Mechanism of loss of consciousness during vascular neck restraint

Jamie R. Mitchell, Dan E. Roach, John V. Tyberg, Israel Belenkie, Robert S. Sheldon

Abstract

Vascular neck restraint (VNR) is a technique that police officers may employ to control combative individuals. As the mechanism of unconsciousness is not completely understood, we tested the hypothesis that VNR simply compresses the carotid arteries, thereby decreasing middle cerebral artery blood flow. Twenty-four healthy police officers (age 35 ± 4 yr) were studied. Heart rate (HR), arterial pressure, rate of change of pressure (dP/dt), and stroke volume (SV) were measured using infrared finger photoplethysmography. Bilateral mean middle cerebral artery flow velocity (MCAVmean) was measured by using transcranial Doppler ultrasound. Neck pressure was measured using flat, fluid-filled balloon transducers positioned over both carotid bifurcations. To detect ocular fixation, subjects were asked to focus on a pen that was moved from side to side. VNR was released 1–2 s after ocular fixation. Ocular fixation occurred in 16 subjects [time 9.5 ± 0.4 (SE) s]. Pressures over the right (R) and left (L) carotid arteries were 257 ± 22 and 146 ± 18 mmHg, respectively. VNR decreased MCAVmean (R 45 ± 3 to 8 ± 4 cm/s; L 53 ± 2 to 10 ± 3 cm/s) and SV (92 ± 4 to 75 ± 4 ml; P < 0.001). Mean arterial pressure (MAP), dP/dt, and HR did not change significantly. We conclude that the most important mechanism in loss of consciousness was decreased cerebral blood flow caused by carotid artery compression. The small decrease in CO (9.6 to 7.5 l/min) observed would not seem to be important as there was no change in MAP. In addition, with no significant change in HR, ventricular contractility, or MAP, the carotid sinus baroreceptor reflex appears to contribute little to the response to VNR.

  • unconsciousness
  • baroreceptor

vascular neck restraint (VNR) is a technique that police officers may employ to control combative individuals. When properly applied, the typical subject loses consciousness within 5–11 s, followed by a full recovery with no lasting medical complications (15, 16). The technique applies lateral compression to the vascular, muscular, and neural structures of the neck, suggesting several physiological mechanisms that might play a role in rendering the subject unconscious. We hypothesized that VNR compresses the carotid arteries and decreases cerebral blood flow sufficiently to produce cerebral hypoxia and loss of consciousness (3, 10, 16). However, the carotid sinus baroreceptor reflex may also cause bradycardia and hypotension and decreased cerebral blood flow. The Valsalva maneuver (characterized by forced strain against a closed glottis) might also contribute to hypotension (5). Increased intrathoracic pressure might increase external constraint to the heart and cause blood to pool peripherally, thereby limiting ventricular filling and reducing stroke volume (SV) and arterial blood pressure (11, 12).

To test the hypothesis that simple compression of the carotid arteries is the main mechanism by which VNR causes loss of consciousness, we measured bilateral middle cerebral artery blood flow velocity (MCAV) and bilateral external neck pressures during VNR in healthy volunteers. To assess the potential contributions of various other mechanisms, we collected noninvasive hemodynamic data and air flow measurements.

METHODS

The study protocol was approved by the Institutional Conjoint Health Research Ethics Board. All subjects provided written informed consent prior to their participation.

