Eur J Appl Physiol (1990) 61:237-245 European Journal of Applied Physiology and Occupational Physiology © Springer-Verlag 1990 Physiological strain due to load carrying Michael Holewijn TNO Institute for Perception, Thermal Physiology Research Group, Kampweg 5, NL-3769 DE Soesterberg, The Netherlands Accepted February 19, 1990 Summary. In an experimental study of load carrying the effects of mass (0, 5.4, 10.4 kg) and the type of sup- port (on the shoulder or on waist) on parameters of physiological strain were quantified to determine the factor(s) which limit carrying time. Four categories of strain were investigated: metabolic (in terms of oxygen uptake), cardiovascular (in terms of heart rate), muscu- lar (in terms of EMG activity) and skin pressure under the shoulder straps. Four young male subjects were tested on a treadmill using different combinations of load and speed. While standing, oxygen uptake was not influenced by the type or mass of the backpack, and averaged 10% maximal oxygen uptake. The heart rate increased significantly by 9 beats per min while stand- ing wearing a backpack, independent of type of support or mass of backpack. While walking both the heart rate and the oxygen uptake were significantly influenced by the mass carried, but both types of strain remained be- low the tolerance limits for prolonged wear. Standing supporting a load did not significantly increase the root mean square value of the EMG signal of the trapezius pars descendens muscle. While walking, load carrying significantly increased the root mean square value, and, converted to force, the largest increase amounted to 2.7% of the maximal force for a load of 10.4 kg sus- pended from the shoulders. This was below levels of force producing fatigue, which was also indicated by an absence of changes in the median power frequency of the EMG signal. The pressure on the skin under the shoulder straps during load carrying on the shoulders was more than a factor of three times higher than the threshold value for skin and tissue irritation. Load transfer to the waist with a flexible frame reduced the pressures on the skin of the shoulder to far below the threshold value. On basis of these results it was con- cluded that even with relatively low loads the limiting factor was the pressure on the skin, if a waist belt did not relieve such pressure on the shoulders. Offprint requests to: M. Holewijn Key words: Electromyography - Load Carrying - Back- pack- Exertion - Shoulder Introduction During leisure or military activities, load-carrying with a backpack is frequently practised. In most of the stud- ies concerning backpacking, the main goal has been to determine the energy cost of walking taking into ac- count a variety of terrains (grade and surface), veloci- ties, and external loads (Datta and Ramanathan 1971; Goldman and Iampietro 1962; Legg and Mahanty 1986; Myles and Saunders 1979; Pandolf et al. 1976) or to determine the level of metabolism, expressed as a percentage of maximal oxygen uptake (VO2 max), which could be maintained without physical fatigue (Epstein et al. 1988; Shoenfield et al. 1977; Evans et al. 1980). A few studies have examined the effects of load-carrying on muscle activity (Cook and Neumann 1987; Bobet and Norman 1982), walking kinematics (Bloom and Woodhull-McNeal 1987; Martin and Nelson 1986), or the effects of load distribution on loss of mobility (Ho- lewijn and Lotens 1987). In this paper the effects of the mode of carrying and the load mass were investigated by simultaneous meas- urement of several physiological strain parameters. Firstly, from a study of the literature different types of strain were identified which could limit the endurance time of walking with a back pack (Holewijn 1986). It was found that in addition to a reduction in physical performance, effects on the metabolic, musculo-skele- tal, and cardiovascular systems, and the skin under- neath the shoulder straps are important. The aim of this study was to quantify all the resulting strains to assess the limiting factor in the endurance time of walking with a normally loaded backpack. It was hypothesized that local strain of the shoulder muscles or pressure on the skin under the shoulder straps could be the cause of frequent complaints during backpacking. Therefore, the load on the shoulder muscles was estimated with
`
`Petitioner Ex. 1061 Page 1
`
`

`
`238 electromyographic techniques (EMG) and the pressure on the skin of the shoulder region was measured at sev- eral locations under the right shoulder strap to assess the distribution of pressure. The oxygen uptake (1202) and the heart rate were also monitored to exclude the possibility that these strains were above the limits of to- lerance. Methods Subjects. Four healthy young male students participated in this study. They all participated regularly in physical activities but were not used to carrying backpacks. The subjects were informed of the purpose and procedures of the study and consented to par- ticipate. The subjects had a mean ag.e of 24 years (range 23-26), mass of 75.1 kg (range 69-81.5) and VO2max of 3.4 l.min -I (range 3.3-3.8). Types of backpack. Two packs were used in this study. One was the backpack in use in the Royal Netherlands Army (Mil), as pack mounted high on the back by straps. The straps ran from the front of the waist belt to the back being attached to the pack on the same side and crossing between the shoulder blades to reach the waist belt on the opposite side. On the shoulders the straps had a width of 5 cm and were of heavy canvas (Fig. la). This