Correlation of neurological level and sweating level of injury in persons with spinal cord injury

Abstract

Objective: Thermoregulatory dysfunction after spinal cord injury (SCI) impairs quality of life and predisposes persons to life-threatening sequela of heat-related illness (HRI) in conditions of high ambient temperature. SCI clinicians currently have no objective way to predict which persons are at greatest risk of HRI. Evaporative cooling via sweating is the body’s most efficient mechanism of heat dissipation. The relationship between the neurological level of injury (NLOI) and the degree of sudomotor dysfunction is not well defined. This study examines the relationship between the NLOI and sweating level of injury (SwLOI). This information can assist SCI clinicians in identifying individuals with SCI who have most impaired sudomotor function and thus highest risk of HRI.

Design: Observational.

Setting: Human physiology laboratory.

Participants: 10 persons with tetraplegia (TP), 14 with paraplegia (PP) and 10 able-bodied (AB).

Intervention: Passive heat stress (1°C rise in core temperature) with sweat responses (SR) quantified with the starch iodine test.

Outcome measures: The most caudal dermatomal level in which sweating was visualized was recorded as the SwLOI, which was compared to the NLOI. Minimum, maximum and median differences between NLOI and SwLOI were calculated.

Results: Persons with tetraplegia demonstrated no SR. Persons with paraplegia demonstrated SR at a median of 1 level below NLOI. Able-bodied controls demonstrated sweating on all skin surface areas.

Conclusions: Persons with motor complete tetraplegia lack evaporative cooling capacity through SR during passive heat stress predisposing them to HRI. Meanwhile, persons with paraplegia sweat on average 1 dermatomal level below their NLOI.

Introduction

Injury to the spinal cord disrupts the afferent neural connection of skin temperature to the hypothalamus from much of the body surface and the efferent autonomic output from hypothalamus to the blood vessels and sweat glands. As a result, core temperature (Tc) regulation under conditions of heat stress is impaired after SCI. Such thermoregulatory dysfunction after SCI predisposes persons to heat related illness (HRI) (i.e. heat cramps, heat exhaustion, heat syncope, heat stroke) under high ambient temperature conditions.^1,2 Symptoms of HRI range from moderate dizziness, nausea, vomiting, headache, weakness and confusion to potentially life-threatening sequelae of cardiac arrhythmias, coma, and seizures. Core temperatures (Tc) in persons with SCI can rise to dangerous levels (>37.8°C) in a short period of time with little self-awareness.^1,3 In addition to the adverse sequelae, persons with SCI self-report more physical discomfort and decreased participation in outdoor activities compared to able-bodied persons in hot environments.⁴

While the potential for this thermoregulatory dysfunction is well known, predicting which persons with SCI, based on their neurological level of injury (NLOI), are most susceptible to HRI is not well defined. Motor and sensory deficits are defined by the International Standards for Neurological Classification after Spinal Cord Injury (ISNCSCI) exam. On the other hand a validated classification system that predicts function and neurological recovery, defining autonomic nervous system (ANS) deficits after SCI is more complicated, as the anatomy of the ANS neurological pathways is more intricate and involves multiple organ systems.^5–7 The International Standards for Autonomic Function after SCI (ISAFSCI) was recently developed,^8–10 but this exam is: (1) not as effective at predicting ANS function or recovery; (2) requires subjective symptom reporting; and (3) is only used by about 50% of clinicians.¹¹ Assessing thermal regulation using the ISAFSCI is rudimentary and consists of only 2 questions: (1) “temperature regulations” (box to check for history of hyper or hypothermia) and (2) “autonomic control of sweating” (box to check for reported hyperhidrosis above the NLOI and hyper or hypohidrosis below the NLOI). In its current form, the validity of this portion of the ISAFSCI is based on the patient’s accuracy as a historian and does not objectively quantify the severity of dysfunction. Furthermore, there is no assessment of hypohidrosis above the LOI – as has been reported in small numbers of persons with tetraplegia (n = 1–6).^12–14 In summary, more comprehensive knowledge of how NLOI predicts individuals at the greatest risk of HRI would directly improve clinical care, education and outcomes in the SCI population.

