ERGONOMICS,
199 1, VOL. 34, NO. 1, 57-66
The effect of arm support on supraspinatus muscle load during simulated assembly work and welding ULFJ ~ V H O L M *GUNNAR , PALMERUD~,
ROLAND KADEFORS~~ and F%~ER HERBERTS~ Department of Orthopaedics, University of Gijteborg, East Hospital, S-416 85 GGteborg, Sweden t Project Lindholmen Industrial Development Centre, PO Box 87 14, S-402 75 Gbteborg, Sweden $ Department of Orthopaedics, University of Gbteborg, Sahlgren Hospital, S-413 45 Giiteborg, Sweden 5 Department of Applied Electronics, Chalmers University of Technology, S-412 96 Geteborg, Sweden Keywords: Arm suspension; Supraspinatus; Intramuscular pressure;
Electromyography. The effect of arm support, by . a suspension device, on muscle load in the supraspinatus muscle was evaluated with simultaneous intramuscular pressure measurement and electromyography (EMG)in nine healthy subjects. Two work situations, a low load assembly type of work, and welding with a higher shoulder muscle load, were simulated in the laboratory. Each subject performed three workcycles of each type, with and without arm support. Arm suspension reduced supraspinatus muscle load in both work situations with reduction in pressure of 34% and 22% respectively, and reduction in normalized EMG of 20% and 17% respectively. The reduction of muscle load was significant, but in the welding situation with arm-suspension 10-1 5 N, average muscle pressure was still higb enough to reduce muscle blood flow.The interpretation of the importance of this toad reduction for the development of work-related shoulder pain is problematic.
1. Introduction Shoulder pain related to work situations with prolonged elevated arm positions is an increasing problem for industrialized society (e-g., Loupajarvi et al. 1979, Hagberg 1984, Aaris et al. 1988).High muscle load and static work situations are believed to be two of the main causative factors (Maeda 1977, Hagberg and Wegman 1987). Two clinical conditions are manifested by shoulder pain in industry. The first is shoulder tendinitis, which has been shown to be related to high muscle load, elevated arm positions, and handling of heavy objects, usually in male industria1 workers. The second is shoulder myalgia, which is more related to monotonous low load work situations like assembly-type work, and more common among women (Hagberg 1984, Herbexts et al. 1984). In evaluation of shoulder muscle load, electromyography (EMG) is most often used. EMG can be applied in evaluation of the relative muscle load in different shoulder muscles (Sigholm et al. 1984).The trapezius muscle is the one most frequently studied, both regarding shoulder muscle load and localized shoulder muscle fatigue (Hagberg 1981 a, Granstr6m et al. 1985, Westgaard 1988). Besides EMG, intramuscular pressure (IMP) can be used in evaluation of local muscle load ( K h e r et al. 1984 b, Parker et a!. 1984, Jhrvholrn et al. 1989). High IMP values have been found in the supraspinatus muscle in elevated arm positions even without hand load (Jiirvholm et a/. 1988 a). IMP 0014-0139/9I S34M 0 1991 Taylor & Francis Ltd.
58
U.Jirvholm et al.
levels exceeding 40 mmHg (5.3 kPa) during contraction reduced the muscle blood flow in the supraspinatus muscle (Jarvholm et a). 1988 b). The ability of arm support to reduce shoulder muscle load has been demonstrated for the trapezius muscle in EMG studies (Erdelyi et al. 1988, Schiildt et al. 1987, Granstrdm et a1. 1985). However, there is still a possibility that arm suspension causes a higher load in other shoulder muscles (Granstrtim et al. 1985), or that the effect of a m suspension in the long run may be unphysiological for human shoulder muscle performance (W inkel 1987). The aim of this study was to evaluate supraspinatus muscle load (a) during light assembly work, and (b) during welding, both with and without arm support. EMG and intramuscular pressure were recorded simultaneously from the supraspinat us, chosen because of its clinical importance and EMG findings of high muscle load and localized muscle fatigue during prolonged arm elevation (Herberts et al. 1976, Kadefors et al. 1976, Hagberg 1981b, Herberts et al. 1984).
2 Materials and methods Nine healthy subjects volunteered for the study. All were male with a mean age of 27 years (range 2 1-50). All subjects were right-handed, and all measurements were done in the right supraspinatus muscle. One subject (the oldest) was a worker in the industry from which the light assembly work was simulated. All others were students who volunteered for the experiment. The experiments were carried out in a laboratory, where working places simulating the two working situations were set-up. Before insertion of the pressure recording catheter, all subjects were familiarized with the working situation, with and without arm support for which a suspension device (K-block@,Mabs Int AB, Norrkoping, Sweden) was used. The suspension force was 1 0 - 15 N, depending on individual preference and comfort. The pressure recording catheter in supraspinatus and the suspension device is shown in figure I. The recordings were carried out from the central part of the right supraspinatus muscle. After local anaesthesia of the skin, an intravenous infusion cannula (VasculonQD,Viggo, Helsingborg, Sweden) was inserted into the muscle. Through the Vasculon@cannula, the EMG electrodes (Stabilohma 1 10, diameter 70 pm,polyurethane isolated; Johnson Matt hey Metals, London) and an intramuscular pressure recording catheter (MyopressQP, Atos Medical, Herby, Sweden) were inserted. This technique of sirnu1taneous intramuscular EMG and IMP recording is described elsewhere (Jarvholm et al. 1989). Intramuscular pressure was measured with the microcapillary infusion technique (MCI) (Styf and Kdrner 1986).With this technique, the pressure is recorded through a saline-filled 1-05mm thick teflon catheter, with four small holes close to its open tip. The infusion is regulated by a microcapillary connected to the teflon catheter and an pressurized' bag of saline. This is a non-constant infusion technique, where the infusion speed of saline is regulated by the pressure difference over a microcapillary. The infusion rate was 1-5ml/h at rest, and thus lower when IMP increased due to muscle con traction. The pressure transducer (Bentley TrantecQ, ~rvine, USA) with a displacement of 3-0x 10- mrn3/kpa, is connected to the system and the signal displayed on a chart recorder and recorded on magnetic tape for later computer analysis. In each test situation the height difference of the pressure transducer relative to the catheter tip was corrected for. The intramuscular EMG electrodes were deinsulated at the tip for approximately 2 mm. The EMG signals were amplified 1000 times, and filtered in such a way that the.
