The following papers were found through my own references and through online searches of the Medline, CINAHL, PsycLIT and Wilson Social Science Index.

Keywords used were: pause, rest break, static load, occupational health, fatigue, Occupational Overuse Syndrome (OOS), Repetitive Strain Injury (RSI), Cumulative Trauma Disorder (CTD), musculoskeletal disorders, data entry, computer users, visual display unit (VDU), visual display terminal (VDT).

They are listed chronologically.

Laporte, W. (1966) The influence of a gymnastic pause upon recovery following Post Office work. Ergonomics 9,6,501-506

This study identified (by interviews and observations) a "critical" point of fatigue in a group of Post Office workers in the afternoon of work. The investigators implemented a 10-minute pause at this time of the day. One group performed a "gymnastic" pause (exercises for the arms, legs, neck and trunk to music) and one group a passive pause. Three psychophysical tests measuring fatigue and a hand grip strength were performed before and after the pauses. They found that the "gymnastic" pause was superior to a passive pause in relieving fatigue. There is no d00escription of the tasks the workers were doing or if there were any reports of musculoskeletal discomfort.

Chaffin,D.B. (1973) Localized muscle fatigue-definition and measurement. Journal of Occupational Medicine 15,4,346-354

A fairly landmark study regarding fatigue. Chaffin asserts that increased muscle tremour after sustained exertion of muscles used to support the forearm and hand in a static posture, results in a great increase in the time it takes a person to perform 2 different types of task requiring precision positioning of the hands. He describes a 0neuromuscular theory for fatigue mechanisms: a sustained contraction of a muscle can result in the fibres losing their tension producing capacity, additional motor units are then needed to maintain the total tension state of the muscle. The increased number of active motor units is accomplished by increased tension in the spindle sensory systems, thus facilitating central nervous system commands to the motor units that produce movement and tension. The paper then discusses what is now commonly known as the "muscle tension theory of OOS". Design recommendations are then given with respect to muscle fatigue. Rest breaks are not mentioned.

Rohmert,W. (1973) Problems of determination of rest allowances. Part 1: Use of modern methods to evaluate stress and strain in static muscular work. Applied Ergonomics 4,2,91-95

This is a classic, landmark study of fatigue and recovery that is often referred to in other papers. The paper asserts:

• No reduction in maximum strength occurs if holding force is limited to 15% of maximum strength.
• The longer the tiring static work lasts the greater the reduction in available maximum strength.
• With the same holding time, the reduction in available maximum strength is greater the heavier the previous tiring holding work was.
• There are no differences between groups of muscles used if forces are related to the individual maximum force. Similarly, there is no difference from worker to worker if forces are related to the individual maximum force.
• Recovery is a function of the degree of fatigue. Compensating for the same decrease in maximal strength takes the same amount of recovery eg a 10% decrease in maximal strength takes 0.72min for recovery.

Rohmert also suggests natural alternations between tiring continuous work and recovering relaxation must only be interrupted by additional rest pauses in those cases in which:

• The intensity of each muscle contraction is too high
• The speed is not optimal
• The relationship between time for relaxation and contractions is too small
• Static muscle components with greater than 15% maximum voluntary contraction (MVC) are included.

The duration of a resting period necessary to remove any remainder of fatigue in static muscle work was considered and found to be dependent on force and duration of muscle contraction, in the manner of an exponential function linked by a multiplication formula. Eg doubling the hold time from 25% to 50% of maximum time (exhaustion time) requires the rest allowance be increased from 150% to 400%.

Rohmert,W. (1973) Problems of determination of rest allowances. Part 2: Determining rest allowances in different human tasks. Applied Ergonomics 4,3,158-162

Physiologically orientated methods have been developed and are presented in this paper to calculate rest allowances for different work tasks. VDU tasks are not included due to the age of this paper. It is, however, asserted that with the exponential increase of fatigue during work the rule of thumb for rest breaks is little and often. This ensures short working periods with a small average degree of fatigue as well as the frequent experience of the high rate of recovery at the beginning of a break. The exception noted for this rule is that breaks do cause a loss on skill, physiological and psychological adaptation to work, and length and frequencies of breaks have to be adjusted accordingly.

Hagberg,M.(1981) Muscular endurance and surface electromyogram in isometric and dynamic exercise. Journal of Applied Physiology 51,1-7

This study aimed to investigate endurance times of skeletal muscle. That is, the time that a muscle can maintain a required force. Endurance times for sustained isometric (static), dynamic and intermittent exercise of the elbow were measured at different contraction levels. The intermittent exercise consisted of 2 seconds contraction and 2 seconds rest. Results showed rapid decrease of endurance times at levels greater than 15-20% MVC for both sustained isometric and dynamic exercise. Endurance times were increased in intermittent isometric exercise compared with the sustained isometric exercise. The extrapolated contraction level that could then be maintained for 1 hour was 8.2% MVC for the sustained isometric and 25.1% MVC for the intermittent exercise. It is suggested that this is because blood flow is impaired in the sustained contraction which leads to an accumulation of metabolites and a subsequent depletion of energy which impairs the contraction process and causes local muscle fatigue. Hagberg also suggests that in the intermittent exercise the blood flow removal of contraction-inhibiting metabolites is facilitated by the rest periods. Rest breaks as short as 2 seconds may then enhance endurance times significantly, as in this study.