During conditions of heat stress, evaporative cooling via sympathetic cholinergic activation of sweat glands in combination with cutaneous vasodilation represent the body’s main physiological defenses against hyperthermia.^15–17 Individuals with SCI have consistently shown an overall impaired capacity for evaporative heat loss.^12,18–20 A frequently stated summary in the literature is that persons with SCI cannot sweat or dilate “below their NLOI,” or in “insensate areas”.^21,22 However, in the few small studies to date, persons with tetraplegia (TP) demonstrate forehead sweating^14,16,23 or complete anhidrosis, even in sensate areas above their NLOI.^12,13,24 Persons with paraplegia (PP) demonstrate more variable sweat responses (SR) in relation to their NLOI, depending on the level and completeness of lesion. Using a crude technique of visual observation of sweat droplets, in the 1970s, Normell found that dermatomes with intact SR extended several centimeters, or one or more dermatomes, below the NLOI in individuals with complete PP.¹² In summary, there is conflicting evidence in the literature regarding the relationship between remaining NLOI of sensorimotor function (defined by the ISNCSCI exam) and remaining sudomotor function. While generalized SR patterns exist during heat stress, the majority of passive heat stress protocols examining SR in individuals with SCI have consisted of relatively small and heterogenous (i.e. variable lesion levels and completeness) numbers of persons with SCI, which limit their power. In addition, none have fully excluded autonomic dysreflexia (AD) as a potential etiology of SR from a neurologically distinct pathway (i.e. sympathetic noradrenergic sweating) not employed during passive heat stress.²⁵

In summary, it is difficult to accurately predict which persons (based on NLOI and completeness) have greatest loss of evaporative cooling capacity during heat exposure, and thus are at the highest risk of HRI. However, it would be of benefit to patients if clinicians were able to educate individuals at increased risk of HRI, given the potential life-threatening sequelae.¹ This study aims to examine the relationship of the NLOI as defined by the ISNCSCI exam,⁷ to dermatomes with intact sudomotor function in persons with motor complete (i.e. AIS A or B) TP and PP during passive heat stress.

We hypothesize that persons with motor complete TP will lack SR and that persons with motor complete PP will demonstrate SR within ±2 dermatomal levels (∼ a few centimeters) of their NLOI.

Methods

Participants

Participants consisted of 24 individuals with motor complete (AIS A or B) SCI: 10 persons with TP (C1–C7), 14 PP (T1–T12) and 10 uninjured able-bodied (AB) age and sex matched controls (Table 1). In participants with SCI the NLOI was determined using the ISNCSCI exam, which was performed at the time of study enrollment by an SCI board-certified physician. At the same time, the ISAFSCI survey was administered to determine self-reported history of thermo-regulatory capacity. Collectively, thirty participants were male and four were female (AB = 2, PP = 2). There were no significant group differences in age or BMI and participants were excluded if they were smokers, had any pressure ulcers, or had any acute illness. Those who used vasoactive medications (e.g. midodrine, sildenafil, etc.) were asked to hold these medications at least 24 h prior to coming into the lab. All participants signed an informed consent.

Table 1. Demographics by group with means and standard deviations.

