Altered Blood Pressure Reflexes in Women With Endometriosis

The data that support the findings of this study are available from the corresponding author upon reasonable request. The institutional Review Board at The Pennsylvania State University (“TxA2 and endometriosis” Study #21851) approved all experimental procedures and protocols. Verbal and written informed consent were voluntarily obtained from all participants before participation and in accordance with the guidelines set forth by the Declaration of Helsinki ( NCT05962034 ). All potential participants underwent screening to include a health history questionnaire, physical examination, medical screening, and a blood chemistry analysis (Chem 24, Quest Diagnostics, Pittsburgh, PA). Twelve women with clinically diagnosed endometriosis (nine diagnosed with laparoscopy, three diagnosed by a gynecologist from symptomology, ultrasound imaging, and family history) and nine healthy women participated in the study. Participants were free of cardiovascular incident history and lacked vascular, neurological, pulmonary, and dermatological disease. Participants were not using over-the-counter or prescription medications with known primary or secondary effects on pressor responses at the time of testing. Participants were instructed to avoid non-steroidal anti-inflammatory use at least 2 weeks prior to each experiment visit. A complete list of medications used by the participants is detailed in Table 1 . All participants were born with a uterus and at the time of testing had at least one intact ovary. Participants were instructed to avoid caffeine and alcohol for 24 hours and strenuous exercise at least 12 hours prior to their experiment visits. Participants were given a capsule of either 650 mg aspirin, a non-selective COX inhibitor, or placebo (cellulose; indistinguishable capsules) with instructions to take the capsule 1 hour prior to their experiment ( 29 , 30 ). Participants were blinded to their intervention until conclusion of their participation; investigators were not blinded. Experiments took place at least 2 weeks apart to minimize possible carryover effects. Upon arrival at the laboratory, participants were instrumented with a three-lead ECG and finometer beat-by-beat blood pressure monitoring system (Finapres ® NOVA, Finapres Medical Systems, Enschede, Netherlands) on the left hand calibrated to the contralateral brachial blood pressure (Connex Spot Monitor, WelchAllyn, Skaneateles Falls, NY). For all procedures, participants were seated upright with feet flat on the floor and instructed not to hold their breath or move their extremities during data collection. The order of the pressor testing was randomized by subject but repeated in the same order within subjects. Baseline data was recorded over a 5-minute period. At the end of the baseline period, participants submerged their right hand in a bucket of ice water (4 – 8 °C) up to the radial styloid process with their hand open and unclenched. At the end of 3 minutes, the participant withdrew their hand which was promptly towel-dried and placed in their lap. Recovery was monitored over the course of 15 minutes. HR and BP were continuously monitored throughout this protocol. Participants were asked to rate their discomfort on a scale of 0–10 using a visual analog scale in the final seconds of baseline and every minute throughout the procedure and recovery until discomfort returned to baseline values. Participants were asked to squeeze a hand grip dynamometer in their right hand as tightly as possible 3 times to determine their maximal voluntary contraction (MVC) and then were allowed to rest. Baseline data was recorded for 5 minutes. At the end of the 5 minutes, participants were instructed to hold a static hand grip of 30% MVC with an allowable range of 28 – 32% MVC for 2 minutes aided with second-by-second visual feedback. The exercise was considered failed if it did not average 25 – 35% MVC over the course of 2 minutes. In the final seconds of this hand grip exercise, a blood pressure cuff was inflated to suprasystolic pressure (~210 – 220 mmHg). The participant released the dynamometer and the cuff remained inflated for 3 minutes. At the end of 3 minutes, the cuff was deflated and recovery was monitored for 5 minutes. HR and BP were continuously monitored throughout this protocol. Participants were asked to rate their discomfort on a scale of 0–10 in the final seconds of baseline and every minute throughout the procedure and recovery until discomfort returned to baseline values. ECG and finometer data were continuously recorded and stored for offline analysis (PowerLab and LabChart, ADInstruments, Bella Vista, NSW, Australia). Change in BP (ΔBP) and change in HR (ΔHR) responses to the CPT were calculated as the difference between ~30-second moving average of the peak response to submersion for the CPT and the moving average of the ~30-second peak BP and HR recorded during baseline. ΔBP and ΔHR responses to the HG+PEI were calculated as the difference between ~30-second moving average of the peak response to hand grip and occlusion and the moving average of the ~30-second peak BP and HR recorded during baseline. ΔSBP, ΔDBP, ΔMAP, ΔHR, and pain responses were analyzed with linear mixed effects model corrected for a priori multiple comparisons with Bonferroni’s method (corrected for direct comparison of ΔSBP, ΔDBP, ΔMAP, and ΔHR between endometriosis and HC during HG and PEI separately; SAS v. 9.4; Cary NC). Time course of MAP response to exercise was calculated as the difference of the averaged 5-second intervals of MAP at the onset of exercise and the average baseline MAP and was analyzed with linear mixed effects model (SAS v. 9.4). Power analysis (G*Power 3.1 Software; α = 0.05, effect size = 0.3, power = 0.95) initially determined 9 subjects per group, but we included additional participants to account for potential withdrawal or failure during testing as well as different methods of diagnosis of endometriosis. Baseline HR, SBP, DBP, and MAP was analyzed with two-way ANOVA (GraphPad Software). Subject characteristics were analyzed with unpaired t-tests.

