Secondary Logo

Journal Logo

Original Research

Predicting One Repetition Maximum Equations Accuracy in Paralympic Rowers with Motor Disabilities

Schwingel, Paulo A; Porto, Yuri C; Dias, Marcelo C M; Moreira, Mônica M; Zoppi, Cláudio C

Author Information
Journal of Strength and Conditioning Research: May 2009 - Volume 23 - Issue 3 - p 1045-1050
doi: 10.1519/JSC.0b013e3181a06356
  • Free



Intensity is one of the major components during an exercise training protocol to make athletes achieve desirable performance. Thus, intensity has to be accurately established and assigned. Resistance training intensity has often being assigned based on one maximum repetition (1RM), operationally defined as the heaviest load that can be moved over a specific range of motion, one time, and with correct performance (21).

One maximal repetition testing procedures elicit several factors, such as specialized professionals, time, and not being recommended for teenagers, cardiac patients, disabled and elderly people because of the risks it can offer (23). During the last few years, some 1RM-predicting equations have been suggested to avoid direct measurement procedures, yet have accurately established individual strength levels. Lander (15) proposed a model that used testing loads and the number of repetitions to determine 1RM, whereas Baechle and Groves (3) recommended standard constant values to be multiplied by testing loads. Validation studies showed conflicting data concerning 1RM-predicting equations, and whereas some equations produce reliable data, predicted values differed between each calculation model (18,30).

Strength levels determine able-bodied rowers performance. However, there are no available data concerning paralympic rowers' performance. It is not known if paralympic rowers tolerate the 1RM test, and still, if so, the same predicting equation developed for able-bodied rowers could be used with this population.

It is known that able-bodied and rowers with spinal cord injuries show remarkable physiological differences concerning peak power output, maximal oxygen uptake, and lactate threshold and production. These differences may rely on the lower muscle mass induced by disease-induced muscle atrophy (13,22). Paralyzed muscles gradually convert to a fast fatigable phenotype (8), and there is still evidence that despite a reduced number of motor units in paralyzed muscles, these muscles can generate higher levels of normal strength, probably caused by an enlargement in motor unit size (31).

Taken together, all these differences, mainly neuromuscular differences, force generation, and strength production could exhibit several patterns between able-bodied and disabled individuals. If this is the case, predicted 1RM equations, which have been developed for able bodied athletes, could not be adequately applied to people with motor disabilities.

Based on this lack of information, in practice, it is difficult to accurately establish strength training loads for rowers with motor disabilities. It is not known if this population has to be submitted to the 1RM measurement test or if the 1RM values could be predicted by former proposed equations regarding at least the main muscle groups involved in the rowing action. Thus, information is required to clarify this issue.

The aim of this study was to analyze the tolerance of paralympic rowers with motor disabilities to 1RM testing procedures and check the accuracy and reliability of two distinct models of equations to predict 1RM, which is recommended for able-bodied subjects and used in paralympic rowers. Our hypothesis was that as athletes, even with their disabilities, rowers would tolerate the 1RM test, and equations would not produce reliable data for this specific population.


Experimental Approach to the Problem

There is no literature reporting whether athletes with motor disabilities could tolerate 1RM testing procedures and if the various existing proposed equations to estimate 1RM values are accurate for this specific population. In this sense, we submitted rowers with motor disability to 1 RM testing procedures and then applied testing loads to a variety of linear and exponential equations to estimate measured 1 RM values in the main muscle groups involved in rowing action. Next, we compared the measured 1 RM with the values obtained by several validated predicting 1RM equations. We also compared the values obtained by linear and exponential equation models. Based on measured and estimated 1RM values we could conclude first if athletes can perform 1RM test and the accuracy of equations as well.


