Computerized Cuff Pressure Algometry as Guidance for Circumferential Tissue Compression for Wearable Soft Robotic Applications: A Systematic Review

Abstract

In this article, we review the literature on quantitative sensory testing of deep somatic pain by means of computerized cuff pressure algometry (CPA) in search of pressure-related safety guidelines for wearable soft exoskeleton and robotics design. Most pressure-related safety thresholds to date are based on interface pressures and skin perfusion, although clinical research suggests the deep somatic tissues to be the most sensitive to excessive loading. With CPA, pain is induced in deeper layers of soft tissue at the limbs. The results indicate that circumferential compression leads to discomfort at ∼16–34 kPa, becomes painful at ∼20–27 kPa, and can become unbearable even below 40 kPa.

Introduction

The vast majority of gait-assistive devices and exoskeletons are manufactured from rigid materials that impart torque assistance to the joints while also transmitting loads to the ground. Recently, attempts are being made to develop lighter, low-profile soft robotic devices, composed of textiles, referred to as soft exoskeletons or exosuits.¹ The majority of hard exoskeletons apply mechanical loads to the user's body via rigid shells (orthoses²), whereas soft exoskeletons apply circumferential forces to the user via connection cuffs.² The physical interactions between the users and hard or soft exoskeletons are quite different. Hard exoskeletons mostly affect weight-bearing sites, which are usually relatively small surface areas, whereas soft exoskeletons interface with larger surface areas, including around limbs. In both cases, excessive mechanical loading can lead to soft tissue injury and cause pressure ulcers (PUs).

PUs are localized areas of soft tissue breakdown^3–6 and are particularly common in individuals who are bedridden, wheelchair bound, or wear a prosthesis or orthosis.³ PUs occur superficially (friction ulcers) or in deep tissues (pressure-related Deep Tissue Injuries [DTIs]). DTIs are mainly caused by sustained compression of the deep muscle layers over bony prominences, and they can be potentially life threatening.^3,5 The pathophysiologic mechanisms of soft tissue breakdown are not completely understood. Theories indicate localized ischemia,^3,7,8 impaired lymphatic drainage,^3,8 elevation of local lactic acid levels,⁷ reperfusion injury, and sustained deformation of cells.^3,5,7,8 Other potentially important contributory factors include malnutrition, age, certain physical conditions, medication, dehydration, circulatory disturbances, and immobility.⁷

Until recently, attempts to establish safe thresholds for the external mechanical loading of soft tissues have been based on interface pressures at load-bearing sites of the body, such as under the ischial tuberosities, the sacrum, the trochanters, and heels. In ergonomics, as well as in clinical practice, skin capillary pressure of 32 mmHg (4.3 kPa), established in 1930 by Landis, is often cited as an exposure above which tissue breakdown could occur.^3,5,9 However, Landis' observations were of an open arterial capillary⁵ within the nail folds.⁹ Later research has set the average capillary pressure at 47 mmHg.⁵ Further, self-regulatory mechanisms cause capillary pressure to stabilize at higher than average values,¹⁰ and capillary closure does not only depend on interface pressures at skin level. Hence, interface pressures well above capillary pressures can be supported by the soft tissues before blood flow is seriously impaired.¹¹ For example, typical interface pressures under the ischium during sitting reportedly range up to 165 mmHg (22 kPa),¹² and a maximum average pressure of 220 mmHg (29.3 kPa) was recorded on the thighs of an able-bodied subject using an exoskeleton.¹³ Hence, the appropriateness of the 32 mmHg threshold criterion for wearable robotics applications is unclear.⁹ Although there continues to be a focus on acceptability of pressure magnitudes, few studies have also considered the importance of pressure Direction, Distribution and Duration (3Ds), in addition to loading cycle frequency.^11,14

The relationship between interface pressure and internal stress is not linear.⁸ Internal stress is highly dependent on the nature of the intervening soft tissues for example, their thickness,^13,15 tone,^15,16 mechanical stiffness,⁴ and integrity,¹⁵ as well as the proximity of bony prominences.^4,5,13,15 Moreover, injury thresholds differ for skin, adipose tissue, and muscle,⁵ with the lowest threshold for muscle.⁵ Thus, a safe threshold based solely on interface pressure is not acceptable.^3–5,8 However, measurement of internal pressure is technically and ethically challenging.⁸ Therefore, several techniques have been used in combination with interface-pressure measurements, for example, measurement of transcutaneous partial pressure of oxygen (TcPO₂),^11,17 transcutaneous partial pressure of carbon dioxide (TcPCO₂),¹⁷ near-infrared spectroscopy (NIRS), and laser Doppler flowmetry (LDF).¹⁸ However, these techniques primarily focus on the perfusion of the most superficial tissue layers.

Pain or discomfort is the most direct reaction of the human body to excessive external loads.¹⁴ Pressure-induced muscle pain is mainly related to strain,¹⁹ and perceived pain is considered a good indicator of potential tissue damage caused by excessive pressure exposures.²⁰ The authors propose that pain and discomfort studies with pressure algometry could be a relevant approach to study tissue-interface exposures for soft robotics applications, and wearable-robots generally. Algometers are used in clinical practice to apply pressure during studies of pressure-induced pain.²¹ Typically, two parameters are measured: the pressure magnitude at which pain occurs (pain detection threshold [PDT]), and the pressure magnitude that causes unbearable pain (pain tolerance threshold [PTT]). The thresholds tend to be measured in kPa, as opposed to interface pressures that are usually measured in mmHg. Traditional algometers are hand-held devices with a 1 cm² probe that applies pressure to a single specific point at a time (Fig. 1a).^20–22 Hand-held algometers have been used extensively in clinics to study changes in the pressure-pain thresholds (PPTs) in fibromyalgia and headache.²³

FIG. 1.

Single-point algometry (a) and cuff pressure algometry (b).

Pons²⁰ describes a study of single-point pain perception at several anatomical sites of the lower limb, typically in contact with wearable devices, indicating that algometry might be a useful tool for establishing acceptable interface-pressure limits. However, Pons also explicitly points out that the reported pressures are caused by punctual, instantaneous forces, which renders the limits unsuitable for sustained external loading. Moreover, single-point pressure algometry is of limited use for soft robotics applications where the forces are transmitted to the body over large contact areas under the connection cuffs, and at anatomical sites with thicker layers of soft tissue (e.g., the thigh and shank).