Participant characteristics are listed in Table 1 . Results are represented as mean ± SD unless otherwise specified. Subjects were comparable between groups for age and body size. Women with endometriosis had lower HDL compared with healthy controls (endo 51 ± 12 vs HC 62 ± 9 mg/dL; p = 0.03). Resting HR (endo 79 ± 11 bpm vs HC 77 ± 7 bpm; p = 0.40) and BP (SBP: endo 109 ± 9 vs HC 106 ± 10 mmHg, p = 0.50; DBP: endo 72 ± 7 vs HC 70 ± 6 mmHg, p = 0.32; MAP: endo 85 5 ± 7 vs HC 82 ± 7 mmHg, p = 0.39) was not different between groups. Data representing pressor and cardioaccelerator responses to CPT are shown in Fig 1 . No participants from the HC group withdrew during the CPT. Among the endo group, 2 participants removed their hand from the cold water prematurely during the placebo and ASA trials, an equipment failure occurred during 1 placebo and 1 ASA trial, and 2 participants withdrew from the study after completing their placebo but before completing their ASA experiment (citing schedule conflicts not related to the study). In total, for CPT analysis, HC placebo n = 9, HC ASA n = 9, Endo placebo n = 9, Endo ASA n = 7. In response to CPT, Women with endometriosis demonstrated lower ΔSBP (p <0.01), ΔDBP (p <0.01), and ΔMAP (p <0.01) in response to cold water hand submersion compared with the healthy control group. There were no differences in heart rate response between groups (p = 0.40). There were no differences in pain ratings between groups (p = 0.55; Fig 3 ). A single dose of 650 mg aspirin did not impact ΔSBP (p = 0.61), ΔDBP (p = 0.65), and ΔMAP (p = 0.57) in either the endometriosis group or the healthy control group. There was no effect of COX inhibition on heart rate response to CPT in either group (p = 0.93). There were no differences in pain ratings between treatments (p = 0.51). Data representing pressor and cardioaccelerator responses to HG+PEI are shown in Fig 2 . Onset of pressor and cardioaccelerator response to exercise is shown in Fig 4 . No participants from the HC group withdrew during HG+PEI, nor were there any failed HG attempts in this group. Among the endo group, 1 participant failed to maintain an HG force output within 25–35% of MVC over 2 minutes, an equipment failure occurred during 1 ASA trial, and 2 participants withdrew from the study after completing their placebo but before completing their ASA experiment (citing schedule conflicts not related to the study). In total, for HG+PEI analysis, HC placebo n = 9, HC ASA n = 9, Endo placebo n = 11, Endo ASA n = 9. There were no differences in MVC between groups (endo 215 ± 64 N v HC 282 ± 84 N; p = 0.07). Women with endometriosis demonstrated attenuated ΔSBP (p < 0.01), ΔDBP (p < 0.01), and ΔMAP (p < 0.01) responses to HG+PEI compared with healthy women. There was no main effect of phase (HG vs PEI) on ΔSBP (p = 0.67), ΔDBP (p = 0.45), and ΔMAP (p = 0.88). However, direct comparisons revealed women with endometriosis demonstrated attenuated ΔSBP in response to HG (p < 0.01) and PEI (p = 0.03). ΔDBP was attenuated in women with endometriosis compared to healthy women in response to HG (p < 0.01) and PEI (p = 0.01). ΔMAP was attenuated in women with endometriosis compared to healthy women in response to HG (p < 0.01) and PEI (p = 0.02). The onset of pressor and cardioaccelerator responses ( Fig 4 ) was different between groups where women with endometriosis did not have an increase in MAP in the first 20 seconds of exercise HG (p < 0.01) nor an increase in HR at exercise onset (p = 0.02). ΔHR was not different between groups (p = 0.18). Pain response was not different between groups (p = 0.15; Fig 3 ). There was no effect of COX inhibition on the ΔSBP (p = 0.07), ΔDBP (p = 0.17), and ΔMAP (p = 0.11) responses to HG+PEI. Both groups showed delayed pressor responses at the onset of HG following ASA (p < 0.01) such that HC did not increase MAP until the 20 second time point, and women with endometriosis showed a decrease in MAP in the first 15 seconds. ASA augmented the cardioaccelerator response at exercise onset (p < 0.01). There was no effect of ASA on ΔHR response (p = 0.48). ASA did not affect pain rating in response to HG+PEI (p = 0.17).