type of backpack puts the load mainly on the shoulders. The second backpack was a custom-made pack (Cust) where most of the load was supported on the hips by means of a flexible frame connected to a padded 10-cm wide waist belt (Fig. !b). The padded shoulder straps were 8 cm wide on the shoulders. Load. Loads of 5.4 kg and 10.4 kg were chosen, representing the fighting and marching order of a Royal Netherlands Army sol- dier, respectively. Measurements without a pack served as control. These loads were applied both while standing and walking on a treadmill at a moderate walking speed of 1.33 m. s-1. Physiological measurements and apparatus. The EMG activity of the descending part of the right trapezius muscle was measured with two surface silver-silver chloride electrodes (PPG, Hellige, FRG), positioned on the distal third of the muscle with an inter- electrode distance of 2 cm parallel to the muscle fibres. The elec- trodes were attached after thoroughly cleaning the skin with alco- hol. The reference electrode was attached on the acromion. The EMG recordings started 1 h after application of the electrodes be- cause by that time the skin impedance had almost stabilised (Zipp et al. 1977). b i Fig. 1. The military a and the custom built backpack b used in the experiment The EMG signals were first passed through a small battery fed pre-amplifier (100x), mounted on a waist belt, and then through an amplifier with a gain of 2 × -50 × and a bandfilter of 5-1,000 Hz (slopes: low pass filter 6 dB. octave -1, high pass filter 12 dB. octave-I). The EMG was then sampled over 1-min periods by an microcomputer (IBM, USA) using a 12 bit A/D board (DT2821, Data Translation, USA) set at a sample frequency of 2,048 Hz, and stored on a hard disk. The root mean square value (rms) of the amplitude was determined on line with a custom built rms detector (AD 637, time constant = 55 ms) and sampled with the same equipment. Post experimental analysis of the EMG consisted of dc-cor- rection and a fast Fourier transform (FFT) with a data analysis software package (Asystant, Macmillan Software Company, USA). From every EMG recording four samples of 1-s duration, equally distributed over the 1-min sample period, were taken for a 2,048 point FFT analysis. The resulting power spectra were aver- aged and from this averaged power spectrum the median power frequency (MPF), i.e. the frequency above and below which the integrated power is equal, was calculated. The rms data of the same four samples were transformed to force values using a pre- viously determined rms versus force relationship. This calibration curve between rms of the EMG of the trapezius muscle and the force produced by this muscle was determined for each subject with two adjustable slings running over the shoulders, one of which was connected to a floor mounted force transducer (Z 2H6, Hottinger Baldwin Messtechnik, FRG). The shoulder was posi- tioned directly above the force transducer. The subject performed three isometric maximal voluntary contractions (MVC), with a 10- min rest period between each contraction, by lifting the shoulders, while sitting with a straight back and with the feet not touching the floor. This posture was chosen to ensure that the force could only be produced by lifting the shoulders and not by other means (leg muscles, leaning forward). The highest force level maintained for 3 s was taken as the MVC. After 30-min rest the rms value was measured for 1 s at force levels of 5%, 10%, 20%, 30%, 50%, and 100% MVC. Between each measurement there was a 10-min rest. By power regression a curve was fitted to the data. The pressure under the shoulder strap on the skin of the right shoulder was measured with a miniature pressure transducer (model 156, Precision Measurement Company, USA) measuring 8.5 x 4 mm and 1 mm thick. The small dimensions made it possible to measure the pressure with a minimal change to the curvature of the shoulder strap thereby introducing a negligible artefact in the recordings. The pressure signal was amplified (MG 3150, Hottinger Bald- win Messtechnik, FRG) and sampled by an IBM microcomputer with a sample frequency of 2,048 Hz and stored on disk. While the subjects were standing the pressure on the skin was measured at five positions under the right shoulder strap and for each position at three locations, i.e. the lateral and the medial edge of the strap and in the middle. The five positions were spaced out equally over the shoulder strap at intervals of 5 cm, position 3 being just on top of the shoulder. During walking the skin pressure was measured only at position 3 on the medial edge of the shoulder strap. The 1202 (1.min-1) was measured with an Oxylog portable system (Morgan Ltd, England) which was mounted on a fixed frame above a treadmill. The VO2 (1-min-1) was normalised with respect to each subject's 1202 .... and with respect to the total load (mass of the subject + load). The 1202max was estimated dur- ing a submaximal treadmill running test, by increasing the run- ning speed at 3% gradient until a heart rate of 160 beats.min -1 was reached. The VO2max was calculated by extrapolating the subject's heart rate versus 1202 relationship to his maximal esti- mated heart rate (,~strand and Rodahl 1986). This method had the advantage that the subjects were not stressed to their limits, but the accuracy was 10%-15% less than a direct measurement of 1202m,~ (Davies 1968). The heart rate was monitored continuously by a custom built cardiotachometer with a charcoal electrode set (Respironics, USA) mounted on the chest.