Computerized cuff pressure algometry (CPA) has been used to stimulate large volumes of deep somatic tissues.^19,21,22,24 In CPA, mechanical tissue compression is achieved by a pneumatically controlled tourniquet cuff wrapped around the limb (Fig. 1b).²² CPA enables an exposure to external loading that is more analogous to the one in soft wearable robotics applications, such as soft exoskeletons. CPA studies can be used to study pain thresholds, stimulus-response functions, and spatial and temporal summation of pain (TSP).¹⁹ CPA has also been found to be less influenced by local pain sensitivity variations, and it is examiner independent.^19,22,24 Further, cuff pressure and intra-arterial pressure under the cuff, and therefore tissue pressure under the cuff, were found to be directly related.²³

The aim of this article is to attempt to establish indicative guidelines for acceptable levels of mechanical tissue compression in humans, for use in the design of soft lower limb exoskeletons. To achieve this aim, we performed a structured systematic review of the literature on pressure-induced pain, specifically of CPA. CPA is more capable of generating mechanical stress in deeper tissues than single-point algometry, and is, therefore, more appropriate for assessing the response of deep tissues to compression as induced by soft exoskeletons.

Methods

Literature search and study selection

A systematic literature search was performed in March 2017 by using EBSCOhost to search the following databases: Medline, Academic Search Complete, AMED, Biomedical Reference Collection, CINAHL Plus, and General Science. The keywords used were “cuff algometry,” and the search was limited to human studies. The initial search yielded a total of 59 articles. After screening, a review of the reference lists of the 12 papers deemed eligible for inclusion identified one additional eligible study,²¹ resulting in a total of 13 studies being included in the review. Figure 2 illustrates the search and screening process. A second reviewer repeated the search and screening process to ensure that the process was accurate and repeatable.

FIG. 2.

Literature search and study selection.

Data extraction and synthesis

Data extracted from the selected studies included: (1) the participants' characteristics (age, sex, and anthropometric characteristics), (2) the assessment methods (tourniquet cuff characteristics and positioning, compression rates and durations, pain-intensity rating, etc.), (3) the variables studied, and (4) the findings of the study.

The relevant independent variables were either pneumatic cuff inflation pressure in CPA or probe pressure in single-point algometry. The relevant dependent variables were the two pressure-induced pain thresholds: PDT and PTT. Other variables, such as pain tolerance limit (PTL), TSP, and interface pressure, were also noted, but they were not filter criteria for inclusion in the review.

Results

Participants

All studies were of healthy participants. Five studies included only male,^{22,23,25–27} one study only female,²¹ and four studies both male and female^24,28–32 participants. Two studies only included 1 participant,^19,22 three studies included 12,^23,25,30 three studies included 16,^21,26,28 two included 20,^27,31 one included 56,²⁹ one included 98,³² and one included 136²⁴ participants. Three studies included participants older than 35 years of age,^21,24,32 in one study the age of the single participant was not reported,¹⁹ but comparison with another study by the same authors with the same data utilized²² led us to believe that the participant was the same. Eight studies reported the mean BMI of the participants,^{22,24,27–32} and three studies reported the mean circumference of the limb studied.^21,30,31

Assessment methods

CPA was performed by using either a 6 cm^19,22,30 or 11 cm^23,25 wide single-chamber tourniquet cuff, or a 13 cm wide double-chamber tourniquet cuff.^{21,24,26–29,31,32} One study³⁰ also used a 5 cm wide cuff with an inner cylindrical chamber filled with water and an outer chamber inflated with air. The cuffs were inflated at a constant rate; one study also assessed the results of increasing the rate of inflation,²³ and two assessed the results of sustained constant compression.^26,31 Participants in all studies rated their pressure-induced pain intensity on an electronic Visual Analogue Scale (VAS), with 0 indicating no pain and 10 cm indicating maximal pain. With the exception of one study,²⁸ rating was performed at 10 Hz. PDT was defined when VAS exceeded 0 cm,^21–29 0.1 cm,³² or 1 cm.³⁰

In three studies, single-point pressure algometry was performed in addition to CPA.^23,24,28 Two studies also utilized MRI and 3D finite element modeling to predict Von-Mises stresses in deeper tissues during external compression.^19,22 Other tests were performed, such as assessment of thermal pain sensitivity,^27,28 cold-pressor test,²⁴ cutaneous pin-prick sensitivity,²⁸ provocation tests with hypertonic saline solution,²⁸ skin and muscle sensitization with capsaicin,²⁵ or selective anesthesia,²⁵ but these are beyond the scope of this systematic review.

All studies reviewed were performed on the lower limb. One was of the thigh as proximally as possible,²⁸ and the remainder were on the lower leg, at the level of the heads of m. gastrocnemius-soleus,^23,25 the heads of the gastrocnemius muscle,^19,22,30 5 cm under the tibial tuberosity,^24,27,29 or at the widest part of m. triceps surae.^31,32 Three CPA studies were also performed on the upper limb.^24,29,32 Participants were tested in the supine^{21,23–26,28,29,31} or seated position.²⁷ Four studies did not detail the tested posture.^19,22,30,32

Variables studied

Of the variables that we find important for soft-robotics application, 1 study only reported PDTs,²⁸ 1 only reported PTTs,²⁹ and 11 studies reported PDTs and PTTs.^{19,21–27,30–32} One study reported PDTs for single-point pressure algometry as well as CPA,²⁸ and two studies reported PDTs for single-point pressure algometry and PDTs and PTTs for CPA.^23,24 Four studies also reported PTL (the rating on VAS at PTT),^21,23,25,32 three studies assessed TSP,^24,29,31 and one study reported mean interface pressure and interface pressure distribution under the cuff.³⁰

Findings

The studies reviewed are chronologically ordered and summarized in Table 1. Table 2 provides a further summary of the pressure levels for PDT and PTT.

Table 1.