Group	TP (n = 10)	PP (n = 14)	AB (n = 10)
Age (years)	39.5 (11.9)	41.3 (10.9)	41.6 (12.3)
BMI	26.1 (5.7)	26.0(5.5)	29.8 (5.5)
Levels of injury	C3-7	T3-T10	na
Motor and sensory complete (AIS A)	6	14	na
Motor complete and sensory incomplete (AIS B)	4	0	na

Heat stress protocol

Participants fasted for at least 3 h prior to participation. To minimize risk of nor-adrenergic sweating from AD,²⁶ participants with SCI were: (1) asked to perform a bowel program at least 24 h prior to testing and (2) emptied their bladder within 15 min of initiating study procedures. The AB participants were also asked to empty their bladder to be sure the heat stress protocol would not be interrupted. Participants were then placed in the supine position in a hospital bed wearing a minimum of clothing (shorts only for men; shorts and bra for women) in a 35°C room. While acclimating to the environment, the iodine starch test was set up to measure SR.^27,28 Based on Normell’s finding that sweating persisted a few centimeters below the NLOI in most cases of PP (no SR seen in the 6 persons with TP),¹² iodine was applied over the skin 3–5 dermatomes above and 5–10 dermatomes below the NLOI (to be more thorough in looking below NLOI) for all persons with SCI. AB persons had iodine applied over the exact same skin surface as their SCI pair. After 3–5 min, which allowed for the iodine to fully dry, starch powder was evenly distributed over the surface using a previously published method.²⁷ This starch-iodine sweat test method relies on a chemical reaction that occurs between iodinated starch powder and sweat, that turns the powder a purple color. After acclimatizing to the room for at least 15 min, a thermocouple was placed sublingually to measure baseline core temperature (Tc), while baseline blood pressure and pulse rate, were measured by Portapres. (Finipres; Enschede, Netherlands). Rate of perceived thermal strain (RPTS) was recorded using a 9-point (0–8) Likert scale at baseline and at end of heat stress protocol.^29–31

After baseline data collection, a water perfusion suit with 42°C water was applied over the entire skin surface area for the AB persons and over sensate skin areas in participants with SCI. Three electrical heating blankets were applied over the face and the entire body surface area on top of the water perfusion suit in the AB group and were applied over the face and sensate skin areas in participants with SCI. The water perfusion suits were not placed over the insensate skin areas of persons with SCI due to risk of burns in these areas. After Tc increased by 1.0°C from baseline (to elicit maximal sweating),²⁸ water perfusion suit and heating blankets were removed.^27,28 The most caudal dermatomal level at which SR occurred bilaterally was recorded as the sweating level of injury (SwLOI), and the number of segmental dermatomes between NLOI and SwLOI was recorded.

Statistical analysis

Descriptive statistics on the average change in RPTS (post heating – baseline) with standard error of the mean (SEM) were calculated. One-way ANOVA was performed to determine the significant effect of neuro status (i.e. TP, PP vs. AB). In addition, minimum, maximum, and median difference in a number of dermatomes between NLOI and SwLOI with SEM were calculated and graphed. Statistical analyses were conducted using GraphPad Prism version 7.04 (GraphPad Software, La Jolla California USA, www.graphpad.com).

Results

During the same absolute rise in Tc of 1.0°C, RPTS of persons with TP, PP and AB rose on average 1.85° ± 0.2892; 2.25° ± 0.2329 and 2.17° ± 0.2357 respectively with no statistical differences among groups (P = 0.51).

Subjective report of “autonomic control of temperature” and “autonomic control of sweating” via the ISAFSCI is reported in Table 2. Overall, persons with TP had more subjective awareness of their impaired ability to thermoregulate. Both TP and PP groups largely reported hypohidrosis below the NLOI. See sample of results of a few participants in Fig. 1.

Table 2. ISAFSCI “autonomic function of sweating” and “temperature regulations” data from TP vs. PP groups subjective reports.

		TP (n = 10)	PP (n = 14)
Autonomic control of sweating
Normal			1
Abnormal	Hyperhidrosis above lesion		1
	Hyperhidrosis below lesion		1
	Hypohidrosis below lesion	10	12
Temperature regulations
Normal		3	6
Abnormal	Hyperthermia	7	7
	Hypothermia	5	4

Figure 1.

