Comparison of Upper-Limb Muscle Activation Levels in Different Physical Fitness Exercises Using Dumbbells and Elastic Tubes

Abstract

Background:

Elastic tubes are used widely in fitness programs because they are lightweight, easy to use, and versatile. However, evidence of their equivalence to other equipment, including dumbbells, remains insufficient.

Hypothesis:

Elastic tubing and dumbbells do not have equivalent loads, although both equipment generate similar symmetry and muscular synergy.

Study Design:

Cross-sectional study.

Level of Evidence:

Level 2b.

Methods:

Thirty physically active people (15 men and 15 women) performed 4 upper-limb exercises (elbow flexion, shoulder abduction, scapular elevation and abduction, and scapular retraction and abduction) with 5 loads (dumbbells ranging from 2 to 6 kg and red, green, blue, black, and silver elastic tubes).

Results:

Only elbow flexion (P = 0.14 and P ≥ 0.99) and shoulder abduction (P = 0.13 and P ≥ 0.99) exercises showed load equivalence in men but not in women. Both types of equipment were symmetrical and generated similar synergies when the load was increased, with no sex differences (P = 0.11). Load equivalence was found only in men and in the 2 exercises, suggesting that this equipment was not interchangeable.

Conclusion:

Dumbbell and elastic tube exercises can be executed in a balanced and symmetrical manner, yielding similar synergies considering the involvement of all muscle groups regardless of sex.

Clinical Relevance:

This study analyzed a higher number of exercises performed by both men and women. The results showed that these 2 pieces of equipment cannot be considered interchangeable, as they generate different loads. However, similar efforts are required for symmetry and muscle synergy.

Resistance training equipment for achieving fitness, sports, and physical therapy goals includes multiple devices. ¹ Dumbbells are used widely in strength-training programs, are included in reference studies on optimizing upper-limb muscle strength, and allow accurate calculation of the load and change the moment of inertia by varying the range of motion and execution speed.^{2 -4} Elastic bands are useful for muscle growth, and are highly versatile as they are light and easy to use, carry, and adapt to different exercises. ⁵

In rehabilitation, elastic bands can improve upper-limb strength, are deemed a viable alternative to conventional dumbbells,^{6 -8} and can even be used in special populations.^9,10 Mechanically, elastic resistance training facilitates multiaxial and functional activation that is adaptable to different goals. ¹¹ Moreover, owing to its specific exercise scale of perceived exertion, ¹² this feature enables a different assessment from other exercise programs supported by surface electromyography (sEMG). In elastic resistance training, eccentric strength due to stretching increases, requiring greater postural control, as shown in health-oriented activities and sports.^{13 -20}

In elastic bands, tension levels are usually associated with color and thickness. However, assessing load throughout the range of motion in kilograms is difficult and rarely compared with other resistance training equipment, including dumbbells and kettlebells.^{21 -23} According to research on elastic bands, the length of Thera-band 4-loop CLX bands does not vary with speed, showing a tension ranging from 0.01 N in the blue band to 2.97 N in the gold band. ²⁴ Nevertheless, this is not expressed in different exercise programs or when compared with free weights. Bergquist et al ⁶ found that elastic bands have a weaker activation effect on agonist muscles, increasing the synergies between muscle groups with similar mechanical actions by increasing instability.

Other studies comparing equipment did not demonstrate the load equivalence of elastic bands to other equipment that increased the load objectively when used together.^{25 -28} Andersen et al. assessed the load similarity between dumbbells and elastic bands, and proposed 3 exercises (shoulder abduction, shoulder external rotation, and wrist extension) for rehabilitation in a sample of exclusively female participants. ²⁹ An equivalence relationship was established between the elastic bands and dumbbells, as proposed by Thera-Band, without including muscle synergies or sex differences. Therefore, they cannot be applied to a fitness field that focuses on muscle growth.

This study aimed to assess whether dumbbells and elastic tubes have equivalent loads in a sample of men and women. We measured bilateral muscle activation during different upper-limb exercises using sEMG. Moreover, we aimed to identify which of the 2 types of equipment generates more tension on agonist muscles and greater contralateral symmetry in muscle contraction, thereby increasing synergies.

Methods

Participants

The participants were physically active undergraduate students of Sports Sciences (15 men and 15 women) aged 20 to 30 years, averaging 2 to 3 days of weekly physical activity and ≥1 year of experience in resistance training. Subjects who did not meet these requirements for being sedentary or were diagnosed with a muscle and/or joint injury in the shoulders, elbows, or wrists were excluded. This research was approved by the ethics committee at the Catholic University of Valencia under code UCV/2021-2022/015.

