avoid caffeine 3 h prior to study participation. Additional inclusion criteria for LBP subjects were persistent LBP of at least 3 mo duration and a typical daily pain intensity of at least 3/10 (0 = “No Pain” and 10 = “Worst Pain Possible”). Subject characteristics are summarized in Table 1.

Table 1: Subject Characteristics

Study Drug

The study drug was yohimbine hydrochloride, a selective pharmacological antagonist of ADRA2 receptors.²⁸ Yohimbine produces effective ADRA2 receptor blockade within 15–30 min after onset of infusion.^28,29 Yohimbine was compounded by the Vanderbilt Investigational Pharmacy (source: Professional Compounding Centers of America, Houston, TX) and administered (0.4 mg/kg) under an Investigational New Drug protocol filed with the Food and Drug Administration. Saline placebo was dispensed from Vanderbilt Investigational Pharmacy stock (Food and Drug Administration approved).

Assessment of Spontaneous BRS

The sequence technique was used to assess spontaneous BRS in the time domain.³⁰ This technique identifies spontaneous ramps in BP (progressive increases or decreases in BP) that are associated with concordant changes in R-R interval. BRS was indexed in this study through simultaneous assessment of continuous beat-to-beat R-R interval via 5-lead electrocardiogram and noninvasive beat-to-beat BP assessment using a Portapres (Finapres Medical Systems BV, Amsterdam, The Netherlands) placed on the middle finger of the non-dominant hand. Output was recorded by the WINDAQ data acquisition system (DI220, DATAQ, Akron, OH, 14 Bit, 500 Hz), visually inspected for artifacts, and processed off-line using software custom written in PV-Wave language (PV-wave, Visual Numerics Inc., Houston, TX). To derive the BRS values, slopes of the linear regression lines between each systolic BP (SBP) value and subsequent R-R interval values were calculated. Analyzable sequences were considered to be those with at least three intervals, and displaying correlation coefficients of r > 0.85, with BRS values calculated as the mean value of the significant slopes obtained. This custom software and BRS assessment technique have been used in several previous studies (e.g., 30).

Procedure

All procedures were performed at the Vanderbilt University General Clinical Research Center and were approved by the university IRB. All subjects provided informed consent before participation. All subjects participated in two laboratory sessions 1 wk apart, at the same time of day to control for circadian rhythms. Subjects remained seated upright in a comfortable chair throughout all laboratory procedures.

During each laboratory session, subjects initially completed a 10-min seated rest period, followed by a 5-min pre-drug resting BRS assessment period (as above). A research nurse under physician supervision then placed an indwelling venous cannula in the participant's dominant arm, followed by a 25-min adaptation period during which subjects rested quietly. A 10 mL dose of normal saline (placebo) or a 0.4 mg/kg dose of yohimbine hydrochloride (in 10 mL saline vehicle) was then infused over a 10-min period using an automated infusion pump.

After a 20-min rest following drug infusion to allow peak ADRA2 antagonist activity of yohimbine to be achieved, subjects participated sequentially in a 1-min pressure pain task using a Forgione-Barber finger pressure stimulator,³¹ a forearm ischemic pain task,³² and thermal pain threshold and tolerance trials using the thermal stimulation device described below. Full results for these non-wind-up stimuli are detailed elsewhere.^10,20

All subjects next participated in a thermal pain task using a Medoc Thermal Sensory Analyzer (TSA-II, Medoc, Inc., Ramat, Israel). This computer-controlled device was used to apply a controlled heat stimulus to the nondominant ventral forearm using a 30 × 30 mm Peltier thermistor probe as reported in prior studies (e.g., 33). Before wind-up assessment, a series of four pain threshold and four pain tolerance trials were conducted with results for the three most reliable trials within each series retained for analyses.²⁰ Subjects then participated in a series of wind-up assessment trials using commercially available software (TPS-CoVAS v3.19, Medoc Inc.) that administered a standardized oscillating thermal stimulation protocol designed specifically to assess C-fiber-mediated temporal summation. This protocol is based on that which has been used to assess temporal summation in several previous studies (e.g., 7).

