Abstract
A system for measuring clothing pressure employing a renewed hydrostatic pressure-balancing method was examined using three calibration methods. All methods revealed an almost perfectly linear Y = X relation for the pressure load (X) and the reading of the system (Y). In the application, the distributions of elastic band pressure were examined on 21 planes from head to foot. The preferred elastic band pressures of the leg and arm were significantly higher than those of the neck and abdomen. These results are due to the large presence of the autonomic nervous system at the surfaces of the neck and abdomen. In the area of the abdomen, the preferred elastic band pressure was higher from the mammilla to the shoulder than for the anteroposterior midlines. The development of compression ware must consider appropriate tightening for each body part.
Keywords
Clothing pressure as an index of clothing comfort has been well studied. Methods of measuring clothing pressure can be classified as direct and indirect. Kirk and Ibrahim 1 proposed an indirect method based on the horizontal and vertical radii of curvature and tension. That calculation method applied Laplace’s expression based on membrane equilibrium theory to clothes and led to the development of the dome method2–4 then the linearizing method. 5 Ishimaru predicted the pressure of worn clothing from physical properties of the cloth and data of the human body shape 6 and proposed a simulation technology that can be used to predict the pressure of worn clothing without first sewing the clothes. 7 Lee et al. 8 calculated clothing pressure accurately by dressing a human body model with clothes affixed to a concentric circle beforehand and obtaining the correct tension and radius of curvature for the main and secondary axes of the stretch of clothing. However, the above-mentioned methods cannot simultaneously measure both the actual pressure applied by the clothing and the feeling of tightening experienced when wearing the clothing. Further comparison is difficult because the actual pressure measurements and subjective feeling are not measured at the same time. Direct methods of measuring clothing pressure were therefore employed to allow such comparison.
In 1991, 9 Mitsuno et al. reported a measurement system based on a hydrostatic pressure-balancing method that uses a water bag for measuring clothing pressure, whereas Ito 10 and Kominami 11 published a contact method of measuring clothing pressure with an air bag. The former study adopted water as a pressure medium in a pressure detector, whereas the latter adopted air. Both media have their advantages and disadvantages. Water is difficult to compress and has a coefficient of thermal expansion lower than that of air, and thus provides much better measurement accuracy. However, the potential energy relating to the height between the water bag and the transducer must be considered. In other words, when using a pressure detector, the pressure transducer must be set at the same height as the water bag. In contrast, there is no problem of potential energy if air as a pressure medium is chosen. However, air is more easily compressed than water. The bag must therefore be sufficiently filled with air so that the opposite sides do not come into contact with each other when the bag is pressed, and the bag of the pressure detector is thus thicker and stiffer than the water bag.
Pressure acting on body parts that are more rigid than the pressure detector is then measurable. However, the pressure values will be erroneously low when the pressure detector is stiff because the pressure detector is buried in a soft body region, such as a region of subcutaneous fat in the abdomen of women. Measurement is impossible with a stiff detector that is unable to bend along the small radius of curvature in a measured body region. In addition, the time constant as an index of the response speed of transient phenomena is large, the air bag does not seem able to follow respiratory movements more quickly than about 0.3 Hz, and the resolving power is poor.
In the scenario of orthopedic medical care for the treatment of a scar, Giele et al. 12 reported that a proposed method of inserting a needle connected to a low-speed pressure transducer under the skin performs better than making a measurement between the skin and boundary surface of a garment. This method does not need to insert a sensor between the skin and a garment, and the wearing conditions are therefore not changed in measuring clothing pressure. However, the method is not suited to the present study because the participants feel pain. Sawada 13 measured the clothing pressure of a pressure suit using a rubber balloon connected to a volume of mercury as a pressure detector. He reported that the relation between the load and the reading of the system was linear from 0 to 88 mmHg (117.3 hPa). However, as Watanabe et al. 14 pointed out, when a pressure detector is made from a stretching material, such as rubber, the contact area is affected by the intensity of pressurization and the relation between the reading of the system and load on the pressure detector is not linear.