We initially hypothesized that, related to increased COX activity, women with endometriosis would demonstrate augmented pressor responses to sympathoexcitatory maneuvers including CPT and HG+PEI. Contrary to our hypothesis, women with endometriosis demonstrated attenuated pressor responses to both sympathoexcitatory maneuvers, and COX inhibition with ASA did not appreciably impact the responses in either group. Beyond this, women with endometriosis showed delayed onset of pressor response to HG+PEI compared with healthy women which persisted following ASA treatment, even dipping in the endometriosis group. Cardioaccelerator responses to both CPT and HG+PEI were not different by group and were unaffected by COX inhibition; however, the cardioaccelerator response to exercise onset was significantly lower in endometriosis and augmented by ASA in both groups. There was no evidence of hyperalgesia to these stimuli as subjective pain responses were not different between groups and were unaffected by ASA. These findings are notable because we, to the best of our knowledge, are the first to provide direct physiological evidence of autonomic dysregulation in response to sympathoexcitatory stimuli in women with endometriosis, a disease considered to be gynecologic in origin and associated with increased lifetime risk of major adverse cardiovascular events. The CPT is a noxious thermal stimulation of cutaneous afferents which induces a reflex arc, increasing muscle sympathetic nerve activity (SNA), resulting in increased BP and HR ( 31 ). Most frequently, this technique is utilized in assessments of pain tolerance in humans ( 32 ) but is a useful non-invasive tool in investigating integrative neurovascular control ( 33 ). Our finding that women with endometriosis displayed attenuated pressor responses to CPT suggests either alteration in sympathetic outflow and/or changes in central integration in this population. It has been postulated that there is a shift in distribution of fiber type in tissue surrounding ectopic endometrium such that there is a dearth of sympathetic innervation and a surplus of afferent innervation related to the inflammatory milieu of the lesions ( 34 , 35 ). It is therefore possible that reduced sympathetic nerve fiber varicosities rather than attenuated SNA alone may explain our pressor findings. In healthy young men, the CPT maneuver increases BP, shifting the homeostatic baroreflex set point, without altering baroreflex sensitivity ( 36 ). It is likely that alterations in central integration of reflex arcs ( 37 , 38 , 39 , 40 ) in women with endometriosis result in this population being unable to alter the homeostatic set point, resulting in attenuated SNA outflow in response to noxious cold exposure compared with healthy individuals. Our findings of altered neural control of the circulation in women with endometriosis is further supported by our finding of delayed pressor and cardioaccelerator response to the onset of HG exercise. Central command, a feedforward mechanism of modulating pressor responses to exercise, increases the homeostatic blood pressure setpoint via the arterial baroreflex ( 41 ). Denervating the sinoatrial nodes of the heart negates the reflex increase in blood pressure and HR at the immediate onset of exercise, ( 42 ) and can even induce a rapid decrease in MAP with no change in MSNA ( 42 , 43 , 44 ). In other investigations interrogating the role of COX in this reflex arc show divergent results. In healthy humans COX has not been shown to play a role in baroreflex sensitivity when inhibited systemically ( 45 ). COX inhibition in isolated carotid sinus preparations from rabbits attenuates baroreflex sensitivity ( 46 ), specifically with indomethacin or high – not low – concentrations of ASA ( 47 ), while perfusing COX metabolites (PGI 2 ) increased baroreflex sensitivity ( 46 , 47 ). We show in the present study both women with endometriosis and healthy women were unable to rapidly mount pressor responses at the onset of exercise following ASA, whereas healthy women alone were able to accomplish this on placebo. We additionally demonstrated a decrease in MAP at the onset of exercise in women with endometriosis only following COX inhibition with ASA. We speculate that enzymatic products of COX play a necessary role in baroreflex resetting at the onset of exercise in young, reproductive-aged women, and that women with endometriosis are unable to implement this adaptation through a pathophysiological mechanism related to COX activity. In the EPR, thinly myelinated group III and IV afferents within the skeletal muscle are stimulated by the mechanical and metabolic actions of muscle contraction, activating a reflex arc that increases SNA, resulting in increased BP and HR ( 48 ). While there was no main effect of phase (HG vs PEI), it is noteworthy that within phase comparisons show the difference in pressor responses to HG exercise is maintained once the mechanical activity of muscle contraction is removed (i.e., isolating the metaboreflex). Further there was no significant difference between groups in MVC, and no relation between force production and peak pressor responses. This indicates either that both the mechanoreflex and the