Arm support and supraspinatus muscle load
59
analysed signal was in the frequency interval of 35-800 H z The signals were recorded on magnetic tape and later computer analysed. Here the root mean square of the signal was calculated. The analogldigital converter had a sampling frequency of 2048 Hz,and the s i pla1 was d,ivided into 0.5 s segments from which the mean values and standard deviatic3ns were calculated. Each segment was submitted to an autoimatic qu~ality - any influence of artifacts in the signal (Arvidsson 19132). The IEJMG control, LU -.Irule out _.__>__~ was normalized for each individual to the mean value of the EMG for a sranaardized test situation consisting of 12 position-load combinations for each subject. The test positions chosen were flexion (F)30 and 60" with a flexed elbow, and abduction (A) 30 and 60"with a straight elbow. In each position the hand was loaded with zero, one or two kg weights. These positions were chosen in order to be able to compare the intramuscular pressure values anived at to other IMP recordings in the supraspinatus muscle, reported previously QQrvholm et al. 1988 a). The test positions were done before each experiment. The mean power firequency of the EMG was automaticaIly calculated by a custom-made spectral rno~ ment ana~lysator(Broman and Kadefors 1979) on line, and evaluated during test coniracaons. r -
-*---A:--
3. Experimental procedure
3.1. Light assembly work This was found in a light industrial setting, where small rings for pens are nickel-plated. The work is a low load, but fairly static work situation, where onIy the amis are used actively. The workers had to hang each ring separately on a small hook in the course of the process. The job was performed in the sitting posture. The position of the a m was a slightly flexed shoulder, and a flexed elbow. The arm was moved from a slightly abducted to a slightly adducted position for each row of hooks. The frame of the small hooks was adjustable in height in order to obtain a more comfortable working position. The working situation is shown in figure 1. There were nine rows of hooks, each row containing 35, a total of 315 hooks. In the experiment each subject completed three frames with and three without arm suspension. Before the work and after each frame,
Figure 1. The assembly work situation, which was simulated in the laboratory. Note the arm suspension device and the I M P catheter in the right supraspinatus muscle.
Figurc 2. The simulated welding work with arm suspension.
the subject performed a test contraction by holding the arm in an abducted position, aiming with the hand at the edge of the frame. This test contraction was included in order to evaluate any mean power frequency changes of the EMG signal induced by the work. signifying localized muscle fatigue. The results were calculated and presented as the mean IMP and mean normalized EMG. for each individual during the whole work cycle with and without arm support. 3.2. Welding situation This was chosen to illustrate a work situation with a higher shoulder load. In this study a welding simulator (Kadefors et 01.1984) was used, making it possibIe to study weEding in a standardized fashion. The shoulder joint was in the situation chosen flexed approximately 60°,the elbow was flexed, and the weight of the 'welding' tool 1-4kg. In this work situation the hand was thus held at shoulder level. The simulated welding is shown in figure 2. Each subject simulated welding three electrodes. The welding was done with and without arm suspension. Each work-cycle was 1.2-1.5 min long. The individual mean valucs of intramuscular pressure and EMG during the welding were calculated for the two test situations. For each subject an amplitude probability distribution function (APDF) diagram for both EMG and IMP values during the two simulated work situations with and without arm support was plotted. The APDF-diagram has been described by Jonsson (1982) for analysis of EMG during dynamic muscle work. In summary the APDF diagram is a cumulative diagram showing the probabilities of the amplitudes of a musclc signal (EMG or IMP) during a work cycle. The diagram can be interpreted so that the steeper the curve-the more static is the muscle Ioad, and a shift of thc curve to the left by a changed work situation-indicates a reduction of the muscIe load. After the simulated work, the IMPduringa maximal test contraction was recorded, in a position of approximately 45" of abduction. 3.3. Statistics The Student's paired t-test was used to compare the IMP values with and without hand Ioad.
Arm support and supraspinatus muscle load
0 1 2
0 1 2
0 1 2
0 1 2
haadlaad (kg)
€PIG i n supraspinatus
handload (kg)
Figure 3. (a)Intramuscularpressure in the supraspinatusat shoulder flexion (F)30 and 60"with a flexed elbow, and shoulder abduction (A) 30 and 60" with a straight elbow. In each position the hand was loaded with 0, 1, or 2 kg. The two separately-shaded columns correspond approximately to the work situation in the assembly type of work (flexion 30" without hand load) and welding work (flexion 60a, 1 kg hand load). The bars indicate one standard deviation (SD).(b) Normalized R M S values of the EMG found in the same test positions shown in figure 3(a).
4. Results The I M P and EMG results in the isometric test positions are shown in figures 3 (a)and (b). As found in previous studies, both IMP and EMG were dependent on arm position and hand load. Both IMP and EMG gave a similar pattern of muscle activation. The correlation coefficient of mean I M P versus mean EMG for the twelve test situations was 0-96.All values of IMP are within the same range as found in our previous studies. The mean values (and SD) of IMP and EMG for the two situations with and without arm support is shown in figures 4 (a)and (b). During the assembly work, mean I M P was 34-9 (SD 12.1)mmHg without, and 22-8 (SD 11-0)mmHg with arm suspension. The time for a subject to mount one frame varied from 2.5-7.5 (mean 52)min. There was no difference in working pace, with or without arm suspension. During 'welding', mean IMP was 65.5 (SD19-5)mmHgwithout, and 51.4 (SD16-8)mmHgwith arm suspension. The average reductions of IMP values for assembly work and welding were thus 34% and 22% respectively. There was a
U.JIirvholm et al.
assmbh) vwk
Figure 4 (a,b). Mean intramuscular pressure (a) and normalized EMG (b),in supraspinatus during assembly and welding work without arm suspension (open columns),and with arm suspension (shaded columns). The bars indicate one standard deviation (SD).
significantly decreased individual mean muscle load in both work situations with arm suspension, compared to no arrn suspension. However, in one individual there was a small increase of IMP values with the arm support during 'welding'. The RMS-values of the EMG (figure 4 (b)) showed a similar pattern as the intramuscular pressure values when the arm was suspended, compared to the non-suspended situation. Here the average reductions for qssembly work and welding amounted to 20% and 17%, respectively. The mean power frequency (MPF) of the EMG signal at the test contractions between each frame of the light assembly work,showed no consistent pattern. There was no evidence of developing localized muscle fatigue. IntramuscuIar pressure at maximal voluntary contraction was recorded in eight of the nine experiments. The IMP was 533 (SD 154)mmHg. The mean IMP during assembly work without arm suspension was 7.5% of MVC,and with arrn suspension 4.8% of MVC. The same calculation for the mean IMP during welding was 12.3% and 9.5% respectively. The APDF-diagrams of both IMP and EMG showed a similar pattern and reduction of load with the arm suspension. Two examples of individual APDFdiagrams are shown in figures S(a)and (b), illustrating one subject with little or no effect of the arm suspension, and one person with a better reduction of load by the arm suspension. The right curve of each APDF-diagram represents the work without arm support and the left curve the work with arm support. The larger difference for subject I1 indicates a more effective reduction of supraspinatus muscle load with the suspension device.