Summary of Studies Reviewed

Reference	Participants	Assessment methods	Variables	Findings
Polianskis et al.²³	Healthy male Age: 21–27 years (median 22 years) n = 12	(A) Single-point pressure algometry (hand-held): 1 cm² probe 5 lower leg sites: (1) Skin-periosteum site (2) Distal medial head of m. gastrocnemius-soleus (3) Distal lateral head of m. gastrocnemius-soleus (4) Middle of medial m. gastrocnemius-soleus belly (5) Middle of lateral m. gastrocnemius-soleus belly 3 repetitions (5 min apart) Supine position (B) CPA: Single-chamber tourniquet cuff (width 11 cm; compressed area: 401 ± 41 cm²) At the level of the heads of m. gastrocnemius-soleus Maximum pressure limit: 180 kPa Electronic VAS (0—no pain, 10—maximal pain), sampling at 10 Hz PDT = VAS >0 Supine position Assessment: Inflation at constant rate: 1.2 kPa/s (3 repetitions, 5 min apart) Inflation at increasing rate: 0.25, 0.5, 1.0 kPa/s	Independent: Cuff inflation pressure Dependent: PDT^* PTT PTL	(A) Single-point pressure algometry (1) PDT = 454 ± 23 kPa (2) PDT = 433 ± 15 kPa (3) PDT = 476 ± 19 kPa (4) PDT = 438 ± 16 kPa (5) PDT = 526 ± 20 kPa (B) CPA at constant compression rate
					Left	Right
				PDT (kPa)	26.0 ± 1.6	25.4 ± 1.8
				PTT (kPa)	60.8 ± 3.8	55.7 ± 3.6
				PTL (cm)	5.4 ± 0.13	5.5 ± 0.13
				CPA at increasing compression rate:The increase of compression rate 0.25–1 kPa/s increased PTT by 164% and decreased PTL by 610%.Direct correlation between the number of compressions, the compression rate, and pain threshold was found.
Polianskis et al.²⁵	Healthy maleAge: 24–26 yearsn = 12	CPA: Single-chamber tourniquet cuff (width 11 cm) At the level of the heads of m. gastrocnemius-soleus Inflation rate: 0.5 kPa/s; maximum pressure limit: 180 kPa Electronic VAS (0—no pain, 10—maximal pain), sampling at 10 Hz PDT = VAS >0Assessment: Before and after skin and muscle selective anesthesia Before and during skin and muscle sensitization (capsaicin) 3 repetitions Supine position	Independent: Cuff inflation pressureDependent: PDT PTT PTL	Baseline pain thresholds (as established from the provided graphs): PDT = 17.1–32.6 kPa PTT = 45.9–62.5 kPa PTL = 5.0–6.26 cmThere is a correlation between pressure intensity and pain intensity after PDT is reached.Muscle anesthesia increased pain thresholds during CPA, and muscle hyperalgesia decreased them.Skin anesthesia did not significantly affect the pain thresholds during CPA, whereas skin hyperalgesia lowered the PDT but did not affect PTT.
Polianskis et al.²⁶	Healthy maleAge: 23–26 yearsn = 16	CPA: Double-chamber tourniquet cuff (width 13 cm) Nondominant leg, at the level of the heads of m. gastrocnemius-soleus Compression: (1) Constant inflation rate: 0.5 kPa/s (2) Constant intensity 10 min at predefined VAS 2, 4, 6 (3) Variable intensity 10 min at constant pain level (VAS 2, 4, 6) Maximum pressure limit: 200 kPa Electronic VAS (0—no pain, 10—maximal pain), sampling at 10 Hz PDT = VAS >0 Supine position Assessment: Pressure-pain sensitivity during single- and double-chamber compression Temporal profile of pain during constant compression Temporal pressure profile during constant level of pain	Independent: Cuff inflation pressure Area under the cuff Time of stimulationDependent: PDT PTT Temporal profile of pain Temporal pressure profile	Medians (interquartile ranges) for pressure-pain sensitivity
					Single chamber	Double chamber
				PDT (kPa)	21 (17–30)	14 (11–19)
				PTT (kPa)	64 (56–77)	50 (37–59)
				Increase in the compressed volume of deep tissues evokes spatial summation of pressure pain.Constant cuff pressure resulted in adaptation.Pain intensity was significantly correlated to the time of constant stimulation.Oscillating pressure evoked constant rather than adapting pain.
Jespersen et al.²¹	Healthy femaleAge: 25–60 years (median 45)Lower leg circumference: 35.7 ± 1.0 cmn = 16	(A) Maximal knee muscle strength in patients, assessed by an isokinetic dynamometer(B) CPA: Double-chamber tourniquet cuff (width 13 cm) Dominant lower leg Inflation rate: 0.5 kPa/s; maximum pressure limit: 100 kPa Electronic VAS (0—no pain, 10—maximal pain), sampling at 10 Hz PPT = VAS >0 Supine position	Independent: Cuff inflation pressureDependent: PPT PTol Pressure-Pain Limit	PDT = 19.9 ± 2.2 kPaPTT = 43.7 ± 4.2 kPaVAS at PTT: 9.4 ± 2.2 cmNo significant correlation between leg circumference and PPT or PTol.CPA parameters were significantly correlated to isokinetic muscle strength.
Izumi et al.²⁸	Healthy: 8 male 8 femaleAge: 28 ± 5 yearsBMI: 23.6 ± 3.2 kg/m²n = 16	(A) Single-point pressure algometry (hand-held): 1 cm² probe Lower limb: (1) 3 cm proximal to the tip of the greater trochanter (2) 3 cm posterior to the posterior edge of the greater trochanter (3) 3 cm distal to the distal edge of the greater trochanter (4) 3 cm anterior to the anterior edge of the greater trochanter (5) 5 cm distal to the tibial tuberosity upper limb(B) CPA: Double-chamber tourniquet cuff (width 13 cm): only one chamber inflated Thigh, as proximally as possible Inflation rate: 1 kPa/s; maximum pressure limit: 100 kPa Electronic VAS (0—no pain, 10—maximal pain), sampling at 0.5 Hz Cuff PPT = VAS >03 repetitions Supine position(C) Cutaneous pin-prick pain sensitivity(D) Thermal pain sensitivity(E) Hip joint pain provocation tests with hypertonic saline solution	Independent: Probe pressure Cuff inflation pressureDependent: PPT Cuff PPT	(A) Single-point pressure algometry (1) PPT = 442 ± 18 kPa (2) PPT = 565 ± 21 kPa (3) PPT = 462 ± 23 kPa (4) PPT = 577 ± 25 kPa (5) PPT = 562 ± 25 kPa(B) CPA: cuff PPT = 31.6 ± 1.1 kPa
Manafi-Khanian al.²²	Healthy maleAge: 31 yearsBMI: 25 kg/m²n = 1	(A) CPA: Single-chamber tourniquet cuff (width 6 cm) Right lower leg at the heads of m. gastrocnemius Inflation rate: 1 kPa/s; maximum pressure limit: 100 kPa Electronic VAS (0—no pain, 10—maximal pain), sampling at 10 Hz PDT = VAS >0 3 repetitions(B) MRI at different stimulation intensities: No pressure Mild (50% PDT): 9.7 kPa Painful (PDT intensity): 19.4 kPa Intense painful stimulation (5 cm on VAS): 30.2 kPa(C) 3D finite element model generation: establishment of Von-Mises stress	Independent: Cuff inflation pressureDependent: PDT PTT	PDT = 19.4 kPaPTT = 46.4 kPaCPA is more capable of generating mechanical stress and strain in deeper tissues than single-point pressure algometry.
Vaegter et al.²⁹	Healthy: 28 male 28 femaleAge: 20–30 yearsBMI: 16.4–38.7 kg/m²n = 56	CPA: Double-chamber tourniquet cuff (width 13 cm): only one chamber inflated Nondominant lower limb: 5 cm under tibial tuberosity Nondominant upper limb: 3 cm proximal to the cubital fossa Inflation rate: 1 kPa/s; maximum pressure limit: 80 kPa Electronic VAS (0—no pain, 10—maximal pain), sampling at 10 Hz PDT = VAS >0 Supine position Assessment of TSP: 10 repeated stimulations: 2 s of PTT intensity, 1 s of 5 kPa 3 mean values calculated: VAS I (stimulation 1–4), VAS II (stimulation 5–7) and VAS III (stimulation 8–10) (A) Before and immediately after 15 min of quiet rest (B) Before, immediately after and 15 min after aerobic bicycling exercise	Independent: Cuff inflation pressureDependent: PTT TSP	Lower-limb pressure pain sensitivity: PTT (kPa)
				Rest		Exercise
				Before	After	Before	After	Before
				68.3 ± 2.1	69.2 ± 2.0	68.3 ± 2.3	70.8 ± 1.9	71.8 ± 1.8
