Relationship between intended force and actual force: comparison between athletes and non-athletes

Introduction

Intended force refers to muscle tension sensations experienced during contraction (Jones & Hunter, 1983). It results from a neuronal process influenced by two factors: the corollary discharge of the central motor command (i.e., the motor signal informs the sensory system about the planned or implemented action) and afferent feedback from the working muscles (Jones & Hunter, 1983; Pageaux, 2016). Indeed, peripheral inputs from the neuromuscular spindles and Golgi tendon organs provide information to higher centers about the in-place contraction (Carson, Riek & Shahbazpour, 2002; Proske & Allen, 2019). Moreover, the cerebellum ensures movement accuracy and sensory perception by communicating with the brain areas and sensory channels (Bhanpuri, Okamura & Bastian, 2012).

Historically, to the extent of standardizing the means of rating the perceived effort at specific muscular forces, scientific literature (Cooper et al., 1979; Jones & Hunter, 1982) focused on the relationship between actual and intended force in different muscle groups. The work by Cooper et al. (1979) suggested that a single motor performance, namely isometric or dynamic contraction, can be perceived with remarkable precision by both small (e.g., adductor pollicis) and large (e.g., quadriceps) muscle groups. However, a perceptual underestimation of force might occur in larger muscle groups, more commonly involved in gross movements, when lower levels of force are required to accomplish fine movements (Jones & Hunter, 1982).

More recently, the scientific interest moved to deepen the relationship between actual and intended force at different intensities of a maximal voluntary contraction. Indeed, in a motor task, the force intensity is influenced by the recruitment and firing rate of the motor unit and by the type of contraction (i.e., ballistic or tonic) (Miyamoto, Kizuka & Ono, 2020). Ballistic contractions are characterized by high firing rates, brief contraction times, and high rates of force development (Zehr & Sale, 1994). Conversely, tonic contractions are sustained muscle contractions to keep persistent muscular tension and generate a maintained force (Bagshaw, 1993). Previous studies pointed out conflicting results on the force intensities and the subjects’ accuracy in matching intended and actual force. In detail, West and colleagues showed that when subjects were asked to produce a specific force level during an isometric knee-extension task (i.e., without maximal voluntary contraction reference), they were more accurate at lower intensities. In contrast, they overestimated the intended force in high-intensity contractions (West et al., 2005). Similarly, Miyamoto, Kizuka & Ono (2020) showed that the actual force did not accurately match the target contraction without visual feedback, with a more precise performance at high rather than low intensities. To the best of our knowledge, none of the above-mentioned studies considered the subjects’ sports experience as a potential influencing variable in force modulation (i.e., ballistic and tonic) at different intensities (Baweja et al., 2009; Miyamoto, Kizuka & Ono, 2020). Indeed, samples were scarcely described without considering that physical activity level and sports experiences could have affected the results. In this regard, athletes are likely to manage better physiological function, muscle contractions, and sports demands than their sedentary counterparts (Thorstensson et al., 1977; Hagberg et al., 1988). Moreover, team sports practice helps athletes more easily govern motor patterns than sedentary individuals due to their higher ability to perform motor tasks under stressful conditions in open-skill contexts (Cortis et al., 2009). Therefore, the primary outcome of this study was to investigate whether athletes and sedentary individuals had similar accuracy in matching the intended and the actual force during ballistic and tonic contractions in an isometric leg extension task. The secondary outcome was understanding at first glance if athletes had a better self-perception of their force production than non-athletes. In detail, hypothesizing that sports practice could positively influence subjects’ ability to modulate force, we expected that the group of athletes could better exert force at the required percentage of the maximal effort with a greater self-perception of the force production.

Materials & Methods

Isometric lower-limb strength

We assessed the maximal isometric strength of the dominant quadriceps for each subject. To achieve this measure, we used a custom-built chair (Fig. 1A) and a load cell (MuscleLab™ 4100e; Ergotest Innovation, Porsgrunn, Norway). The load cell was placed three centimeters above the lateral malleolus of the dominant limb, perpendicular to the subject’s shank. Subjects had to sit maintaining the knee angle at 90 deg. Straps on chest and legs avoided any additional movement during the strength test. After a standardized warm-up (i.e., a 10-min walk on a treadmill at 5 km/h and 10 repetitions of half-squat exercise), each subject was familiarized with the experimental setup by performing ten sub-maximal isometric quadriceps contractions with real-time visual feedback on the level of the exerted force. Specifically, the familiarization protocol consisted of five contractions at a submaximal self-selected intensity for both BAL and TON conditions. Then, after a 5-minute rest, subjects performed the test without any external feedback. Subjects were required to exert, for each of the two contractions (i.e., BAL and TON), first the maximal voluntary isometric contraction (MVIC) and then the three submaximal intensities (i.e., 25%, 50%, and 75% of the MVIC). Regardless of the condition and the intensities, three consecutive trials were performed for each intensity. A recovery of 40 s was adopted among trials, while a 5-minute rest occurred between the two contraction modalities. In the BAL condition, the subjects were asked to reach the required submaximal force intensity as fast as possible (Fig. 1B). Conversely, in the TON condition, the subjects had to hit an accelerometer (MuscleLab™ 4100e; Ergotest Innovation, Porsgrunn, Norway) to define the moment from which their actual force was supposed to correspond to the intended force (Fig. 1C). Load cell and accelerometer signals were synchronously recorded at 100 Hz with the dedicated software (MuscleLab™ 4100e; Ergotest Innovation, Porsgrunn, Norway).