Ferguson et al. 15 reported on Flexi-Force. This is a sensor made from two layers of polyester film buried in conductive silver materials and filled with pressure-sensitive ink and it does not respond to pressurization with water. A sheet sensor is similar to a Flexi-Force sensor. Therefore, neither the Flexi-Force sensor nor sheet sensors, which are rigid thin sensors buried in soft materials without distortion, respond to a soft material such as water.
Flexible pressure sensors advocated by Wang et al. 16 can measure pressure in a range (0–8 MPa) wider than that of clothing pressures experienced in everyday life and were developed to measure the shock transmitted through a protective suit. Such sensors have a slightly larger diameter of 35 mm and thickness of 3 mm. For example, the comfort pressure acting on the abdomen when a woman is standing is low, at 5 hPa, and a measuring range of approximately 0–90 hPa is sufficient even for someone who is working, except when measuring the pressure imparted by shoes. The measuring systems described above are unsuitable for the present study owing to their size and hardness. Our ultimate goal is to obtain basic information that can serve as a reference when designing comfortable compression products.
Experimental method
Measuring system for clothing pressure
Renewed system
The present study therefore renewed the measuring system, employing a hydrostatic pressure-balancing method for clothing pressure.9,17 In addition, a pressure detector (Pouch) was constructed because of the lack of available options. A high-performance measuring system for clothing pressure was rebuilt, and how band pressure depends on the whole body was examined when using this system.
A diagram of the renewed system for measuring clothing pressure is shown in Figure 1. A pressure detector (Pouch) was constructed. The change in pressure detected by the Pouch is converted to an electric signal by a blood pressure transducer (Nihon Kohden, P10EZ-1). The electrical signal is amplified by a sensor coupler (PP-101H) and amplifier (AP-100H) and saved to a personal computer during monitoring using a polygraph system (RMT-1000 software and DC-300H hardware). The system employs a small flat bag (25 mm × 15 mm × 0.1 mm). The Pouch pressure detector is filled with water as shown in Figure 2. Pouch was made from thin (20 µm) and high-density polyethylene film, allowing flat contact with the measuring region. In other words, the radii of curvatures of both contact surfaces were infinite and the pressure could be measured without regard to tension in either surface. The Pouch was filled with distilled water through a double-lumen tube, thus ensuring the distilled water was enclosed within the system. In addition, two T-shaped stopcocks were used to remove air bubbles. The respiratory movement of participants was monitored using a thermistor (TR-861 T), and the girth change was examined using a thoracic pickup sensor (TR-851 T). The thoracic change was recorded simultaneously to examine which parts of the output wave pattern from the thermistor corresponded to exhale and inhale phases. The electrical signals were amplified by an amplifier unit (PP-104H, AP-100H) and simultaneously recorded with the clothing pressure.
Overview of the measuring system. Photographs showing the pressure detector (Pouch), a thermistor for monitoring respiration, and a thoracic pickup for measuring the girth change. The Pouch in the photograph had already been used for approximately 200 h.

Calibration methods
Pressure generated between the human body and clothes is affected by the hardness (i.e. rigidness or softness) of the measuring region of the body and clothing materials. The three calibration methods described below were thus considered. Two are based on water pressure: one uses the contact surface between subcutaneous fat and flexible clothes as soft materials, whereas the other uses a spherical shell to represent bone and a balloon to represent subcutaneous fat and flexible clothing. The third method uses a plastic piece and weight,
9
with both contact surfaces being rigid; e.g. a stiff belt and a rigid region, such as bone.
Devices for calibrating clothing pressure.