`
`Petitioner Ex. 1061 Page 2
`
`

`
`239 Experimental procedure. Prior to the load-carrying sessions, each subject's VOem~ was estimated, followed by the measurement of the force versus rms calibration curve. After a rest period of 30 rain the load carrying sessions started. Each carrying session consisted consecutively of 20-min standing, 10-rain rest and 20-min walking on the treadmill at a velocity of 1.33 m-s -~. While standing and walking 1202 and heart rate were re- corded continuously on a chart recorder. Both while standing and walking the pressure, the EMG and the rms were measured at the 1st, 10th, and 20th min. Off-line, the average EMG and rms values were calculated for each l-rain measurement period. This cycle was repeated five times, with a 20-rain rest between each cycle. The five carrying conditions (no backpack, Mil and Cust back- packs, each with a 5.4- and 10.4-kg load) were administered ac- cording to a balanced design. Statistics. The data were assessed by analysis of variance (ANOVA) with the Systat computer programme (Systat Inc, USA) after checking normality of the data (Kolmogorov-Smirnov test) and the homogeneity of variance. If significant F values were found (P<0.05) the differences between levels within an effect were analysed for significance by a Newman Keuls post hoc test (e<0.05). Results 120 110 'c- 100 d~ P 9O 80 70 --]standing ~wotking - ° • ,:! N ~.:, i:.:.~ .~ i:~:-a ' il 11 j_ i " no 5./,kg 5./,kg lO.4kg lO./,kg Load mit cust mit cust Fig. 3. The effect of different combinations of load and backpack type on the heart rate (mean and SD) while standing and walking (speed = 1.33 m. s-~). Mil, Military backpack; Cust, custom built backpack Oxygen uptake There was no significant difference in metabolism be- tween the 1st, 10th, and 20th min in any of the carrying sessions. Therefore, the data were averaged over the three measuring points. In Fig. 2 the effect of the type of backpack on the relative metabolic strain (% 1?O2 m~x) is shown. While standing the relative metabolic strain was not significantly influenced by the load and aver- aged 10% VO2 .... However, while walking differences between loads were evident. Walking with the 5.4-kg load caused a significant increase of 1.5% 1702 .... and with the 10.4-kg load the increase was 4.8% VO2 .... resulting in absolute 1702 of 0.96 1. min- ~ and 1.08 1. min- 1 respectively. No signifi- cant difference was found between the two types of backpack. Comparing 1?O2 for the two loads, the energy cost necessary for displacement of body mass and load sep- arately can be calculated. The average energy cost dur- ing walking without a load amounted 4.2 W.kg -1 of body mass. However, the average energy cost per kg load at first decreased (1.1 W-kg -1 for the first 5.4 kg) but then increased (6.3 W-kg-~ for the next 5 kg) with increasing loads. The average energy cost per kg load for the first 5 kg was thus lower than for a kg of body mass, but increased steeply with increasing loads. x 30 12 E 0 > ~ 2o 40 "i-7 standing ~] walking i '//~ ///.~ 1/// /I// ,11, i;"~ ¢//~ no load ;55; ,,,/., ;5;; ;f.;; j .... 