Secondary Logo

Journal Logo

Original Research

Comparison of Two-Hand Kettlebell Exercise and Graded Treadmill Walking

Effectiveness as a Stimulus for Cardiorespiratory Fitness

Thomas, James F.; Larson, Kurtis L.; Hollander, Daniel B.; Kraemer, Robert R.

Author Information
Journal of Strength and Conditioning Research: April 2014 - Volume 28 - Issue 4 - p 998-1006
doi: 10.1519/JSC.0000000000000345
  • Free



The origin of the kettlebell (KB), which resembles a cannonball with an affixed handle, can be traced to Russia, circa early 1700s (12). There has been a recent increase in the popularity of KB exercises, which have been purported to improve muscular strength, muscular endurance, aerobic capacity, and to reduce body fat (12). Moreover, KB exercise has recently been suggested for consideration as a form of training for members of the armed services (9) and for its potential benefit in a rehabilitation setting (1,15). Kettlebells are potentially useful for improving muscular strength and cardiorespiratory fitness, in that, they use ballistic full-body movements using accumulated forces from numerous muscle groups (5,7). Biomechanically, training with a KB is associated with the generation of high peak forces of the posterior muscle chain, mediated by cyclic deceleration–acceleration, in an explosive manner (2).

A few studies have investigated the effects of KB exercise on cardiovascular stress and fitness. Jay et al. (5) concluded that KB training did not improve aerobic fitness. The study consisted of KB training 3 d·wk−1 for 8 weeks. However, the training protocol from that study only consisted of 10–15 minutes of interval training 3 d·wk−1 for 8 weeks. A recent study by Farrar et al. (2) reported moderately high V[Combining Dot Above]O2 (≈34 ml·kg·min−1) across a KB protocol consisting of as many KB swings as could be performed in 12 minutes. The KB protocol used in the study, however, may not be practical for people of different ages who wish to use KB exercises on a regular basis to improve cardiorespiratory fitness, muscular strength, and muscular endurance. Another recent study by Hulsey et al. (4) compared the metabolic demand of a KB swing routine with treadmill (TM) running at equivalent rating of perceived exertion (RPE). Thirteen subjects completed a 10-minute KB routine consisting of 35-second exercise (KB swings) intervals followed by 25-second rest intervals. V[Combining Dot Above]O2, metabolic equivalents, pulmonary ventilation, and caloric expenditure were significantly higher for TM (25–39%) than for KB. The investigators concluded that the metabolic responses for KB met the standards of American College of Sports Medicine to improve cardiorespiratory fitness, and the KB protocol could provide sufficient exercise stress to improve aerobic capacity (3). It appears that the use of a specific RPE for exercise intensity to compare cardiorespiratory responses of TM and KB exercise for 10 minutes, resulted in cardiorespiratory responses of heart rate (HR), blood pressure (BP), and V[Combining Dot Above]O2 that were higher in the TM than KB activity. The subjects in this study were not familiar with KB activity, which could have affected the results. Moreover, the required rest intervals during the KB but not the TM sessions could have influenced the results.

Walking is the most common form of exercise in the United States (11). Moreover, the current American College of Sports Medicine (ACSM) recommendations for exercise assert that adults should perform a minimum of 30 minutes of brisk walking daily as a part of a healthy active lifestyle (3). We designed a protocol that used continuous submaximal full-body KB exercise to simulate the cyclic continuous movements, involved in brisk walking in an effort to develop a KB protocol that would offer practical application for effective improvement of cardiorespiratory and muscular fitness in different populations. We have chosen 2 full-body KB exercises that use lower-body movements that correspond with those used during walking, coupled with upper-body movements that allow the KB exercises to be comparable to walking exercise. The KB exercises in this study were performed in accordance with standard “hardstyle” KB form (13), which ensured consistency, reproducibility, and safety of KB movements, an important component of a well-designed repeatable KB protocol. The purpose of this study was to determine whether continuous prolonged KB activity could be used to produce cardiovascular stress similar to that of brisk walking. It was hypothesized that a moderate-intensity, continuous KB protocol using standardized KB exercise technique, would produce similar metabolic and cardiorespiratory responses to that of a rigorous bout of graded TM walking, but that KB would elicit greater RPE than TM exercise.