				Upper-limb pressure pain sensitivity: PTT (kPa)
				Rest		Exercise
				Before	After	Before	After	Before
				75.9 ± 1.5	74.9 ± 1.6	75.8 ± 1.5	76.5 ± 1.4	76.2 ± 1.4
				VAS III was significantly higher than VAS II, and VAS II was significantly higher than VAS I.Hypoalgesia was induced by aerobic and isometric exercises to different degrees.Isometric exercises reduced TSP.
Graven-Nielsen et al.²⁴	Healthy male: 51 younger:Age: 24.6 ± 5.0 yearsBMI: 23.3 ± 2.4 kg/m²17 middle-aged:Age: 53.0 ± 8.1 yearsBMI: 26.9 ± 3.5 kg/m²Healthy female:51 younger:Age: 24.2 ± 5.5 yearsBMI: 22.6 ± 3.4 kg/m²17 middle-agedAge: 53.9 ± 5.2 yearsBMI: 23.6 ± 3.5 kg/m²n = 136	(A) Single-point pressure algometry (hand-held): 1 cm² probeLower limb: 20 cm proximal to the base of patella Upper limb: 10 cm above the cubital fossa 2 repetitions Supine position(B) CPA: Double-chamber tourniquet cuff (width 13 cm) Nondominant lower limb: 5 cm under the tibial tuberosity Nondominant upper limb: 3 cm proximal to the cubital fossa Inflation rate: 1 kPa/s; maximum pressure limit: 80 kPa Electronic VAS (0—no pain, 10—maximal pain), sampling at 10 Hz cPPT = VAS >0 Assessment of TSP: 10 repeated stimulations: 2 s at cPTT intensity, 1 s at 5 kPa 3 mean values calculated: VAS I (stimulation 1–4), VAS II (stimulation 5–7), VAS III (stimulation 8–10) Supine position(C) Cold-pressor test	Independent: Probe pressure Cuff inflation pressureDependent: PPT cPPT cPTT cPTL TSP	Pressure pain sensitivity (kPa)
					Upper limb		Lower limb
					1	2	1	2
				PPT	367 ± 160	334 ± 162	543 ± 264	509 ± 243
				cPPT	30.4 ± 15.1	34.5 ± 15.8	26.7 ± 12.9	27.4 ± 11.8
				cPTT	69.1 ± 16.1	70.6 ± 15.4	58.4 ± 18.4	60.6 ± 19.5
				cPTL	6.1 ± 2.6	5.6 ± 2.9	6.5 ± 2.5	6.7 ± 2.9
				cPPT and cPTT were higher in men than women.cPPT was higher and cPTT lower in middle-aged subjects than in younger subjects.VAS III was significantly higher than VAS II, and VAS II was significantly higher than VAS I.TSP was larger in women than men.
Manafi-Khanian et al.³⁰	Healthy: 6 male 6 femaleAge: 23–33 yearsLower leg circumference: 31–36 cmBMI: 18.8–25.5 kg/m²n = 12	CPA: Single-chamber air tourniquet cuff (width 6 cm) Water tourniquet cuff (width 5 cm) Right leg just below the heads of m. gastrocnemius Inflation rate: 1 kPa/s; maximum pressure limit: 100 kPa Electronic VAS (0—no pain, 10—maximal pain), sampling at 10 Hz PDT = VAS >1 3 repetitions	Independent: Cuff inflation pressureDependent: PDT PTT IP IPD			Air cuff	Water cuff
				PDT (kPa)		18.3 ± 3.5	28.0 ± 10.1
				PTT (kPa)		42.7 ± 11.6	72.0 ± 11.6
				IP (kPa)	At PDT	18.4 ± 3.6	20.8 ± 5.8
				IP (kPa)	At PTT	41.5 ± 10.6	41.3 ± 10.0
				PDT and PTT were significantly lower when assessed with air cuff compared with water cuff.IP was not significantly different from inflation pressure of the air cuff, but significantly lower than inflation pressure of the water cuff. IP was not significantly different between the cuffs.IPD was significantly more homogeneous with the water cuff. Homogeneity of IPD was significantly correlated with pain sensitivity.
Manafi-Khanian et al.¹⁹	Healthy subjectn = 1	(A) CPA: Single-chamber tourniquet cuff (width 6 cm) Right lower leg at the heads of m. gastrocnemius Inflation rate: 1 kPa/s Electronic VAS (0—no pain, 10—maximal pain) 3 repetitions(B) MRI at different stimulation intensities: No pressure Mild (50% PDT): 9.7 kPa Painful (PDT intensity): 19.4 kPa - Intense painful stimulation (5 cm on VAS): 30.2 kPa(C) 3D finite element model generation: establishment of Von-Mises stress	Independent: Cuff inflation pressureDependent: PDT PTT	PDT = 19.4 ± 2.9 kPaPTT = N/ACPA is less capable of generating strain in tissues around bones than in more superficial muscle tissue.
Kvistgaard Olsen et al.³¹	Healthy10 male10 femaleAge: 24.8 ± 2.2 yearsBMI: 25.6 ± 4.1 kg/m²Lower leg circumference: 37.1 ± 3.3 cmn = 20	CPA: Double-chamber tourniquet cuff (width 13 cm) Dominant lower limb: widest part of m. gastrocnemius Inflation rate: 1 kPa/s Electronic VAS (0—no pain, 10—maximal pain) PT = VAS >0 3 repetitions Supine positionAssessment of TSI: Constant stimulation 10 min at individually defined compression level	Independent: Cuff inflation pressureDependent: PT PTol TSI	Pressure pain sensitivity (kPa)
					Assessment
					1	2	3
				PT	27.4 ± 9.0	26.0 ± 10.9	30.2 ± 13.4
				PTol	68.8 ± 28.5	62.0 ± 21.3	64.8 ± 22.2
Vaegter et al.²⁷	Healthy maleAge: 24.4 ± 2.0 yearsBMI: 24.8 ± 2.1 kg/m²n = 20	(A) CPA: Double-chamber tourniquet cuff (width 13 cm) Lower leg: 5 cm under tibial tuberosity Inflation rate: 1 kPa/s; maximum pressure limit: 100 kPa Electronic VAS (0—no pain, 10—maximal pain), sampling at 10 Hz cPPT = VAS >0 Assessment before and immediately after: (A) 15 min quiet rest (B) 3 min isometric knee extension at 30% MVC Seated position(B) Assessment of MVC(C) Assessment of heat pain sensitivity	Independent: Cuff inflation pressure MVCDependent: cPPT cPTT	MVC = 455.0 ± 86.7 NPressure pain sensitivity (kPa)
					Rest		MVC
					Before	After	Before	After
				cPPT	20.6 ± 8.5	21.7 ± 9.0	24.4 ± 11.2	26.3 ± 11.7
				cPTT	63.7 ± 18.4	64.2 ± 18.3	63.1 ± 18.3	74.2 ± 18.3
				Isometric exercise significantly increased PTT.
Lemming et al.³²	Healthy48 male: 26 normally active men (NAM) 22 highly active men (HAM)Age: 33.6 ± 11.1 yearsBMI: 24.7 ± 2.3 kg/m²50 female: 23 normally active women (NAW) 27 highly active women (HAW)Age: 35.2 ± 10.6 yearsBMI: 23.1 ± 2.7 kg/m²n = 98	CPA:Double-chamber tourniquet cuff (width 13 cm): one or both chambers inflatedLower limb: mid-portion of m. triceps suraeUpper limb: heads of m biceps and m. tricepsInflation rate: 1 kPa/s; maximum pressure limit: 100 kPaElectronic VAS (0—no pain, 10—maximal pain), sampling at 10 HzPDT = VAS >0.1	Independent: Cuff inflation pressureDependent: PDT PTT VAS-PTT SR	Lower-limb pressure pain sensitivity (kPa)
					HAM	NAM	HAW	NAW
				PDT_s	34.1 ± 21.0	19.7 ± 11.1	24.9 ± 17.4	19.2 ± 12.3
				PDT_d	24.1 ± 15.7	16.8 ± 10.4	20.6 ± 14.9	16.3 ± 11.2
				PTT_s	90.5 ± 18.0	90.0 ± 19.3	83.1 ± 22.5	73.5 ± 22.9
				PTT_d	81.1 ± 23.8	81.2 ± 24.9	67.6 ± 28.9	52.6 ± 21.9
				VAS-PTT_s	5.7 ± 3.6	7.2 ± 2.5	7.6 ± 2.8	8.4 ± 2.1
				VAS-PTT_d	7.5 ± 2.5	8.8 ± 1.6	8.3 ± 2.3	8.4 ± 2.2
				Upper-limb pressure pain sensitivity (kPa)
					HAM	NAM	HAW	NAW
				PDT_s	30.5 ± 17.4	28.4 ± 12.9	21.2 ± 12.5	19.6 ± 13.6
				PDT_d	34.5 ± 20.6	30.5 ± 18.0	29.9 ± 20.4	20.4 ± 13.3
				PTT_s	94.9 ± 14.2	98.8 ± 5.4	90.2 ± 15.7	80.0 ± 22.9
				PTT_d	92.2 ± 1.4	96.3 ± 10.9	84.3 ± 23.6	75.2 ± 27.0
				VAS-PTT_s	4.6 ± 3.2	5.9 ± 2.7	7.0 ± 2.9	7.4 ± 2.5
				VAS-PTT_d	5.1 ± 3.1	6.2 ± 2.6	7.3 ± 2.8	8.1 ± 2.1
				PDT for single-chamber stimulation was significantly higher in the highly active group.PTT for single- and double-chamber stimulation was significantly lower in women. No significant differences were found with respect to activity level.Higher cuff pain sensitivity was found in the lower limb than in the upper limb. Overall, 65–69% of the subjects reached the maximum pressure limit in the upper limb, and 29–54% in the lower limb.VAS-PTT scores were significantly higher in women, but they did not differ significantly with respect to activity level.SR was significantly higher in women, but did not differ significantly with respect to activity level.