Arm support and supraspinatus muscle load "Assembly work"
Subject I
Subject I
(4
(b)
Subject I1
.
Subject LI
Figure 5 (a, b). Two individual amplitude probability distribution function (APDQ diagrams of IMP and EMG during simulated assembly work (a), and simulated welding (b),one illustrating a subject (I) with little or no effect of arm suspension and the other a subjbct 01) with a more effective reduction of muscle load with the arm suspension.
64
U.Jtirvholm et al.
5. Discussion In the complex field of work-related shoulder pain, reduction of shoulder muscle load during work, is one important step to prevent work-induced disease and impairment. The use of arm suspension has been shown to reduce the load on the trapezius muscle in several studies (Erdelyi et al. 1988, Schiildt et al. 1987, Granstrom et al. 1985). The trapezius muscle has been monitored in most EMG studies, since it is easy to record EMG from, and the load on the trapezius has been shown to fairly well reflect total shoulder muscle load (Hagberg 1981 a). Here we have chosen to study the supraspinatus muscle, since many EMG studies have shown that this muscle is heavily loaded in elevated arm positions, and EMG signs of localized muscle fatigue were often found (Hagberg 1981 b, Kadefors et al. 1976, Sigholm et al. 1984). Furthermore, many clinical problems in shoulder pain are d a t e d to the supraspinatus tendon (Herberts et al. 1984). In our previous studies of intramuscular pressure (IMP)in shoulder muscles, the supraspinatus has generated very high IMP-values in abducted arm positions, and the IMP in the supraspinatus has been shown to give equal information concerning relative muscle load as does the EMG (Jiirvholm et al. 1989). It has also been shown that muscle blood flow in the supraspinatus during contraction was reduced at an IMP of 40 mmHg (Jarvholm et al. 1988b). Hence the supraspinatus muscle is well documented in many respects. The assembly type of work situation chosen for our study was found in an industry where arm support was used to reduce shoulder muscle load. Many of the employees had experienced shoulder pain related to the work situation. Our study showed that this type of work situation induced low load on the supraspinatus muscle as measured with the intramuscular pressure measurement. This was expected since the position during the assembly type of work most closely resembled a situation with only 30" of shouIder flexion and without hand load, as shown in figure 3(a). The load on the supraspinatus muscle was significantly reduced by.arm suspension. However, this small reduction may be unimportant, since it has beem argued that the monotony of this type of work situation rather than muscular exertion is the most important problem (Westgaard et al. 1988). This type of work is often associated with trapezius myalgia (Hagberg and Wegman 1987). The APDF diagrams of both IMP and EMG showed a similar pattern of reduction of load by the arm suspension for each individual. The APDF diagram has previously mostly been used for EMG-analysis of dynamic work situations, and has provided valuable information of muscle load patterns (Jonsson 1982). Our study confirms this and also show that the same type of diagram can be used on the IMP values recorded during a dynamic work situation. Since the absolute IMP values are related to muscle blood perfusion, this type of analysis may give information useful for determining muscle load in relation to muscle fatigue. The temporal aspect of the work is in this respect also very important to analyse. There were no consistent changes in mean power frequency of the EMG signal in the test contractions used in this experiments. This indicates that there was no fatigue induced in this work situation in the supraspinatus muscle. The low pressure values observed support this conclusion, and furthermore both work situations were of short duration, during which time localized muscle fatigue may not occur. KBrner et al. (1984a) found that the spectra1 changes of the EMG signal from the biceps were affected already at 20 mmHg. In our study most subjects had a muscle pressure below 20 mmHg some time during the work. This may allow for fatigue recovery. It could also be argued that our method using bipolar intramuscular electrodes is less sensitive in
Arm support and supraspinatus muscle load
65
detecting the low-frequency part of the EMG signal, which increases at localized muscle fatigue (Kadefors 1978). During the simulated welding work, the IMP was higher due to the higher degree of shoulder flexion and higher hand load. However the IMP did not amount to more than 10-1 2% of the I M P at maximal voluntary contraction. The arm support significantly reduced the IMP in supraspinatus also in this work situation. The reduction was however not large enough to reduce the mean IMP below 40 mmHg. Also in this work situation there was some dynamic muscle activity, but somewhat less marked compared to the assembly type work situation. It is known that the I M P varies within a muscle, and that superficial parts of a muscle generate lower , IMP values compared to deeper parts (Kirkebn and Wisnes 1982, Sejersted et al. 1984). It is thus possible that other parts of the supraspinatus muscle have lower IMP values than the reported here, since we measured I M P in the central part of the muscle. Additional reduction of muscle load in the supraspinatus in the welding work situation could possibly be achieved by increasing the suspending force on the arm or add some suspension on the tool to reduce hand load. There is probably no benefit from a total suspension of the arm making it 'weightless', since this may reduce the muscular control and precision of the work situation.