Nine male, international level paralympic rowers (age, 30 ± 7.9 years; height, 175.1 ± 5.9 cm; weight, 69 ± 13.6 kg; body mass index, 22.8 ± 4.2 kg/m2) who took part at the world championships of adaptive rowing held in 2007 in Munich, Germany and scored among the best 50 times for their specific categories participated in the present study. The sample consisted of 7 one-legged amputated and 2 cerebral paralyzed subjects. Athletes trained 5 times per week (2 sessions of resistance training at least 2 hours a day). Subjects experienced specific rowing training for at least 2 years and resistance training for at least 1 year. During the period of data acquisition, athletes were at the competition phase of their training program and had already reached their performance peak. Training schedules during this phase were 2 sessions of resistance training per week. The first resistance session had loads of 60% 1RM, and the other resistance training session had loads of 80% 1RM. Specific training was conducted on water, where each athlete trained within a specific competition boat (2 sessions per week), and 1 session per week was conducted using the Concept2 rowing ergometer. Specific boat and ergometer training intensity varied between 80 and 120% of individual anaerobic threshold.

After subjects were fully informed about any risk factors and discomfort associated with the procedures, they gave written consent to participate in the study. Before beginning the study, all subjects reported to the lab for testing familiarization purposes. After familiarization was completed for all of the procedures, definitive measurements were conducted. All experimental procedures were approved by the Medicine School Ethics Committee of Bahia's University, Salvador, Brazil.

Testing Procedures

Subjects were submitted to 1RM testing for lying T-bar row. This test was performed with the subject positioned on a T-bar row machine with the chest lying on a bench, and the subject grasped the bar handles with each hand using an overhand grip, detaching the bar from its locked position and letting it hang straight down so that their arms were fully extended. Next, the subject was asked to lift or row the bar up onto the stomach area, returning then to the start position. The flat barbell bench press procedure was performed with the subject lying on a bench with their shoulder blades pinched together. Feet were kept flat at the end of the bench, and the buttocks were always in contact with the bench. The weight was gripped with hands equidistant from the center of the bar with the elbows bent to 90° and the elbows beneath the wrists. The movement was started by lifting the bar off of the pins and lowering it until it touched the chest. The weight was then pushed off of the chest, terminating when the arms were straight, from which point the weight could be lowered again. The 45-degree plate-loaded leg press was performed with the subject seated on the machine with their gluteus and spinal column on the padded support. Subjects with cerebral palsy were placed with their feet on the center of the platform and approximately shoulder width apart, whereas individuals with amputated limbs placed their feet exactly in the center of the platform (medial body line). The legs were extended and safety released, lowering the weight by flexing the knees until the same were at a 90-degree angle in relation to their hips. The subject then pushed back up to the starting position without locking their knees (Righetto Fitness Equipment; Atrex Line, Campinas, SP, Brazil). Athletes were asked not to exercise during the day before the experiments took place, and all of them reported to have remained resting.

One Repetition Maximum Testing Procedures

Testing procedures were conducted with all methodological caution with respect to athletes' disabilities, following the American College of Sports Medicine Guidelines for the 1RM test (2). Warm up and initial 1RM testing loads were obtained from athlete's regular resistance training protocol and applied to the coefficients described by Baechle and Groves (3). Testing load enhancement was also established by Baechle and Groves' (3) coefficients and applied to the number of repetitions with predicted 1RM values.

ACSM guidelines (2) for 1 RM testing consisted of a warm-up of a set of 5-10 repetitions at 40-60% of the predicted 1RM. Subjects then rested for 1 minute and performed some light stretching. Thereafter, 3-5 repetitions were performed with 60-80% of the predicted 1RM. After a 3-minute rest, 3 subsequent attempts were made to determine the 1 RM value, with 3-5 minutes of rest between each lift trial. In the present study, no more than 3 attempts were required for all subjects of the sample to reach 1RM values. No injuries were observed during the 1 RM testing; therefore, the subjects presented the proper technique and the complete range of motion required for this measurement.

Predicting One Repetition Maximum Equations

We have compared the 1RM-measured values with linear and exponential equation models, recommended in current literature for 1RM-predicting values. Tested equations are described below:

Linear Equations

Kg represents the load in kilograms and rep the number of repetitions for the given kilogram load.

Exponential equations:

Kg represents the load in kilograms and rep the number of repetitions for the given Kg load.