Physical properties of participants and the experimental band
Participants
Participants were 15 females aged 21.3 ± 1.2 years, whose physical characteristics (mean ± SD) were a height of 158.4 ± 3.5 cm, weight of 51.1 ± 5.5 kgf, chest girth of 79.3 ± 6.1 cm, top bust girth of 79.0 ± 6.1 cm, under bust girth of 69.8 ± 4.1 cm, waist girth of 65.3 ± 4.8 cm, lower waist girth of 72.0 ± 7.0 cm, hip girth of 89.9 ± 3.5 cm, groin girth of 53.2 ± 3.8 cm, and body mass index of 20.4 ± 1.9 kgf·m−2. The bodies of the participants ranged widely from thin to obese. The participants slept for 7 hours on the night prior to the experiment and ate 2 hours before the measurement. After entering a room with an artificial climate (Artificial Climate Experimental System of Shinshu University; room temperature, 24.5 ± 0.3℃; relative humidity, 50% ± 0.5%; air current, 0.11 ± 0.16 m/s; illumination, 827 ± 27 lx), the clothing pressure and pressure feeling were measured for all body parts of all participants over 1 hour; further detail is given below and in Figures 4 and 5. The participants were able to evaluate a perfect-fitting feeling without sweating or shaking because they were in a comfortable environment at a neutral temperature. The participants acclimatized for at least 1 hour after entering the climatic chamber before the experiment began.
Measuring planes and points for the band pressure. Questionnaire on the pressure felt when wearing the experimental band.

Elastic band
The textile structure of the experimental elastic band was plain weave, the weft materials were rayon and natural rubber, the warp material was rayon, and the textile density was 14 ends·cm−1, 16 picks·cm−1. The elastic band (width, 2.5 cm; thickness, 1.29 ± 0.02 mm, measured with a dial thickness gage (Tokyo Seimitsu Kogyo); weight, 0.07 gf·cm−1) was examined in a tensile test under conditions of a grab width of 100 mm and strain rate of 50 mm/min. The samples were strained to 50 mm (50%) and returned promptly by a tensile strength tester (Shimazu, Autograph S-100). The maximum value of stress was 28.71 ± 4.54 gf/cm (n = 9), which shows the elastic band was stretchable. A trial calculation of the tensile load was made using Kirk’s equation for the clothing pressure (P) as a function of tensile load and radii of curvature (P = TH/ρH + TV/ρV, where ρH and ρV are the horizontal and vertical radii of curvature and TH and TV are horizontal and vertical tensions, respectively). The radii of curvature were measured using a small sliding gauge. Results show that the tensile load was a maximum of 15 gf/cm, as mentioned below.
Measuring points and planes for clothing pressure
Measuring points
Participants repeatedly tightened and loosened a band with a length almost the same as their girth. The heights of the band were decided according to JIS Z 8500:2002. Participants adjusted the length of the band to achieve “a perfect fit”, corresponding to a comfortable pressure feeling. The band length was fixed using a double clip exerting 5.5 gf by the experimenter. The participants prepared their posture (standing in an upright position) and evaluated the feeling of tightness, and the clothing pressure generated between the skin and elastic band was then measured. The measuring points of clothing pressure are shown in Figure 4. The right figure shows the cross-section of each plane having four or 12 measuring points. F (B) is the cross-point between the anterior (posterior) middle line and the each girth line, whereas RA (LA) is the point of 3 cm to the right (left) of F on the each girth line, RB (LB) is the cross-point of the right (left) mammilla line and the each girth line, RC (LC) is the cross-point of the right (left) side line and the each girth line, RD (LD) is the cross-point between the right (left) shoulder line and the each girth line, and RE (LE) is the point of 3 cm to the right (left) of B on the each girth line.
Measuring planes
The preferred elastic band pressure (EBP) was measured on 21 planes (i.e. two planes on the head, seven on the trunk, and six on each of the upper and lower limbs) corresponding to the 138 measuring points of clothing pressure shown in Figure 4. The right figure shows the cross-section of each plane having four to 12 measuring points. First, the transducer was attached to a pole at the same height as the Pouch and the measuring plane, and zero pressure was confirmed. Second, the Pouch was inserted into the gap between the skin and band using a thin spatula that was inserted into the Pouch pocket. The experiment was carried out in the follicular phase of the menstrual cycle of each participant, because the participant would be more sensitive to clothing pressure than during the luteal phase. 18 The data for each plane obtained in the experiments were analyzed by a one-factor repeated measures of variance. The clothing pressure in eight groups of measuring planes was analyzed by multiple comparison employing the Tukey method.
