Competitive swimmers may perform as many as 500,000 arm strokes a year in a pulling motion at the shoulder (3). This degree of repetition could amount to more than 1,500 shoulder flexion/adduction movements per day in an 11-month training schedule and may predispose many swimmers to an overuse injury (2,35). The likelihood of injury seems to be increased in those athletes who rely solely on work in the water as their predominant means of performance enhancement (36). This approach seems to be more prevalent in causing an instability of the glenohumeral joint (3,40) and a tight posterior shoulder capsule (40). Strength and flexibility exercises have been recommended as adjunct training approaches to prevent injuries (12,23) and enhance swim performance (18).
Previous research has produced mixed results when determining the interchangeable nature between resistance exercises and similar calisthenic exercises. The majority of these studies have centered on comparing maximum push-up repetitions to 1RM bench press (14,17,26,27). Typically, push-up performance is only moderately correlated with 1RM bench press (r = 0.5 to 0.6), but the addition of variables such as body mass (BM), height, BM × push-ups, and lean body mass (LBM) × push-ups can increase the correlation slightly (R = 0.7 to 0.8). However, the standard errors of estimate (SEE) in those studies were so large that the authors concluded that 1RM bench press performance could not be estimated accurately using push-up performance.
Few studies have been done with other analogous exercises such as pull-ups and lat-pulls. In several studies, lat-pull repetitions appear to have the capability of predicting 1RM lat-pull strength with reasonable accuracy in college men (4), male athletes (7), and adult women (5,24). Only 1 study has been identified in which pull-up repetitions were measured in women (22). In that study, none of the average college women could perform a single pull-up repetition against BM; thus, a counter-weighing device was used to assess the 1RM pull-up capability. The moderately trained women were able to perform a 1RM pull-up against a load equivalent to 73% of BM (22). However, when evaluated for 1RM lat-pull, the women could lift a load equivalent to only 55% of BM, with a moderate correlation between the 2 measures (r = 0.44) (22).
Elite women athletes, such as swimmers, train extensively using their arm and shoulder muscles in a variety of resistance exercises and exhibit higher-than-normal levels of shoulder strength (10). It is common for them to perform assisted and free-hanging pull-ups to increase shoulder strength in a pulling motion. Because only 1 study directly compared pull-ups and lat-pulls and because that study used moderately trained women who were not capable of performing a free-hanging pull-up, it would be advantageous to the strength and conditioning specialist to determine the degree of association between these 2 exercises in elite athletes to evaluate the feasibility of interchanging them in a training program. Therefore, the purposes of this study were to determine the relationships among pull-ups, lat-pull repetitions, and 1RM lat-pulls in elite women swimmers and to assess the effect of various anthropometric dimensions on each exercise.
The lat-pull and pull-up are 2 commonly used shoulder exercises that are often assumed to be analogous to one another (11,41) and thus are used interchangeably in a training program (11). There appears to be limited information in the research literature on these shoulder pulling exercises, especially among elite women athletes. We sought to evaluate the similarities and differences among the 2 exercises and to determine the effect of body composition on their performance in a group of highly trained women.
Subjects were 28 highly trained collegiate women who were members of the 3-time National Collegiate Athletic Association (NCAA) Division II national championship team. Each subject volunteered to participate after the study was explained and a consent form was signed. The study was approved by the university Institutional Review Board. All subjects had identical workout loads for the week preceding the study. Pull-ups and lat-pulls were a regular part of the subjects' training routine, and all subjects were familiar with the proper execution of each exercise. Demographic information for the subjects is provided in Table 1.
All participants completed an identical 15-minute warm-up emphasizing shoulder muscle movements and flexibility before the pull-up test. Pull-ups were performed on a standard horizontal bar (diameter = 2.6 cm). Subjects were instructed to complete the maximum number of free-hanging pull-ups with no pause between. To be counted as a complete pull-up, the subject was required to lift the BM from a stationary full-arm extension hanging position (pronated grip) until the chin was above the bar. Halfrepetitions were counted if the subject was able to reach a position of 90 degrees of flexion at the elbow. The distance between the inner edges of each index finger was measured to determine the grip width to be used for the 1RM lat-pull and lat-pull repetitions. The reliability of a typical pull-up test has been reported to be between 0.92 and 0.96 (4,5).
