Year : 2018  |  Volume : 22  |  Issue : 3  |  Page : 138--143

Upper extremity muscular strength in push–pull tasks: Model approach towards task design

Joydeep Majumder1, Sanjay M Kotadiya1, Lokesh Kumar Sharma1, Sunil Kumar2,  
1 Division of Physiology and Ergonomics, ICMR-National Institute of Occupational Health, Ahmedabad, Gujarat, India
2 Former Director-in-Charge & Scientist G, ICMR-National Institute of Occupational Health, Ahmedabad, Gujarat, India

Correspondence Address:
Joydeep Majumder
Division of Physiology and Ergonomics, ICMR-National Institute of Occupational Health, Ahmedabad - 380 016, Gujarat


Background: Pushing and pulling in workplaces are common actions. Repetitive forceful exertions in long-duration works lead to increased risk of musculoskeletal disorders and injuries. Aim: To investigate the upper extremity strength in generic push–pull modes while using hand tools and forecasting the limits of the workers while frequent or continuous operation. Settings and Design: The study was conducted among men workers in Ahmedabad city, India, and the design was cross-sectional study. Materials and Methods: In all, 100 men were recruited (Group 1: 18–39 years and Group 2: 40–60 years). Upper extremity muscular strength (isometric mode) testing of the preferred hand during push–pull type of manual hand-tool operations was carried out for 60 s. Forecasting of strength to generate predictions for future events (120 s) based on known past events (measured 60 s) was carried out using Holt–Winters time-series model. Statistical Analysis: Descriptive statistics was used for analysis. For prediction model evaluation, WEKA 3.8.2 was used. Results: Anthropometric parameters of both groups were similar, having no effect on generated strength. Largely, pull strength was recorded to be higher than push strength, wherein Group 2 men generated slightly higher strength. Seated strength was also higher than standing exertion. Forecasting reveals steady strength values for Group 1 men, whereas steep decline among Group 2 men with increasing duration of trial. Conclusion: The strength data generated would aid in work schedule design. Strength forecasting model would assist in developing engineering guidelines in the design of tools at workplace.

How to cite this article:
Majumder J, Kotadiya SM, Sharma LK, Kumar S. Upper extremity muscular strength in push–pull tasks: Model approach towards task design.Indian J Occup Environ Med 2018;22:138-143

How to cite this URL:
Majumder J, Kotadiya SM, Sharma LK, Kumar S. Upper extremity muscular strength in push–pull tasks: Model approach towards task design. Indian J Occup Environ Med [serial online] 2018 [cited 2019 Jan 21 ];22:138-143
Available from:

Full Text


Pushing and pulling are common actions while using hand tools in workplaces and homes, requiring repetitive usage by an operator with variable allowance of shoulder, arm, whole hand, and only finger manipulations. These two activities in generic mode are far more common than they are generally assumed, both unidirectional and bidirectional torque applications. Researchers have documented that continuous or frequent maneuvering of hand tools require high torque generation by the upper extremity during pushing–pulling results in musculoskeletal complaints of the lower back and upper extremities.[1],[2],[3] Inappropriate selection or use of hand tools such as saw, chisel, and drill in carpentry; spanner, rod bender, screwdriver, and trowel in construction; sickle, trowel, and maize sheller in agricultural operations demanding force beyond an operator's capacities may result in loss of control, muscle fatigue, or localized discomfort and injury.[2] Repetitive forceful exertions in endurance are also associated with increased risk for musculoskeletal disorders and injury.[4],[5]

Literature reports that particularly for push–pull type of activities in dynamic postures, the musculoskeletal system gets over exerted and there remains an increased risk of accidents due to tripping/slipping because of forward or backward inertia.[3],[6] However, for hand tools requiring straight translation motions or rotational motions, though the risk of tripping/slipping is less, muscular loading of the upper extremities is inevitable.