CPA, computerized cuff pressure algometry; cPPT, cuff pressure pain threshold; cPTT, cuff pressure pain tolerance; cuff PPT, cuff algometry pain threshold; IP, mean interface pressure; IPD, interface pressure distribution; MVC, maximal voluntary muscle contraction; PDT, pain detection threshold; PDT^*, pressure-pain detection threshold; PPT, pressure-pain threshold; PT, pressure-pain threshold; PTL, pain tolerance limit; PTol, pressure-pain tolerance; PTT, pain tolerance threshold; SR, summation ratio; TSI, temporal summation index; TSP, temporal summation of pain; VAS-PTT, score on the Visual Analogue Scale at pain tolerance threshold.

Table 2.

Summary of PDT and PTT Data from the Review

		CPA
Reference	n	Special	PDT (kPa)	PTT (kPa)	PTL (cm)	Single-point algometry PPT (kPa)
Polianskis et al.²³	12	Left	26	61	5.4	433–526
		Right	25	56	5.5
Polianskis et al.²⁵	12	—	17–33	46–63	5.0–6.3	—
Polianskis et al.²⁶	16	Single chamber	17–30	56–77	—	—
		Double chamber	11–19	37–59
Jespersen et al.²¹	16	—	20	44	9.4	—
Izumi et al.²⁸	16	—	32	—	—	442–577
Manafi-Khanian al.²²	1	—	19	46	—	—
Vaegter et al.²⁹	56	—	—	68–72	—	—
Graven-Nielsen et al.²⁴	136	—	27	58–61	6.5–6.7	509–543
Manafi-Khanian et al.³⁰	12	Air	18	43	—	—
		Water	28	72
Manafi-Khanian et al.¹⁹	1	—	19	—	—	—
Kvistgaard Olsen et al.³¹	20	—	26–30	62–69	—	—
Vaegter et al.²⁷	20	—	21–26	63–74	—	—
Lemming et al.³²	98	Single chamber	19–34	74–91	5.7–8.4	—
Double chamber	16–24	53–81	7.5–8.8

Overall minimal and maximal values are highlighted in bold for each variable.

The pressure-pain sensitivity assessed by CPA indicates a common pattern across the studies. At the lower limb, PDT levels ranged from 16.3 ± 11.2 to 34.1 ± 21.0 kPa,³² but the majority of the reported PDTs were under 30 kPa (∼20–27 kPa). PTT levels ranged from 42.7 ± 11.6³⁰ to 90.5 ± 18.0 kPa.³² At the upper limb, PDT levels ranged from 19.6 ± 13.6 to 34.5 ± 20.6 kPa,³² and PTT levels ranged from 69.1 ± 16.1²⁴ to 98.8 ± 5.4 kPa.³² Several factors were found to significantly influence CPA-assessed pain thresholds, such as tourniquet-cuff properties, pattern of compression, inter- and intra-individual differences, and exercise.