During wind-up assessment, three sequences of 10 heat pulses each were applied to the dominant ventral forearm, with the thermode in a fixed position throughout each sequence. The thermode was moved to a different non-overlapping site in the same area for each subsequent sequence of trials, with a 2-min interval between sequences. Within each pulse sequence, 10 successive pulses (0.5 s duration) were administered from a baseline temperature of 40°C at a frequency of 0.5 Hz, a frequency known to elicit C-fiber mediated wind-up in the dorsal horn.² Three sequences with different target temperatures were conducted (47°C, 49°C, and 50°C in randomized order). At the peak of every heat pulse within each sequence, subjects were asked to provide a verbal numeric pain intensity rating using a 0–100 scale with which they had previously been familiarized (anchored with 0 = “No Pain or Warmth” and 100 = “Worst Possible Pain”). Because wind-up by definition is evidenced by increasing perceived pain intensity during repeated brief stimuli at a constant intensity (reflecting temporal summation), the slope of the line fitted to the series of 10 stimuli at each temperature was an optimal index of wind-up (Fig. 1 shows an example of placebo condition wind-up results in one healthy control and one LBP subject at 49°C). To derive such a measure for use in primary analyses, within-subject regressions were conducted regressing the series of 10 pain ratings during each wind-up trial on a dummy coded variable with a fixed interval of one (range = 1–10). This produced a standardized slope (β) for each individual at each temperature reflecting degree of wind-up. These standardized slopes were used as the dependent variables in primary analyses described below.

Figure 1.: Placebo condition wind-up data for one healthy control subject and one low back pain (0 subject over 10 trials with constant stimulus intensity of 49°C.

Statistical Analysis

All analyses were conducted using the SPSS for Windows Version 13 statistical package (SPSS, Inc., Chicago, IL). To reduce the number of analyses and minimize possible family-wise error rate issues, analyses were restricted to resting SBP and BRS data derived for sequences in which SBP was increasing given that the literature indicates that resting SBP and increases in SBP have stronger effects on pain responses than do comparable diastolic measures.^5,12 Primary analyses used a series of univariate General Linear Model (GLM) analyses to examine the main and interactive effects of resting BRS or SBP, subject type, and the study drug on the wind-up measures. Each analysis included resting pre-drug BRS or SBP (using the appropriate baseline for each drug condition), Subject type, and drug condition, and the two- and three-way interactions of these variables as the independent measures, with the dependent measure being the standardized slope of pain ratings during wind-up evaluation at each temperature. To control for the within-subject nature of the drug manipulation, a dummy coded subject ID variable was included as a control variable. Preliminary analyses indicated that drug order was significantly (P < 0.05) associated with magnitude of the effects of yohimbine on wind-up and therefore was included as a covariate in the primary analyses.

Significant interaction effects were followed-up with post hoc simple effects tests (i.e., GLM analyses within each subject type) to determine whether the individual effects contributing to the interaction were themselves significant. For the purpose of graphically portraying the source of significant interaction effects, median splits were conducted on mean baseline SBP and BRS variables and these were used to generate estimated marginal means in order to portray wind-up slopes by relatively higher and lower BRS (Mean BRS: High = 17.0 ± 1.04 ms/mm Hg, Low = 7.0 ± 0.49 ms/mm Hg) and SBP (Mean SBP: High = 112.6 ± 1.07 mm Hg, Low = 97.0 ± 1.17 mm Hg). These dichotomized cardiovascular variables were used solely for graphical presentations and were not used in primary analyses. BRS data were unavailable for nine subjects due to inability to detect a signal sufficient for reliable sequence analysis. All probability values reported are two-tailed with a P < 0.05 criterion for significance.

RESULTS

Preliminary Analyses

Table 2 summarizes results for placebo and yohimbine condition wind-up assessment by subject type. In the placebo condition, the LBP group displayed significantly larger wind-up slopes at all three target temperatures. These larger positive slopes indicate greater temporal summation of pain in subjects with chronic pain than in pain-free subjects. In the yohimbine condition, differences like those above were significant only for wind-up at 49°C.

Table 2: Mean (±se) Wind-Up Slopes Across Drug Conditions and Subject Types

Preliminary GLM analyses were also conducted to examine the relationship between resting SBP and resting BRS across subject types. These analyses indicated that SBP and BRS were significantly associated in the LBP group [F(1,38) = 5.52, P < 0.05] but not in controls (P > 0.10). This former association reflected diminished BRS in LBP subjects with higher BP (r = −0.36). This difference across groups approached significance [Subject Type X SBP interaction on BRS levels: F(1, 81) = 2.78, P < 0.10].