Twenty-four healthy police officers (27–40 yr of age, 3 women) who had no prior relevant medical history and who were free of medications were recruited to attend one clinical laboratory session. Exclusion criteria included a history of orthostatic hypotension, cardiac or cerebrovascular disease, or cardiac rhythm disturbances. Subjects refrained from consuming caffeine-containing foods or beverages after the previous midnight. Height, weight, and neck circumference were recorded and body mass index (BMI) was calculated. Arterial pressure was measured from the middle finger of the nondominant hand referenced to mid-heart level with an infrared finger photoplethysmograph while heart rate (HR) was measured with an electrocardiogram (Nexfin, model 1, BMEYE Cardiovascular Intelligence, Amsterdam, The Netherlands). Blood flow velocity was measured in the right (R) and left (L) MCA by transcranial Doppler ultrasound (TCD; TCD100M, Spencer Technologies, Seattle, WA) using probes positioned bilaterally against the temporal skull above the zygomatic arch. Once a satisfactory velocity signal was obtained, the probes were fixed in position using a cranial harness. Bilateral external neck pressures were measured using flat, fluid-filled balloons taped in place over the carotid bifurcations and connected to pressure transducers (model P23 ID, Statham Gould, Oxnard, CA). The balloons were placed at the projected intersection of the superior border of the thyroid cartilage and the lower angle of the mandible (8). In 12 subjects, additional balloons were taped in place over the largest projection of the L and R sternocleidomastoid muscles (SCM). To monitor respiration, nasal airflow was recorded with an airflow pressure sensor connected to a nasal cannula (Ultima Dual Airflow Pressure Sensor, model 0585, Braebon Medical, Carp, ON).

Experimental protocol.

After instrumentation, baseline data were collected for 60 s (baseline 1). A single certified VNR instructor from the Calgary Police Service Skills and Procedures Unit positioned himself behind the seated subject placing his right (dominant) arm in VNR position without applying any pressure. As the positioning of the instructor may have changed baseline conditions, additional data were collected for 60 s (baseline 2). The subject was then instructed to focus visually on a fluorescent pen while an investigator seated directly in front of the subject moved the pen from side to side. After data collection was started, VNR was applied maximally. Data collection was continuous during VNR and for 60 s following its release. VNR was stopped 1) if the subject tapped the instructor's arm indicating the desire to terminate; 2) 1–2 s after ocular fixation (indication of unconsciousness) (16); or 3) the instructor stopped after a prolonged period (∼20 s) with no ocular fixation. Data were then collected for 60 s every 2 min during a recovery period lasting 15 min. The time at which the subject indicated his/her wish to stop the VNR, the time it took to induce ocular fixation, or the time at which the instructor chose to stop the VNR was recorded.

Data analysis.

The conditioned bilateral external neck pressure signals were amplified, passed through a low-pass filter (100 Hz), and digitized at 200 Hz (Sonometrics, London, ON). The digitized data were analyzed using software (CV Works, Calgary, AB) developed in our laboratory. The Nexfin monitor uses a combination of the volume clamp method and Physiocal criteria as previously described (14, 20) to measure finger arterial pressure. To overcome the most important drawback of the early Finapres devices (6), it reconstructs the distal finger arterial pressure into the more central brachial arterial pressure waveform (2) that was recently validated (17). The Nexfin uses the successor of the Modelflow algorithm as previously described (21) to estimate real time SV from the arterial waveform and a three-element windkessel model of LV afterload (1). This next generation algorithm was recently validated in intensive care unit patients (1) and heart failure patients (18).

PMDViewer software from Spencer Technologies (Seattle, WA) yielded TCD velocity waveform data sampled at 125 Hz. Custom Matlab (Mathworks, Natick, MA,) software used these waveform data to produce interbeat interval (IBI) sequences, which were matched to IBI sequences produced by Nexfin. MCA blood flow velocities were averaged over the cardiac cycle and expressed as mean velocity (MCAVmean; as defined by the matching beat times in Nexfin) and were considered to be a reflection of cerebral blood flow.

In cases where minimum MCAVmean were not equal on both sides during VNR, time to minimum MCAVmean was selected from the side with the lowest value. Except where indicated, data are expressed as means ± SE obtained during the last 30 s of the indicated period except for during VNR, which are mean values during 5 s of minimum MCAVmean. Percent changes during VNR were calculated as changes from baseline 2 values.

Statistical analysis.

Statistical comparisons were performed using SigmaPlot (Systat Software, 2008). Linear correlations were calculated for all indicated variables (y = y0+a*x). A paired t-test was used to test for the significance of changes between baseline 2 and VNR data. A t-test was used to test for the significance of changes between groups. Repeated-measures ANOVA (Student-Newman-Keuls method) was used to test for the significance of changes at different points in time; a P value <0.05 was considered statistically significant.

RESULTS

Table 1 lists subject characteristics and anthropometric measurements by sex.

View this table:
Table 1.