Statistical Analyses

Data were first applied to the Kolmogorov-Smirnoff test and Bartlett's criterion to analyze sample distribution and choose between parametric or nonparametric tests. The statistical power of our sample (n = 9) was 0.60 with a type I error of 0.5 and type II error of 0.75, and a Pearson coefficient of variation (d = 0.7) with an ICC coefficient (0.6) were obtained using PIFACE (Java Applets for Power and Sample Size [computer software]. Available at:∼rlenth/Power). and SPSS software (SPSS, Chicago, Ill, Release 16.0.2, 2008). One-way analysis of variance (ANOVA) with Tukey post-hoc test was performed using GraphPad InStat 3.06 for Windows (GraphPad Software, San Diego, Calif) to determine significant differences between the measured and predicted equations 1RM values. As confidential limit, it was defined as p ≤ 0.05.


Quantitative data are summarized in Tables 1 and 2. Qualitative analysis of our experiments showed that despite specific motor disabilities of our subject population, all of them were able to tolerate 1RM testing protocols.

Table 1
Table 1:
Measured and linear equation predicted 1 repetition maximum (RM) values. Data are presented as mean ±SD (n = 9). *p < 0.001 compared with measured 1RM values.
Table 2
Table 2:
Measured and exponential equation predicted 1 repetition maximum (RM) values. Data are presented as mean ±SD (n = 9). *p < 0.001 compared with measured 1RM values.

Our results showed that all equations provided accurate and reliable data concerning maximal upper body strength. However, in most of cases, the predicted values remained less than the measured ones. Lying T-bar row and flat barbell bench press values were not significantly different (p = 0.84 and 0.23 for lying T-Bar row and flat barbell bench press, respectively) among all predicting equations and measured 1RM values. Yet, for the 45-degree plate-loaded leg press exercise, all predicting equations significantly (p < 0.001) underestimated the measured values.


Predicting equations are often used to determine maximal strength and thus assign adequate training loads, mainly for nonathletes and subjects with disabilities, because 1RM procedures require training experience. Reynolds et al. (23) had established equations of 1RM prediction for the flat barbell bench press and plate-loaded leg press exercises and quantified loads of 1, 5, 10, and 20 repetitions to failure in 70 adult subjects of both sexes using an ample and diversified population. The authors found strong correlations, with the greatest prediction accuracy, between the values of 1RM and the equation of prediction for 5RM (leg press: R = 0.974 and bench press: R = 0.993) and stated that the dynamic muscular strength (1RM) can be predicted with precision through tests of smaller multiple repetitions.

Current literature has a wide range of predicting 1RM equations for several uses. Pereira and Gomes (21) stated that submaximal tests, despite their high correlation with measured values and acceptable SEs, often under- or overestimate measured values. Our results are in agreement with this statement because all estimated values were less than the measured ones. Another question concerning predicting equations, discussed by Wood et al. (30) and Reynolds et al. (23), is the number of submaximal repetitions used by each equation. According to these authors, exponential equations are more accurate when submaximal tests with more than 10 maximal repetitions are used.

We have tested equations that apply 10-25 maximal repetitions. Concerning upper body strength, all of them resulted in reliable predicted 1RM values. However, for the lower limbs, neither of them was able to generate accurate results. We have to take into account, when considering our data, that although our statistical power and analyzes did not reach high levels of statistical standards, they remained inside acceptable values for exploratory studies, as this one. Furthermore, regarding this small specific population, we believe that our results will contribute useful findings to the literature and practice coaches who deal with this kind of athlete.

Whisenant et al. (29), verifying the validity of 11 equations of 1RM prediction for the bench press exercise, stated that the effectiveness of the equations depends on the number of executed repetitions and that a raised precision is proportional to a smaller amount of repetitions. However, Reynolds et al. (23) state that even using a larger number of repetitions, one can present relative precision; furthermore, protocols with more repetitions could be a better tool to be used in populations unable to perform tests with huge loads.

Most of prediction equations were established from only one or two exercises, generally the bench press and squat, and when used for other exercises, they present low precision. This statement is supported by the results of Hoeger et al. (14), who analyzed the number of repetitions performed with 40, 60, and 80% of predicted 1RM in 7 different exercises in trained and untrained subjects of both sexes. They have shown that the number of repetitions with distinct loads was different for each exercise, and the total amount of executions with the same intensity was different between the groups. In this sense, the authors suggest that specific submaximal tests cannot be generalized for 1RM prediction of unspecific exercise.