6. Conclusion Our study has shown that arm suspension will reduce muscle load in the supraspinatus as measured with intramuscular pressure and EMG,both in an assembly-type work situation and during welding with a higher shoulder muscle load. The reduction of muscle load was significant, but in the welding situation with arm suspension l S l 5 N, average muscle pressure was still high enough to reduce muscle blood flow. There is no evidence from this study that arm suspension will reduce the amount of work-related shoulder pain and impairment, since monotony, work organization, and socioeconomic factors are as important factors to consider, particularly in light assembly work. In the case of welding work, more effective support than suspension force l(r15 N is needed in order to reduce the IMP sufficiently. References AARAS,A., W E S T G A ~ DR., H. and STUNDEN,E. 1988, Postural angles as an indicator of postural load and muscular injury in occupational work situations, Ergonomics,31, 915933. ARWDSSON, A. 1982, A statistical method for detection of disturbances in physiological signals, Technical Report, Research Laboratory of Medical Electronics, Chalmers University of Technology, Giiteborg, Sweden, 6. BROMAN,H. and KADEFOR~ R. 1979, A spectral moment analyser for quantification of electromyograms, Proceedings o j the 4th Congress of ISEK (Boston, USA), 90-91. ERDELYI, A., SIHVONEN, T., HELIN,P. and H~NNINEN, 0. 1988, Shoulder strain in keyboard workers and its allevation by arm supports, Int. Arch. Occup. Environ. Health, 60,ll9-124. GRANSTR~M, B., KVARNSTR~M, S. and T ~ N B A C K E F. R ,1985, Electromyographyas an aid in the prevention of excessive shoulder strain, Appl. Ergonomics, 16,49-54. HAGBERG, M. 1981a, Work load and fatigue in repetitive arm elevations, Ergonomics, 24, 543-555. HAGBERG, M. 1981b, Electromyographic signs of shoulder muscle fatigue in two elevated arm positions, Am. J. Phys. Med., 60, 111-121. HAGBERG, M. 1984, Occupational and musculoskeletal stress and disorders of the neck and shoulder: a review of possible pathophysiology, Int. Arch. Occup. Environ. Health, 52, 269-278.
HAGBERG,M. 1987, Shoulder pain: pathogenesis, in Clinical Concepts in Regiorurl Musculoskeletal Illness, ed. N. M.Hadler (Grune & Stratton Inc., New York), 191-200.
66
Arm support and supraspinatus muscle load
HAGBERG, M. and WEGMAN,D. H. 1987, Prevalence rates and odds ratios of shoulder-neck diseases in different occupational groups, Br. J. Ind. Med., 44, 602-610. HERBERTS, P. and KADEFORS,R. 1976, A study of painful shoulders in welders, Acta Orthop. Scand., 47, 381-387. HERBER'IS, P., KADEFORS, R., H~GFORS, C. and SIGHOLM,G. 1984, Shoulder pain and heavy manual labour, Clin. Orthop., 191, 166178. JARVHOLM,U., PALMWUD, G., S m , J., HERBERTS, P. and KADEFORS, R. 1988% Intramuscular pressure in the supraspinatus muscle, J . Orthop. Res., 6, 230-238. U., STYF, J., SUURKULA, M. and HERBWTS,P. 1988 b, Lntramuscular pressure and JKRVHOLM, muscle blood flow in supraspinatus, Ew. J. Appl. Physiol., 58, 219-224. JARVHOLM, U., PALMERUD, G., HERBERTS, P., H~GFORS, C. and KADEFORS, R. 1989,Intramuscular pressure and electromyography in the supraspinatus muscle at shoulder abduction, Clin. Orthop., 245, 102-1 09. JONSSON, B. 1982, Measurement and evaluation of local muscular strain in the shoulder during constrained work, J. H m n Ergol., 11, 73-88. KADEFORS,R. 1978, Applications of electromyography in ergonomics: new vistas, Scond. J. Rehub. Med., 10, 127-133. P. 1976, Muscular reaction to welding work: an KADEFORS,R., PEIE&N, I. and HEIRBERIS, electromyographic investigation, Ergonomics, 19, 543-558. KADEFORS, R., PETERS^, 1. and HERBER-rs, P. 1984, Shoulder pain in industry: new aspects, Bull. Hist. Mar. Trop. Med. (Adynia), 35, 99-109. KIRKES~, A. and WISNES,A. 1982, Regional tissue fluid pressure in rat calf muscle during sutained contraction or stretch, Acta Physiol. Scand., 114, 551-556. KORNER,L., PARKER,P., ALMSTR~M, C., HERBERTS, P. and KADEFORS,R. 1984a, The relation between spectral changes of the myoekctric signal and the intramuscular pressure of human skeletal muscle, Eur. J . Appl. Physiol., 52, 202-206. K ~ R N E RL.,, PARKER,P., ALMSTR~M, C., ANDERSSON, G. B: J., HERBERIS,P., KAI)EFDRS, R., PALMERUD, G. and Zarn\~m, C. 1984b, Relation of intramuscular pressure to the force output and myoelectric signal of skeletal muscle, J. Orthop. Res., 2, 289-296. LOUPN~RVI, T., KOURINKA,J., VIROLAINEN,M. and HOLMBERG, M. 1979, Prevalence of tenosynovitis and other injuries of the upper extremities in repetitive work, Scand. J. Work, Environ. & Health, 5, 48-55. MAEDA, K. 1977, Occupational cervicobrachial disorders and its causative factors, J. Human Ergol., 6, 193-202. PARKER,P. A., K ~ R N E RL., and KADEFORS,R. 1984, Estimation of muscle force from intramuscular total pressure, Med. & Biol. Eng. & Cornput., 22, 453-457. K., EKHOLM, J., HARMS-RINGDAHL., K., N m m ~G. ,and ARBORELIUS, U. P. 1987,Effect . SCH~LDT, of arm support or suspension on neck and shoulder muscle activity during sedentary work, Scand. J . Rehab. Med., 19, 77-84. SEJERSTED, 0.M., HARGENS, A. R., KARDEL, K. R., B ~ MP.,, JENSEN, 0.and HERMANSEN, L. 1984, Intramuscular pressure during isometric contraction of human skeletal muscle, J. Appl. Physiol., 56, 287-295. G., HERBERTS, P., ALMSTR~M, C. and KADEFORS, R. 1984, Electromyographic analysis SIGHOLM, of shoulder muscle.load, J. Orthop. Res., 1, 379-386. S ~ PJ., R. and K ~ R N E RL., M. 1986, Microcapillary infusion technique for measurement of intramuscular pressure during exercise, Clin. Ort hop., 207, 253-262. WESTGAARD, R. H. 1988, Measurement and evaluation of postural load in occupational work situations, Eur. J. Appl. Phys., 57, 291-304. WINKEL,J. 1987, On the significance of physical activity in sedentary work, in Work with Display Units 86, ed. B. Knave and P. G. Widebick (Elsevier, Amsterdam), 229-235. Manuscript received 16 March 1990. Manuscript accepted 10 July 1990.