After a 3-day recovery period from resistance training and dry-land activities, the subjects completed a 1RM lat-pull test. The 1RM lat-pull was performed using a seated lat-pull machine (Prostar Lat/Row Combo, Kansas City, MO). Each subject's grip width was marked on the lat-pull bar using Velcro strips. Subjects performed a 10-repetition warm-up at a self-selected weight of 30-50% of estimated 1RM. For the 1RM lat-pull, the seat was adjusted to allow full-arm extension during the eccentric phase of the exercise, and an adjustable padded bar was positioned over the mid-thigh for stabilization. For an attempt to be acceptable, the bar was required to reach an anterior inferior position to the subject's chin with the subject maintaining an upright position. Subjects were allowed to pick their initial weight based on past experience. If the attempt was successful, the weight was increased by 2.3 to 6.8 kg depending on the ease of completing a single repetition. If the attempt was not successful, weight was removed and another attempt was given. A minimum of 3 minutes of recovery was given between attempts (6). This procedure was continued until a complete repetition was not possible. The 1RM was typically reached in 4 to 6 attempts. The reliability of this testing method has previously been noted to range from 0.94 to 0.99 (4,5,20).
Following a 2-day recovery period, lat-pull repetitions were performed on the same device using 80% of 1RM load. The same procedure was used for hand placement, body position, and warm-up as for the 1RM lat-pulls. Subjects completed the maximum number of repetitions until failure at a self-regulated pace with a maximum 2-second pause between repetitions (25). Considering the precision of the weight increments, all loads lifted by the subjects were within 0.5 kg of the required 80% 1RM.
An experienced investigator determined skinfold thickness at the triceps, suprailiac, abdominal, and thigh sites using Harpenden calipers (John Bull Company, Birmingham, England). The Jackson-Pollock 4-site equation was used to estimate body density, and the Lohman equation was used to convert density to %fat as recommended by Hayward and Stolarczyk (19). Fat mass (FM) was calculated as BM × %fat/100, and lean body mass (LBM) was determined as BM - FM (19).
Limb measurements included arm and forearm length. Three measurements were taken using an anthropometer (Lafayette Instruments, Lafayette, Indiana, model 01290), and the average was used to represent each site. Arm length was measured from the acromion process to the olecranon process on the right arm (30). Forearm length was measured from the olecranon process to the styloid process of the radius of the right arm (30).
Pearson correlation coefficients were calculated among the variables. Partial correlations were used to remove the effect of selected variables from the relationship between other variables. Simple and multiple linear regression analyses were used to estimate pull-up and 1RM lat-pull performances from various combinations of independent variables. In multiple regression analysis, standardized beta weights were used to evaluate the relative contribution of each independent variable in explaining the overall variance.
Five of the 9 anthropometric dimensions were significantly related to pull-up performance (Table 1). Only 2 body size variables (BM and LBM) were significantly related to 1RM lat-pull (Table 1). These 2 variables appeared to exert opposite effects on pull-up and 1RM lat-pull performance, while contributing no effect to lat-pull repetitions. None of the anthropometric variables was significantly related to lat-pull repetitions (Table 1). Because neither arm length was significantly related to any strength performance, differences among the subjects were not considered to contribute to pulling performance.
Pull-up performance had stronger exponential relationships than linear relationships with BM (r = 0.64), LBM (r = −0.56), and FM (r = −0.56) (Figures 1-3). If LBM and FM were both included in a power expression to predict pull-up performance, the resulting equation (R = 0.62, SEE = 2.2 repetitions) was as follows:
An analysis of differentials of this equation indicated that the negative effect of gains in FM on pull-up performance was 1.5 times greater than that of the negative effect of gains in LBM.