Apparently, the dynamic working situation may last from minutes to hours, wherein the strength generated and sustained may vary. Therefore, model forecasting assists in predication of the sustainable strength on time scale based on observation data. Forecasting models have been reported to be applied successfully in time-series research in the literature. The Holt–Winters time series forecasting has been reportedly used in medical research,[7],[8],[9] productivity, production planning, and so on. In this study, Holt–Winters time-series forecasting is used for upper extremity strength prediction on the time scale. This model uses univariate simple exponential smoothing to forecast,[10] wherein more weights are assigned to recent time-series observations than older observations. The Holt–Winters model builds on the simplest form of exponential smoothing and takes into account the possibility of a time series exhibiting some form of trend or seasonality, both of which are updated by simple exponential smoothing.[11]

Although literature reports human strength data for variable work profile and postures, longitudinal forecasting of human strength for designing an optimum work–rest schedule was not found. Determination of human strength capabilities is an important consideration in the development of ergonomic guidelines for pre-employment screening of workers performing manual materials handling jobs. Therefore, this study attempted to systematically investigate the upper extremity strength in generic push–pull modes while using hand tools and forecasting the limits of the workers while frequent or continuous operation.

 Materials and Methods

The experiment was conducted on 100 male healthy and physically active volunteers (18–60 years age) who voluntarily agreed to participate in the laboratory simulated strength testing experiment during push–pull type of manual hand-tool operations. The volunteers were categorized into two groups as per their age (Group 1: 18–39 years and Group 2: 40–60 years). A written informed consent form was obtained from the participants. Anthropometric parameters were measured for all the participants.

For the quantification of hand–arm–shoulder complex strength, isometric strength testing of the preferred hand in concentric and eccentric movements was recorded at 0 Nm torque for the dynamometer of Humac Norm – Testing and Rehabilitation System (CSMi, USA). The push–pull handle coupler (0.3 m) was affixed to the dynamometer shaft at the pre-set anatomical zero position. The virtual range of motion of both the levers was set at +5° for concentric movement and −5° for eccentric movement. Since the focus of the study was quantification of strength in isometric mode, mechanical stops were fixed at +2° (concentric movement) and −2° (eccentric movement). This was done as, in case of any mechanical failure of the instrument, there remains no chance of any injury to the volunteer. The experiment was performed for 60 s without any verbal encouragement, wherein each volunteer was instructed to generate maximum voluntary strength and maintain it till the next instruction (after 60 s). Every volunteer was given a trial before the start of the experimentation process.

The study focused on maximum voluntary strength of only hand–arm–shoulder complex, thus the inadvertent exertion of force from other body parts, or allowing the body weight to influence greater strength by any volunteer was restricted. The height of the seating bench was also adjusted individually for each volunteer, so as to allow optimal grip of the feet on the floor surface, as well as restrict the gravitational back force influencing strength generation [Figure 1]. Data obtained from the instrument software were recorded as torque/ms. Data were entered in the Excel sheet and descriptive statistics were used to analyze data using SPSS 16.0 (Chicago, IL, USA).{Figure 1}

Time-series forecasting was carried out using Holt–Winters model, which estimates the predicted values on previous event data. The model work flow for the time-series forecasting[12] is depicted in [Figure 2]. The forecast modeling was performed on WEKA 3.8.2 environment.[13] The model was evaluated using the mean absolute error (MAE) and root mean square error (RMSE) of the predictions, direction accuracy, and relative absolute error. Time was taken as periodic attribute, and sit pull, stand pull, sit push, and stand push were taken as target attributes. In this study, it was assumed that the strength is reversed with progression of application of force, that is, strength decreases in cases of frequent or continuous hand-tool operation.{Figure 2}


[Table 1] presents the mean, standard deviation, and 95% confidence interval of the anthropometric parameters for both groups of workers. It is observed that all the parameters for both the groups of volunteers were similar. This implies that anthropometrics may not have influenced the strength generated by both the group of workers during the experimental trials.{Table 1}

As observed in [Table 2], the strength generated by Group 2 workers was slightly higher (~3%–13%) than their younger counterparts. The mean pull strength in both the groups was recorded to be higher than the push strength. In addition, the ability to generate and sustain strength in seated posture was higher than the standing posture (~32%–42%).{Table 2}