metaboreflex or the metaboreflex alone are attenuated in endometriosis. Regarding the mechanoreflex component, it is possible there are alterations in the mechanically-gated channels in the groups III and IV sensory muscle afferents such that afferent responsiveness to a given mechanical perturbation of muscle contraction is less likely to occur. It has been previously shown that transient receptor potential (TRP) receptor activation produces a tachyphylactic effect on the mechanoreflex ( 49 ). TRP messenger RNA as well as expression of its agonists, endocannabinoids, are upregulated systemically in women with endometriosis ( 50 ). Regarding the metaboreflex component, multiple potential explanation exists, the most likely decreased transduction of sympathetic nerve activity to the vasculature (efferent output). There may also be decreased intramuscular metabolite generation/accumulation and/or differences in muscle fiber type distribution that impact afferent stimulation ( 51 , 52 , 53 ). In our interpretation of the results of the HG+PEI testing, it is important we acknowledge that, although it did not meet statistical significance, women with endometriosis displayed slightly lower MVC than healthy women despite nearly identical BMI between groups. A previous study examining physical performance in women with endometriosis reported lower MVC despite greater BMI in an endometriosis group compared with women without endometriosis ( 54 ). The added context of similarly attenuated pressor responses to the CPT suggests this was not an influential factor in the attenuated pressor responses demonstrated by women with endometriosis. There was also no relation across groups between force production during the HG test and peak pressor responses ( Fig 5 ). However, this data may reflect attenuated end-organ responsiveness in women with endometriosis. It is possible that sensory input, central integration, and sympathetic outflow are all well-maintained in endometriosis, but alterations of expression or sensitivity of β-adrenergic receptor in the vasculature ( 55 , 56 ) and nicotinic acetylcholine receptors at the neuromuscular junction ( 57 ) exist such that vasoconstriction and muscle contraction are relatively impaired in this population. The finding that women with endometriosis did not rate their perceived pain to either maneuver differently from that of the healthy groups is surprising given previous findings in animal models of endometriosis ( 15 , 58 , 59 ) and that women with endometriosis demonstrate hyperalgesia ( 6 , 60 ). Immunohistochemical evidence suggests that endometriosis lesions directly sensitize peripheral nerve fibers resulting in augmented perceived pain ( 61 ). The present finding of indistinguishable pain rating between groups may be related to the mode of pain induction. Women with endometriosis and associated myofascial pain symptoms were reported to experience lowered pressure-pain thresholds in distant somatosomes ( 60 ). Thermal sensitivity to both hot and cold in those experiencing chronic pelvic pain appears to be heightened, however this was not specific to endometriosis, and was tested only in the lower abdomen, lower back, and on the dominant lower limb near the abdominal trunk ( 62 ). Data from pre-clinical models of endometriosis relate hyperalgesia to hot noxious stimuli rather than cold, potentially explaining the discrepancy between the animal models and the human data collected here. Taken together, this may suggest any alteration in the perception of pain in women with endometriosis is specific to regions surrounding lesion innervation or perhaps the mode of noxious stimuli, but elevated perceptions of pain in endometriosis is not the result of altered central integration at higher brain regions alone. It is unsurprising there was no effect of COX inhibition on the magnitude of pressor responses in the healthy group ( 63 , 64 , 65 , 66 ). The lack of impact of ASA treatment on pressor or pain responses in the present clinical group may be related to the diverse enzymatic products of COX. With a non-specific inhibition of COX activity giving an acute oral dose of ASA, we may have been preventing the production of the vasodilator prostacyclin in addition to blocking action of thromboxane and prostaglandins on sensory afferents. It is additionally possible that, if sympathetic nerve fibers are so greatly downregulated as to attenuate the ability to perform systemic sympathetic vasoconstriction, increases in afferent sensitivity induced by COX activity may be proportionally missed. In other words, a COX-mediated increase in sensory afferent sensitivity would increase SNA responses to sympathoexcitatory stimuli, but