Although pull-ups alone were not significantly correlated with 1RM lat-pull (r = 0.34, p = 0.08), combining them with BM or LBM increased the accuracy of predicting 1RM lat-pull substantially (R = 0.70, SEE = 6.4 kg, CV = 9.7%). The regression coefficients for BM and LBM were positive and contributed similarly (52% and 54%, respectively) to the common variance (R2 = 0.49). The equations to predict 1RM lat-pull were as follows:
The addition of FM or %fat did not contribute significantly to the prediction of 1RM lat-pull. Furthermore, replacing the individual variables with the product of the 2 (i.e., BM × pull-ups or LBM × pull-ups), as has been done when predicting bench press with push-ups (14,26,27), was not effective for prediction of 1RM lat-pull (R2 < 0.14).
Pull-ups were significantly related to 1RM lat-pull/kg (Table 1). The zero-order correlation (r = 0.69, p <0.01) was similar to the partial correlation between 1RM lat-pull and pull-ups when BM was controlled statistically (r12.3 = 0.64). However, the low correlation between lat-pull reps and pull-ups (r = 0.07) was not enhanced when the effect of BM, LBM, FM, %fat, or various combinations of these variables were controlled by partial correlation (r12.3 ≤ 0.13).
Lat-pull repetitions were not significantly correlated with any anthropometric dimensions (Table 1) or any of the combinations of anthropometric dimensions and pull-up performance. Therefore, no combination of variables produced a significant multiple regression estimate of lat-pull repetitions performance. If the subjects were dichotomized relative to their 1RM lat-pull/kg BM, those with a relative strength greater than 1.0 kg/kg BM were significantly different on 6 of 8 body dimensions and on pull-up repetitions (Table 2).
Few studies have observed the relationship between the calisthenic exercise of pull-ups and the resistance exercise of lat-pulls. Ball (4) found only a moderate correlation between pull-ups and 1RM lat-pull (r = 0.40) in a mixed-gender sample. Using a wide range of submaximal loads (55-95% of 1RM), Ball et al. (5) noted that lat-pull repetitions performed to fatigue could estimate 1RM lat-pull with acceptable precision (r = 0.98, SEE = 4.4 kg) in college women. Kuramoto and Payne (24) also noted a strong relationship (r = 0.95) between 1RM lat-pull and lat-pull repetitions in women of diverse ages (20-70 yrs), which allowed accurate prediction of the 1RM (SEE = 1.85 kg). In a study of athletic men, Chandler et al. (7) found that lat-pull repetitions to fatigue at 60% of 1RM were only moderately related to 1RM lat-pull (r = 0.46) and pull-ups (r = 0.50). Furthermore, 1RM lat-pulls could be predicted accurately (R = 0.72) only when pull-up repetitions were combined with BM (7). The lack of research comparing pull-ups to lat-pulls in women is most likely a result of the inability of many women to perform unassisted pull-ups (32).
In the current study, the lack of association between lat-pull repetitions and pull-ups in a highly trained athletic sample might be partially explained by the manner in which the lat-pull repetition test was conducted. Perhaps a greater association between the 2 variables might have been evident if the lat-pull repetitions had been done with a weight equivalent to BM, although holding weight constant in the relationship between pull-ups and lat-pull repetitions did not seem to indicate that (r12.3 = 0.13). Although all of our sample could perform at least part of a pull-up, it is possible in a diverse sample of college women athletes that a large majority of them might not be able to perform a single pull-up and yet have a diverse range of pulling strength as measured by a lat-pull test (8,22). The concept of a counter-weighted pull-up may have great application in women athletes and deserves further study (22).
The lack of a strong association between pull-ups and lat-pull repetitions in the current study supported the contention that 1 of these exercises should not be substituted for the other in a resistance training regimen. The correlations between pull-ups and anthropometric dimensions suggested that the ability to perform pull-ups may be heavily influenced by body composition factors (22,38) and may not be a good indicator of absolute shoulder pulling strength (34). The fact that greater amounts of both LBM and FM were exponentially detrimental to pull-up performance agreed with the findings of other investigators (22,38) and suggests that greater size does not always mean greater strength, especially among women. As was expected, FM had a greater negative effect on pull-up performance than did LBM. Obviously, this could be a greater problem in women than men because women tend to have 80-150% more FM and a 40-50% greater FM-to-LBM ratio than men (28,29).