[Figure 3]a shows a decreasing plateau in push strength for both sitting and standing postures, although the deteriorating trend is more prominent in seated posture. As seen in the graph, preferred hand push–pull activity for men above 40 years of age is more productive and safer in standing posture than seated posture. [Figure 3]b shows that the decreasing plateau in pulling activity is similar for seated and standing posture. Pull strength was observed to be substantially higher in sitting posture than while in standing posture, among men above 40 years of age. Among men in 20–39 years of age, although push strength has a steep decremental trend during the 1-min trial, the predicted trend for the next 3 min reaches a plateau [Figure 3]c. However, the strength application during pushing in standing posture remained almost similar during experimentation. Similar pattern of strength application was observed during pulling activity among men in the age group of 20–39 years [Figure 3]d. For all the four conditions, both push and pull activities in seated posture generated higher strength than the standing posture.{Figure 3}

For push strength in seated posture, men from both age groups reflected a similar trend of depreciating plateau after 30 s of strength application. However, prediction trend for the next 120 s reveals that men in Group 1 would maintain the depreciating plateau, whereas for men in Group 2 would result in steep decline in strength with increasing time [Figure 4]a. For push strength in standing posture, Group 2 men exerted more strength when compared with men in Group 1, although both the groups reflected similar pattern of decremental trend in strength application [Figure 4]b. For pull strength too, similar results were observed as of push strength [Figure 4]c. Similar to the push strength in standing posture, Group 2 men exerted more strength when compared with Group 1 men while pulling in standing posture. Although a sharp decline in strength application was observed for both the groups within 10 s of onset of the trial, a decremental plateau was evident, thereafter in both the groups. Prediction, however, reveals steady strength values for Group 1 men, whereas steep decline among Group 2 men with increasing duration of trial [Figure 4]d.{Figure 4}

[Table 3] shows the evaluation of prediction model. Absolute error, as mentioned, is the absolute value of the difference between forecasted value and observed value. As seen, the range of MAE is 0.831–3.438; the predictor is near to observed value. The model reached 100% accuracy except in the case of Group 1 stand pull. The values for RAE and RMSE also show the forecasted value adaptability, overall revealing model forecast with high accuracy.{Table 3}


Literature reports that for most of the push–pull operations, muscular strength and anthropometric parameters play a pivotal role. However, in this study, upper extremity muscular strength was not influenced by anthropometric variables. This was in contrast to earlier studies reported in literature.[3],[14] The possible reasons may be in the dynamic postures adopted in the earlier studies, which allow higher strength generation, which is suitable for specific task. While this study ventured into static posture to comprehend the maximum strength that a worker can develop and sustain, that is, generic strength of the upper extremity muscles, as well as when dynamic posture could not be adopted, namely, confined space. Workability under restricted condition would need the involvement of the upper extremity muscles only, wherein a work bout may also demand longitudinal time. Therefore, this study focused only on the push–pull strength of the group of muscles of the upper extremity. For both operations, the strength took the peak within the first few seconds, and thereafter a sharp decline was observed till the end of the bout at 60 s. This is in consistence with previous studies reporting exponential decline in exertion force with endurance time.[15],[16] As any task requiring push–pull strength may be longitudinal or periodic, it is imperative to comprehend the sustainability of strength. Therefore, model approach has been applied to the observed data so as to understand the pattern of strength that can be sustained with increase in duration of force exertion at work.

Our study depicted higher pull strength vis-à-vis the observed push strength as ~74 and 67 N, respectively, in seated posture. This is in tandem to the earlier reported study on Indian agricultural workers, wherein the isometric pull and push strength of males was 92 and 77 N, respectively.[17] A similar study from Canada reported isometric both arm pull and push strength of males as 109 and 89 N, respectively.[18] Furthermore, as we tried to comprehend the influence of age on the strength generated, we see that Group 2 depicted higher strength in pull and push operation under both seated and standing postures. Several studies have shown that muscular strength increases till the middle age and then decreases with further progressive age.[19],[20],[21] In addition, training, particularly endurance training, increases the muscular strength of the upper extremities till the middle age.[19] This corroborates with the finding of this study with regard to midlife age and endurance training. All the volunteers in our study were working professionals, wherein Group 1 had lesser endurance training (working experience) than Group 2 workers, thereby consistent with the previous study iterating skilled workers have longer endurance time than nonskilled workers.[22]