SwLOI vs. NLOI in a person with (a) C5 AIS A*; (b) T4 AIS A and (c) T8 AIS A SCI after rise in core temperature of 1°C. Arrows pointing to SwLOI (level at which SR seen bilaterally). *Note for Fig. 1a: C6 to T1 dermatomes of the arm and hand are not shown in this picture frame but were also examined and demonstrated the same anhidrotic pattern as the chest pictured.

Of the 24 participants with SCI, all of the PP participants demonstrated bilateral SR, while none of the TP participants displayed SR over any skin surface areas, including forehead, as such, we were unable to assign a SwLOI in participants with TP. The range of dermatomes between the NLOI and SwLOI in persons with PP are seen in Fig. 2 with median of –1 denoted by the dotted line. On average, persons with PP were able to sweat 1 dermatome level below their NLOI. All 10 AB persons demonstrated diffuse SR over all skin surface areas examined, as anticipated (data not shown).

Figure 2.

Difference between NLOI vs. SwLOI dermatomal levels calculated by NLOI-SwLOI. Note negative values denote caudal levels of sweating compared to NLOI (i.e. sweating persists below NLOI). Positive values denote SwLOI is more cephalad to NLOI (i.e. anhidrosis above NLOI). Minimum difference +5 with maximum difference of –6, mode range +1 to –2 with median = –1 dermatome difference (indicated by dotted line).

Incidentally, 3 persons with TP experienced symptomatic AD during the study, which was diagnosed by typical AD symptoms of headache, facial flushing and rise in systolic blood pressure >20 mmHg from baseline (all >150 mmHg). Two episodes resulted from bladder distension (as SBP dropped immediately after bladder drainage) and 1 resulted from bowel distension. During AD, these TP participants demonstrated profuse SR on forehead (n = 2) and chest (n = 1). In each of these cases, the study was terminated at the onset of AD and was repeated on another day. SR did not occur in the absence of AD.

Discussion

While thermoregulatory dysfunction, and more specifically impaired evaporative cooling capacity via sympathetic cholinergic sweating after SCI is a known sequela, predicting individuals at a heightened risk for HRI based on their NLOI has not objectively been defined.

Evaporative cooling via sweating is controlled by the sympathetic nervous system. The sudomotor pre-ganglionic fibers travel in the intermediolateral cell column within the lateral horn of the thoracolumbar (T1-L2) spinal cord. These sympathetic neurons synapse at three locations: (1) the same level ganglion (2) a cephalad ganglion or (3) a caudal ganglion and give rise to post-ganglionic neurons which travel to and innervate the sweat glands.³²

Sudomotor innervation of the cervical dermatomes (i.e. upper extremities) originates in the thoracic cord (i.e. below the cervical lesion). Thus, based on this anatomic arrangement, a person with a cervical complete (AIS A) lesion may have impaired sympathetic function (i.e. sweating), which could be impaired in areas of intact sensation above the NLOI. Meanwhile persons with PP, would be anticipated to have intact sympathetic sudomotor neurons above their NLOI, allowing for more appropriate SR during heat stress.

Measurement of sudomotor activity in past investigations has been captured via whole-body sweat rates, sweat capsules, and skin temperature as a surrogate SR index, i.e. sweating decreases skin temperature. The majority of previous studies measured evaporative cooling are primarily small pilot studies (N = 1–10) in persons with SCI with varying injury characteristics under protocols that employed a myriad of heat stressors (e.g. passive vs. exercise induced).^12,13 While the general statements that sweating is impaired “below the lesion” or in “insensate areas” were commonly made,^21,22 only one prior study had the specific objective of correlating NLOI with a sweating level using visual detection of water droplets over large skin surface areas under a plastic sheet.¹² Using this technique, Normell found “dissociation between areas with loss of cutaneous thermoregulatory sudomotor responses and areas of loss of somatic sensibility.”³³ He also reported that in most of individuals “the dissociation between sudomotor responses and somatic sensation was only a few centimeters or … the width of one or more somato-sensory dermatomes.”³³ Thus, data suggests that the sensory and sweating level are not always equivalent, which parallels the anatomic arrangement of the sympathetic sudomotor versus sensorimotor tracts.