Experimental Approach to the Problem

Each subject completed a preliminary session to determine body composition and V[Combining Dot Above]O2max and to familiarize participants with proper standardized KB exercise technique. Subjects subsequently completed 2 experimental sessions: a KB exercise session and a TM exercise session. The sessions were separated by at least 36 hours. In the first experimental session, subjects completed a 30-minute KB session that included 3 continuous 10-minute sets of 10 KB swings followed by 10 sumo deadlifts, with 3-minute rests between 10-minute exercise periods. The second experimental session was a 30-minute TM regimen that began at the walking speed and 4% grade that matched the V[Combining Dot Above]O2 from the KB session and included 3-minute rest intervals after 10-minute TM exercise periods. Metabolic and cardiorespiratory responses for the 2 experimental sessions were compared.


Ten subjects (5 women, 5 men between 21 and 31 years of age) completed the study. All subjects were volunteers and provided informed written consent before participation. Each prospective subject completed a medical history questionnaire and was excluded for any of the following: cardiovascular disease, metabolic disease, or exercise-limiting injuries including musculoskeletal injuries of the shoulder, back, pelvis, knee, or ankle.

Each subject was instructed to maintain normal dietary habits and to have eaten 2 hours before their participation in activity sessions. Subjects were instructed to hydrate the night before and day of the testing sessions. Subjects were told to refrain from ingestion of alcohol and caffeine the night before and day of testing. Finally, each subject was instructed to be well rested before all activity sessions with the suggestion of 8 hours of sleep the previous night.

Descriptive characteristics of the subjects are shown in Table 1. The study was approved by the Southeastern Louisiana University Institutional Review Board, and all subjects gave written consent to participate.

Table 1
Table 1:
Anthropometric characteristics of subjects.


Figure 1 shows the protocol for the preliminary and exercise sessions. Before each session, duplicate calibrations were performed on the flow meter with the use of a 3.0 L calibration syringe (Model 5530; Hans Rudolph Inc., Kansas City, MO, USA). Gas analyzers of the metabolic system (ParvoMedics TrueOne 2400 Metabolic Measurement System, Salt Lake City, UT, USA) were calibrated with gases of known composition.

Figure 1
Figure 1:
Protocol timeline. KB = kettlebell; HR = heart rate; BP = blood pressure.

Session 1

In the first session, each subject completed a medical history questionnaire and signed an informed consent to participate in the study. Body composition was calculated using 7-site skinfold measurements (14). After sitting for approximately 10 minutes, resting HR and BP measurements were taken. The BP measurements were taken with an Omron Automatic Blood Pressure Monitor Model HEM-703CP (Omron Healthcare, Inc., Vernon Hills, IL, USA). Subjects sat for 3 minutes before testing for baseline metabolic measurements. To determine V[Combining Dot Above]O2max, each subject completed a graded exercise test on a TM (Quinton Instrument Company Q55 or Q65 series, Seattle, WA, USA) to volitional exhaustion using the Kraemer TM Protocol (6) that began at 2.5 miles per hour and a 4% incline. The speed was increased by 1 mile per hour every 2 minutes until the subjects reached volitional exhaustion. Metabolic measures were continually recorded using a metabolic cart. Throughout the test, the RPE and HR measurements were recorded every minute using the Borg's (15 point) scale and HR monitor (Polar Electro E30, OY, Finland), respectively. Upon completion, immediate and 5-minute postexercise BP and HR measurements were recorded. After subjects reached a HR at or below 120 b·min−1, a Russian KB certified (RKC)—level 2, KB expert conducted a 10-minute familiarization lesson for the KB exercises to be performed in session 2. The familiarization was to ensure proper technique and form for the KB swings and sumo deadlifts.