The degree to which the difference in muscle distribution between the genders may affect pull-up and lat-pull performances is not currently known. Transverse magnetic resonance imaging has indicated that the largest peak skeletal muscle mass is observed in the upper portion of the thigh for women and at the level of the shoulder for men (1). The gender difference in upper body skeletal muscle mass suggests that average women have less upper body muscle mass relative to their total mass than men (15,21). However, information on the relative distribution of total muscle mass in specific athletic groups is limited. Analysis of world-class weightlifters has suggested that contractile tissue comprises approximately 30% less BM in women (13). Weightlifting is a sport in which greater muscle mass in both the upper and lower body may promote better performance within a weight classification. To the contrary, swimming is predominately a shoulder activity where size and strength in that region may predispose an athlete to better performance. Recently, van der Tillaar and Ettema (39) noted that gender differences in strength among average men and women were eliminated when considered relative to LBM. Indeed, the exponential factor for both BM (PU α BM−6.86) and LBM (PU α LBM−7.40) in this study were similar to the value for PU α BM−7.91 noted in young athletic men (38). This would seem to indicate that in very fit women athletes, the ability to perform pull-ups may be affected by body composition components in a similar manner as in men. Additional research needs to be done to confirm this speculation.
A major reason for athletes to perform resistance training is to deter injury, especially those incurred from overuse (12,23). Furthermore, there is evidence that training in swimming does not promote bone mineral density in young women (10), whereas vigorous resistance training does (9). Add to this the possibility that dry-land resistance training can improve swim performance more than swim training alone and hence offer a potential time savings for young athletes (16), and these exercises offer great promise when judiciously incorporated in a swimmer's total training program.
There is no doubt that swimming is predominantly an arm-pulling sport, but what remains in question is the degree to which performances in pull-ups and lat-pulls would be associated with improvements in swim times. Trappe and Pearson (37) compared groups performing a counter-weighted pull-up training routine to a group utilizing lat-pulls and found that neither training method enhanced nor detracted from swim performance. Girold et al. (16) found that dry-land strength training produced similar increases in 50-m swim performance to those produced by resistive in-water training but noted that both these techniques were more efficient than traditional swim training alone despite equal training times. Because all of our subjects performed each of these exercises, it was not possible to determine whether training with 1 exercise would have a greater effect on swim performance than training with the other exercise. Because Richardson et al. (35) has noted that the backstroke, butterfly, and freestyle have similar mechanics, it would seem reasonable that additional strengthening of the major arm-pulling muscles (e.g., latissimus dorsi) would contribute positively to power production in swimming (16,18). Further research in these areas could have ramifications for the development of resistance training programs, especially for women swimmers. Determining these relationships might allow construction of more specific resistance training programs to enhance performance and prevent injury.
This is one of the first studies to investigate factors related to full-lever pull-ups in women athletes. Few studies have been done on women because of the inability of the vast majority of them to perform a single pull-up. This inability to perform unassisted pull-ups may center around their greater overall proportion of FM and on their lower proportion of muscle mass in the shoulder region (1). In highly trained women athletes such as swimmers, body composition still appears to play a part in their ability to complete pull-up repetitions, with the addition of FM producing a greater negative effect than the addition of LBM. In either case, a larger BM, even if it is mostly LBM, may invoke a certain penalty when performing pull-ups (38) but could aid in performing lat-pulls. The part that body size plays in swim performance remains to be investigated.
Although women may be able to perform lat-pull repetitions using a submaximal weight, this exercise may not be analogous to pull-ups as a shoulder muscle exercise. Body composition and structural factors appear to make different contributions to the performance of 1RM lat-pull and pull-ups in elite performers. From the current results, it appears that the lat-pull exercise offers a wider variety of resistance applications to women athletes than does the pull-up. Further research may be required to determine the best exercises for evaluating and strengthening the shoulder-pulling strength of women for enhancing both athletic and occupational performance.