5.4kg 5.4kg Mil Cust 777~/ 7777 ,//, e/// .... t.,/e .... ¢/.,J ,/// ///, ,//, ,//, ,//, ¢/.,/ ~//, ,/1, ///, ,//~ ¢/// / / / / ,-//t r//, r//, I//) .... /// .... ¢/1. lO.Z, kg lO.Z. kg Mil Cust Fig. 2. The effect of different combinations of load and backpack type on the relative metabolic strain [%maximal 02 uptake (% 1202 .... ) while standing and walking (speed= 1.33 m-s-l). Mil, Military backpack; Cust, custom built backpack Heart rate While standing the average heart rate increased signifi- cantly by 9 beats, min-~ with load carrying. There was no significant difference between the four load carrying conditions (Fig. 3). While walking the heart rates during control meas- urements remained significantly lower than the heart rates during the load-carrying conditions (Fig. 3). The 5.4-kg load caused a significant increase of 8 beats.rain -1 and the 10.4-kg load added a further sig- nificant increase of 6 beats.min-1. The type of back- pack had no significant effect on the heart rate. Electromyographic activity of the trapezius muscle Amplitude. The relationship between rms and the lifting force of the descending part of the trapezius muscle, measured during test contractions prior to the load- carrying sessions, was fitted with a second order curvil-
`
`Petitioner Ex. 1061 Page 3
`
`

`
`240 I D standing 30 [TAwatking z20 t 0 no 5.~ki" 54kg 10.4~ lO.~.kg load mit cust mit cust Fig. 4. The force level (mean and SD) of the descending part of the trapezius muscle while standing and walking without a load and using two types of backpack carrying 5.4- and 10.4-kg loads. Mil, Military backpack; Cust, custom built backpack inear line, having an average correlation coefficient of 0.99. The ANOVA revealed that there was no signifi- cant change in rms over time. In further analyses the three measurements were averaged. Converting rms of the EMG of the trapezius muscle to force resulted in the force levels shown in Fig. 4. Although the force level while standing increased when a load was carried, the effect was not significant. The force averaged 5.4 N. While walking the force lev- els increased significantly during load carrying, but not in the control situation. The Cust with a 5.4- and 10.4-kg load and Mil with the 5.4-kg load resulted in similar force levels of 15 N (l.6%MVC), 17 N (1.7%MVC), and 19 N (1.9%MVC), respectively. The Mil, however, containing a 10.4-kg load resulted in a force level of 27 N (2.7% MVC), which was significantly higher than in the other three load conditions. The force level was significantly dependent on the subject, in particular for the heavy load. This explains in part the variation in force level. With Mil the increase in force, comparing standing and walking, was significantly higher than with Cust. Mean power frequency The MPF did not decrease significantly with time while standing or walking with a backpack. The average MPF /.o 30 ~ 20 & military backpack 5.4 kg & II custom build backpack 5.l, kg )[ ~ position mid. lat.1 location /.o custom build backpack 10+4 kg ~ 30 o, 20 10 0med. mid. Lot.1 Tmed. mid. Lot3 location tocation Fig. 5. The distribution of average skin pressure of the right shoulder measured at five positions under the shoulder strap while standing carrying two different loads and using two types of backpack. For each position the pressure was measured in the middle of the strap (mid) and at the medial and lateral edge (med. and lat.)