Results

All the subjects completed the study. Results of the WFQ-R questionnaire showed that thirty-nine subjects were right dominant and one was left dominant. Unpaired t-test for baseline comparisons showed statistically significant differences between NON-ATHL and ATHL groups for MET and MVIC of both BAL and TON contractions (Table 1).

Strength values expressed at 25%, 50%, and 75% MVIC in both BAL and TON contractions are presented in Table 2.

The AC results are presented in Fig. 2. The three-way ANOVA analysis showed a significant main effect of contractions (p < 0.001; F = 19.83; ${\eta }_{\mathrm{p}}^2$ = 0.34), force intensities (p < 0.001; F = 63.41; ${\eta }_{\mathrm{p}}^2$ = 0.63), and an interaction between them (p < 0.001; F = 14.10; ${\eta }_{\mathrm{p}}^2$ = 0.27). The Bonferroni post hoc comparisons are shown in Fig. 2 and highlight the significant differences among force intensities. A significant main effect of the group (p < 0.05; F = 4.58; ${\eta }_{\mathrm{p}}^2$ = 0.10) was also observed. The between-group (ATHL vs. NON-ATHL) differences in percentage showed a greater accuracy of the ATHL: (i) BAL contraction (25% MVIC: 3.79%, 50% MVIC: 33.54%, and 75% MVIC: 9.60%) and (ii) TON contraction (25% MVIC: 27.98%, 50% MVIC: 30.36%, and 75% MVIC: 26.76%).

The results for the aggregate Likert scores in the 25%, 50%, and 75% MVIC in both BAL and TON contractions are presented in Fig. 3. The three-way ANOVA analysis showed a significant main effect of contractions (p < 0.001; F = 48.85; ${\eta }_{\mathrm{p}}^2$ = 0.56) and a significant interaction between force intensities and groups (p < 0.05; F = 3.49; ${\eta }_{\mathrm{p}}^2$ = 0.08).

Discussion

This study aimed to investigate the relationship between intended and actual force, comparing athletes and sedentary individuals during BAL and TON contractions. The main finding confirmed our hypothesis, showing that the ATHL group could exert force at the required MVIC percentage more accurately than the NON-ATHL group, independently of the type of contraction. Since the recruitment and firing rate of the motor units and the contraction type could affect both force production (Enoka & Fuglevand, 2001) and the relationship between intended and actual force (Miyamoto, Kizuka & Ono, 2020), we can assume that sports practice could enhance muscle recruitment strategies (Enoka & Fuglevand, 2001) and the ability to modulate force.

Similarly, the higher levels of physical activity and training of ATHL account for the higher MVIC values. Therefore, the baseline results supported the hypothesis that the subjects’ sports practice could be an influencing variable in force production and claimed for group division when investigating the relationship between intended and actual force. Regarding the force values recorded in the present study, it is interesting to note that, under most conditions and independently from the group, the actual force tended to overestimate the intended force (Table 2). Indeed, this underproduction of force was in line with previous results performed on isometric knee extension (Cooper et al., 1979), isokinetic knee extension (Jackson et al., 2006), isometric elbow flexion and extension (John, Liu & Gregory, 2009), and isometric thumb adduction (Cooper et al., 1979). Although the causes of errors in estimating muscular effort have yet to be completely understood, some possible speculations to account for low levels of accuracy could be addressed. Indeed, it has been theorized that during voluntary motor task performance, descending motor command signals include physiological noise derived from neuronal and synaptic action potential discharges, leading to muscle force production variability (Harris & Wolpert, 1998). In detail, neuronal noise increases with the magnitude of the motor command (i.e., the intensity of a motor task) (John, Liu & Gregory, 2009). Moreover, transient post-contractile potentiation of spinal reflex pathways may summate with previously set motor commands, producing errors in effort perception (Hutton, Enoka & Suzuki, 1984).

Even though the actual force most undershot the intended force, the deviation from the actual force was less marked in the ATHL than in the NON-ATHL group. It is generally accepted that force modulation is accomplished by combining the rate coding of individual motor units and the recruitment of more or fewer motor units (Clamann, 1993). Moreover, muscle adaptations (e.g., following training) at both neural and morphological levels underpin the enhanced ability of force regulation in skeletal muscle (Folland & Williams, 2007; Carroll et al., 2009). Thus, chronic exposure to training and exercises with repetitive movements against loads for a long time, as athletes regularly do, could have substantially increased their ability to modulate force (Maden-Wilkinson et al., 2020).