199 1, VOL. 34, NO. 1, 57-66
The effect of arm support on supraspinatus muscle load during simulated assembly work and welding ULFJ ~ V H O L M *GUNNAR , PALMERUD~,
ROLAND KADEFORS~~ and F%~ER HERBERTS~ Department of Orthopaedics, University of Gijteborg, East Hospital, S-416 85 GGteborg, Sweden t Project Lindholmen Industrial Development Centre, PO Box 87 14, S-402 75 Gbteborg, Sweden $ Department of Orthopaedics, University of Gbteborg, Sahlgren Hospital, S-413 45 Giiteborg, Sweden 5 Department of Applied Electronics, Chalmers University of Technology, S-412 96 Geteborg, Sweden Keywords: Arm suspension; Supraspinatus; Intramuscular pressure;
Electromyography. The effect of arm support, by . a suspension device, on muscle load in the supraspinatus muscle was evaluated with simultaneous intramuscular pressure measurement and electromyography (EMG)in nine healthy subjects. Two work situations, a low load assembly type of work, and welding with a higher shoulder muscle load, were simulated in the laboratory. Each subject performed three workcycles of each type, with and without arm support. Arm suspension reduced supraspinatus muscle load in both work situations with reduction in pressure of 34% and 22% respectively, and reduction in normalized EMG of 20% and 17% respectively. The reduction of muscle load was significant, but in the welding situation with arm-suspension 10-1 5 N, average muscle pressure was still higb enough to reduce muscle blood flow.The interpretation of the importance of this toad reduction for the development of work-related shoulder pain is problematic.
1. Introduction Shoulder pain related to work situations with prolonged elevated arm positions is an increasing problem for industrialized society (e-g., Loupajarvi et al. 1979, Hagberg 1984, Aaris et al. 1988).High muscle load and static work situations are believed to be two of the main causative factors (Maeda 1977, Hagberg and Wegman 1987). Two clinical conditions are manifested by shoulder pain in industry. The first is shoulder tendinitis, which has been shown to be related to high muscle load, elevated arm positions, and handling of heavy objects, usually in male industria1 workers. The second is shoulder myalgia, which is more related to monotonous low load work situations like assembly-type work, and more common among women (Hagberg 1984, Herbexts et al. 1984). In evaluation of shoulder muscle load, electromyography (EMG) is most often used. EMG can be applied in evaluation of the relative muscle load in different shoulder muscles (Sigholm et al. 1984).The trapezius muscle is the one most frequently studied, both regarding shoulder muscle load and localized shoulder muscle fatigue (Hagberg 1981 a, Granstr6m et al. 1985, Westgaard 1988). Besides EMG, intramuscular pressure (IMP) can be used in evaluation of local muscle load ( K h e r et al. 1984 b, Parker et a!. 1984, Jhrvholrn et al. 1989). High IMP values have been found in the supraspinatus muscle in elevated arm positions even without hand load (Jiirvholm et a/. 1988 a). IMP 0014-0139/9I S34M 0 1991 Taylor & Francis Ltd.
58
U.Jirvholm et al.
levels exceeding 40 mmHg (5.3 kPa) during contraction reduced the muscle blood flow in the supraspinatus muscle (Jarvholm et a). 1988 b). The ability of arm support to reduce shoulder muscle load has been demonstrated for the trapezius muscle in EMG studies (Erdelyi et al. 1988, Schiildt et al. 1987, Granstrdm et a1. 1985). However, there is still a possibility that arm suspension causes a higher load in other shoulder muscles (Granstrtim et al. 1985), or that the effect of a m suspension in the long run may be unphysiological for human shoulder muscle performance (W inkel 1987). The aim of this study was to evaluate supraspinatus muscle load (a) during light assembly work, and (b) during welding, both with and without arm support. EMG and intramuscular pressure were recorded simultaneously from the supraspinat us, chosen because of its clinical importance and EMG findings of high muscle load and localized muscle fatigue during prolonged arm elevation (Herberts et al. 1976, Kadefors et al. 1976, Hagberg 1981b, Herberts et al. 1984).
2 Materials and methods Nine healthy subjects volunteered for the study. All were male with a mean age of 27 years (range 2 1-50). All subjects were right-handed, and all measurements were done in the right supraspinatus muscle. One subject (the oldest) was a worker in the industry from which the light assembly work was simulated. All others were students who volunteered for the experiment. The experiments were carried out in a laboratory, where working places simulating the two working situations were set-up. Before insertion of the pressure recording catheter, all subjects were familiarized with the working situation, with and without arm support for which a suspension device (K-block@,Mabs Int AB, Norrkoping, Sweden) was used. The suspension force was 1 0 - 15 N, depending on individual preference and comfort. The pressure recording catheter in supraspinatus and the suspension device is shown in figure I. The recordings were carried out from the central part of the right supraspinatus muscle. After local anaesthesia of the skin, an intravenous infusion cannula (VasculonQD,Viggo, Helsingborg, Sweden) was inserted into the muscle. Through the Vasculon@cannula, the EMG electrodes (Stabilohma 1 10, diameter 70 pm,polyurethane isolated; Johnson Matt hey Metals, London) and an intramuscular pressure recording catheter (MyopressQP, Atos Medical, Herby, Sweden) were inserted. This technique of sirnu1taneous intramuscular EMG and IMP recording is described elsewhere (Jarvholm et al. 1989). Intramuscular pressure was measured with the microcapillary infusion technique (MCI) (Styf and Kdrner 1986).With this technique, the pressure is recorded through a saline-filled 1-05mm thick teflon catheter, with four small holes close to its open tip. The infusion is regulated by a microcapillary connected to the teflon catheter and an pressurized' bag of saline. This is a non-constant infusion technique, where the infusion speed of saline is regulated by the pressure difference over a microcapillary. The infusion rate was 1-5ml/h at rest, and thus lower when IMP increased due to muscle con traction. The pressure transducer (Bentley TrantecQ, ~rvine, USA) with a displacement of 3-0x 10- mrn3/kpa, is connected to the system and the signal displayed on a chart recorder and recorded on magnetic tape for later computer analysis. In each test situation the height difference of the pressure transducer relative to the catheter tip was corrected for. The intramuscular EMG electrodes were deinsulated at the tip for approximately 2 mm. The EMG signals were amplified 1000 times, and filtered in such a way that the.