If an objective correlation between NLOI and SR can be made, it would provide support for modification of the “temperature regulation” and “autonomic regulation of sweating” section of the ISAFSCI to be more clinically useful. With the intention of correlating SwLOI with NLOI in individuals with SCI categorized by the ISNCSCI exam as TP (C1-C8) and PP (T3-T10), and using a more sophisticated techniques (compared to visual inspection of water droplets), we studied SR in the largest group of persons with motor complete TP, paired with a group of persons with motor complete PP and AB persons during passive heat stress.

SR in tetraplegia (TP)

In the few small studies (N = 1–6) in individuals with TP, minimal forehead SR or complete absence of SR over the entire body surface area have been reported.³⁴ Guttman reported complete loss of sweating in 4 persons with motor complete TP via the quinizarin dye test.¹³ Later in the 1970s, Normell reported the absence of sweat droplets (via visual exam) on six participants with motor complete TP during passive ambient heat stress.¹² Of note however, Normell did not examine forehead SR. Subsequently, Gass reported 2 of 3 persons with TP had no forehead SR (1 complete and 1 incomplete) during exercise while Webborn reported 2 persons with incomplete tetraplegia to have only minimal forehead SR via body mass changes.^14,35

The presence of forehead sweating reported in a few individuals with incomplete TP lesions may be explained by (1) an incomplete autonomic lesion with remnant functioning pre-ganglionic sympathetic nerves, (2) re-innervation changes in spinal circuitry, or (3) noradrenergic sweating, which can also be stimulated during an episode of AD.²⁵ Interestingly, the density of dopamine beta-hydroxylase (converts dopamine to norepinephrine) on facial autonomic fibers is greater than the density on the rest of the body, which may explain why sweating occurred only on the face during noradrenergic sweating (i.e. AD).³⁶ It is important to note that AD was never considered as an explanation for SR in any of the aforementioned studies, nor was blood pressure measured/reported, so whether the forehead sweating was noradrenergic (pathologic) or cholinergic (physiologic) is unknown.

In this study examining SR over a large body surface area in the largest cohort of motor complete persons with SCI to date, the ten persons with motor complete TP demonstrated complete anhidrosis (including forehead) after a rise in core body temperature of 1°C. Our findings confirm that of Normell with the added finding of anhidrosis in the face and forehead of persons with TP.¹² These objective findings confirm the accuracy of the subjective TP ISAFSCI data where all persons with TP reported “hypohidrosis below the lesion.” In conclusion, complete anhidrosis seen in response to heat stress in persons with TP (1) challenges the statement that persons with SCI are only anhidrotic in insensate areas and (2) suggests the option of “hypohidrosis above the LOI” should be added to the “autonomic control of sweating” section of the ISAFSCI to capture the pathophysiology seen in persons with TP and (3) informs clinicians of the high risk of HRI in persons with TP so appropriate education and prevention can be implemented.

SR in paraplegia

Many investigators have attempted to define a level of paraplegia at which sweating and thermoregulation is impaired compared to able-bodied controls. Paraplegic subjects show decreased sweating when compared to AB controls however no statistical significance in sweat rates has been found within persons of paraplegia of varying levels (T1-T12).^20,23,37