Subject characteristics

Table 2 describes the responses to VNR. A majority of the subjects (16/24) lost consciousness; of the subjects who did not stop prematurely (Ocular Fixation Group and Instructor Stop Group), VNR was applied effectively in 80% (16/20). Of note; one subject within the Subject Stop Group signaled termination just before the instructor released the restraint at 22 s; time to stoppage for the three others was 7.3 ± 0.8 s.

View this table:
Table 2.

Responses to VNR

Figure 1A shows the placement of the VNR instructor in relation to the subject, while Fig. 1B shows bilateral balloon pressure transducers over the L and R carotid bifurcations and SCMs.

Fig. 1.

A and B: positioning of the vascular neck restraint (VNR) instructor in relation to the subject and the bilateral positioning of the balloon pressure transducers over the left (L) and right (R) carotid bifurcations and sternocleidomastoid (SCM), respectively.

Hemodynamic effects of VNR.

Figure 2, A and B, illustrates representative responses in beat-to-beat changes before, during, and after VNR in the Ocular Fixation and Instructor Stop Groups, respectively. The vertical hatched lines indicate the start and release of VNR (not indicated is ocular fixation occurring at 7.8 s after start of VNR in Fig. 2a). Figure 2a shows pressures recorded over the L and R SCM and L and R carotid bifurcations during VNR. Pressure over the R SCM rapidly approached 600 mmHg in the Ocular Fixation Group example and exceeded that measuring limit over the L SCM in both examples. Also, pressure over the R carotid bifurcation reached 200 and 400 mmHg in the Ocular Fixation and Instructor Stop Groups examples, respectively, and was slightly greater than over the L carotid. Figure 2b shows R and L MCAVmean decreasing to 0 cm/s for a sustained period during VNR in the Ocular Fixation Group example, while the L MCAVmean only decreased to 30 cm/s in the Instructor Stop Group example. Figure 2c illustrates the consistent MAP during the entire period in the Ocular Fixation Group example while MAP initially decreased then increased during VNR before returning to baseline after VNR release in the Instructor Stop Group example. Figure 2d shows decreased SV during VNR with an immediate recovery to pre-VNR levels on release while HR remained unchanged throughout in the Ocular Fixation Group example, while the opposite occurred in the Instructor Stop Group example. Figure 2e shows a small decrease in dP/dt during VNR which immediately recovered to pre-VNR levels on release in the Ocular Fixation Group example, while dP/dt remained consistent in the Instructor Stop Group example. Figure 2f shows the phasic changes in air flow during respiration. Immediately after the start of VNR, a final inspiratory effort occurred as indicated by the positive airflow deflection that was followed by no airflow for the duration of VNR in the Ocular Fixation Group example. The Instructor Stop Group example showed small respiratory efforts (small airflow deflections) during VNR. On VNR release, phasic inspiration and expiration resumed in both examples.

Fig. 2.

Ocular Fixation Group (A) and Instructor Stop Group (B) illustrating representative responses in beat-to-beat changes before, during, and after VNR. The vertical hatched lines indicate the start and release of VNR. MCAVmean, mean middle cerebral arterial velocity (cm/s); MAP, mean arterial pressure (mmHg); HR, heart rate (beats/min); SV, stroke volume (ml); dP/dt, rate of change of pressure (mmHg/s).

Figure 3, A and B, shows that R and L MCAVmean decreased in both the Ocular Fixation and Instructor Stop Groups during VNR, compared with baseline 2 values. On release of VNR, R and L MCAVmean promptly returned to baseline 2 levels and remained constant. There was no change in MAP (Fig. 3C). In the Instructor Stop Group, HR decreased during VNR and remained depressed compared with baseline 2 values (Fig. 3D). SV decreased during VNR in the Ocular Fixation Group (Fig. 3E).

Fig. 3.

Hemodynamic data from baseline to 5 min after VNR was released in both the Ocular Fixation and Instructor Stop Groups. R (A)and L (B) MCAVmean decreased in both the Ocular Fixation and Instructor Stop Groups during VNR, compared with baseline 2 values. On release of VNR, MCAVmean returned to baseline 2 levels. There was no change in MAP (C). In the Instructor Stop Group, HR (D) decreased during VNR and remained depressed compared with baseline 2 values. SV (E) decreased during VNR in the Ocular Fixation Group. *P < 0.05 compared with baseline 2.