Evaluation of the pressure feeling
The pressure feeling was evaluated with the ratio scale. 19 The participants evaluated themselves in terms of the pressure they felt when wearing the elastic band by completing a questionnaire (see Figure 5). Before putting on the elastic band, participants estimated what they considered a “loose-fitting feeling” and “perfect-fitting feeling”. They then evaluated the pressure stimulation provided by the band under two different conditions; one was wearing the experimental band in the measuring plane without tightening, whereas the other was wearing the band with a loose-fitting feeling. The participants were asked to indicate individual pressure feelings under both conditions to provide base feelings, and to mark points on a line in the questionnaire corresponding to a loose-fitting feeling. In the same way, they were asked to mark points for a perfect-fitting feeling using loose-feeling and tight-feeling conditions as a basis. Participants who liked a tight band marked their loose-fitting feeling more toward the right whereas participants did not like a tight band marked their perfect-fitting feeling more toward the right. The participants then tightened the experimental band so they felt it fit perfectly, and marked a point on a line in the questionnaire for that feeling.
The length of the line connecting the points marked by the participants to indicate the loose- and perfect-fitting feelings was measured. Numerical values for the pressure feeling were obtained by supposing that the feeling without tightening had a score of zero and a perfect-fitting feeling had a score of 1. 19 In other words, the numerical value exceeded 1 if participants felt the experimental band was tight, whereas it was less than 1 if they felt the band was loose.
Results and discussion
Calibration of the measuring system for clothing pressure
Calibrations of the measuring system for clothing pressure employing the three methods described earlier are shown in Figure 6. The reduced scales of the vertical axis are separated to confirm the details of each figure. Figure 6(a) presents the calibration results obtained employing water pressure when the Pouch is immersed in a viscous liquid.
Calibration of the measuring system for clothing pressure.
Figure 6(b) presents the results obtained employing water pressure when the Pouch is inserted between an elastic balloon as a soft material and a plastic spherical shell as a hard material. Figure 6(c) presents the results obtained when Pouch is inserted between a plastic board and a weight. The results are shown in red for an increasing load and blue for a decreasing load. The results for increasing and decreasing loads coincide in Figure 6(c).
Figure 6(a) presents a calibration for the pressure range 0–35 hPa. Figure 6(b) presents a calibration for the range 20–370 hPa. (In the case of low water pressure of approximately 20 hPa, coinciding with the height of the plastic spherical shell, the balloon was not filled with water and it was thus not possible to measure the pressure accurately.) Figure 6(c) presents a calibration for the range 0–120 hPa. The relations between the applied load (X) and the reading of the measuring system (Y) were Y = 1.00X + 0.25 (Figure 6(a)), Y = 1.04X (Figure 6(b)), and Y = 1.00X + 0.23 (Figure 6(c)). Straight lines that had a gradient of almost 1.0 and passed approximately through the origin could be fitted to the data. In addition, no hysteresis was recognized for any line. The calibration results reveal that, irrespective of whether the Pouch was surrounded by soft or rigid material, within the range 0–370 hPa, the linear relation between the applied load (X) and the reading of the system (Y) had a gradient of approximately 1. The system is thus suitable for the measurement of clothing pressure.
Measurements of EBP, respiratory movement, and girth of the abdomen
Simultaneous measurements of the EBP, respiratory movement, and girth of the abdomen when the participants wore the elastic band are shown in Figure 7. The four traces in the figure are, from top to bottom, two traces of clothing pressures on the right sides (i.e. the measuring point RC in Figure 4) of planes 7 and 10, one trace of respiratory movement measured by a thermistor set to a nostril, and one trace of the abdominal girth measured by a thoracic pickup set to the cross-point of the right-side line of plane 6. The clothing pressure changed with the respiratory movement and change in abdominal girth. At the measuring points, the clothing pressure of the abdomen (on planes 3–8) was affected by respiratory movement even when the participant stood still, but it did not change on other planes. The pressure data obtained three times for 10 respirations as an index of respiratory movement were averaged for each plane, and statistically significant differences were calculated in a paired t-test, because the preference for tightness is highly individual.