1. Abe, T, Kearns, CF, and Fukumaga, T. Sex differences in whole body skeletal muscle mass measured by magnetic resonance imaging and its distribution in young Japanese adults. Br J Sports Med
37: 436-440, 2003.
2. Allegrucci, M, Whitney, SL, and Irrgang, JJ. Clinical implications of secondary impingement of the shoulder in freestyle swimmers. J Orthop Sports Phys Ther
20: 307-318, 1994.
3. Bak, K. Nontraumatic glenohumeral instability and coracoacromial impingement in swimmers. Scand J Med Sci Sports
6: 132-144, 1996.
4. Ball, TE. The predictability of muscular strength and endurance from calisthenics [Abstract]. Res Q Exerc Sport
64: A-39, 1993.
5. Ball, TE, Van Vleet, C, Lahey, RJ, and Glass, AL. The relationship of relative muscular endurance to maximal lifting capacity in college men and women [Abstract]. J Strength Cond Res
9: 277, 1995.
6. Bilcheck, HM, Kraemer, WJ, Maresh, CM, and Zito, MA. The effect of isokinetic fatigue on recovery of maximal isokinetic concentric and eccentric strength in women. J Strength Cond Res
7: 43-50, 1995
7. Chandler, T, Ware, JS, and Mayhew, JL. Relationship of lat-pull repetitions and pull-ups to 1-RM lat-pull strength in male athletes. J Human Mov Studies
41: 25-37, 2001.
8. Clemons, JM, Duncam, CA, Blanchard, OE, Gatch, WE, Hollander, DB, and Doucet, JL. Relationship between the flexed-arm hand and select measures of muscular fitness. J Strength Cond Res
18: 630-636, 2004.
9. Egan, E, Reilly, T, Giacomoni, M, Redmond, L, and Turner, C. Bone mineral density among female sports participants. Bone
38: 227-233, 2006.
10. Emslander, HC, Sinaki, M, Muhs, JM, Chao, EY, Wahner, HW, Bryant, SC, Riggs, BL, and Eastell, R. Bone mass and muscle strength in female college athletes (runners and swimmers). Mayo Clin Proc
73: 1151-1160, 1998.
11. Fahey, TD. Basic Weight Training for Men & Women
(2nd ed.). Mountain View, CA: Mayfield, 1994. pp 102-104.
12. Fleck, SJ and Falkel, JE. Value of resistance training for the reduction of sports injuries. Sports Med
3: 61-68, 1986.
13. Ford, LE, Detterline, AJ, Ho, KK, and Cao, W. Gender-and height-related limits of muscle strength in world weightlifting champions. J Appl Physiol
89: 1061-1064, 2000.
14. Frost, RW, Devlin, BS, Soukup, MA, and Timson, BF. Relationship between pushup performance and 1-RM bench press strength in high school students. MO J Health Phys Educ Rec Dance
6: 53-59, 1996.
15. Gallagher, D and Heymsfield, SD. Muscle distribution: Variations with body weight, gender, and age. Appl Radiat Isot
49: 733-734, 1998.
16. Girold, S, Marin, D, Dugue, B, Chatard, JC, and Millet, G. Effect of dry-land vs. resisted- and assisted-sprint exercises on swimming sprint performance. J Strength Cond Res
21: 599-605, 2007.
17. Hart, CL, Ward, TE, Mayhew, JL, and Ball, TE. Pushups as a measure of upper body strength in males. J Human Mov Studies
18: 287-294, 1990.
18. Hawley, JA, Williams, MM, Vickovic, MM, and Handcock, PJ. Muscle power predicts freestyle swimming performance. Br J Sports Med
26: 151-155, 1992.
19. Hayward, VH and Stolarczyk, LM. Applied Body Composition Assessment
Champaign, IL: Human Kinetics, 1996.
20. Hollander, DB, Kraemer, RR, Kilpatrick, MW, Rsmadian, ZG, Reeves, GV, Francois, M, Hebert, EP, and Tryniecki, JL. Maximal eccentric and concentric strength discrepancies between young men and women for dynamic resistance exercises. J Strength Cond Res
21: 34-40, 2007.