In both NON-ATHL and ATHL groups, a better accuracy was observed at lower (i.e., 25% MVIC) than higher (50% and 75% MVIC) force intensities. Thus, our finding suggested that differences between intended and actual force appeared to be intensity-dependent. Previous studies showed that the presence of an anchoring process (i.e., MVIC as reference), as in our study, addressed results in different directions. West et al. (2005) reported that when the experimental task was performed before the MVIC, the actual force matched the intended force at the intermediate level (i.e., 50% MVIC) while without the anchoring process, the most accurate match occurred at 25% MVIC. In contrast with our results, other studies on precision grip (Kumar, Narayan & Chouinard, 1997), power grip (Kumar, Narayan & Chouinard, 1997), and chest press (Jackson & Dishman, 2000) showed that the actual force matched the intended force at intermediate level (i.e., 40–50% MVC).

In our study, the matching between the intended and actual force was lower at higher intensities (i.e., 50% and 75% MVIC). Thus, the intended force resulted in a more marked perceptual overestimation in these conditions. Hence, our results suggested that a perceptual mistake occurred at higher force intensities even with previous knowledge of the maximal effort (i.e., the MVIC) as an anchoring process. Even though our data cannot explain why subjects produced the most accurate match at the lower force intensity (25% MVIC), two possible speculations may be acknowledged. First, West et al. (2005) explained that subjects subconsciously under-produced force at higher intensities as a protective mechanism. Indeed, the perceptual overestimation occurred to maintain a reserve capacity and prevent mechanical and metabolic damages, which may arise in further force generation. Second, lower force percentages are used in everyday contexts, most for simple motor tasks (e.g., walking or getting up from a chair). Therefore, subjects could have performed a better perceptual estimation because of the proximity to the levels of force they are accustomed to, regardless of the group.

Our findings highlighted that the relationship between intended and actual force was more accurate in the BAL than in the TON condition. Since the BAL contraction occurred rapidly and without afferent sensory feedbacks (Vincken, Gielen & Van der Gon, 1984), it was mostly controlled by central predetermined regulatory mechanisms through a feedforward strategy (Hanneton et al., 1997). Conversely, TON contraction had a longer duration, and during this time, information from sensory receptors could interfere with the ongoing task (Desmurget & Grafton, 2000). Indeed, feedback control mechanisms are predominant during the skill acquisition process and essential for constructing a newly learned movement (Seidler, Noll & Thiers, 2004). In our case, the isometric contraction at leg extension represented a basic motor behavior, where corrective feedback used in complex motor skills was mainly unnecessary. In this regard, Lohse, Sherwood & Healy (2011) found that an internal focus of attention (i.e., focusing on muscle contraction) decreased accuracy in basic motor behavior (i.e., 30% MVIC) compared to an external focus of attention (i.e., focusing on pushing over the force platform). Similarly, subjects in the longer TON contraction could have internally directed their attention on muscles, calling into play counterproductive corrective feedback. This hypothesis could account for the worse accuracy in the TON than BAL contraction, where feedforward mechanisms were revealed to be more effective.

Finally, subjects’ self-perception, deepened through the Likert scale rating at the end of each contraction, highlighted no differences between groups. This result was unexpected since the ATHL group was more accurate in force production, implying a greater awareness of force modulation. However, the motor task studied in this research was a simple one. Thus, we can speculate that a more sport-oriented task could have differentiated the self-confidence performance between athletes and non-athletes. Moreover, a main effect of the contraction type was observed in the Likert aggregate scores, with a greater self-confidence towards the TON against the BAL contraction. Since the BAL contraction was the most accurate, we can speculate that the longer time the subjects had to modulate the force led to a perceptual mistake in their self-perceived accuracy in the TON condition. Nonetheless, further research should deepen the self-perception of force production, considering also the between-trial variability of the Likert scores.

The present study has some potential limitations to be acknowledged. Although both groups included male and female subjects, it was impossible to perform a sex analysis because the relatively small number of males and females in both groups would not have allowed sufficient statistical power. Then, athletes practiced different sports. Thus, recruiting athletes from similar sports with the predominant use of lower limbs could have enhanced the differences with non-athlete peers. Moreover, a dynamic or sport-oriented task could have improved accuracy and self-confidence, being closer to the infield practice.

Conclusions

The present study expanded the previous results on the relationship between the intended and the actual force by investigating whether sports practice could be an influencing factor. Our findings showed that athletes could exert force at the required MVIC percentage more accurately than non-athletes, independently of the type of contraction. Since the accuracy between intended and actual forces seems to decrease as a function of the increment of the force intensity, our results suggested that coaches and practitioners should focus on training the force accuracy at high intensity. In most situations, sports actions require fast movements and a high level of strength; thus, the accuracy at a high force level in ballistic movements is supposed to be an important factor in achieving performance. Indeed, repetitive training should aim to improve the accuracy of force production at high force levels. Differences in accuracy between ballistic and tonic contraction suggest different control mechanisms. Further work is needed to explore these hypotheses and to deepen the perceptual overestimation detected at the higher isometric force intensities.

Supplemental Information