Figure 4 shows percent changes in hemodynamic measurements resulting from the application of VNR. CO decreased similarly in both the Ocular Fixation and the Instructor Stop Groups. However, note that these decreases in CO resulted from different changes in HR and SV. MAP and dP/dt were not different between groups and did not change during VNR. Both groups had reductions in R and L MCAVmean but L MCAVmean decreased more in the Ocular Fixation Group compared with the Instructor Stop Group.

Fig. 4.

Percent changes in hemodynamic measurements resulting from the application of VNR. In both the Ocular Fixation and the Instructor Stop Groups, cardiac output (CO; %) decreased equally; however, they were caused by different changes in HR and SV. MAP and dP/dt were not different between groups and did not change during VNR. Both groups had reductions in R and L MCAVmean but L MCAVmean decreased more in the Ocular Fixation Group. *P < 0.05 compared with baseline 2, †P < 0.01 between groups.

External bilateral neck pressure.

Figure 5 shows pressures over the carotid arteries and the SCM muscles in both groups. In the Ocular Fixation Group, pressure over the R carotid was greater than over the L carotid while pressure over the L SCM was greater than over the R SCM. Similar trends were seen in the Instructor Stop Group. There were no differences between the groups at either the carotid artery or SCM muscle locations. Note that pressures over the SCM are underestimated because values could only be measured up to 600 mmHg; when pressure exceeded this value, a pressure of 600 mmHg was recorded.

Fig. 5.

Pressures over the carotid arteries and the SCM muscles in both groups. In the Ocular Fixation Group, pressure over the R carotid was greater than over the L carotid while pressure over the L SCM was greater than over the R SCM. Similar trends were seen in the Instructor Stop Group. There were no differences between the groups at either the carotid artery or SCM muscle locations. *P < 0.05 between R and L pressures.

Anthropometric relationships.

Figure 6, A and B, respectively, shows the relations between time to ocular fixation and neck circumference and BMI for each subject in the Ocular Fixation Group. There was a tendency (P = 0.08) for those subjects with a larger neck circumference to fixate more quickly (Fig. 6A). There was an inverse relation between time to ocular fixation and BMI (Fig. 6B).

Fig. 6.

Relations between time to ocular fixation and neck circumference (A) and body mass index (BMI; B), for each subject in the Ocular Fixation Group. There was a tendency for those subjects with a larger neck circumference to fixate more quickly (A). There was an inverse relation between time to ocular fixation and BMI (B).

There was no relationship between the changes in R and L MCAVmean and BMI or external carotid pressures (data not shown).

DISCUSSION

Our results demonstrate that the most important mechanism causing unconsciousness during VNR is decreased cerebral blood flow due to bilateral carotid artery compression. The substantial decrease in MCAVmean (Ocular Fixation Group: R 83%, L 80%) can be presumed to translate into an acute period of cerebral ischemia sufficient to render subjects unconscious. Thus 11 subjects in the Ocular Fixation Group had no detectable flow on at least one side while six subjects had no detectable flow on both sides. The small decrease in CO (9.6 to 7.5 l/min) observed would not seem to be important as there was no change in MAP. In addition, with no significant change in HR, ventricular contractility, or MAP, the carotid sinus baroreceptor reflex appears to contribute little to the response to VNR.

Carotid artery compression.

Interruption of cerebral blood flow rapidly produces cerebral hypoxia, and loss of consciousness can occur within seconds (3, 10, 16). Cerebral oxygenation has been related to cerebral perfusion as indicated by transcranial Doppler flow velocity (10, 19). Njemanze (13) determined that if mean flow velocity is reduced by more than 50%, unconsciousness would ensue. In our study, 16 subjects (Ocular Fixation Group) were rendered unconscious with an 83% and 80% reduction in R and L MCAVmean, respectively. The four subjects who did not lose consciousness (Instructor Stop Group), had a 74% reduction of R MCAVmean, but only a 40% reduction in L MCAVmean. Thus our data support the findings of Njemanze (13) in that the critical lower limit of MCAVmean of 50% below baseline must be met bilaterally to induce unconsciousness.