GLM analyses were also conducted to examine the degree to which wind-up was associated with measures of thermal pain threshold and tolerance. Main effects (across subject types and drug conditions) for wind-up slope at 47°C indicated that degree of wind-up significantly predicted both thermal pain threshold [F(1,98) = 11.31, P < 0.001] and thermal pain tolerance [F (1,98) = 7.38, P < 0.01]. Follow-up correlational analyses indicated that greater wind-up at 47°C was associated with both lower pain threshold (r = −0.31) and lower pain tolerance (r = −0.34). No other significant main or interaction effects were noted (P > 0.10). Similar analyses for wind-up at 49°C and 50°C revealed no significant effects (P > 0.10).

Associations Between Resting SBP and Wind-Up

A significant subject type X SBP interaction was found for wind-up slope at 47°C [F(1,98) = 4.81, P < 0.05]. This interaction is portrayed in Figure 2. Follow-up simple effects analyses indicated that higher resting SBP levels were associated with significantly lower wind-up in healthy controls [F(1,53) = 4.91, P < 0.05], but with significantly higher wind-up in LBP subjects [F(1,43) = 4.34, P < 0.05]. Other main and interaction effects in this analysis were nonsignificant (P > 0.10).

Figure 2.: Thermal wind-up at 47°C by resting systolic blood pressure level and subject type.

SBP analyses for wind-up slope at 49°C indicated a significant main effect of SBP [F(1,99) = 3.90, P < 0.05]. Examination of mean wind-up by relatively higher and lower SBP groups indicated that across subject types and drug conditions, individuals with higher resting SBP showed less wind-up (0.69 ± 0.05) than individuals with lower resting SBP (0.80 ± 0.05). Other main and interaction effects in this analysis were nonsignificant (P > 0.10), as were all effects in wind-up analyses at 50°C (P > 0.05).

Associations Between Resting BRS and Wind-Up

The relationship between spontaneous resting BRS and degree of wind-up was also examined. A significant subject type X BRS interaction was found for wind-up slope at 49°C [F(1,81) = 5.21, P < 0.05]. The source of this interaction is presented in Figure 3. Follow-up simple effects analyses indicated that greater spontaneous BRS was associated with significantly lower wind-up in healthy controls [F(1,42) = 10.16, P < 0.01], but BRS was not associated significantly with wind-up in LBP subjects (P > 0.10). Other main and interaction effects in this analysis were nonsignificant (P > 0.10). Similar BRS analyses for wind-up at 47°C and 50°C revealed no significant effects (P > 0.10).

Figure 3.: Thermal wind-up at 49°C by spontaneous baroreflex sensitivity and subject type.

DISCUSSION

Our results indicated that in healthy controls, higher resting SBP and greater BRS were associated with significantly lower thermal wind-up. These findings suggest that the hypoalgesia associated with elevated BP and baroreflex activation described previously^6–10 may arise, not only from greater activation of descending pain inhibitory pathways, but also from diminished activation of pain facilitatory pathways and lower central hyperexcitability.

These findings are not surprising in light of known functional links between brain structures contributing to both pain modulation and autonomic control. Key areas in the midbrain and the brainstem, such as the periaqueductal gray and the RVM allow bidirectional modulation of spinal cord activity through descending pain inhibitory and facilitatory networks.^23,24 The RVM may serve a protective role by increasing its inhibitory output or conversely, become maladaptive and permit long-lasting abnormal pain. The RVM is involved in the maintenance of tactile and thermal hypersensitivity, as observed in central sensitization,²³ and plays a pivotal role in the baroreflex.³⁴ Given overlapping physiological substrates, associations between cardiovascular variables and central sensitization (reflected in wind-up) are not surprising.