Arm support and supraspinatus muscle load
59
analysed signal was in the frequency interval of 35-800 H z The signals were recorded on magnetic tape and later computer analysed. Here the root mean square of the signal was calculated. The analogldigital converter had a sampling frequency of 2048 Hz,and the s i pla1 was d,ivided into 0.5 s segments from which the mean values and standard deviatic3ns were calculated. Each segment was submitted to an autoimatic qu~ality - any influence of artifacts in the signal (Arvidsson 19132). The IEJMG control, LU -.Irule out _.__>__~ was normalized for each individual to the mean value of the EMG for a sranaardized test situation consisting of 12 position-load combinations for each subject. The test positions chosen were flexion (F)30 and 60" with a flexed elbow, and abduction (A) 30 and 60"with a straight elbow. In each position the hand was loaded with zero, one or two kg weights. These positions were chosen in order to be able to compare the intramuscular pressure values anived at to other IMP recordings in the supraspinatus muscle, reported previously QQrvholm et al. 1988 a). The test positions were done before each experiment. The mean power firequency of the EMG was automaticaIly calculated by a custom-made spectral rno~ ment ana~lysator(Broman and Kadefors 1979) on line, and evaluated during test coniracaons. r -
-*---A:--
3. Experimental procedure
3.1. Light assembly work This was found in a light industrial setting, where small rings for pens are nickel-plated. The work is a low load, but fairly static work situation, where onIy the amis are used actively. The workers had to hang each ring separately on a small hook in the course of the process. The job was performed in the sitting posture. The position of the a m was a slightly flexed shoulder, and a flexed elbow. The arm was moved from a slightly abducted to a slightly adducted position for each row of hooks. The frame of the small hooks was adjustable in height in order to obtain a more comfortable working position. The working situation is shown in figure 1. There were nine rows of hooks, each row containing 35, a total of 315 hooks. In the experiment each subject completed three frames with and three without arm suspension. Before the work and after each frame,
Figure 1. The assembly work situation, which was simulated in the laboratory. Note the arm suspension device and the I M P catheter in the right supraspinatus muscle.
Figurc 2. The simulated welding work with arm suspension.
the subject performed a test contraction by holding the arm in an abducted position, aiming with the hand at the edge of the frame. This test contraction was included in order to evaluate any mean power frequency changes of the EMG signal induced by the work. signifying localized muscle fatigue. The results were calculated and presented as the mean IMP and mean normalized EMG. for each individual during the whole work cycle with and without arm support. 3.2. Welding situation This was chosen to illustrate a work situation with a higher shoulder load. In this study a welding simulator (Kadefors et 01.1984) was used, making it possibIe to study weEding in a standardized fashion. The shoulder joint was in the situation chosen flexed approximately 60°,the elbow was flexed, and the weight of the 'welding' tool 1-4kg. In this work situation the hand was thus held at shoulder level. The simulated welding is shown in figure 2. Each subject simulated welding three electrodes. The welding was done with and without arm suspension. Each work-cycle was 1.2-1.5 min long. The individual mean valucs of intramuscular pressure and EMG during the welding were calculated for the two test situations. For each subject an amplitude probability distribution function (APDF) diagram for both EMG and IMP values during the two simulated work situations with and without arm support was plotted. The APDF-diagram has been described by Jonsson (1982) for analysis of EMG during dynamic muscle work. In summary the APDF diagram is a cumulative diagram showing the probabilities of the amplitudes of a musclc signal (EMG or IMP) during a work cycle. The diagram can be interpreted so that the steeper the curve-the more static is the muscle Ioad, and a shift of thc curve to the left by a changed work situation-indicates a reduction of the muscIe load. After the simulated work, the IMPduringa maximal test contraction was recorded, in a position of approximately 45" of abduction. 3.3. Statistics The Student's paired t-test was used to compare the IMP values with and without hand Ioad.
Arm support and supraspinatus muscle load
0 1 2
0 1 2
0 1 2
0 1 2
haadlaad (kg)
€PIG i n supraspinatus
handload (kg)
Figure 3. (a)Intramuscularpressure in the supraspinatusat shoulder flexion (F)30 and 60"with a flexed elbow, and shoulder abduction (A) 30 and 60" with a straight elbow. In each position the hand was loaded with 0, 1, or 2 kg. The two separately-shaded columns correspond approximately to the work situation in the assembly type of work (flexion 30" without hand load) and welding work (flexion 60a, 1 kg hand load). The bars indicate one standard deviation (SD).(b) Normalized R M S values of the EMG found in the same test positions shown in figure 3(a).
4. Results The I M P and EMG results in the isometric test positions are shown in figures 3 (a)and (b). As found in previous studies, both IMP and EMG were dependent on arm position and hand load. Both IMP and EMG gave a similar pattern of muscle activation. The correlation coefficient of mean I M P versus mean EMG for the twelve test situations was 0-96.All values of IMP are within the same range as found in our previous studies. The mean values (and SD) of IMP and EMG for the two situations with and without arm support is shown in figures 4 (a)and (b). During the assembly work, mean I M P was 34-9 (SD 12.1)mmHg without, and 22-8 (SD 11-0)mmHg with arm suspension. The time for a subject to mount one frame varied from 2.5-7.5 (mean 52)min. There was no difference in working pace, with or without arm suspension. During 'welding', mean IMP was 65.5 (SD19-5)mmHgwithout, and 51.4 (SD16-8)mmHgwith arm suspension. The average reductions of IMP values for assembly work and welding were thus 34% and 22% respectively. There was a
U.JIirvholm et al.
assmbh) vwk
Figure 4 (a,b). Mean intramuscular pressure (a) and normalized EMG (b),in supraspinatus during assembly and welding work without arm suspension (open columns),and with arm suspension (shaded columns). The bars indicate one standard deviation (SD).
significantly decreased individual mean muscle load in both work situations with arm suspension, compared to no arrn suspension. However, in one individual there was a small increase of IMP values with the arm support during 'welding'. The RMS-values of the EMG (figure 4 (b)) showed a similar pattern as the intramuscular pressure values when the arm was suspended, compared to the non-suspended situation. Here the average reductions for qssembly work and welding amounted to 20% and 17%, respectively. The mean power frequency (MPF) of the EMG signal at the test contractions between each frame of the light assembly work,showed no consistent pattern. There was no evidence of developing localized muscle fatigue. IntramuscuIar pressure at maximal voluntary contraction was recorded in eight of the nine experiments. The IMP was 533 (SD 154)mmHg. The mean IMP during assembly work without arm suspension was 7.5% of MVC,and with arrn suspension 4.8% of MVC. The same calculation for the mean IMP during welding was 12.3% and 9.5% respectively. The APDF-diagrams of both IMP and EMG showed a similar pattern and reduction of load with the arm suspension. Two examples of individual APDFdiagrams are shown in figures S(a)and (b), illustrating one subject with little or no effect of the arm suspension, and one person with a better reduction of load by the arm suspension. The right curve of each APDF-diagram represents the work without arm support and the left curve the work with arm support. The larger difference for subject I1 indicates a more effective reduction of supraspinatus muscle load with the suspension device.