Our sample was composed by international level rowers, all of then with motor disabilities. These subjects, despite being athletes, had motor disabilities and, thus, did not apply to any of the before-mentioned population. The literature lacks data as to whether some of the available equations could be applied to this specific group. Our data demonstrated that despite 2 athletes with severe cerebral paralysis, lying T-bar row and flat barbell bench press 1RM could be predicted by any of the tested equations depending on the equation model and number of submaximal repetitions. However, the 45-degree plate-loaded leg press exercise did not reach the same results, and leg press predicted values for all of the tested equations reached values significantly less than the measured levels.

These results could be explained by the type of disability found in our sample. Most of the subjects had one leg amputated, and thus, this factor could be the reason why leg press predicted values failed to achieve similar levels as those measured.

The few available equations in literature for the leg press exercise do not apply to subjects with motor disabilities, and hence, studies that compared muscular activation and biomechanics execution of lower limbs exercises during maximum strength tests, performed through unilateral and bilateral form, showed significant differences (9,19,25). Therefore, leg press results could be attributed to a different pattern of movement execution.

Practical Applications

We have shown that, despite small but tolerable errors for lying T-bar row and flat barbell bench press exercises, all linear and exponential tested equations were accurate and produced reliable results. Thus, these equations could be applied to rowers with motor disabilities for 1RM prediction. However, regarding the 45-degree plate-loaded leg press, none of them provided accurate results with this specific population, so it should not be used in practice to predict lower-limb maximal strength for this kind of athlete. Concerning lower limb strength, it is more accurate to directly measure 1RM values of amputated and cerebral paralyzed rowers.


The authors thank Fundação de Amparo a Pesquisa do Estado da Bahia (FAPESB), Brazilian agency for financial support (Process Number: 5013/2006). Yuri C. Porto, Monica M. Medeiros and Marcelo C.M. Dias are recipient of IC grants from FAPESB and to the athletes who took part of this study. All authors disclose here any conflict of interest, with no professional relationship with companies or manufacturers who will benefit from the results of the present study. Results of the present study do not constitute endorsement by the authors or the NSCA.