We further investigated whether the presence of chronic pain altered the inverse associations between cardiovascular measures and wind-up present in healthy subjects. Our findings indicated that in contrast to the significant inverse association between resting BP and wind-up in controls, chronic pain subjects displayed a significant positive association. Chronic pain appeared to affect associations with baroreflex measures similarly. A significant inverse relationship between spontaneous BRS and wind-up observed in pain-free controls was entirely absent in the chronic pain group. These findings parallel reported alterations in BP- and BRS-related hypoalgesia in chronic pain.^10,19–21 Mechanisms underlying observed changes in associations between cardiovascular variables and wind-up remain speculative. Tissue injury leading to persistent afferent input induces peripheral and central sensitization. The increased neuronal barrage at the spinal level activates spinal projection neurons and can lead to descending supraspinal facilitation in part via the RVM, which in turn plays a role in cardiovascular regulation.³⁴ Findings of impaired BRS in chronic pain patients^10,16 could further augment abnormal pain as a result of reduced baroreceptor-mediated analgesia.

Our results demonstrated that higher SBP was associated with lower BRS only in chronic pain subjects. These findings may have important prognostic implications for the pathogenesis of cardiovascular disease in individuals suffering from chronic pain who have increased hypertension risk, as prior work suggests.³⁵ Individual differences in baroreflex-mediated hypoalgesia predict subsequent increases in resting BP.³⁶ Depressed baroreflex function has been associated with increased cardiovascular morbidity,^37–39 including pathogenesis of hypertension.⁴⁰ Our results suggest that resting BP in chronic pain patients is increased to the extent that central sensitization processes are activated, whether this relationship is causal is yet to be determined. A genetically determined mechanism that leads to reduced BRS may also put a subgroup of chronic pain individuals at higher risk for cardiovascular disease, hypertension and end-organ damage depending upon their level of baroreflex and autonomic balance.^37,41–43 Acute musculoskeletal injury results in a shift in the autonomic nervous system toward a sympathetic dominance and significantly lower BRS,^44–46 and such findings may be relevant to long-term hypertension risk in chronic pain as well.⁴⁰

A final aim of this study was to test for noradrenergic modulation of wind-up and its contribution to associations with cardiovascular function. Both opioidergic and noradrenergic neurons have been shown to modulate C-fiber central hyperexcitability²² and to have analgesic synergy.^47,48 Moreover, both neurotransmitter systems are involved in cardiovascular regulation.⁵ In this study, we specifically tested the possible influence of ADRA2 mechanisms (which impact both pain and cardiovascular systems) on human wind-up. Yohimbine, an ADRA2 antagonist, did not alter wind-up in this study, nor did it contribute to the observed links between cardiovascular measures and wind-up. The expected increase in wind-up with yohimbine in the chronic pain group may not have occurred because of selective drug effects on the C-fiber-mediated input, rather than directly on the process of wind-up. Yohimbine could have exerted antinociception through serotonin1A receptors such that no change in wind-up was perceived.⁴⁹

The following study limitations are acknowledged. Relationships between cardiovascular measures and wind-up were observed at 47°C and 49°C, but not at 50°C. The neural processes that contribute to wind-up might evoke such profound neural discharges to the initial heat pulse at this latter target temperature that additional increases in neural discharges evoked by subsequent heat stimuli were difficult to achieve. That is, this higher temperature may have elicited a stronger wind-up response in all subjects, even those with relatively little central sensitization, which may have reduced wind-up variability. Another limitation is sample size and, therefore, statistical power issues may have contributed to some of the nonsignificant findings. Power issues could also have contributed to the lack of ADRA2 blockade effects, whereas examination of observed effect sizes suggest this is unlikely. A related issue is that blood levels of yohimbine and its metabolites were not obtained. Therefore, the degree of ADRA2 blockade at the time of wind-up testing could not be confirmed, whereas yohimbine's primary metabolite exhibits significant ADRA2 blockade actions with a 5 h half-life.²⁸ Finally, the issue of elevated Type I error rate due to multiple statistical tests must be considered. Although the number of primary analyses was not large (six), use of the bonferroni correction would have eliminated significant study findings. Although the bonferroni correction is considered by some to be overly conservative and may substantially inflate the risk of Type II error,^50,51 conclusions drawn from the current findings would clearly be strengthened by replication.

In summary, elevated resting BP and greater spontaneous BRS inhibit wind-up (an index of central pain sensitization) in healthy individuals, but not in individuals with chronic pain. The absence or reversal of normal interactions between overlapping systems modulating cardiovascular function and pain may contribute to impaired cardiovascular regulation and increased hypertension and cardiovascular risk in chronic pain patients.