the vasoconstrictor effect of increased SNA may not be observed if there are fewer distal sympathetic innervations. Future research should investigate the possibility of alterations in peripheral afferent sensitivity with afferent nerve activity recordings to investigate this possibility. There are a few limitations to the present study to be considered. First, we excluded few participants on the basis of medications. The two women with endometriosis taking norepinephrine-dopamine reuptake inhibitor medications were asked not to take this medication the day of testing. Based on the mechanism of action of these drugs ( 67 ) we would anticipate an augmented pressor response not an attenuated response when the reuptake of NE is inhibited. None of the other medications utilized by the endometriosis group have known effects on pressor responses to sympathoexcitatory stimuli. Second, more women with endometriosis were utilizing oral contraceptive pills (OCP) than in the healthy group. While this was not ideal, the data examining the impact of OCP on pressor responses is contradictory ( 68 , 69 ), but suggests OCP augment pressor responses in healthy women ( 70 , 71 ), specifically in the present HG+PEI protocol ( 72 ). Additionally, the variety of pharmacologic formulations for OCPs is not ideal for rigorous investigation. We posit including these women in this sample adds to the ecologic validity of these data without meaningfully altering the results. Third, we also must acknowledge the limitation of utilizing a non-specific COX inhibitor when the COX-2 isoform was the basis of our hypothesis. We chose ASA intervention as a proof-of-concept approach that maintained ecologic validity as an over-the-counter pain reliever that targeted our proposed mechanism of action. Exploring the effect of COX-2 selective inhibitors is worth exploring in future research. Fourth, because the HG exercise resulted only in a low total ΔHR response for both groups, we feel interpreting the findings of our onset of cardioaccelerator is limited; future research should utilize a more robust exercise mode to investigate this phenomenon. Fifth, we did not directly measure sympathetic nerve activity or conduct a thorough investigation into the baroreflex. The CPT and EPR require the totality of the reflex arc – sensory afferent stimulation, central integration, sympathetic outflow, end-organ response – to produce increases in blood pressure. Therefore, it is necessary for us to discuss possible sites of alterations related to endometriosis to explain our data; however, future research should directly investigate the exact mechanism creating this discrepancy between healthy women and women with endometriosis. The findings of the present study support the notion that endometriosis systemic disease that is not limited to the reproductive system. In fact, this study suggests autonomic dysregulation in endometriosis, whether as a cause or a consequence of the pathophysiology is unknown. This finding of attenuated pressor responses may also explain previous reports of orthostatic intolerance in this population ( 73 ) as well as the common comorbidity of endometriosis with postural tachycardia syndrome ( 74 ). Clinicians should consider the possibility of neurologic symptoms in this population related to this disease process. Our results indicate that women with endometriosis were unable to mount similar blood pressure responses to sympathoexcitatory maneuvers such as cold exposure and exercise compared to a healthy group of women. Despite attenuated pressor responses, women with endometriosis rated their perceived pain to these maneuvers similar to that of healthy groups. COX activity, while not a mediator of the magnitude of the pressor response in this group, may play a role in the mechanism of attenuated pressor response via altered baroreflex resetting and attenuated sympathetic outflow or end-organ responsiveness. Future research is needed to delineate the mechanisms of impaired pressor responses in women with endometriosis. Endometriosis is a systemic inflammatory condition that appears to have pathophysiological manifestations in the autonomic nervous system. Women with endometriosis respond to common sympathoexcitatory maneuvers with attenuated blood pressure responses. Future research should investigate the precise mechanisms attenuating the reflexive increase in blood pressure following sympathoexcitatory stimulation in women with endometriosis. This information can be used to inform larger clinical trials and where relatively simple non-invasive autonomic testing might be incorporated as a component of diagnostics and multidisciplinary management of endometriosis care.