As operations at a workplace demand variety of postural orientations (seated or standing), we tried to look into the model prediction of the seated and standing postures in upper extremity strength for 120 s beyond the observed 60-s trial bout [Figure 3]a, [Figure 3]b, [Figure 3]c, [Figure 3]d. We found that single-hand push activity for Group 2 men was more productive and safer in standing posture than seated posture. However, men in Group 1 depicted a steep decremental trend in strength during the 1-min trial; the predicted trend for the next 2 min reaches a plateau for both push and pull operations. Our results showed that in both the age groups, seated operation generates more strength[18] and sustains it in predictive endurance. Although some earlier studies contradict such results,[23],[24] this trend may be due to greater exertion stability in seated posture.

Furthermore, we tried to comprehend the age difference in the model prediction of strength for 120 s beyond the observed 60-s trial bout [Figure 4]a, [Figure 4]b, [Figure 4]c, [Figure 4]d. Although stability in generated strength is observed, the trend is not uniform for Group 2 men, particularly in pulling operation. This may be due to their progression toward midlife, as also reported by Keller and Engelhardt.[25] While predicting for the sustainability of strength for 120 s after the observed 60-s bout, the age was considered in model calculations. This might have affected the trend as seen in [Figure 4]a, [Figure 4]b, [Figure 4]c, [Figure 4]d. The effectiveness of the models is observed in [Table 3], revealing high accuracy in model forecasts.

The trend in strength observed, although inconsistent, may not be alarming, until any extenuating factor(s) make the work difficult, for example, any disease/disorder restricting physical performance. Thus, this is one of the strengths of the study, as this study only recruited healthy volunteers having unrestricted physical performance. Furthermore, volunteers with any recent (within 1 month) disease/disorder restricting physical performance were also excluded from the study, thus limiting recall bias. Although recall bias of the self-reported symptoms is recognized, self-report is a reliable source for determining work-related hazards.[26],[27] However, this study has certain limitations. Women workers were not considered in the study, so as greater variability in age group classification could be considered. The sample size was also moderate. In addition, cause and effect could not be addressed in this cross-sectional study, in which the trends associated with age may simply be due to the individual capability of the volunteers participated. Variability in strength generation and sustainability may also be affected by any number of environmental and developmental factors.

The findings suggest that the upper extremity strength in push–pull modes while using hand tools is an important criterion for work schedule design. Furthermore, forecasting of strength data suggests limits of the workers while frequent or continuous operation. In addition to job design on the workers' front, strength forecasting would assist in developing engineering guidelines in the design of tools at workplace.


The authors thank the Director, ICMR-National Institute of Occupational Health, Ahmedabad, for permitting to conduct this study. The authors are indebted to Mrs. BG Shah, Mr. RC Patel, and Mr. DS Kshirsagar for recruitment of participants and assistance during the study.

Financial support and sponsorship


Conflicts of interest

There are no conflicts of interest.