Simultaneous measurements of the elastic band pressure, respiratory movement, and girth of the abdomen.
Evaluation of the pressure feeling
Supposing that the pressure feeling when the participants had not tightened the experimental band has a score of zero whereas the perfect-fitting feeling has a score of 1, a score of the loose pressure feeling of 0.22 ± 0.11 (mean ± SD) and a score of the tight pressure feeling of 1.67 ± 0.25 were calculated. The score for a perfect-fitting feeling was almost 1 (1.06 ± 0.09). The experiments were thus confirmed to be carried out under conditions estimated to provide an approximately perfect-fitting feeling.
Distribution of EBP over the whole body
Vertical direction
Preferred elastic band pressures of 138 body regions (unit: hPa)
The differences in preference for the vertical clothing pressure at 12 measuring points on planes 3–8 were examined. Figure 8 shows clothing pressure for the vertical direction. The data for each body part on 12 vertical lines were analyzed by one-factor repeated-measures analysis of variance, and after significance was confirmed, were analyzed by multiple comparison employing the Tukey method. The clothing pressures of the anteroposterior midline were significantly lower than those of the body parts between the mammilla line and shoulder line. This result relates to the body surface shape. The radius of curvature from the mammilla line to shoulder line (B-D) is small, whereas there is no curvature at point F and the curvature at point B is concave.
Preferred elastic band pressures in the vertical direction (cross-section).
Horizontal direction
The average of the distribution of EBP over the whole body is shown in Figure 9 whereas the significant difference between planes is given in Table 2. An asterisk shows a significant combination. Light (dark) shading in the table shows that the plane given in the top row had a higher (lower) EBP than the plane given in the left column. The preferred clothing pressure for the neck plane was lower than the pressure for all other planes. The neck contains blood vessels and a vagus nerve near the surface, which increase the sense of tightening and slows or stops blood flow when constricted. The preferred clothing pressure for the waist plane was lower than that for all other planes except the neck. The preferred clothing pressure was lower for the under bust and waist planes than for the chest, lower waist and hip planes. The skin is not equally sensitive all over the body because the density of mechanoreceptors differs for each part.
20
It is possible the surface of the body from the under the bust to the waist has more mechanoreceptors than the chest, lower waist and hips.
Average elastic band pressure in each measuring plane. Clothing pressure of each plane and significant differences in pressure
Furthermore, the skin temperature at the palm surface significantly decreased when subjects were constricted at the waist by a band stimulated twice for two minutes at 4.0 to 7.4 mmHg (5.3 to 9.9 hPa), which was larger than the perfect fitting pressure (3.3 to 4.3 mmHg: 4.4 to 5.7 hPa). This suggests the frequency of the regular, sustained spikes discharged by the sympathetic nerve in conjunction with the blood flow at the surface of the skin decreased. 21 The clothing pressure for abdomen affects the function of the autonomic nervous system. A nerve plexus of the autonomic nervous system is distributed over the body surface from under the bust to the waist where muscle is thinner than that at the chest and hips. These parts were thus more sensitive to tightening.
The preferred clothing pressure was lower for waist and under-bust planes than for leg and arm planes except for the uppermost arm and elbow planes. The waist, armpit, and inside of elbow contain blood vessels and nerves near the surface and thus have lower preferred tightening than the leg. It is therefore suggested that females do not like a tight feeling on body parts that contain more blood vessels and nerves distributed at the surface. In particular, it is necessary to consider pressure acting on the body region from under the bust to the waist because of the difference in the number of mechanoreceptors and the location of a nerve plexus of the autonomic nervous system.