Significantly higher PDT and PTT were found during single-chamber compression compared with double-chamber compression, indicating spatial summation of pressure-induced pain.^26,32 Significantly higher PDT and PTT were also assessed by a water cuff compared with an air cuff, owing it to a larger homogeneity of the interface pressure distribution of the former. Further, with the water cuff, the interface pressure was significantly lower than the inflation pressure.³⁰

The number of compressions, compression rate, and pain thresholds were strongly correlated.²³ The increase of compression rate increased PTT and decreased PTL.²³ Pain intensity was significantly correlated to the time of constant stimulation.²⁶ Constant cuff pressure resulted in pain adaptation, whereas oscillating pressure did not.²⁶

PDT and PTT were higher for men than women,^24,32 and PDT was higher and PTT was lower in older participants.²⁴ PDT and PTT were significantly correlated to isokinetic muscle strength.²¹ PDT for single-chamber stimulation was significantly higher in a group of highly active participants, but no significant differences in PTT were found with respect to participants' activity level.³² Isometric^27,29 and aerobic²⁹ exercises directly before CPA increased pain thresholds.

Discussion

Pressure-pain sensitivity of soft tissues

A key finding of the current review is that previous examples of interface pressures of seated individuals (22 kPa¹²) and exoskeleton users (29.3 kPa¹³) involve pressures exceeding PDTs identified in this review. A risk curve due to deformation, ischemia, and other factors has been proposed by Stekelenburg et al.,⁷ indicating that the risk for tissue damage depends not only on the magnitude of external loading but also on its duration. By adapting this risk curve, we hypothesize the probability of deep tissue injury when loaded with pressure-induced pain thresholds (Fig. 3). The curve indicates two boundary values for external loading, where the risk for tissue damage depends on the loading duration. Above the upper extreme, tissue damage occurs instantly, and below the lower extreme, no damage occurs, independently of the loading duration.⁷ The absolute pressure values at these points are not proposed, but considering pain to be an indicator of potential tissue damage, and the correlation between pressure intensity and pain intensity after PDT is reached,²⁵ we hypothesize that: (1) the upper extreme occurs at pressures that cause the worst pain imaginable (i.e., PTL = 10 cm); (2) the minimal value occurs below PDT; and (3) PDT and PTT fall near the lower and the higher extremes, respectively. We propose that compression at PDT is likely to induce deep tissue damage over time, depending on the duration and pattern of external loading (Fig. 3B), and that compression at PTT is likely to induce near instantaneous damage (Fig. 3A); therefore, it should be avoided outright for soft exoskeletons.

FIG. 3.

Hypothetical tissue damage curve based on the external pressure and loading duration, with indicative position of PTT and PDT pressure studies on the tissue damage curve. (A) Low risk loading duration at PTT; (B) low risk loading duration at PDT. PDT, pain detection threshold; PTT, pain tolerance threshold.

To establish a safety standard for external loading, the absolute value of the minimal external pressure magnitude needs to be identified. We hypothesize that it falls between the reported potentially injuring stresses for muscle (35 kPa³³) and the compression that is considered to be beneficial, such as that produced by compression garments. According to the RAL-GZ standard³⁴ for medical compression stockings, the highest compression class stockings exert more than 49 mmHg (6.5 kPa) of compression at the ankle. However, it has been reported that even compression exceeding 30–40 mmHg (4.0–5.3 kPa) can cause discomfort.³⁵ Higher cuff pain sensitivity was found in the lower limb than in the upper limb,³² which indicates that data for the limbs cannot be used interchangeably for wearable robotics design without substantiating research.

Comparison of cuff pressure algometry and single-point algometry

There was a significant difference between the PPTs obtained by single-point algometry and CPA. The PDTs for single-point pressure algometry are site dependent and tend to be 20 times that of CPA. PDT thresholds for single-point algometry in the reviewed studies ranged from 433 ± 15 to 526 ± 20 kPa on the lower leg,²³ from 509 ± 243 to 543 ± 264 kPa at the thigh,²⁴ and from 442 ± 18 to 577 ± 25 kPa at the hip.²⁸

As mentioned earlier, higher cuff pain sensitivity was found in the lower limb than in the upper limb,³² which is in contrast to the inverse relationship established by single-point pressure algometry.²⁴ It has been reported that sustained external pressure corresponding to 50% PDT becomes painful in a few minutes.²⁰ In instances such as this, it is of great importance to distinguish between the two methods for assessing pressure-pain sensitivity. In single-point pressure algometry, 50% PDT corresponds to ∼140–300 kPa, whereas in CPA, it corresponds to about 9–15 kPa. In two studies discussed,^19,22 50% PDT (9.7 kPa) was considered as “mild pressure.” Although this is still considerably lower than the thresholds established by single-point pressure algometry, it is more than twice as high as the recommended limit for interface pressure (4.3 kPa), cited previously and widely believed to be outdated.

Application of CPA data to gait-assistive devices

Sustained constant pressure was shown to result in adaptation to pain in healthy adults, whereas oscillating pressure caused an increase in pain intensity with time.^24,26,29 This presumably happens due to central modulation of pain, most importantly TSP. TSP is defined as gradually increasing the perception of pain that occurs when a series of identical painful stimuli is applied with a frequency above 0.3 Hz (Fig. 4a).^{24,29,36–39} This is very important for gait-assistive devices where cyclical movements are involved. The fact that higher compression rate not only facilitates TSP but also decreases PTL²³ plays an important role when considering gait velocity, that is, faster walking could become intolerably painful at lower pain intensities. On the other hand, exercise-induced hypoalgesia^27,29 could act as a mitigating factor due to the aerobic nature of walking.

FIG. 4.

(a) Temporal and (b) spatial summation of pain.

Pain was found to develop earlier and faster during constant compression with a wider tourniquet cuff than with a narrower one.^32,40 This may seem counterintuitive, as distribution of force over a larger area increases the homogeneity of compression, which consequently increases the tolerability of pain.³⁰ However, previous studies report that wider cuffs eliminate arterial blood flow at lower pressures⁴¹ and without total collapse of the arteries.⁴² Moreover, the percent of cuff pressure reaching the deep tissue near the bone was found to be much higher for wider cuffs compared with narrow ones.⁴¹ Most importantly, wider cuffs subject a greater mass of tissue to compression,⁴¹ thereby activating a larger number of nociceptors. This leads to Spatial Summation of Pain defined as increased perception of pain at the same magnitude of mechanical stimulation when larger, compared with smaller areas of body tissue that are stimulated (Fig. 4b).^43,44 Thus, there seems to be a need to establish an efficient relationship between the cuff width and the force transmitted to the body, when circumferential compression is used to actuate movement. Interestingly, the interface pressure did not differ significantly from the inflation pressure of a 6 cm wide cuff.³⁰

Finally, TSP is more prominent in nonhealthy individuals, and pressure-induced pain thresholds are achieved at lower pressures in less active people^21,32 and people with lower isokinetic muscle strength,²¹ who are the most probable users of gait-assistive devices. There is a need for further research on CPA perceptions of pain and discomfort, as well as tissue responses, for patients in addition to healthy people. Further, there is a need to study CPA for these users, considering temporal aspects to reflect short to medium use of soft exoskeletons, including during gait patterns.