`
`Petitioner Ex. 1061 Page 4
`
`

`
`while standing (35 Hz) was not significantly different from the MPF while walking (34 Hz). Neither the type of backpack nor the mass of the load significantly in- fluenced the MPF while standing and walking. Pressure of the shoulder straps on the skin While standing the pressure distribution on the skin un- der the right shoulder strap measured in each of the 15 sites is graphically represented in Fig. 5, showing the differences between the two backpacks and the effect of increasing the load. The two backpacks showed pressure increasing from the back at the lower edge of the scapula to the top of the shoulder and a sharp decrease on the front side of the shoulder. Carrying the loads using Mil caused a peak pressure on the acromion (location = 3, position = lateral) and another on the upper edge of trapezius muscle (location = 3, position = medial). The former peak was also found using Cust, but smaller in amplitude. The peak pressure on the medial side was not present with Cust. The peak skin pressures using 241 Mil were significantly higher than the skin pressures us- ing Cust. The maximal pressures amounted to 20 kPa (150 mm Hg) (Mil, 5.4 kg), 27 kPa (203 mm Hg) (Mil, 10.4 kg), 2 kPa (15 mm Hg) (Cust, 5.4 kg and 10.4 kg). At most positions the pressure on the edges of the shoulder strap was higher than in the middle of the strap. The statistical analysis showed further that increas- ing the load from 5.4 to 10.4 kg in Mil caused a signifi- cant increase in the skin pressure of 36%, whereas no significant effect was found with Cust. The form of the pressure distribution did not appear to be significantly influenced by the load level. While walking the pressure showed sinusoidal fluc- tuations about 0.2 s out of phase with rms of the EMG of the trapezius muscle (Fig. 6). Similar to measure- ments made while standing, the skin pressure while walking was significantly dependent on mass and the type of backpack (Fig. 7). The post hoc test showed that Cust had a significantly lower skin pressure on the top of the shoulder than Mil. Further, the skin pressure us- ing Mil increased significantly with increase in the load from 5.4 kg to 10.4 kg. 25, i i i ~ i ~ I "pressure & 15 i i i i i i 0.2 0.6 1.0 1./* 18 time(s) 0.25 0.15 tn 005 Fig. 6. A typical example of the sinusoidal variations of the skin pressure and root mean square (RMS) of the electromyogram of the trapezius pars descendens muscle while walking (speed= 1.33 m.s -I) usingthe military backpack with a load of 5.4 kg 30 1 .[ r w-"-"'- 20 m o r'l r 1 .2g m 5./. kg 10.4kg 5.4kg 10./.kg cust. mit. Fig. 7. The average skin pressure on the top of the shoulder while walking using two types of backpack and carrying two different loads. Mil, Military backpack; Cust, custom built backpack Discussion Metabolic and cardiovascular strain In this study, in contrast to other studies, it was found that carrying a backpack had a significant effect on the heart rate while standing (Borghols et al. 1978; Pierry- nowski et al. 1981; Pimental and Pandolf 1979). A pos- sible explanation may be that in this study the time taken standing was more than a factor of two longer than in other studies and in combination with a differ- ent type of backpack this may have resulted in signifi- cant effects on the cardiovascular system. This increase in heart rate has been commonly observed during static muscular exercise. Kilbom (1976) has concluded in his review that the resulting increase in cardiac output dur- ing static contractions is mainly directed towards the peripheral parts of the body and only a small part is supplied to the myocardium. In this study standing while carrying a backpack, however, required no signif- icant extra metabolic energy which is in agreement with other studies (Borghols et al. 1978; Pierrynowski et al. 1981). Thus, the. relationship normally found between heart rate and VOz during dynamic exercise was dis- rupted during static contractions. Pandolf et al. (1977) and Pimental and Pandolf (1979) formulated an empirical energy prediction equa- tion relating metabolic weight, body mass and load mass, walking velocity, grade and terrain: MR = 1.5m+ 2(m+ L)(L.m-1)2 + n(m+ L)(1.5vZ +O.35v. G ) (1) where MR is metabolic rate (W), m is subject mass (kg), L is external load (kg), n is terrain coefficient, v is walk- ing velocity (m-s-l), and G is grade (%).