Kogi, K. (1982) Finding appropriate work-rest rhythm for occupational strain on the basis of EMG and behaviour changes. In: P.A. Buser, W.A Cobb, T. Okuma (eds) Kyoto Symposium: Electroencephanolography and clinical neurophysiology, Amsterdam, Elsevier, Biomedical Press pp 738-49

Types of fatigue indicators are discussed in this paper, the need to take rests and the redesign of work-rest cycles to take these indicators into account. Kogi asserts that the maximum endured time in a static contraction and time onset of local pain has a linear relationship with sustained force as a percentage of maximum voluntary contraction (MVC). He found that even for a 10-20 % MVC pain could be felt within several minutes in a sustained elbow flexion task. Fatigue was seen to be avoided by spontaneous alternation of active muscles during an overhead tracking task using a lever when EMG recordings were studied. The author's EMG studies on light industry tasks lead him to recommend that the work-rest cycles need to be adjusted at two different levels. Firstly, the contraction cycle of each contraction must be appropriate to ensure brief intra-work pauses without static load. Secondly, there must be break after a certain period of continuous work. He also suggests that common systems with 1-2 breaks and lunch are not appropriate for repetitive tasks. He asserts work schedules need to be changed to ensure spontaneous pauses are taken frequently in order to prevent excessive muscle fatigue. Kogi feels EMG data is useful in helping to determine the most appropriate schedules.

Hagberg, M. and Kvarnstrom, S. (1984) Muscular endurance and electromyographic fatigue in myofascial shoulder pain. Archives of Physical Medicine and Rehabilitation 65,522-525

An endurance test was conducted on patients with at least one-year history of diffuse shoulder-neck pain. Each patient had to hold both arms in a static position of 30° flexion for as long as possible (the endurance time). EMG recordings were made. All patients demonstrated shorter endurance on their painful side or more painful side. EMG showed muscle fatigue on both sides. EMG signs of fatigue were faster on the painful side for the trapezius muscle but not supraspinatus muscle, but not significant. The findings suggest that the short endurance time in these patients is an effect of localised fatigue and that the patients' disorders had changed the work capacity of the shoulder muscles.

Zwahlen,H.T; Hartmann,A.L. and Rangarajulu,S.L. (1984) Effects of rest breaks in continuous VDT work on visual and musculoskeletal comfort/discomfort and on performance. In: G. Salvendy (Ed) Human-Computer Interaction, Elsevier Science, Amsterdam

An experiment was conducted with typists working on a VDU with mixed tasks: a data entry task and a task correcting errors on the screen. The tests were carried out durin0g 2 full working days. A 15-minute rest break was given in the morning and afternoon and there was a 45-minute lunch break. At the beginning and end of each work period a discomfort rating was completed on screen by the typists. The typists were on an incentive pay scheme. The results showed that the rest breaks were beneficial in reducing musculoskeletal discomfort. The scores for visual discomfort were not as high as musculoskeletal scores and did not decrease as much with breaks. A considerable cumulative effect was seen over the course of the day for both musculoskeletal and visual discomfort. Keystroke rate was slightly higher after breaks but there were no error trends evident. The authors concluded that 2 rest breaks in addition to lunch appeared not to be sufficient to adequately reduce musculoskeletal and visual stress in continuous VDU work. However, sample size was small (n=6).

Janaro,R.E and Bechtold, S.E. (1985) A study of the reduction of fatigue impact on productivity through optimal rest break scheduling. Human Factors 27,4,459-4660

This study compared work output when using a self-determined rest break policy and when using a rest break policy model that was developed. Fatigue was operationally defined as the "decrease in performance due to prolonged activity". The model is based on the assumptions that: work rate is a continuous function of time; that work rate decreases linearly through out periods of work; no work occurs during rest periods but work rate potential increases linearly throughout; rest periods end at the instant full recovery is reached and a time penalty is associated with each rest period (to return to work station or resume work etc). The model uses an algorithm to determine the "best" number of rest breaks. The experimental subjects rode a bicycle ergometer in 4 separate sessions. The first was for diminishing apprehension, the second for data collection of decay and recovery rates, the third the subjects rode using self-determined breaks and in the fourth were instructed to rest according to the model. Results showed that output was greater and there were more rest breaks with the model policy than the self-determined policy. Extrapolation of the results into work situations and in particular VDU tasks is questionable.