Results for eight blocks (i.e. head, neck, chest, abdomen, upper leg, lower leg, upper arm, and lower arm) are now summarized in Figure 10. The preferred EBP was lower for the neck than for the upper and lower legs and lower arm, and higher for the lower leg than for any block. The surfaces of legs do not contain many mechanoreceptors or a nerve plexus of the autonomic nervous system, thus the legs are highly tolerant to pressure; compression wear covering the leg is viable. Wong
22
reported the discrimination of garment pressure on the body below the waistline for 10 females. The threshold of pressure comfort was 22.0 hPa at the posterior thigh, 21.2 hPa at the posterior knee, and 20.6 hPa at the posterior calf. In contrast, the results of the present study increase centrifugally from plane 9 to plane 15, except for plane 12 (corresponding to the upper knee), which has a rigid bone; values are 15.8 hPa at the posterior thigh, 22.4 hPa at the posterior knee, and 19.7 hPa at the posterior calf (see Figure 10). The disagreement requires further investigation considering differences in the material, width and thickness of the band and the measuring device.
Preferred clothing pressure in eight blocks over the whole body.
Low clothing pressure over the whole body was measured in this study. A single band was used to generate a tightening feeling on the body of each participant, who then evaluated a perfect fit. For planes 15 and 21, blood vessels and nerves are protected with tarsal and carpal bones and the participants’ tolerance for a tightening feeling was high. Because clothing applies pressure to different body areas, the preferred pressure level will be lower than that reported here. We cannot propose a numerical value recommendation for computer-aided design systems just through the data of this paper. However, we can say an insufficient value in the grading of the body surface shape was not enough to decide the comfort of wear. This difference will be explored further in future work.
Conclusion
Clothing pressure as an index of comfort has been well studied. Various approaches, such as indirect methods and the use of a predictive equation, have been studied in research on clothing pressure. A direct method was chosen in the present study to compare the clothing pressure with the feeling of wearing it. Pressure generated between the human body and clothes is affected by the hardness (i.e. rigidness or softness) of the measuring region of the body and clothing materials. A system for measuring clothing pressure employing a renewed hydrostatic pressure-balancing method, which has a thin and pliable soft sensor, was examined using three calibration methods. All methods revealed an almost perfectly linear Y = X relation for the pressure load (X) and the reading of the system (Y). In the application, the distributions of EBP were examined on 21 planes from head to foot. The preferred clothing pressure for the neck plane was lower than that for all other planes. The neck contains blood vessels and a vagus nerve near the surface, which increases the sense of tightening and a reduction in blood flow when constricted. The preferred clothing pressure for the waist plane was lower than that for all other planes except the neck. The preferred clothing pressure was lower for the under-bust and waist planes than for the chest, lower waist, and hip planes. The skin is not equally sensitive all over the body because the density of mechanoreceptors differs for each part. It is possible the body surface from under the bust to the waist contains more mechanoreceptors than the chest, lower waist, and hips. The clothing pressure for the abdomen affects the function of the autonomic nervous system. A nerve plexus of the autonomic nervous system is distributed over the body surface from under the bust to the waist, where the muscle is thinner than that at the chest and hips. These parts were thus sensitive to tightening. The preferred EBP was higher for the lower leg than for any other plane. The surfaces of legs do not contain a nerve plexus of the autonomic nervous system and are unaffected by respiratory movement, thus the legs are highly tolerant to pressure. In the area of the abdomen, the preferred EBP was higher from the mammilla to the shoulder than for the anteroposterior midlines. The development of compression wear must consider appropriate tightening for each body part irrespective of whether that body part is sensitive to tightening. Only one band was used in the present study. Each participant's tightening feeling will increase with the area covered. There should be a greater number of active pressure sensors in future work and the pressure level should be lower than that reported here. This difference and participants’ movement, different posture, and pressure acclimatization will be explored in future work.
Footnotes
Declaration of conflicting interests
The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: A part of this work was supported by the Japanese Society for the Promotion of Science KAKENHI Grant Number 17H01954.
Trial registration
This study was carried out after the review of the Ethical Review Board for Humans of Shinshu University (approval number H27-6).