Carotid sinus baroreceptor reflex.

The lack of an important decrement in contractility (dP/dt), SV, and CO implies little or no important contribution of the carotid sinus baroreceptor reflex to loss of consciousness during VNR. The absence of correlations between changes in HR, SV, MAP, and CO and the decrement in MCAVmean or time to ocular fixation supports this conclusion (data not shown).

We could not confirm the findings of Reay and Holloway (15) in which two of five subjects who had the greatest decrease in HR also had the shortest time to minimum flow, suggesting an important vasovagal response to VNR. We found no significant correlation between decreased HR and time to minimum MCAVmean (r = 0.44, P = 0.09) or time to ocular fixation (r = 0.07, P = 0.79; data not shown). Perhaps the mandatory termination of VNR before ocular fixation and use of facial blood flow velocity as a surrogate for brain perfusion in their study account for the different results.

The Valsalva maneuver.

Reduced venous return sharply lowers SV and blood pressure during Phase II of a Valsalva maneuver (5). Although most normal subjects do not lose consciousness during a Valsalva maneuver, decreased blood pressure could contribute to a loss of consciousness. Despite a properly applied VNR protecting the patency of the trachea, we measured zero airflow in most of our subjects during VNR. This could be explained by a lack of respiratory effort and therefore a breath hold occurred during the application. In the small number of subjects who did exhibit respiratory efforts during VNR, the presence of airflow indicates that the trachea was patent and that, in the majority of subjects, breath holding likely occurred. Since we did not measure intrathoracic pressure, we cannot directly address the question of how changes in intrathoracic pressure might have affected hemodynamics. However, the absence of changes in MAP and the small decrease in CO observed suggests that a Valsalva maneuver did not contribute substantially to loss of consciousness.

Clinical effects of VNR.

Ocular fixation with VNR was commonly associated with brief periods of myoclonic jerking and, in some subjects, the eyes turned upward. Many subjects reported a narrowing of the visual field with color changes or having entered into a dreamlike state. Importantly, recovery occurred almost immediately on release with no observed or reported negative side effects. Nine subjects sensed mild tracheal compression and one subject noted slight residual neck pain at the conclusion of the experiment. As previously described, despite most subjects showing no air flow during VNR, some subjects had infrequent respiratory efforts with measured airflow, thus confirming tracheal patency.

Practical implications.

We were able to quantify the pressure applied by a police instructor over the carotid bifurcation and, in 12 subjects, simultaneously measured pressure over the largest portion of the SCM muscle. For consistency, we purposely used the same right hand-dominant instructor to apply all VNRs. This translated into a greater pressure over the R vs. L carotid and L vs. R SCM (statistically significant in the Ocular Fixation Group; Fig. 5). This pattern of pressure measurements may vary among police officers. Since pressures over the R and L carotids and SCM did not correlate with the decrease in MCAVmean (greater applied pressure did not result in a greater decrease in MCAVmean) in either artery (data not shown), we do not anticipate that the effectiveness of a properly applied VNR to change if similar pressures are achieved, i.e., if mean bilateral external neck pressures are maintained at greater than ∼100 mmHg, there should be an effective decrement in MCAVmean leading to unconsciousness in most subjects.

Relations between VNR and body habitus.

There was a negative correlation between time to ocular fixation and BMI (Fig. 6B). One might expect those with a large neck (large mass) to have an abundance of tissue (e.g., muscle and connective tissue) that could play a role in compressing the carotid arteries. However, the absence of a significant correlation between both BMI and bilateral external neck pressure and decreased MCAVmean (data not shown) does not support that speculation. Although it is possible that vertebral arterial blood supply was compromised during VNR, we find this unlikely. Licht et al. (9) measured blood flow at the unprotected arterial region midway between the origin of the vertebral artery from the subclavian artery and its entry into the foramen of the sixth transverse cervical process and found no significant volume flow changes for four multiple rotational head positions. What appears to be more relevant to VNR is a forward tilting of the head in the mid-sagittal plane. Jargiello et al. (7) found no instances of vertebral artery compression using power Doppler imaging during head anteflexion in 428 patients when symptoms of vertebrobasilar insufficiency were brought on by specific head positioning. Although the above studies only measured vertebral arterial flow with different head positions without external neck compression, it would seem that it would be difficult to compromise vertebral arterial flow during VNR.