Terminology and definition of pressure-induced pain thresholds

Single-point algometry traditionally only assesses one parameter, called PPT, which corresponds to PDT in CPA studies. PTT, on the other hand, is a parameter introduced by CPA, and in this method, both PDT and PTT are referred to as PPTs.

Based on this systematic review, we have identified a need to standardize terminology, as well as the definitions of measured parameters in CPA. Namely, PDT, as classically defined, is also dubbed Pressure-Pain Detection Threshold (PDT²³), Pressure-pain Threshold²¹ (PT,³¹ or cPPT²⁷), and cuff Algometry Pain Threshold (cuff PPT²⁸). PTT is also called Pressure-pain Tolerance²¹ (PTol³¹). Moreover, PDT is usually defined as pain intensity exceeding 0 on the VAS scale, but one study³⁰ defined it as the intensity exceeding 1 cm on the VAS scale. The definition of PTT varies, being defined as maximal pain intensity of 10 cm on the VAS scale²² or at the point a test subject induced termination of the experiment.^{21,24,27,29,30} In one study,²⁶ the authors explain that the pain intensity strong enough to make one feel like stopping the stimulation does not comply with the classical description of pain tolerance, therefore they provisionally labeled that threshold as the PTL.

Limitations

The present review summarizes pain-inducing pressure thresholds achieved by CPA with healthy participants. We hypothesize that deep somatic tissue pain indicates excessive external loading, and can, thus, be useful in studying and possibly setting safe thresholds for circumferential compression of the lower limbs. However, the applicability of these thresholds to the design of soft lower limb exoskeletons still needs to be established.

Moreover, most of the reviewed studies included less than 20 participants, and all participants were healthy. Potential exoskeleton users, however, are understood to be patients with specific pathologies that can change pain perception, which warrants a separate research review. Also, the variability of the cuffs used in the studies renders it difficult to compare all the results, as PDT depends on cuff width⁴⁰ and shape.⁴⁵ Also, the anatomical sites for testing differed, and pain intensity at PDT was not consistently defined.

Conclusions

For this review, we identified 13 studies, where CPA was performed on healthy adults at the lower and upper limb. Higher cuff pain sensitivity was found in the lower limb. PDT levels ranged from 16.3 ± 11.2 to 34.1 ± 21.0 kPa at the lower limb, and from 19.6 ± 13.6 to 34.5 ± 20.6 kPa at the upper limb. PTT levels ranged from 42.7 ± 11.6 to 90.5 ± 18.0 kPa at the lower limb, and from 69.1 ± 16.1 to 98.8 ± 5.4 kPa at the upper limb. We propose that the levels of PDT, in particular, are of primary interest for soft exoskeleton use as they relate to initial detection of discomfort. It is to be expected that levels corresponding to PDT will result in tissue damage during prolonged use and should, therefore, be avoided. Factors that significantly influenced pain thresholds were tourniquet-cuff properties, pattern of compression, inter- and intra-individual differences, and exercise, which explains some of the sources of variation.

The results of single-point algometry showed higher pain sensitivity in the upper limb, and the thresholds were about 20 times higher than those obtained by cuff pressure algometry. Further, sustained constant pressure resulted in adaptation to pain, whereas oscillating pressure caused an increase in pain intensity with time. Also, participants tolerated higher pain intensities at lower compression rates.

The results acquired by cuff pressure algometry give important insight into the relationship between external loading and discomfort or pain. This can be useful in studying and possibly setting safe thresholds for circumferential compression of the lower limbs, which may occur when soft robotics are used for wearable assistive device applications. However, in the absence of laboratory studies under the specific circumstances during soft-exoskeleton use, we can only provide an approximate range for maximal loading that corresponds to the PDT levels identified in this review for healthy adults (i.e., 16–34 kPa). More research is needed in studying PDT and tissue physiological response for cyclical temporal loading using CPA to provide more detailed safety and comfort guidance for soft exoskeleton contact with the body.

Footnotes

Acknowledgments

This research was completed as part of the XoSoft project, which has received funding from the European Union's Horizon 2020 framework programme for research and innovation under grant agreement number 688175.

Author Disclosure Statement

No competing financial interests exist.

References

Asbeck

, De Rossi

SMM

, Galiana

, Ding

, Walsh

. Stronger, smarter, softer: next-generation wearable robots. IEEE Robot Autom Mag, 2014; 21:22–33.

De Rossi

SMM

, Vitiello

, Lenzi

, Ronsse

, Koopman

, Persichetti

, et al. Sensing pressure distribution on a lower-limb exoskeleton physical human-machine interface. Sensors (Basel, Switzerland), 2010; 11:207–227.

Bouten

, Oomens

, Baaijens

, Bader

. The etiology of pressure ulcers: skin deep or muscle bound?. Arch Phys Med Rehabil, 2003; 84:616–619.

Oomens

CWJ

, Loerakker

, Bader

. The importance of internal strain as opposed to interface pressure in the prevention of pressure related deep tissue injury. J Tissue Viability, 2010; 19:35–42.

Agam

, Gefen

. Pressure ulcers and deep tissue injury: a bioengineering perspective. J Wound Care, 2007; 16:336–342.

Black

, Baharestani

, Cuddigan

, Dorner

, Edsberg

, Langemo

. National Pressure Ulcer Advisory Panel's updated pressure ulcer staging system. Urol Nurs, 2007; 27:144.

Stekelenburg

, Gawlitta

, Bader

, Oomens

. Deep tissue injury: how deep is our understanding?. Arch Phys Med Rehabil, 2008; 89:1410–1413.

Reenalda

, Jannink

, Nederhand

, Ijzerman

. Clinical use of interface pressure to predict pressure ulcer development: a systematic review. Assist Technol, 2009; 21:76–85.

Swain

, Bader

. The measurement of interface pressure and its role in soft tissue breakdown. J Tissue Viability, 2002; 12:132–146.

10.

Defloor

. The risk of pressure sores: a conceptual scheme. J Clin Nurs, 1999; 8:206–216.

11.

Bader

. The recovery characteristics of soft tissues following repeated loading. J Rehabil Res Dev, 1990; 27:141.

12.

Shen

, Parsons

. Validity and reliability of rating scales for seated pressure discomfort. Int J Ind Ergon, 1997; 20:441–461.

13.

Tamez-Duque

, Cobian-Ugalde

, Kilicarslan

, Venkatakrishnan

, Soto

, Contreras-Vidal

. Real-time strap pressure sensor system for powered exoskeletons. Sensors, 2015; 15:4550–4563.

14.