Hagberg, M. and Sundelin,G. (1986) Discomfort and load on the upper trapezius muscle when operating a wordprocessor. Ergonomics 29,12,1637-1645

This study measured discomfort and load on the trapezius muscle by EMG during VDU work. The effects of introducing short pauses were evaluated. The types of work periods used were:-

1. 5 hours continuous with coffee breaks and lunch.
2. 3 hours continuous with coffee breaks.
3. 3 hours of work with short pauses 10 times per hour reminded by soft music playing for 15 secs every 6^th minute.

The operators were instructed to put their hands on their laps and try to relax their shoulders and arms during the pauses. Operators were allowed spontaneous short pauses also. Musculoskeletal and visual discomfort was rated by the operators before and after VDU work periods. The number of pauses was noted from video reco0rdings. Electromyographies (EMG) of the upper trapezius muscles were taken. MVC of shoulder elevation was measured before each work period. The discomfort ratings increased after all 3 work periods. This was less where short pauses were introduced. Spontaneous short pauses were fewer when operators were reminded by the music to pause. EMG recordings showed no significant difference in load levels in the 3 work periods. There was a negative correlation between the number of spontaneous short pauses taken by the operators and the static median loads on the upper trapezius. The investigators were not able to demonstrate any effect of the introduced short pauses on muscle load, though the discomfort rate decreased. The paper reports that physiological studies indicate that pauses can be short (2s) but need to be taken frequently to enhance endurance and decrease fatigue.

Jonsson,B. (1988) The static load component in muscle work. European Journal of Applied Physiology 57,305-310

It is suggested in this paper that static loading may be the most important factor in muscular fatigue during work. Past research is reviewed, including the research that suggests that static load should not exceed 5% MVC and that fatigue will appear after one hour at 5% MVC. It has been suggested that job rotation helps to decrease static load. In this paper EMG recordings were taken from electronic assembly tasks, glass blowing and mining tasks involved in job rotation schemes. In some of the less dynamic tasks there was very little difference in muscle load between the tasks. This suggests that if seeking to decrease static load by varying tasks, the type of tasks included in the rotations is important.

Henning,R.A; Sauter,S.L; Salvendy,G and Krieg,E.F (1989) Microbreak length, performance, and stress in a data entry task. Ergonomics 32,7,855-864

The effects of brief rest pauses were evaluated in terms of performance and well being for a repetitive data entry task. The evaluation was in a simulated office environment and consisted of 6 40-minute work periods. There was a 10-minute break away from the workstation between work periods and a 45-minute lunch break. The subjects (n=20) were given a microbreak after 20-minutes of work. The length of this was self-regulated and the subjects were instructed to resume work when they felt ready to continue. A mood survey (to measure tension, fatigue, irritation, boredom and perceived stress) was computer administered before and after each 40-minute work period. Results showed that the mean microbreak length was 27.4 seconds, keystroke output was significantly less and the correction rate greater in the second half of the work period. Microbreak length was predictive of correction rate and heart rate eg a long microbreak predicted low correction rate and a lower heart rate in the second half of the work period. A high level of "fatigue" and "boredom" were associated with long microbreaks. 'Tension', 'fatigue' and 'irritation' levels significantly elevated above resting levels throughout the sessions. The authors proposed that the decreased output and increased correction rates in the second half suggest that discretionary rest breaks were not effective in controlling fatigue as the worker terminated the break before performance recovery was completed. This may be done to avoid task interruption and loss of adaptation to work. Also, it was suggested that a single microbreak might not provide sufficient time to allow adequate recovery. They concluded that microbreaks are instrumental in reducing fatigue and associated performance decrements but in order for workers to receive the full benefits of a 00scheduled break the length may need to be more closely controlled.

Swanson, N.G., Sauter, S.L., and Chapman, L.J. (1989) The design of rest breaks for video terminal work: a review of the literature. In A. Mital (Ed), Advances in Industrial Ergonomics and Safety I: Proceedings of the Industrial Ergonomics and Safety Conference, Taylor and Francis, London, pp ? 895-899.

This paper is a literature review of rest breaks in light, repetitive work and in VDU/ office work to 1989. Support is outlined for short, frequent rest breaks in terms of increased productivity, worker satisfaction and fatigue studies. Studies advocating activity or exercises during rest breaks are cited. The authors conclude that the evidence suggests frequent rest breaks benefit production and comfort in VDU work. (Most of the more recent studies referenced are included in this Extended Bibliography Version 2.0 and also in WMSD and Rest Breaks).