Procedure

A descriptive, comparative study was conducted. The subjects performed 4 common upper-limb exercises for physical fitness, always bilaterally. Each exercise was performed with 5 loads, adding 1 (2-6) kg to the dumbbells and changing the color of the elastic tube (red to silver), and 20 exercises per piece of equipment. Activity was measured in the biceps brachii (BB), acromial deltoids (AD), cervical (CT), and dorsal (TD) trapezius muscles. The data were analyzed to assess the equivalence between the dumbbells and elastic tubes after applying different loads. Participants were asked to complete 2 days of measurements - 1 per piece of equipment - and a standardized warm-up protocol.

In the first session, the weight (Seca 750), height (Seca 213), and biacromial distance of each subject were assessed (Table 1).

Table 1.

Descriptive data

Parameter	Value
N	30
Sex
Male	15 (50 %)
Female	15 (50 %)
Age, y	24.9 ± 2.62
Weight, kg	66.22 ± 10.88
Height, m	1.7 ± 0.09
BMI, kg/m2	22.87 ± 2.33
Laterality
Right-handed	17 (56.67 %)
Left-handed	13 (43.33 %)
Biacromial distance, cm	39.23 ± 3.34
Experience, y	1.77 ± 1.68

Data are presented as n (%) or mean ± SD. BMI, body mass index.

The subjects completed a standardized warm-up before placing electrodes in each muscle group according to SENIAM and Criswell to measure muscle activation as follows: BB, longitudinally in the medial and external ventral humerus (channels 1 and 2); AD, downwards towards the deltoid “V” in the lateral and outermost part at the confluence of the scapular spine and clavicle (channels 3 and 4); CT, in parallel, in the central space between the cervical vertebra C7 and the acromion (channels 5 and 6); and TD, in parallel, at the midpoint between the scapular vertex and the spine (channels 7 and 8).^30,31

Electromyographic recordings were performed on an 8-channel sEMG device (Megawin ME6000-T8, Bittium Corp) at a sampling frequency of 1 kHz for 20 s. The data were converted from analog to digital format (12-bit; DAQCard-700; National Instrument) using Megawin software, and saved on a hard drive for subsequent analysis in files with a .ASC extension. Signal analysis was performed using MATLAB (R2023a) (Mathworks Inc). A fourth-order Butterworth bandpass filter was first applied from 20 to 400 Hz. The signal, or root mean square (RMS), was then calculated. Finally, segmentation was performed by collecting the central 16 seconds. Data were collected on average signal amplitudes and smoothed.

Four exercises were performed with different pieces of equipment each day, in a counterbalanced manner, with 5 increasing loads: red, green, blue, black, and silver elastic tubes (Thera-Band tubing) and 2, 3, 4, 5, and 6 kg dumbbells (Bodytone Dumbbell). The elastic tubes were 1.35 m long, reaching 1.20 m with grips, as per the manufacturer’s information, marked at the midpoint.

In each exercise, the subjects performed 6 repetitions at the pace marked by a metronome set to 45 pulses per minute and standing up with their feet separated at their biacromial distance, stepping on an elastic tube with a 2-minute break between exercises.

In the second session (1 week later), the subjects exercised with a piece of equipment that was not used in session 1, following the same order. On both days, the exercises were performed as follows (Figure 1):

Figure 1.

Execution of the study exercises and images of the pieces of equipment.

(1) Elbow flexion: standing up with slight knee flexion, flexing the elbows bilaterally until the hands reach shoulder height. (2) Shoulder abduction: standing up with slight knee flexion and bilateral shoulder abduction until the hands reach shoulder height. (3) Scapular elevation (trapezius): standing up with slight knee flexion and shoulder abduction with elbow flexion until the hands reach shoulder height. (4.1) T-bar row (dorsal muscle) with elastic tube: standing up with slight knee flexion, shoulders flexed at 90º, elbows extended, forearm in a neutral position, extension and external rotation of the shoulders with elbow flexion until reaching shoulder abduction, bringing the hands to shoulder height, and keeping the elbows behind the vertical line of the trunk. (4.2) T-bar row (dorsal muscle) with dumbbells: lying on a stretcher in a prone position, with protruded shoulders in passive flexion towards the floor and elbows extended, holding a dumbbell in each hand, extension and external rotation of the shoulders with elbow flexion until reaching shoulder abduction, bringing the hands to shoulder height, and keeping the elbows behind the vertical line of the trunk. These 2 exercises have the greatest variation due to the unequal effect of gravity. The characteristics of the device do not allow the design of the exercise to have mechanics similar to those of other exercises.