Each familiarization session was conducted using a one-on-one format between the subject and KB expert because of the novice KB training status of the subjects. During this session, the proper technique and movement phases of each KB exercise were demonstrated, explained, and practiced by the subject. The KB expert employed corrective drills to facilitate each subject's acquisition of proper technique and form throughout the activity sessions. Those drills were: face-the-wall squat for proper base during the swing and sumo deadlift, RKC hip bridges for hip drive and stabilization, a crease drill for the hip hinge action of the loading phase of swing, and finally naked or unweighted versions of both exercises for proper timing. In addition, the KB expert taught each subject instructional cue phrases that would indicate a need to adjust the exercise form during the activity sessions. Those cues included: “drive your hips”; “straight spine”; “arms long/big chest”; “pop and squeeze.” This was especially important for the maintenance of correct form during the KB session, because the subjects performed the exercise with a nose clip in place and a mouth piece connected to the metabolic cart.

Session 2

The second session consisted of the KB protocol. Before testing began, the electrode belt for a HR monitor was adjusted/fastened around the chest of each subject. After sitting for 10 minutes, resting HR and BP measurements were recorded. This was followed by a repeated demonstration of the 2-hand swing and sumo deadlift KB exercises to be performed (Figures 2 and 3) and an explanation of the protocol to be followed. The protocol consisted of three 10-minute activity periods, alternating between 10 repetitions of the 2-handed KB swings and 10 repetitions of the KB sumo deadlifts.

Figure 2
Figure 2:
Two-hand swing kettlebell exercise (A, B, C).
Figure 3
Figure 3:
Sumo deadlift kettlebell exercise (A, B).

After each 10-minute activity period, the subjects received a 3-minute seated rest period. Requiring the subjects to perform each repetition in time with an 80-hertz beat from a metronome, controlled the rate of exercise. The 2-handed swing was a 2 beat movement: 1 beat up, 1 beat down. The sumo deadlift was a 4 beat movement: 2 beats up, 2 beats down. Throughout the exercise period, subjects were verbally prompted to speed up or slow down to maintain the appropriate speed, and verbal cues were made to each subject regarding their technique to ensure proper form. The weights used for the exercises were graded appropriately for gender; male subjects preformed with a 35 lb (16 kg) KB and female subjects performed with a 25 lb (12 kg) KB. If a subject's form began to deteriorate or the subject chose to “drop” down in weight, he or she was given a lighter KB for the next 10-minute exercise period. If this option was used, men were given a 30 lb KB and women a 20 lb KB. V[Combining Dot Above]O2, respiratory exchange ratio (RER), HR, and RPE were continually monitored throughout the exercise session and recorded every minute. Blood pressure was taken before exercise, during rest 1, during rest 2, immediately after exercise, and 5-minute after exercise.

Session 3

Before beginning the third session, 5-minute averages for the percentage of V[Combining Dot Above]O2 during the KB session was calculated for each subject. The third session consisted of a graded TM walking protocol. Before testing began, subjects were asked to sit for 5 minutes. Resting HR and BP measurements were taken. The format of the third session was identical to the KB routine: three 10-minute periods with a 3-minute rest period in between each bout. This protocol matched the timing and activity duration of KB activity in session 2. As in session 1 testing, the grade was set and maintained at a 4% incline. The speed was adjusted as needed, so that the average V[Combining Dot Above]O2 per 5-minute interval would stay within ± 3–5 ml·kg·min−1 of the calculated averages from the KB routine. Respiratory exchange ratio, %V[Combining Dot Above]O2max, relative V[Combining Dot Above]O2, HR, and RPE were determined every minute. Blood pressure was taken before exercise, during rest periods, and immediately and 5-minute after exercise.

Statistical Analyses

A 2 × 2 × 30, gender × trial × time, repeated measures analysis of variance (ANOVA) was performed to compare changes over time between experimental trials and gender for V[Combining Dot Above]O2, RER, and RPE. Two-minute averages for HR were recorded, and a 2 × 2 × 15, gender × trial × time, ANOVA was conducted for HR. A 2 × 2 × 5, gender × trial × time, ANOVA was conducted for systolic and diastolic BP. The repeated measures design employed in this study used a Bonferroni's correction factor to account for the low number of subjects. Subsequent partial eta squared and power analyses were conducted. It was determined that eta squared and mean power for V[Combining Dot Above]O2, kcal·min−1, RER, and HR were 0.71, 0.61; 0.64, 1.0; 0.54, 1.0; and 0.24, 0.54. For systolic BP (SBP), diastolic BP (DBP), and RPE, the values were 0.59, 0.72; 0.28, 0.87; and 0.69, 1.0. These values indicate ample power with the small number of subjects. The statistical analyses were performed using the SPSS PASW Statistics 20 (IBM Corp., Somers, NY, USA) program with an alpha level set at p ≤ 0.05.