Arm support and supraspinatus muscle load "Assembly work"
Subject I
Subject I
(4
(b)
Subject I1
.
Subject LI
Figure 5 (a, b). Two individual amplitude probability distribution function (APDQ diagrams of IMP and EMG during simulated assembly work (a), and simulated welding (b),one illustrating a subject (I) with little or no effect of arm suspension and the other a subjbct 01) with a more effective reduction of muscle load with the arm suspension.
64
U.Jtirvholm et al.
5. Discussion In the complex field of work-related shoulder pain, reduction of shoulder muscle load during work, is one important step to prevent work-induced disease and impairment. The use of arm suspension has been shown to reduce the load on the trapezius muscle in several studies (Erdelyi et al. 1988, Schiildt et al. 1987, Granstrom et al. 1985). The trapezius muscle has been monitored in most EMG studies, since it is easy to record EMG from, and the load on the trapezius has been shown to fairly well reflect total shoulder muscle load (Hagberg 1981 a). Here we have chosen to study the supraspinatus muscle, since many EMG studies have shown that this muscle is heavily loaded in elevated arm positions, and EMG signs of localized muscle fatigue were often found (Hagberg 1981 b, Kadefors et al. 1976, Sigholm et al. 1984). Furthermore, many clinical problems in shoulder pain are d a t e d to the supraspinatus tendon (Herberts et al. 1984). In our previous studies of intramuscular pressure (IMP)in shoulder muscles, the supraspinatus has generated very high IMP-values in abducted arm positions, and the IMP in the supraspinatus has been shown to give equal information concerning relative muscle load as does the EMG (Jiirvholm et al. 1989). It has also been shown that muscle blood flow in the supraspinatus during contraction was reduced at an IMP of 40 mmHg (Jarvholm et al. 1988b). Hence the supraspinatus muscle is well documented in many respects. The assembly type of work situation chosen for our study was found in an industry where arm support was used to reduce shoulder muscle load. Many of the employees had experienced shoulder pain related to the work situation. Our study showed that this type of work situation induced low load on the supraspinatus muscle as measured with the intramuscular pressure measurement. This was expected since the position during the assembly type of work most closely resembled a situation with only 30" of shouIder flexion and without hand load, as shown in figure 3(a). The load on the supraspinatus muscle was significantly reduced by.arm suspension. However, this small reduction may be unimportant, since it has beem argued that the monotony of this type of work situation rather than muscular exertion is the most important problem (Westgaard et al. 1988). This type of work is often associated with trapezius myalgia (Hagberg and Wegman 1987). The APDF diagrams of both IMP and EMG showed a similar pattern of reduction of load by the arm suspension for each individual. The APDF diagram has previously mostly been used for EMG-analysis of dynamic work situations, and has provided valuable information of muscle load patterns (Jonsson 1982). Our study confirms this and also show that the same type of diagram can be used on the IMP values recorded during a dynamic work situation. Since the absolute IMP values are related to muscle blood perfusion, this type of analysis may give information useful for determining muscle load in relation to muscle fatigue. The temporal aspect of the work is in this respect also very important to analyse. There were no consistent changes in mean power frequency of the EMG signal in the test contractions used in this experiments. This indicates that there was no fatigue induced in this work situation in the supraspinatus muscle. The low pressure values observed support this conclusion, and furthermore both work situations were of short duration, during which time localized muscle fatigue may not occur. KBrner et al. (1984a) found that the spectra1 changes of the EMG signal from the biceps were affected already at 20 mmHg. In our study most subjects had a muscle pressure below 20 mmHg some time during the work. This may allow for fatigue recovery. It could also be argued that our method using bipolar intramuscular electrodes is less sensitive in
Arm support and supraspinatus muscle load
65
detecting the low-frequency part of the EMG signal, which increases at localized muscle fatigue (Kadefors 1978). During the simulated welding work, the IMP was higher due to the higher degree of shoulder flexion and higher hand load. However the IMP did not amount to more than 10-1 2% of the I M P at maximal voluntary contraction. The arm support significantly reduced the IMP in supraspinatus also in this work situation. The reduction was however not large enough to reduce the mean IMP below 40 mmHg. Also in this work situation there was some dynamic muscle activity, but somewhat less marked compared to the assembly type work situation. It is known that the I M P varies within a muscle, and that superficial parts of a muscle generate lower , IMP values compared to deeper parts (Kirkebn and Wisnes 1982, Sejersted et al. 1984). It is thus possible that other parts of the supraspinatus muscle have lower IMP values than the reported here, since we measured I M P in the central part of the muscle. Additional reduction of muscle load in the supraspinatus in the welding work situation could possibly be achieved by increasing the suspending force on the arm or add some suspension on the tool to reduce hand load. There is probably no benefit from a total suspension of the arm making it 'weightless', since this may reduce the muscular control and precision of the work situation.