Statistical Analysis

Descriptive Analysis

The data were described using means and standard deviations, along with medians and interquartile ranges for continuous quantitative variables and proportions for qualitative variables.

Linear regression models; agonist muscle activation

According to preliminary data analysis, agonist muscle activation, expressed as a percentage of total microvolts, did not vary substantially with exercise intensity. Changes between exercise intensities were identified by measuring muscle activation during maximum voluntary isometric contraction. To mitigate the possible effects of the participants’ dominant side, which might have contributed to the differences in muscle activation between the left and right arms, the average agonist muscle activation was calculated in both arms.

Asymmetry Index

An asymmetry index was calculated to assess the symmetry of the exercise execution, and is defined as the logarithm of the ratio of left and right agonist muscle activation (uV). To eliminate laterality effects, the numerator of the ratio was always the maximum agonist activation value in microvolts in both arms.

$$Asymmetry=log\frac{maximumagonistactivation}{minimumagonistactivation}$$

This index ranges from 1 to infinity, and values close to zero indicate perfect symmetry.

Exercise Impact

The exercise impact was assessed by calculating the percentage of antagonist muscle activation and averaging the activation of the right and left muscles to avoid laterality effects.

Analysis

The results were first compared using the Wilcoxon signed-rank test to determine whether exercises with elastic tubes and dumbbells activated muscles similarly. We disregarded the specific requirements of each exercise to ascertain whether the 2 pieces of equipment globally induced similar muscle activation. This parameter can help to understand whether exercises with dumbbells and elastic tubes are comparable. The muscle activation generated by different dumbbell weights and elastic tube colors was compared using the pairwise Wilcoxon signed-rank test to determine whether varying exercise demands translated into significant changes in the study variables.

After the initial comparisons, we fitted a mixed-effects linear regression model to identify the differences between exercises performed with dumbbells and elastic tubes. Measurements were normalized in each exercise according to the protocol by Zuur and Ieno, which starts by determining the error structure of the data beyond the optimal model. ³² We determined the random structure of the model by including random factors. Finally, we fitted the fixed structure of the model using the maximum likelihood method, iteratively removing noninformative interactions until the model stopped improving. The below-optimal model included exercise type and intensity as fixed factors, as well as the participant’s sex, height, biacromial distance, and experience. In addition, all interactions among exercise type, intensity, and sex were included, and all quantitative variables were scaled.

$$\begin{array}{l}Y~\mathrm{Muscle}\times \mathrm{Exercise}\_\mathrm{weight}\times \mathrm{Sex}+\mathrm{Height}+\\ \mathrm{Biacromial}\_\mathrm{distance}+\mathrm{Age}+\mathrm{Experience}\end{array}$$

We compared all nested models using the corrected Akaike information criterion calculated using MuMIn Version 1.46.0. ³³ In all cases, the assumptions of the linear models were tested by visually inspecting the residuals and DHARMa residuals using the packages DHARMa Version 0.4.6, and the performance package Version 0.10.3.^34,35 The models were fitted using lme4 Version 1.1-30 and lmerTest Version 3.1-3.^36,37 All tests were performed with α = 0.05 in R Version 4.2.2. ³⁸ Data in the tables were read using the package openxlsx Version 4.2.5 (for .xlsx files) and/or with haven Version 2.5.0 (for .sav files).^39,40 The data were plotted using ggplot2 Version 3.3.6, ggpubr Version 0.4.0, and other functions integrated into the packages.^41,42

Results

The data presented in Table 2 show the equivalence in muscle activation between the dumbbells and elastic tubes, except for the T-bar row (dorsal trapezius). In this case, the elastic tubes generated less activity, significantly differing (P < 0.001) in terms of load.

Table 2.

Average muscle activation (uV) with dumbbells and elastic bands

Muscle	Dumbbells	Elastic tubes	V	P value	FDR
Biceps brachii	247.73 ± 116.18	191.06 ± 95.00	206	0.04	0.084
Acromial deltoids	436.99 ± 114.49	401.6 ± 109.24	248	0.23	0.306
Cervical trapezius	333.5 ± 141.83	321.61 ± 126.99	296	0.76	0.758
Dorsal Trapezius	306.67 ± 98.15	146.3 ± 66.39	26	<0.001	1.1 × 109

Data are presented as mean ± SD. FDR, false discovery rate; V, Wilcoxon signed-rank test.

The Wilcoxon signed-rank test showed that each intensity generated different levels of muscle activation, and a 3-way analysis of variance (ANOVA) indicated that the resulting force was lower when the participants were more experienced, with significant sex differences (Table 3).

Table 3.