There were no time × gender effects for HR, RPE, SBP, DBP, V[Combining Dot Above]O2, RER, kcal·min−1, and thus we have presented collapsed data points with both men and women (n = 10) in the figures. For HR, there was a significant main effect for time (F(14,42.20) = 62.57, p = 0.00) and a significant time × trial interaction (F(14,42.20) = 3.66, p = 0.00) with an increase over time with greater HR during the KB trial (Figure 4).

Figure 4
Figure 4:
Heart rate during kettlebell and treadmill sessions. Values are mean (±SE).

Analysis of RPE revealed a significant main effect for time (F(29,745.39) = 42.26, p = 0.00) and a significant time × trial interaction (F(29,745.13) = 7.27, p = 0.00). A post hoc pair-wise comparison revealed greater RPE in session 2 (KB trial) vs. session 3 (TM trial) (Figure 5).

Figure 5
Figure 5:
Ratings of perceived exertion during kettlebell and treadmill sessions. Values are mean (±SE).

There was a significant time effect for SBP (F(4,44.65) = 23.31, p = 0.00) but no time × trial interactions (F(4,44.65) = 1.81, p = 0.16). There was also a significant time effect for DBP (F(4,36.34) = 6.30, p = 0.003) but no time × trial interactions (F(4,36.43) = 1.18, p = 0.70) (Figure 6). Statistical analysis of V[Combining Dot Above]O2 revealed a significant main effect for time (F(29,60.92) = 39.66, p = 0.00) but not a significant time × trial interaction for V[Combining Dot Above]O2 (F(29,60.92) = 2.45, p = 0.058) (Figure 7). For RER, there was a significant main effect for time (F(29,74.13) = 18.82, p = 0.00) but no significant time × trial effect (F(29,74.13) = 1.37, p = 0.15) (Figure 8). The pattern of change for both trials over time was a reduction in RER. This was the same for kcal·min−1 with a significant main effect for time (F(14,60.94) = 31.7, p = 0.000) but no significant time × trial effect (F(14,60.94) = 1.03, p = 0.39) (Figure 9). Treadmill running speed that was adjusted to duplicate oxygen cost in the KB trial increased and stabilized overtime (Figure 10).

Figure 6
Figure 6:
Blood pressure during kettlebell and treadmill sessions. Values are mean (±SE).
Figure 7
Figure 7:
Oxygen consumption during kettlebell and treadmill sessions. Values are mean (±SE).
Figure 8
Figure 8:
Respiratory exchange ratio during kettlebell and treadmill sessions. Values are mean (±SE).
Figure 9
Figure 9:
Kilocalories per minute during kettlebell and treadmill sessions. Values are mean (±SE).
Figure 10
Figure 10:
Treadmill speed during session 3. Values are mean (±SE).


The hypothesis that a moderately intense KB protocol would produce metabolic responses similar to those produced by a brisk walking protocol was confirmed for V[Combining Dot Above]O2, RER, kcal·min−1, and BP in this study. Moreover, as hypothesized, KB elicited greater RPE than TM. However, contrary to our hypothesis, KB produced higher HR than TM exercise. The data collected in this study indicate that when regulated for frequency of work, a KB routine consisting of 2-hand swings, and sumo deadlifts will elicit similar metabolic responses to those from a moderate-intensity TM walking protocol designed for the improvement of aerobic fitness.