1Kuiper JI, Burdore A, Verbeek JH, Frings-Dresen MH, van der Beak AJ, Vhkari-Juntura ER. Epidemiologic evidence on manual materials handling as risk factors for back disorders: A systemic review. Int J India Ergon 1999;24:389-404.
2Lin JH, Radwin RG, Fronczak FJ, Richard TG. Forces associated with pneumatic power screwdriver operation: Statics and dynamics. Ergonomics 2003;46:1161-77.
3Tiwari PS, Gite LP, Majumder J, Pharade SC, Singh VV. Push/pull strength of agricultural workers in central India. Int J India Ergon 2010;40:1-7.
4Bureau of Labor Statistics. Employer-Reported Workplace Injuries and Illnesses – 2016; 2016. Available from: [Last accessed on 2018 Jun 05].
5National Research Council. Musculoskeletal Disorders and the Workplace: Low Back and Upper Extremities. Washington DC: National Academies Press; 2001.
6Chaffin DB. Manual materials handling and the biomechanical basis for prevention of low back pain in industry – An overview. Am India Hyg Assoc J 1987;43:989-96.
7Pereira A. Performance of time-series methods in forecasting the demand for red blood cell transfusion. Transfusion 2004;44:739-46.
8Medina DC, Findley SE, Guindo B, Doumbia S. Forecasting non-stationary diarrhea, acute respiratory infection, and malaria time-series in Niono, Mali. PLoS One 2007;2:e1181.
9Buczak AL, Baugher B, Moniz LJ, Bagley T, Babin SM, Guven E, et al. Ensemble method for dengue prediction. PLoS One 2018;13:e0189988.
10Newberne JH. Holt-Winters, Forecasting: A Study of Practical Applications for Healthcare Managers. Mike O'Callaghan Federal Hospital Las Vegas NV; 2006. Available from: [Last accessed on 2018 May 07].
11Walters A, Cai Q. Investigating the use of holt-winters time series model for forecasting population at the state and sub-state levels. Demographics and Workforce Section, Weldon Cooper Center for Public Service, University of Virginia; 2008.
12Mohammed EA, Naugler C. Open-source software for demand forecasting of clinical laboratory test volumes using time-series analysis. J Pathol Inform 2017;8:7.
13Hall M, Frank E, Holmes G, Pfahringer B, Reutemann PH, Witten IH. The WEKA data mining software: An update. ACM SIGKDD Explor 2009;11:10-8.
14Agrawal KN, Tiwari PS, Gite LP, Pharade S, Majumder J, Bhusanababu V. Isometric cranking and steering strength of agricultural workers in central India. J Agric Eng 2009;46:1-8.
15Nag PK. Endurance limits in different modes of load holding. Appl Ergon 1991;22:185-8.
16Lee TH. Endurance time, muscular activity and the hand/arm tremor for different exertion forces of holding. Int J Occup Saf Ergon 2016;22:71-6.
17Gite LP, Majumder J, Mehta CR, Khadatkar A. Anthropometric and Strength Data of Indian Agricultural Workers for Farm Equipment Design. Bhopal: CIAE; 2009.
18Das B, Wang Y. Isometric pull-push strengths in workspace: 1. Strength profiles. Int J Occup Saf Ergon 2004;10:43-58.
19Karavirta L, Häkkinen A, Sillanpää E, García-López D, Kauhanen A, Haapasaari A, et al. Effects of combined endurance and strength training on muscle strength, power and hypertrophy in 40-67-year-old men. Scand J Med Sci Sports 2011;21:402-11.
20Rantanen T, Masaki K, He Q, Ross GW, Willcox BJ, White L, et al. Midlife muscle strength and human longevity up to age 100 years: A 44-year prospective study among a decedent cohort. Age (Dordr) 2012;34:563-70.
21Didomenico A, Nussbaum M. Measurement and prediction of single and multi-digit finger strength. Ergonomics 2003;46:1531-48.
22Rose L, Ortengren R, Ericson M. Endurance, pain and resumption in fully flexed postures. Appl Ergon 2001;32:501-8.
23Yates JW, Karwowski W. Maximum acceptable lifting loads during seated and standing work positions. Appl Ergon 1987;18:239-43.
24Mital A, Faard H. Effects of sitting and standing, reach distance and arm orientation on isokinetic pull strengths in the horizontal plane. Int J India Ergon 1990;6:241-8.
25Keller K, Engelhardt M. Strength and muscle mass loss with aging process. Age and strength loss. Muscles Ligaments Tendons J 2013;3:346-50.
26Vyas H, Das S, Mehta S. Occupational injuries in automobile repair workers. Ind Health 2011;49:642-51.
27Majumder J, Shah P, Bagepally BS. Task distribution, work environment, and perceived health discomforts among Indian ceramic workers. Am J Ind Med 2016;59:1145-55.