Mak

AFT

, Zhang

, Boone

. State-of-the-art research in lower-limb prosthetic biomechanics-socket interface: a review. J Rehabil Res Dev, 2001; 38:161–173.

15.

Sangeorzan

, Harrington

, Wyss

, Czerniecki

, Matsen

. Circulatory and mechanical response of skin to loading. J Orthop Res, 1989; 7:425–431.

16.

Linder-Ganz

, Gefen

. Stress analyses coupled with damage laws to determine biomechanical risk factors for deep tissue injury during sitting. J Biomech Eng, 2009; 131:11003.

17.

Makhsous

, Lin

, Hanawalt

, Kruger

, LaMantia

. The effect of chair designs on sitting pressure distribution and tissue perfusion. Hum Factors, 2012; 54:1066–1074.

18.

Mak

AFT

, Zhang

, Tam

EWC

. Biomechanics of pressure ulcer in body tissues interacting with external forces during locomotion. Annu Rev Biomed Eng, 2010; 12:29–53.

19.

Manafi-Khanian

, Arendt-Nielsen

, Graven-Nielsen

. An MRI-based leg model used to simulate biomechanical phenomena during cuff algometry: a finite element study. Med Biol Eng Comput, 2016; 54:315–324.

20.

Pons

. Wearable Robots: Biomechatronic Exoskeletons. Hoboken, N.J: Wiley; 2008, pp. 154–156.

21.

Jespersen

, Dreyer

, Kendall

, Graven-Nielsen

, Arendt-Nielsen

, Bliddal

, et al. Computerized cuff pressure algometry: a new method to assess deep-tissue hypersensitivity in fibromyalgia. Pain, 2007; 131:57–62.

22.

Manafi-Khanian

, Arendt-Nielsen

, Frøkjær

, Graven-Nielsen

. Deformation and pressure propagation in deep somatic tissue during painful cuff algometry. Eur J Pain, 2015; 19:1456–1466.

23.

Polianskis

, Graven-Nielsen

, Arendt-Nielsen

. Computer-controlled pneumatic pressure algometry—a new technique for quantitative sensory testing. Eur J Pain, 2001; 5:267–277.

24.

Graven-Nielsen

, Vaegter

, Finocchietti

, Handberg

, Arendt-Nielsen

. Assessment of musculoskeletal pain sensitivity and temporal summation by cuff pressure algometry: a reliability study. Pain, 2015; 156:2193.

25.

Polianskis

, Graven-Nielsen

, Arendt-Nielsen

. Pressure-pain function in desensitized and hypersensitized muscle and skin assessed by cuff algometry. J Pain, 2002; 3:28–37.

26.

Polianskis

, Graven-Nielsen

, Arendt-Nielsen

. Spatial and temporal aspects of deep tissue pain assessed by cuff algometry. Pain, 2002; 100:19–26.

27.

Vaegter

, Hoeger Bement

, Madsen

, Fridriksson

, Dasa

, Graven-Nielsen

. Exercise increases pressure pain tolerance but not pressure and heat pain thresholds in healthy young men. Eur J Pain, 2017; 21:73–81.

28.

Izumi

, Petersen

, Arendt-Nielsen

, Graven-Nielsen

. Pain referral and regional deep tissue hyperalgesia in experimental human hip pain models. Pain, 2014; 155:792–800.

29.

Vaegter

, Handberg

, Graven-Nielsen

. Isometric exercises reduce temporal summation of pressure pain in humans. Eur J Pain, 2015; 19:973–983.

30.

Manafi Khanian

, Arendt-Nielsen

, Kjær Petersen

, Samani

, Graven-Nielsen

. Interface pressure behavior during painful cuff algometry. Pain Med (Malden, Mass), 2016; 17:915–923.

31.

Kvistgaard Olsen

, Fener

, Waehrens

, Wulf Christensen

, Jespersen

, Danneskiold-Samsøe

, et al. Reliability of pain measurements using computerized cuff algometry: a DoloCuff Reliability and Agreement Study. Pain Pract, 2017; 17:708–717.

32.

Lemming

, Börsbo

, Sjörs

, Lind

, Arendt-Nielsen

, Graven-Nielsen

, et al. Cuff pressure pain detection is associated with both sex and physical activity level in nonathletic healthy subjects. Pain Med (Malden, Mass), 2017; 18:1573–1581.

33.

Gefen

, Gefen

, Linder-Ganz

, Margulies

. In vivo muscle stiffening under bone compression promotes deep pressure sores. J Biomech Eng, 2005; 127:512.

34.

RAL German Institute for Quality Assurance and Labelling. Medical compression hosiery—Quality assurance. In: Requirements, Compressive Behaviour, Compression Classes, vol. RAL-GZ 387/1. Sankt Augustin, Germany: RAL German Institute for Quality Assurance and Labelling, 2008, p. 13.

35.

Chan

, Fan

. Effect of clothing pressure on the tightness sensation of girdles. Int J Cloth Sci Tech, 2002; 14:100–110.

36.

Graven-Nielsen

, Arendt-Nielsen

. Translational aspects of musculoskeletal pain: from animals to patients. In: Arendt-Nielsen

(Ed). Fundamentals of Musculoskeletal Pain. Seattle, WA: ASP Press, 2008, pp. 347–366.

37.

Anderson

, Craggs

, Bialosky

, Bishop

, George

, Staud

, et al. Temporal summation of second pain: variability in responses to a fixed protocol. Eur J Pain, 2013; 17:67–74.

38.

Latremoliere

, Woolf

. Central sensitization: a generator of pain hypersensitivity by central neural plasticity. J Pain, 2009; 10:895–926.

39.

Staud

. The important role of CNS facilitation and inhibition for chronic pain. Int J Clin Rheumtol, 2013; 8:639–646.

40.

Estebe

, Le Naoures

, Chemaly

, Ecoffey

. Tourniquet pain in a volunteer study: effect of changes in cuff width and pressure. Anaesthesia, 2000; 55:21–26.

41.

Crenshaw

, Hargens

, Gershuni

, Rydevik

. Wide tourniquet cuffs more effective at lower inflation pressures. Acta Orthop Scand, 1988; 59:447.

42.

Moore

. Wide tourniquets eliminate blood flow at low inflation pressures. J Hand Surg, 1987; 12:1006.

43.

Defrin

, Ronat

, Ravid

, Peretz

. Spatial summation of pressure pain: effect of body region. Pain, 2003; 106:471–480.

44.

Greenspan

, Thomadaki

, McGillis

. Spatial summation of perceived pressure, sharpness and mechanically evoked cutaneous pain. Somatosens Mot Res, 1997; 14:107.

45.

John

, Narracott

, Morris

, Woodcock

, Lawford

, Hose

. Influence of intermittent compression cuff design on interface pressure and calf deformation: experimental results. Conf Proc IEEE Eng Med Biol Soc, 2007; 2007:2122–2125.