A recent study by Hulsey et al. (4) reported higher values for V[Combining Dot Above]O2, metabolic equivalents (METS), pulmonary ventilation, and caloric expenditure during a bout of TM running compared with KB exercise performed at the same RPE. In that study, subjects completed 10 minutes of separate KB and TM exercises at an intensity equivalent to an RPE of ≈15. In this study, we report greater RPE for KB compared with TM exercise at the same V[Combining Dot Above]O2. This finding would explain the greater V[Combining Dot Above]O2 for TM compared with KB exercise reported in the previous study (4). The protocol of Hulsey et al. (4) tasked subjects to complete a 10-minute KB swing routine consisting with 35 seconds intervals of activity and 25 seconds of rest. The subjects were asked to complete the maximum number of swings possible during the activity periods and rest during the allotted 25 seconds in activity periods. Our experimental protocol was designed in consideration of the constraint of the Hulsey et al. (4) study, but also to be in accordance with ACSM recommendations for aerobic conditioning, and to ensure the practicality of the protocol and its applicability to later implementation by novice fitness enthusiasts or trainers (3).

Only 1 investigation has examined the effects of KB training on improvement in aerobic fitness. Jay et al. (5) examined the effects of 8 weeks of KB training on aerobic fitness. Subjects complete 10 intervals of 30-second KB exercise with 1 minute (for 4 weeks) and 30 seconds (for the final 4 weeks). The exercise progressed across 8 weeks beginning with an unweighted swing to a deadlift with KB, to a 2-handed swing with KB, to a 1-handed swing with KB. They estimated aerobic fitness using an Astrand submaximal cycle test and reported no improvement in aerobic fitness. It is possible that error from submaximal cycling tests to predict V[Combining Dot Above]O2max played a role in the findings of the study. Moreover, there is no indication that the subjects were required to complete a specific number of KB swings in a controlled regimented manner. In the current investigation, a new well-controlled KB protocol was developed and compared with a controlled walking protocol, matched for V[Combining Dot Above]O2 that would meet American College of Sports Medicine recommendations for exercise intensity and duration for the improvement of aerobic fitness (3). Both were evenly matched for time and work rate in an effort to equalize the metabolic stress between KB exercise, and the most commonly prescribed aerobic modality for novice fitness enthusiasts, walking. This study provides the first evidence that a continuous KB protocol can produce the same metabolic cost of moderate-intensity TM walking at 4 miles per hour and 4% grade and potentially have a positive effect on the improvement of aerobic fitness.

In addition to cardiovascular benefits that can be produced from KB exercise protocols, previous investigations provide evidence that KB workouts produce significant strength gains. A study examining the translational effect of a 10-week KB training program on strength, power, and endurance reported that KB training significantly improved strength, power, and endurance for Olympic style barbell lifts and bodyweight exercises (7). Another study investigating the effectiveness of a worksite intervention using KB training to improve musculoskeletal and cardiovascular health found that KB training increased muscle strength of the trunk extensors and lower back. However, no improvements of aerobic fitness were shown (4). This study suggests when the proper KB technique and protocols are used, KB exercise can produce the same metabolic stress and greater cardiovascular responses than brisk TM walking at 4 miles per hour and 4% grade. This suggests that suggest this form of exercise should lead to significant improvement in aerobic fitness. Moreover, findings from previous studies suggest that this form of exercise should also improve muscular strength and endurance (7,8,10).

Farrar et al. (2) reported the oxygen cost of KB swings to be ∼34 ml·kg·min−1 in a protocol in which subjects performed as many swings as possible over the course of 12 minutes. In this study, we designed a protocol that used controlled timed KB swings and sumo lifts with maintained lifting technique throughout the 30 minutes of KB exercise, with 2 and 3 minutes of rest periods. Using both KB exercises, subjects' steady V[Combining Dot Above]O2 was ≈29 ml·kg·min−1, which would correspond to ≈ 60% of their V[Combining Dot Above]O2max.

In conclusion, the KB protocol developed for this study matched the metabolic responses V[Combining Dot Above]O2, METs, and RER from an equal duration of graded TM walking. It also exceeded cardiovascular and psychological responses of HR and RPE walking. Thus, KB exercise protocol used in this study shows promise as an effective exercise modality to improve and maintain cardiorespiratory fitness. Future studies should examine the effectiveness of the KB protocol to increase V[Combining Dot Above]O2max and musculoskeletal strength, endurance, and rehabilitation from injury.