Cardiovascular disease (CVD) is the leading cause of death worldwide. Endometriosis is a female-specific risk factor for CVD; specifically, women with endometriosis are at 50% increased risk of myocardial infarction and 39% increased risk of stroke ( 1 , 2 ). Endometriosis, estimated to be present in ~10% of women ( 3 , 4 , 5 ,), is a gynecologic condition characterized by invasive extrauterine endometriotic lesions, chronic pain, hyperalgesia, and systemic inflammation ( 6 , 7 , 8 , 9 , 10 , 11 ,). Women with endometriosis demonstrate marked endothelial dysfunction, characterized decreased nitric oxide (NO)-mediated vasodilation, in both the macrovasculature ( 12 . 13 ) and microvasculature ( 14 , 13 ). This vascular dysfunction is not improved following a broad knockdown of NF-κB-mediated inflammatory pathways ( 13 ). However, the activity of the inducible form of cyclooxygenase-2 (COX-2), a key enzyme in inflammation upregulated by NF-κB, is upregulated in endometriotic lesions, resulting in an increase in the downstream metabolite thromboxane A2 (TxA2; 15 ). Increased expression of COX-2 and TxA2 activity is a pathological mechanism identified in both clinical populations and pre-clinical models of CVD and is associated with sensory afferent sensitization. The upregulation of this pathway results in an augmented exercise pressor reflex (EPR; 16 , 17 , 18 , 19 , 20 , 21 ), or the reflex increase in blood pressure following the initiation of physical activity. The EPR is a reflex arc wherein the mechanical activity of muscle contraction (mechanoreflex) and the metabolic perturbation of muscle contraction (metaboreflex) stimulate intramuscular afferents to reflexively increase heart rate and blood pressure necessary to support adequate perfusion of skeletal muscle during activity. An augmented EPR is associated with increased risk of mortality ( 22 ), specifically increasing the risk of myocardial infarction and stroke. In human patients with CVD as well as pre-clinical models of CVD, COX inhibition attenuates augmented pressor responses ( 17 , 18 , 21 ). A polymorphism for the COX-2 gene has been identified in women with endometriosis from distinct populations (Brazilian and Korean women; 23 , 24 ). Increased expression of COX-2 has been identified in ectopic endometriotic lesions, eutopic endometrium, and peritoneal macrophages in women with endometriosis ( 25 , 26 , 15 , 27 ). In a pre-clinical model of endometriosis, upregulation of the COX-2 pathway is related to hyperalgesia and knocking down COX-2 is associated with increased tolerance to a noxious stimulus ( 15 ). It is well established that endometriosis-related pain is often neurally-mediated ( 28 ); however, the relation between COX activity and integrative neurovascular function in patients with endometriosis is yet undetermined. Therefore, the purpose of this study was to determine if the inflammatory milieu characteristic of endometriosis impacts integrative neurovascular control in this population. To examine this, we performed multiple tests designed to activate different components of pressor reflexes in women with endometriosis and healthy women with placebo and following aspirin, a non-selective COX inhibitor. We hypothesized that women with endometriosis would demonstrate augmented pressor reflexes to sympathoexcitatory stimuli compared with healthy women, and that this difference would be mediated, at least in part, by COX activity.