1. Adams, GM. Exercise physiology: laboratory manual. Madison, WI: Brown & Benchmark, 1994.
    2. American College of Sports Medicine. Acsm's guidelines for exercise testing and prescription. Baltimore: Lippincott Willians and Wilkins, 2005.
    3. Baechle, TR and Groves, BR. Weight training: steps to success (2nd ed.). Champaign, IL: Human Kinetics, 1992.
    4. Berger, RA. Determination of resistance load for 1rm and 10rm. J Assoc Phys Ment Rehabil 51: 100-110, 1961.
      5. Berger, RA. Relationship between dynamic strength and dynamic endurance. Res Q 41: 115-116, 1970.
      6. Brown, C, Abadie, B, Boling, R, and O'nan, D. Cross validation of an existing regression equation to predict one repetition maximum (1-Rm) strength. Res Q Exerc Sport A-32, 1994.
        7. Brzycki, H. Strength testing: predicting a one-rep max from reps-to-fatigue. J Phys Educ Recreation Dance 64: 88-90, 1993.
          8. Crameri, RM, Weston, AR, Rutkowski, S, Middleton, JW, Davis, GM, and Sutton, JR. Effects of electrical stimulation leg training during the acute phase of spinal cord injury: a pilot study. Eur J Appl Physiol 83: 409-415, 2000.
          9. Cresswell, AG, and Ovendal, AH. Muscle activation and torque development during maximal unilateral and bilateral isokinetic knee extensions. J Sports Med Phys Fitness 42: 19-25, 2002.
          10. Dohoney, P, Chromiak, JA, Lemire, D, Abadie, BR, and Kovacs, C. Prediction of one repetition maximum (1-Rm) strength from a 4-6 Rm and a 7-10 Rm submaximal strength test in healthy young adult males. J Exerc Physiol 5: 54-59, 2002.
            11. Epley B. Poundage Chart. Boyd Epley Workout. Lincoln, NE: University of Nebraska Press, 1985.
              12. Hagerman, FC. Applied physiology of rowing. Sports Med 1: 303-326, 1984.
              13. Hagerman, F, Jacobs, P, Backus, D, and Dudley, GA. Exercise responses and adaptations in rowers and spinal cord injury individuals. Med Sci Sports Exerc 38: 958-962, 2006.
              14. Hoeger, WWK, Hopkins, DR, Barette, SL, and Hale, DF. Relationship between repetitions and selected percentages of one repetition maximum: a comparison between untrained and trained males and females. J Appl Sport Sci Res 4: 47-54, 1990.
              15. Lander, J. Maximums based on reps. Natl Strength Cond Assoc 6: 60-61, 1985.
              16. Lombardi, VP. Beginning Weight Training. Dubuque, IA: W.C. Brown, 1989.
                17. Mayhew, JL, Ball, TE, Arnold, MD, and Bowen, JC. Relative muscular endurance performance as a predictor of bench press strength in college men and women. J Appl Sport Sci Res 6: 200-206, 1992.
                  18. Mayhew, JL, Prinster, JL, Ware, JS, Zimmer, DL, Arabas, JR, and Bembem, MG. Muscular endurance repetitions to predict bench press strength in men of different training levels. J Sports Med Phys Fitness 35: 108-113, 1995.
                  19. Monteiro, WD and Simão, R. Is there bilateral deficit in the practice of 10rm in arm and leg exercises? Rev Bras Med Esporte 12: 115-118, 2006.
                  20. O'Connor, BJ, Simmons, J, and O'Shea, P. Weight Training Today. St. Paul, MN: West Publishing, 1989.
                    21. Pereira, M and Gomes, P. Muscular strength and endurance tests: reliability and prediction of one repetition maximum. Rev Bras Med Esporte 9: 325-335, 2003.
                    22. Porto, YC, Almeida, M, Schwingel, PA, Sá, C, and Zoppi, CC. Anthropometric and physical characteristics of motor disabilited paralympic rowers. Res Sports Med 16: 203-212, 2008.
                    23. Reynolds, JM, Gordon, TJ, and Robergs, RA. Prediction of one repetition maximum strength from multiple repetition maximum testing and anthropometry. J Strength Cond Res 20: 584-592, 2006.
                    24. Shaw, CE, Mccully, KK, and Posner, JD. Injuries during the one repetition maximum assessment in the elderly. J Cardiopulm Rehabil 15: 283-287, 1995.
                    25. Vandervoort, AA, Sale, DG, and Moroz, J. Comparison of motor unit activation during unilateral and bilateral leg extension. J Appl Physiol 56: 46-51, 1984.
                    26. Wathen, D. Load assignment. In: Essentials of Strength Training and Conditioning. T.R. Baechle, ed. Champaign, IL: Human Kinetics, 1994. pp. 435-446.
                      27. Welday J. Should you check for strength with periodic max lifts? Scholas Coach 57: 49, 1988.
                        28. Wentworth, MC and Abadie, BR. Prediction of one repetition maximum shoulder press strength from a 5-10 repetition submaximal strength test in college-aged females. Res Q Exerc Sport 1: 45-49, 1999.
                          29. Whisenant, MJ, Panton, LB, East, WB, and Broeder, CE. Validation of submaximal prediction equations for the 1 repetition maximum bench press test on a group of collegiate football players. J Strength Cond Res 17: 221-227, 2003.
                          30. Wood, TM, Maddalozzo, GF, Harter, RA. Accuracy of seven equations for predicting 1-Rm performance of apparently healthy, sedentary older adults. Meas Phys Educ Exerc Sci 6: 67-94, 2002.
                          31. Yang, JF, Stein, RB, Jhamandas, J, and Gordon, T. Motor unit numbers and contractile properties after spinal cord injury. Ann Neurol 28: 496-502, 1990.

                          bench press; leg press; lying t-bar row; maximal strength; strength training

                          © 2009 National Strength and Conditioning Association