21. Janssen, I, Heymsfield, SD, Wang, ZM, and Ross, R. Skeletal mass and distribution in 468 men and women 18-88 yr. J Appl Physiol
89: 81-88, 2000.
22. Johnson, D, Lynch, J, Mayhew, J, Nash, K, and Cygan, J. Relationship of one repetition maximum lat-pull and pull-up performances to relative muscular endurance and body composition [Abstract]. J Strength Cond Res
14: 364, 2000.
23. Johnson, JE, Sim, FH, and Scott, SG. Musculoskeletal injuries common in competitive swimmers. Mayo Clin Proc
62: 289-304, 1987.
24. Kuramoto, A and Payne, V. Predicting muscular strength in women: A preliminary study. Res Q Exerc Sport
66: 168-172, 1995.
25. Lachance, PF and Hortobagi, T. Influence of cadence on muscular performance during push-up and pull-up exercise. J Strength Cond Res
8: 76-79, 1994.
26. Mayhew, JL, Ball, TE, Bowen, JC, and Arnold, MD. Pushups as a measure of upper body strength in females. J Osteopath Sports Med
4: 11-14, 1990.
27. Mayhew, JL, Ball, TE, Arnold, MD, and Bowen, JC. Push-ups as a measure of upper body strength. J Appl Sport Sci Res
5: 16-21, 1991.
28. Mayhew, JL, Hancock, K, Rollison, L, Ball, TE, and Bowen, JC. Contributions of strength and body composition to the gender difference in anaerobic power. J Sports Med Phys Fitness
41: 33-38, 2001.
29. Mayhew, JL, Prather, LM, Mcguire, K, Ball, TE, and Bowen, JC. Validation of gender difference in the Margaria-Kalamen anaerobic power test. J Human Mov Studies
36: 289-301, 1999.
30. Norton, K, Whittingham, N, Carter, L, Kerr, D, Gore, C, and Marfell-Jones, M. Measurement techniques in anthropometry. In: Anthropometrica
. Norton, K and Olds, T, eds. Sydney, Australia: New South Wales Press, 1996. pp. 61-62.
31. Nuber, GW, Jobe, FW, Perry, J, Moynes, DR, and Antonelli, D. Fine wire electromyography analysis of muscle of the shoulder during swimming. Am J Sports Med
14: 7-11, 1986.
32. Pate, RR, Burgess, ML, Woods, JA, Ross, JG, and Baumgartmer, T. Validity of field tests of upper body muscular strength. Res Q Exerc Sport
64: 17-24, 1993.
33. Pink, M, Perry, J, Browne, A, Scovazzo, ML, and Kerrigan, J. The normal shoulder during freestyle swimming: An electromyographic and cinematographic analysis of twelve muscles. Am J Sports Med
19: 569-576, 1991.
34. Ricci, B, Figura, F, Felici, F, and Marchetti, M. Comparison of male and female functional capacity in pull-ups. J Sports Med Phys Fitness
28: 168-175, 1988.
35. Richardson, AB, Jobe, FW, and Collins, HR. The shoulder in competitive swimming. Am J Sports Med
8: 159-163, 1980.
36. Salis, RE, Jones, K, Sunshine, S, Smith, G, and Simon, L. Comparing sports injuries in men and women. Int J Sports Med
22: 420-423, 2001.
37. Trappe, SW and Pearson, DR. Effects of weight assisted dry-land strength training on swimming performance. J Strength Cond Res
8: 209-213, 1994.
38. Vanderburgh, PM and Edmonds, T. The effect of experimental alterations in excess mass on pull-up performance in fit young men. J Strength Cond Res
11: 230-233, 1997.
39. Van Der Tillaar, R and Ettema, G. Effect of body size and gender in overarm throwing performance. Eur J Appl Physiol
91: 413-418, 2004.
40. Weldon, EJ and Richardson, AB. Upper extremity overuse injuries in swimming: A discussion of swimmer's shoulder. Clin Sports Med
20: 423-438, 2001.
41. Wescott, WL. Strength Fitness: Physiological Principles and Training Techniques
(1st ed.). Boston: Allyn and Bacon, 1982. pp 126-128.