Practical Applications

The KB protocol developed for this study resulted in substantial metabolic cost that was comparable to that of the graded TM walking protocol employed in this study. The alternation of 10 KB swings and 10 KB sumo deadlifts with loads that were used represent a reasonable sustainable resistive exercise regimen that will elicit substantial cardiorespiratory stress that was shown to be equal to or greater than brisk graded walking. For those individuals whom are unable to withstand the ballistic stress produced during traditional aerobic modalities (brisk walking, jogging, running, jumping rope, and the others), employment of the KB protocol from this study may provide not only the desired amount of aerobic conditioning but also muscular strength improvements, not possibly gained through traditional aerobic fitness regimens. Finally, it is important that the kind of instruction we provided our subjects in this study be included in other KB regimens. As with other forms of resistance exercise, previous injuries, training state, and adherence to proper lifting techniques are all very important considerations to avoid injury.


1. Brumitt J, En Gilpin H, Brunette M, Meira EP. Incorporating kettlebells into a lower extremity sports rehabilitation program. N Am J Sports Phys Ther 5: 257–265, 2010.
2. Farrar RE, Mayhew JL, Koch AJ. Oxygen cost of kettlebell swings. J Strength Cond Res 24: 1034–1036, 2010.
3. Garber CE, Blissmer B, Deschenes MR, Franklin BA, Lamonte MJ, Lee I, Nieman DC, Swain DP. Position stand. Quantity and quality of exercise for developing and maintaining cardiorespiratory, musculoskeletal, and neuromotor fitness in apparently healthy adults: Guidance for prescribing exercise. Med Sci Sports Exerc 43: 1334–1359, 2011.
4. Hulsey CR, Soto DT, Koch AJ, Mayhew JL. Comparison of kettlebell swings and treadmill running at equivalent rating of perceived exertion values. J Strength Cond Res 26: 1203–1207, 2012.
5. Jay K, Frisch D, Hansen K, Zebis MK, Andersen CH, Mortensen OS, Andersen LL. Kettlebell training for musculoskeletal and cardiovascular health: A randomized control trial. Scand J Work Environ Health 37: 196–203, 2011.
6. Kraemer RR, Francois M, Sehgal K, Sirikul B, Khan T, Valverde R, Castracane VD. Amylin and selective glucoregulatory peptide alterations during long-term exercise. Med Sci Sports Exerc 43: 1451–1456, 2011.
7. Lake JP, Lauder MA. Kettlebell swing training improves maximal and explosive strength. J Strength Cond Res 26: 2228–2233, 2012.
8. Manocchia P, Spierer DK, Minichiello J, Braut S, Castro J, Markowitz R. Transference of kettlebell training to traditional strength, power, and endurance. J Strength Cond Res 27: 477–484, 2013.
9. O'Hara RB, Serres J, Traver KL, Wright B, Vojta C, Eveland E. The influence of nontraditional training modalities on physical performance: Review of the literature. Aviat Space Environ Med 83: 985–990, 2012.
10. Otto WH III, Coburn JW, Brown LE, Spiering BA. Effects of weightlifting vs. kettlebell training on vertical jump, strength, and body composition. J Strength Cond Res 26: 1199–1202, 2012.
11. Rafferty AP, Reeves MJ, McGee HB, Pivarnik JM. Physical activity patterns among walkers and compliance with public health recommendations. Med Sci Sports Exerc 34: 1255–1261, 2002.
12. Tsatsouline P. Enter the Kettlebell. St. Paul, MN: Dragon Door Publications, Inc, 2006.
13. Tsatsouline P. Russian Kettlebell Challenge. Instructor Manual. St. Paul, MN: Dragon Door Publications, Inc, 2008.
14. Whaley M, Brubaker PH, Otto RM, eds. ACSM's Guidelines for Exercise Testing and Prescription (7th ed.). Philadelphia, PA: Lippincott Williams & Wilkins, 2006. pp. 62–63.
15. Zebis MK, Skotte J, Andersen CH, Mortensen P, Petersen MH, Viskær TC, Jensen TL, Bencke J, Andersen LL. Kettlebell swing targets semitendinosus and supine leg curl targets biceps femoris: An EMG study with rehabilitation implications. Br J Sports Med 47: 1192–1198, 2013.

oxygen consumption; caloric expenditure; respiratory exchange ratio

Copyright © 2014 by the National Strength & Conditioning Association.