Sundelin,G. and Hagberg, M. (1989) The effects of different pause types on neck and shoulder EMG activity during VDU work. Ergonomics 32,5,527-537

EMG recordings of the trapezius and levator scapulae muscles and discomfort ratings of operators (n=12) were studied during 3 30-minute work periods of VDU work. Three different types of pauses were introduced and compared: an active pause with pause gymnastic movements (no more specifics given); passive pause (the operator relaxed her arms into her lap, leaned backwards in her chair and closed her eyes) and a diverting pause (the operator left the work station for a walk). Each pause lasted 15-20 seconds and operators were reminded of it every 6 minutes with soft music. Results showed the static muscular load was low and did not change with different kinds of pauses, though it is questionable whether 30 minutes is long enough to develop a measurable increase in muscle tension. High contraction levels differed significantly between the three types of pauses, reflecting the activity during the pause. There was greater muscle activity pattern variation with active pauses compared to passive pauses, though the percentage of reference voluntary contraction (measured before the work periods) was lower with passive pauses. Discomfort ratings tendered to be higher with passive pauses and operators tended to find pauses with activity more relaxing than passive pauses. The operators generally regarded the introduced pauses as disturbing to their work routines.

Byström,S.E.G; Mathiassen,S.E. and Fransson-Hall,C. (1991) Physiological effects of micropauses in isometric hand grip exercises. European Journal of Applied Physiology 63,405-411

These researchers studied the physiological response to a continuous and intermittent hand grip exercise to exhaustion at 25% MVC. The intermittent condition consisted of a 10-second pause every 3 minutes. They found that the maximal endurance time was 43% longer in the condition with pauses than the continuous exercise condition. The subjects also reported less fatigue with the intermittent exercise. Subjects in the pause group also returned more rapidly to their normal MVC. Blood flow was significantly higher during the rest pauses than during the contractions. However, the results did show that there were greater downward shifts in electromyography (EMG) frequency after intermittent exercise, indicating that more motor units were being recruited to maintain the same level of tension. This change lasted 24 hours compared with 4 hours in the continuous exercise group. There was no physiological evidence to explain the differences in endurance times and the authors suggest that the longer times in the pauses group could be due to differences at a sensory level and also due to psychological factors. They propose pauses could give an immediate sense of relief and therefore postpone the subjective threshold of fatigue. Due to this, they warned against using pauses to prolong muscle work exposure time and therefore risk musculoskeletal disorders. The study, however, was conducted at 25% MVC and to exhaustion so extrapolation of this risk into the VDU workplace is tenuous.

Dul,J; Douwes,M. and Smitt,P. (1991) A work-rest-model for static postures. In: Y.Queinnec and F.Daniellou (Eds) Designing for Everyone: Proceedings of the 11^th Congress of the International Ergonomic Association, volume 1, New York, Taylor and Francis, pp 93-95

This paper discusses the development of a mathematical work-rest-model to predict the course of musculoskeletal fatigue and recovery during work with static postures and rest. It asserts that supplying sufficient and properly distributed rest pauses during work can prevent musculoskeletal fatigue. The model combines studies on muscle fatigue and recovery during static isometric contractions. It defines the muscle effort (% MVC) as the muscle load expressed as a percentage of the critical muscle group (the muscle with the largest effort in a given posture) strength. The work time is the duration of the muscle effort of the critical muscle group. The rest time of a work-rest period is the duration that the critical muscle group can relax after loading during the work time. The model then calculates the absolute maximum work time, which is the maximum possible holding time of the critical muscle group's muscle effort and the maximum work time, which is the remaining maximum holding time of muscle effort until exhaustion. The major output of the model is muscle fitness, which is the maximum work time as a percentage of the absolute maximum work time. Other experiments have shown that muscle fitness is linearly related to body part discomfort eg 100% muscle fitness = "no discomfort". This model, for static muscle work, predicts that for a given total work time and total rest time, many short work-rest periods are better than a few long work-rest cycles given that work periods are constant. Eg for a muscle effort of 20% and a total work time of 16 minutes and a rest time of 16 minutes, increasing the number of rest breaks from 2 to 4 to 8 to 16, the minimum muscle fitness increases from about 0 to 40 to 60 to 70% respectively. The model was trialed with subjects holding a weight with an elevated arm and showed relatively good agreement with a biomechanical model, individual muscle strength measurements and body discomfort scale. A computer program was developed and trialed by 13 Occupational Health Officers in the Netherlands to develop and support recommendations for changing work-rest times during monotonous, static posture work. The negative points of this model are commented on in Mathiassen and Winkel (1992).

Kuorinka,I. (1992) Prevention of VDU-related musculoskeletal problems; a shift from a specialist's to a participatory approach. In H.Luczak, A.Cakir and G. Cakir (eds) Work with Display Units 92: Proceedings of the third International Scientific Conference on work with Display Units, Technische Universitat Berlin, 217-220