6. Conclusion Our study has shown that arm suspension will reduce muscle load in the supraspinatus as measured with intramuscular pressure and EMG,both in an assembly-type work situation and during welding with a higher shoulder muscle load. The reduction of muscle load was significant, but in the welding situation with arm suspension l S l 5 N, average muscle pressure was still high enough to reduce muscle blood flow. There is no evidence from this study that arm suspension will reduce the amount of work-related shoulder pain and impairment, since monotony, work organization, and socioeconomic factors are as important factors to consider, particularly in light assembly work. In the case of welding work, more effective support than suspension force l(r15 N is needed in order to reduce the IMP sufficiently. References AARAS,A., W E S T G A ~ DR., H. and STUNDEN,E. 1988, Postural angles as an indicator of postural load and muscular injury in occupational work situations, Ergonomics,31, 915933. ARWDSSON, A. 1982, A statistical method for detection of disturbances in physiological signals, Technical Report, Research Laboratory of Medical Electronics, Chalmers University of Technology, Giiteborg, Sweden, 6. BROMAN,H. and KADEFOR~ R. 1979, A spectral moment analyser for quantification of electromyograms, Proceedings o j the 4th Congress of ISEK (Boston, USA), 90-91. ERDELYI, A., SIHVONEN, T., HELIN,P. and H~NNINEN, 0. 1988, Shoulder strain in keyboard workers and its allevation by arm supports, Int. Arch. Occup. Environ. Health, 60,ll9-124. GRANSTR~M, B., KVARNSTR~M, S. and T ~ N B A C K E F. R ,1985, Electromyographyas an aid in the prevention of excessive shoulder strain, Appl. Ergonomics, 16,49-54. HAGBERG, M. 1981a, Work load and fatigue in repetitive arm elevations, Ergonomics, 24, 543-555. HAGBERG, M. 1981b, Electromyographic signs of shoulder muscle fatigue in two elevated arm positions, Am. J. Phys. Med., 60, 111-121. HAGBERG, M. 1984, Occupational and musculoskeletal stress and disorders of the neck and shoulder: a review of possible pathophysiology, Int. Arch. Occup. Environ. Health, 52, 269-278.
HAGBERG,M. 1987, Shoulder pain: pathogenesis, in Clinical Concepts in Regiorurl Musculoskeletal Illness, ed. N. M.Hadler (Grune & Stratton Inc., New York), 191-200.
66
Arm support and supraspinatus muscle load
HAGBERG, M. and WEGMAN,D. H. 1987, Prevalence rates and odds ratios of shoulder-neck diseases in different occupational groups, Br. J. Ind. Med., 44, 602-610. HERBERTS, P. and KADEFORS,R. 1976, A study of painful shoulders in welders, Acta Orthop. Scand., 47, 381-387. HERBER'IS, P., KADEFORS, R., H~GFORS, C. and SIGHOLM,G. 1984, Shoulder pain and heavy manual labour, Clin. Orthop., 191, 166178. JARVHOLM,U., PALMWUD, G., S m , J., HERBERTS, P. and KADEFORS, R. 1988% Intramuscular pressure in the supraspinatus muscle, J . Orthop. Res., 6, 230-238. U., STYF, J., SUURKULA, M. and HERBWTS,P. 1988 b, Lntramuscular pressure and JKRVHOLM, muscle blood flow in supraspinatus, Ew. J. Appl. Physiol., 58, 219-224. JARVHOLM, U., PALMERUD, G., HERBERTS, P., H~GFORS, C. and KADEFORS, R. 1989,Intramuscular pressure and electromyography in the supraspinatus muscle at shoulder abduction, Clin. Orthop., 245, 102-1 09. JONSSON, B. 1982, Measurement and evaluation of local muscular strain in the shoulder during constrained work, J. H m n Ergol., 11, 73-88. KADEFORS,R. 1978, Applications of electromyography in ergonomics: new vistas, Scond. J. Rehub. Med., 10, 127-133. P. 1976, Muscular reaction to welding work: an KADEFORS,R., PEIE&N, I. and HEIRBERIS, electromyographic investigation, Ergonomics, 19, 543-558. KADEFORS, R., PETERS^, 1. and HERBER-rs, P. 1984, Shoulder pain in industry: new aspects, Bull. Hist. Mar. Trop. Med. (Adynia), 35, 99-109. KIRKES~, A. and WISNES,A. 1982, Regional tissue fluid pressure in rat calf muscle during sutained contraction or stretch, Acta Physiol. Scand., 114, 551-556. KORNER,L., PARKER,P., ALMSTR~M, C., HERBERTS, P. and KADEFORS,R. 1984a, The relation between spectral changes of the myoekctric signal and the intramuscular pressure of human skeletal muscle, Eur. J . Appl. Physiol., 52, 202-206. K ~ R N E RL.,, PARKER,P., ALMSTR~M, C., ANDERSSON, G. B: J., HERBERIS,P., KAI)EFDRS, R., PALMERUD, G. and Zarn\~m, C. 1984b, Relation of intramuscular pressure to the force output and myoelectric signal of skeletal muscle, J. Orthop. Res., 2, 289-296. LOUPN~RVI, T., KOURINKA,J., VIROLAINEN,M. and HOLMBERG, M. 1979, Prevalence of tenosynovitis and other injuries of the upper extremities in repetitive work, Scand. J. Work, Environ. & Health, 5, 48-55. MAEDA, K. 1977, Occupational cervicobrachial disorders and its causative factors, J. Human Ergol., 6, 193-202. PARKER,P. A., K ~ R N E RL., and KADEFORS,R. 1984, Estimation of muscle force from intramuscular total pressure, Med. & Biol. Eng. & Cornput., 22, 453-457. K., EKHOLM, J., HARMS-RINGDAHL., K., N m m ~G. ,and ARBORELIUS, U. P. 1987,Effect . SCH~LDT, of arm support or suspension on neck and shoulder muscle activity during sedentary work, Scand. J . Rehab. Med., 19, 77-84. SEJERSTED, 0.M., HARGENS, A. R., KARDEL, K. R., B ~ MP.,, JENSEN, 0.and HERMANSEN, L. 1984, Intramuscular pressure during isometric contraction of human skeletal muscle, J. Appl. Physiol., 56, 287-295. G., HERBERTS, P., ALMSTR~M, C. and KADEFORS, R. 1984, Electromyographic analysis SIGHOLM, of shoulder muscle.load, J. Orthop. Res., 1, 379-386. S ~ PJ., R. and K ~ R N E RL., M. 1986, Microcapillary infusion technique for measurement of intramuscular pressure during exercise, Clin. Ort hop., 207, 253-262. WESTGAARD, R. H. 1988, Measurement and evaluation of postural load in occupational work situations, Eur. J. Appl. Phys., 57, 291-304. WINKEL,J. 1987, On the significance of physical activity in sedentary work, in Work with Display Units 86, ed. B. Knave and P. G. Widebick (Elsevier, Amsterdam), 229-235. Manuscript received 16 March 1990. Manuscript accepted 10 July 1990.