`
`Petitioner Ex. 1061 Page 5
`
`

`
`242 10 8 2 l i = i i l ~ 2:dW.dkg -1- l:W" kg -1 °o lb 2b 3'0 5b & 70 load (kg] Table 1. The predicted external load (kg) at which the metabolic limit is reached at a walking speed of 1.33 m.s -1 for different age-groups and sexes. It should be noted that this load is the sum of the mass of the backpack, and the mass of the clothing/equip- ment and the footwear Age External mass (kg) (year) Man Woman 20 32 19 40 17 4 60 5 -- For a woman aged 60 years the metabolic tolerance limit is al- ready exceeded without an external weight Fig. 8. The predicted relationship between load and metabolic cost for a subject with a body mass of 75 kg and walking at a speed of 1.33 m.s -J. Line 1, Metabolic cost per kg total mass (W-kg-1); Line 2, extra metabolic cost per added kg load (6M.fkg -~) Predictions using Eq. (1) have proved to be in agreement with experimental results for walking at a normal speed and with loads up to 50 kg (Myles and Saunders 1979; Pierrynowski et al. 1981; Pimental and Pandolf 1979; Pimental et al. 1982). In most studies it has been assumed that the metabolic cost per kg load is not dependent on the total mass for loads carried cen- trally on the body (Goldman and Iampietro 1962; Datta et al. 1973; Myles and Saunders 1979; Pierry- nowski et al. 1981 ; Pimental and Pandolf 1979; Soule et al. 1978). Partial differentiation of Eq. (1) with respect to the external load yields Eq. (2): 6MR - 2 .(2m .L -1 + 3). (L.m-1)2+ 6L n.(1.5vz +o.35v. G ) (2) Equation (2) shows that the metabolic cost per kg load (W. kg-1) increases with increasing loads (Fig. 8, line 2). It can be seen in Fig. 8 that the metabolic cost aver- aged over the total mass (body mass + load) has a pa- rabolic form with a minimum at a load of around 16-kg. However, the differences in the values for the different loads are so small that it, may be expected that these would become submerged in experimental noise. A more sensitive parameter for investigation of the meta- bolic rate per kg added load is given in line 2 in Fig. 8, which shows a sharp increase with increasing load. This means, for example, that the extra metabolic cost for the first kg of added mass is less than for the 10th kg of added mass. Qualitatively, the predicted increase is in agreement with the data from this study, but, quantita- tively, there is a discrepancy. A reasonable explanation for this is as yet lacking. Several studies have shown that a VO2 around 40% lkO2max and a heart rate around 110 beats.min -1 can be maintained for periods of less than 2 h (Evans et al. 1980, 1983; Grandjean 1967; Michael et al. 1961; Nag et al. 1980; Nag and Sen 1979; Rutenfranz 1985). The measured metabolic and cardiovascular strains in the present study were below these levels so it may be con- cluded that for a young male population loads up to 10.4 kg would not limit the endurance time of walking. However, for a different age-group or sex these toler- ance limits may be exceeded (Jorgensen 1985).. Com- b!ning the data of the effect of age and sex on VO2max (Astrand and Rodahl 1986) with Eq. (1) one can predict the external mass at which the metabolic tolerance limit is reached (Table 1). It should be noted that these predictions are based on body masses of 60 and 70 kg for an average woman and man and a surface consisting of a flat hard top road. Other conditions will result in different predicted external loads. In particular, different terrains will in- crease the metabolic strain (Pandolf et al. 1976). Muscular strain Since Rohmert (1966) published the endurance curves for static contractions of the arm muscles, it has long been assumed that contractions below 15% MVC could be maintained indefinitely. The results of recent studies have shown that static contractions should be around 5% MVC in order to avoid the effects of fatigue after 1 h (BjOrkstrn and Jonsson 1977; Hagberg 1981; Jons- son 1978; Sjogaard et al. 1986). In this study standing with the two loads resulted in average force levels for the descending part of the trapezius muscle well below I%MVC. No differences existed between the two types of carrying system, although using Mil most of the load was supported on the shoulders, in contrast to Cust. A possible explanation may be that while standing with a backpack the shoulder girdle is, for the major part, rest- ing on the ribcage, needing only small muscle forces to stabilize the shoulder girdle and, therefore, no differ- ences between the two backpacks should be found. However, walking with a load significantly increases the force level generated by the descending part of tra- pezius muscle, without surpassing the 5% MVC limit. It can be concluded that the force level generated by the descending part of the trapezius muscle during the carrying condition was below the static force level which can be maintained for a long period.