Two case reports have attempted to defined an exact “cut off” NLOI at which sweating capacity impairs thermoregulation. Downey et al. reported that a person with T6 paraplegia demonstrated similar sweat losses, as indicated by weight loss, to an AB participant.¹⁹ The similar weight losses from a reduced number of active sweat glands suggests a greater output of sweat from the available sweat glands, possibly facilitated by the greater thermal drive during heat stress. However, it was further noted that the increase in sweat rate was primarily due to a large increase in forehead sweat rate, which is considered to be an inefficient method as sweat that drips before evaporation is not effective for heat dissipation. As a result, despite similar sweat losses, Tc of the persons with T6 paraplegia rose higher than the AB person, demonstrating comparably impaired thermoregulation. Meanwhile, Guttman et al. measured rise in Tc under passive heat stress of one AB, 4 persons with SCI lesions above T8 (C6-T4 NLOI) to one below T8. They reported that T8 was the “cut off level” at which persons with SCI could thermoregulate similarly to AB persons despite anhidrosis in the lower extremities.¹³

Our study of 14 persons with PP found intact SR on average 1 dermatomal level below their NLOI with the majority [∼70%] exhibiting some degree sweating below their sensory level of injury, which corroborates to the findings of Normell and colleagues.¹² As in persons with TP, these objective findings support the accuracy of the subjective findings from the ISAFSCI, of a majority (12 of 14) of persons with PP who checked “hypohidrosis below the lesion”.

Skin surface area consideration

In the burn literature, it is consistently reported that individuals with burns encompassing greater than 40% total body surface area (TBSA) have the most compromised thermoregulatory function during heat stress and heat acclimation; while those with <40% TBSA burns have thermoregulatory capacities similar to AB persons.^38–42

Guttumans’ T8 cut off level of intact thermoregulation would be consistent with the burn literature as persons with spinal cord lesions at T8 and below will likely have intact SR on greater than 40% TBSA, validation of this cut off requires further testing.

Until the “cut off” NLOI at which thermoregulation is unimpaired is determined by larger studies, the authors’ recommend that persons with a NLOI above T7 (who may sweat down to T8 based on our data) or T8 under heat stress, apply artificial sweat (e.g. water spray) over at least 40% of their TBSA, especially on the face in persons with TP, and anhidrotic areas,⁴³ and/or utilize ice vests before exposure to heat stress to prevent Tc rise.⁴⁴

Limitations

Limitations of this study include the use of sensory incomplete persons (i.e. AIS B) who may have more “autonomic incompleteness” than persons with sensorimotor complete (AIS A) injuries. Reorganization of sympathetic preganglionic neurons can occur after SCI and may provide for some autonomic integrity despite motor complete classification by ISNCSCI. Unfortunately, there is currently no objective validated and comprehensive measure of autonomic completeness, so the ISAFSCI was used as the best surrogate in this study. Furthermore, 50 percent of the participants with SCI were actively taking anticholinergic (Ach) medications for management of neurogenic bladder, which has potential to decrease cholinergically mediated SR. Fortunately, the few studies that have examined sympathetic (specifically vasomotor or sudomotor) levels/completeness post-SCI in persons taking Ach reported that Ach only slightly decreased, but did not abolish sympathetic responses (e.g. sympathetic skin responses), while others stated that limited available data precludes the ability to predict the effects of on sweating responses.^45,46

Conclusion

This study demonstrates that persons with motor complete TP fully lack evaporative cooling capacity via sympathetic cholinergic sweating under passive heat stress conditions putting them at high risk of HRI. This study challenges the statement that all persons with SCI (specifically those with TP) are anhidrotic only over insensate areas or below the lesion. It is likely more accurate, and consistent with anatomy, that persons with motor complete TP are anhidrotic over sensate and insensate areas, which is new and noteworthy. Thus our findings call for modification to the current ISAFSCI exam to include “hypohidrosis above the LOI” to the “autonomic control of sweating section”. Meanwhile, persons with motor complete paraplegia usually sweat an average of one dermatome level below their NLOI. This information helps clinicians objectively predict which patients are at increased risk of HRI, so appropriate preventative education efforts can be targeted towards such persons.

Disclaimer statements

Contributors None.

Funding This work was supported by a Veteran’s Affairs Rehabilitation Research and Development Service [award number 1 IK2 RX001805-01A2].

Conflict of interest The authors declare no financial conflicts of interest.