Limitations.

There are several limitations to the study. First, we recognize that the SV measurements were obtained noninvasively and not validated against an invasive, calibrated flow measurement. However, Harms et al. (4) showed nonsignificant differences throughout a range of SV changes induced by an orthostatic stress when comparing invasive thermodilution-determined SV to noninvasive Modelflow SV finger pressure measurements. Second, we used mean flow velocity as a surrogate for cerebral perfusion assuming laminar flow and that the vessel diameter did not change enough to explain the profound decreases in flow velocity. Third, the recorded minimum MCAVmean for the 5-s period did not necessarily occur exactly at the time of ocular fixation and was taken to accurately represent the period (mean time 9.5 ± 0.4 s; Table 2). Most subjects did not experience minimum flow for at least 1–2 s after VNR application, implying that the 5-s period does represent the important 6- to 7-s period during VNR. Moreover, Fig. 2A shows a representative example of beat-to-beat changes that demonstrate minimal flow changes occurring exactly at ocular fixation (1–2 s preceding VNR release). Last, we could not explain why four subjects were resistant to loss of consciousness, or more specifically, why they only had a 40% reduction in L MCAVmean. Presumably, anatomical differences in neck structures may have accounted for the differences in MCAV, despite no significant differences in applied neck pressure. Moreover, since we could not measure vertebral artery blood flow, we cannot exclude between-group differences. We also cannot explain why the Instructor Stop Group had a significant decrease in HR and not SV while the Ocular Fixation Group showed the opposite changes. The small number of subjects in the Instructor Stop Group make such comparisons difficult.

Conclusions.

Although many police agencies have adopted VNR, there are no studies describing the physiology of VNR with only two early limited investigations (15, 16). Our results indicate that the most important mechanism in causing unconsciousness was decreased cerebral blood flow due to carotid artery compression. The slight decrease in SV and CO should not be considered physiologically important enough to cause unconsciousness. Since there were no significant changes in HR, ventricular contractility, or arterial pressure, the carotid sinus baroreceptor reflex appears to contribute little or nothing to the response to VNR. With the majority of subjects rendered unconscious and, importantly, no serious adverse events in our subjects, we conclude that VNR is a safe and effective force intervention; however, outcomes could vary in different populations (i.e., unhealthy or older subjects).

DISCLOSURES

No conflicts of interest, financial or otherwise, are declared by the authors.

AUTHOR CONTRIBUTIONS

Author contributions: J.R.M. and R.S.S. conception and design of research; J.R.M. performed experiments; J.R.M. and D.E.R. analyzed data; J.R.M., D.E.R., J.V.T., I.B., and R.S.S. interpreted results of experiments; J.R.M. and D.E.R. prepared figures; J.R.M. drafted manuscript; J.R.M., D.E.R., J.V.T., I.B., and R.S.S. edited and revised manuscript; J.R.M., D.E.R., J.V.T., I.B., and R.S.S. approved final version of manuscript.

ACKNOWLEDGMENTS

The authors appreciate the technical assistance of Cheryl Meek, Ryan Richardet, Rozsa Sas, and Andrea Smailys, as well as Dr. Andrew Demchuk, Eileen Stewart, and Caroline Stephenson from the Calgary Stroke Program. The authors extend a special recognition to the Calgary Police Service for encouraging the advancement of scientific knowledge in the area of subject restraint and allowing its members to participate in this study. A special recognition is extended to S/Sgt. Chris Butler and Cst. John Warin and the 25 police volunteers—their enthusiastic participation and collective expertise was an essential component of this study.

REFERENCES

  1. 1.
  2. 2.
  3. 3.
  4. 4.
  5. 5.
  6. 6.
  7. 7.
  8. 8.
  9. 9.
  10. 10.
  11. 11.
  12. 12.
  13. 13.
  14. 14.
  15. 15.
  16. 16.
  17. 17.
  18. 18.
  19. 19.
  20. 20.
  21. 21.
View Abstract