`
`Petitioner Ex. 1061 Page 6
`
`

`
`243 The cyclic variations in force while walking (Fig. 6) can be explained from the kinematics of walking. Ac- cording to Stokes et al. (1989) the pelvis rotates in the frontal plane opposite to the shoulder girdle during most of the stride cycle. The shoulder on the opposite side to the leg striking the ground is lifted and rotated forward. When this movement of the shoulder is im- peded due to a load supported by the shoulder, the tra- pezius muscle has to generate a higher force in order to overcome this. From the data on the force generated by the descending part of the trapezius muscle during load carrying using Mil, it can be seen that with a load of 5.4 and 10.4 kg the extra absolute force level (above walk- ing with no load) was 8.4 and 17 N per shoulder, re- spectively. The extra force generated by the descending part of the trapezius muscle doubled with a doubling of the load carried in Mil. The force level during load carrying using Cust was not significantly influenced by the load level, indicating that most of the load had been transferred to the hips by the flexible frame, leaving on the shoulder only a constant load that was needed for stabilization. Muscle fatigue is generally accompanied by a shift of the MPF to the lower frequencies (e.g. Anderson et al. 1977; Hagberg 1981; Kadefors et al. 1968; Moritani et al. 1986). Such a decline in time of the MPF was not found in this study which is in agreement with the low force levels generated by the descending part of the tra- pezius muscle. Skin pressure According to Husain (1953), a skin pressure of more than 14 kPa (105 mm Hg) results in irritation, redness and, for an exposure time of 2 h, in subcutaneous oed- ema and inflammation of the dermis and underlying tissue. Besides the amount of pressure, the duration of the pressure is also an important factor in the develop- ment of the symptoms (Kosiak 1958; Stobbe 1975). Re- cently Sangeorzan et al. (1989) reported that for skin over muscle, a pressure of less than 10 kPa (75 mm Hg) reduces the transcutaneous partial pressure of oxygen, used as an index for local circulation, to 0. Therefore, it is assumed that in order to avoid these effects, the pres- sure applied on the skin under the shoulder straps should be below 10 kPa (75 mm Hg). The Mil exceeded this limit on the top of the shoulder for both loads. The Cust caused skin pressures far below this limit. High pressures can probably cause arm muscle weakness due to temporary failure of the superficial nerves in the plexus brachialis as found by Funaski (1978) and Rothner et al. (1975). As the skin pressure was the only strain clearly exceeding the tolerance lim- it, it is concluded that the frequent complaints during walking using Mil, with relatively light loads, were caused by pressure on the skin under the shoulder straps. From the pressure measurements it can be seen that in Cust the frame transferred a considerable part of the mass to the hips. This is not just displacing the Table 2. The expected vertical and calculated vertical forces (N) working on one shoulder while standing Load Backpack Force (kg) (N) calculated expected 5.4 Mil 75 54 Cust 11 ? 10.4 Mil 105 104 Cust 12 ? ?, Not known; Mil, military backpack; Cust, custom built back- pack problem to the waist, since the contact area between the waist belt and the hips can be quite large, thereby re- ducing the pressure. The disadvantage is that to prevent the waist belt slipping off the hips it must be pulled tight, thereby increasing the pressure on the skin. Data of Scribano et al. (1970) has shown, however, that the hips are less sensitive to pressure by a factor of three, so it may be concluded that load bearing by the hips is preferable to load carrying on the shoulders. Assuming that the average skin pressure over the three locations was representative for a section of the strap (= position), pressure data can be integrated over the contact surface with the shoulder to obtain the total force working vertically on the shoulder, after correc- tion for the angle (Table 2). The calculation of the ver- tical force and the expected vertical force is shown in detail in the Appendix. The calculated vertical force for Mil was in reason- able agreement with the expected force value. For Cust no expected force value has been given as no informa- tion was available on the amount of load transferred to the hip by its frame. The design of Mil is particularly unfavourable because of the front straps running to the front belt. With straps running over the shoulders and under the arms to the back again, the load on the shoulders would be halved (Appendix, Fig. 10b). This would also solve the problem of the hip belt being pulled under the rib cage causing complaints of im- peded ventilation. Appendix Calculation of the vertical force The pressure data can be converted to a vertical force by integrat- ing the pressure over the area under the shoulder strap. It is assumed that the average pressure at each position is rep- resentative for that particular section of the strap. As the five po- sitions were equally spaced 5 cm apart, the surface area per posi- tion is 5 cm times the width of the shoulder strap, amounting for Mil to 25 cm 2 and for Cust to 40 cm 2. It is also assumed that the shoulder can be represented by a cylinder with a radius of 5 cm, and that only the skin pressures of position 2, 3, and 4 contribute significantly to the vertical force on the shoulder (Fig. 9).
`
`Petitioner Ex. 1061 Page 7
`
`

`
`244 $3 J Fig. 9. A simplified cross-section of the shoulder represented as a circle. Positions 2, 3 and 4 are indicated (pr