Relationships between intensity, experience, and sex

	χ2	df	P value
Exercise/weight	193.417	9	>0.001
Experience	17.556	1	>0.001
Muscle/sex	10.035	1	0.002
Exercise-weight/sex	14.646	9	0.10
Muscle/exercise-weight/sex	14.31	9	0.11

In most exercises, the asymmetry coefficient was similar, regardless of the equipment or load. Figure 2 shows that no significant differences were found among the 4 exercises (P = 0.30), except for deltoids (P = 0.03).

Figure 2.

Normalized activation, asymmetry index, and synergies by muscle group and sex.

No significant differences in synergies produced in each exercise were found between the sexes (P = 0.11), muscle groups (P = 0.51), or weight/tension (P = 0.77) according to 3-way ANOVA. The most striking effect was an increase in exercise completion in terms of muscle activation with the subject’s experience (P = 0.02) (Figure 3).

Figure 3.

Relationship between subject’s experience and agonist muscle activation in the exercises.

Discussion

Selecting the appropriate training equipment is crucial and requires defining criteria for distinguishing the load of each piece of equipment. Therefore, not all equipment is interchangeable, and they yield different levels of impact. ⁴³ The results showed equivalence between elbow flexion (P values from 0.14 to ≥0.99) and shoulder abduction (P values from 0.13 to ≥0.99) exercises only in men. Accordingly, the load equivalence results were not generalizable to either sex (Table 4).

Table 4.

Load equivalence between dumbbells and elastic tubes in men

Exercise	Elbow flexion (biceps)		Shoulder abduction (deltoids)
Equipment	Dumbbell	Elastic tube	Dumbbell	Elastic tube
Load	2 kg	Red	2 kg	Red-green
	3 kg	Green	3 kg	Red-green
	4 kg	Green	4 kg	Blue-black-grey
	5 kg	Blue-black-grey	5 kg	Blue-black-grey
	6 kg	Blue-black-grey	6 kg	Blue-black-grey

In women, changing equipment entailed modifying the exercise and its level, which showed a significantly lower activation value when using elastic tubes, with P values ranging from <0.001 to 0.04 in biceps, mainly at intermediate loads (3-5 kg), and from <0.001 to 0.01 in deltoids, at the highest loads (4-6 kg). This is important because good adherence during the initial training phase is crucial for continuity. ⁴⁴ Although this equipment has benefits, ⁵ caution should be exercised when generalizing its use, as muscle activation decreases with an increase in the subject’s experience (P < 0.001). Both types of equipment were effective in specifically affecting agonist muscles, with no significant differences, regardless of sex. Dumbbells have traditionally shown higher activation percentages in exercises and more effectiveness in isolation exercises, ⁴⁵ including in hypertrophy training, whereas elastic bands require more balance and postural control in single-joint exercises where the body position and work plane affect muscle activation; conditions of each exercise must be analyzed without generalizing them.^6,46 Elastic tubes can provide greater symmetry during shoulder abduction (P = 0.03) without sex differences (P = 0.30) than other upper-limb exercises, which require greater balance as they involve more stabilizers.^47,48 Accordingly, functional capacity can be improved with different equipment while preventing potential injuries due to overload by minimizing unnecessary compensatory motions.^11,49 In any event, when returning to the initial position of the elastic band, smooth eccentric motion requires good postural control. Thus, experience was fundamental in controlling this equipment (P < 0.001). With both types of equipment, muscle activation showed matching synergies, without significant differences in biceps (P = 0.10), shoulder (P = 0.54), or CT (P = 0.55). The elastic tubes generated a significantly higher (P > 0.001) force than the dumbbells only in the dorsal trapezius. In both sexes, as experience increases, so do synergies, with an increase of 0.16 ± 0.06 points per year. Experience positively impacts exercises demanding a high range of motion with muscle compensation, changing the role depending on the context.^50,51

Load equivalence between kilograms and elastic tube colors was found in only 2 exercises performed by men: elbow flexion (BB) and shoulder abduction (AD). The results showed equivalence between the 2 devices in both sexes, with similar agonist muscle activation, exercise symmetry, and synergies. These findings enabled us to indiscriminately select any of this equipment for compensated and precise muscle work. However, their load equivalences cannot be generalized, other than for specific exercises and in men.

Limitations

Among the limitations that can be described are the need to have a sample of different age ranges to observe the effect of each material according to its increase or to introduce a greater number of exercises. ⁵² Other elastic devices, such as flat or closed bands, could be added to achieve a more global view. On the other hand, lower-limb or global exercises, which are very common in physical activity programs, could be included. In addition, other anthropometric variables could have been included that could have provided more information to the study, such as the calculation of wingspan or the symmetry between the size of the upper and lower limbs.