Fatigue and perceptual responses of heavier- and lighter-load isolated lumbar extension resistance exercise in males and females

Experimental approach to the problem

The relationship between, load, sex, and fatigue response was examined across 3 testing conditions using an ILEX machine (MedX, Ocala, FL, USA). A fatigue response test (FRT) was performed for the following conditions; heavy load (HL; 80% MVC), light load (LL; 50% MVC) and control (CON; no training). The conditions were separated by no less than 72 h, with the condition type (HL, LL, CON) being randomized to minimize order effect.

Participants

Approval was granted from the University Health, Exercise, and Sport Science (HESS) ethics committee at the first authors’ institution (ID No. 687). Previous work from our group using knee extension revealed a between-condition effect size (ES) of 1.86 (between HL and LL). Sample estimate was based upon this to determine participant numbers (n) calculated using G * Power (Faul et al., 2007; Faul et al., 2009). The calculations showed that a minimum of four participants were necessary to meet the required power of 0.8 at an alpha level of p < 0.05 for comparison between loading conditions. However, recruitment for each sex group (males and females) was increased to n = 10 to provide greater power for secondary analysis of between-sex comparisons. Recreationally active asymptomatic males (n = 10) and females (n = 10) with no previous training experience of the lumbar extensors were recruited (see Table 1 for characteristics). Prior to testing, participants completed an informed consent and a physical activity readiness questionnaire (PARQ). For the purpose of this study, the exclusion criteria included: individuals who suffered from a heart condition, history of CLBP, any contraindications to exercise identified on the participant PARQ, or any knee or hip conditions which prevented the use of the machine restraints.

Testing procedures

Prior to testing, all participants attended a familiarization session where they were assessed for lumbar range of motion using a goniometer built in to the ILEX machine, and performed isometric testing using the ILEX machine every 12° beginning at 72° (full lumbar flexion) through 60°, 48°, 36°, 24°, 12°, and 0° (full lumbar extension) to allow them to experience the technique required and reduce any learning effect. Details of the full test protocol using the MedX ILEX machine and its restraint system have been documented elsewhere (Graves et al., 1990). In brief, participants were provided a specific dynamic warm-up of the lumbar extensors for ∼60 s using ∼27 kg for males and ∼20 kg for females. An MVC was then performed at seven different joint angles (as described above) by the participant gradually building force up to maximal effort over 3 s, and then relaxing over a further 3 s. Participants were provided with verbal encouragement to ensure maximal effort and were permitted ∼10 s recovery between testing angles. The ILEX machine and restraint system is shown in Fig. 1 and prevents rotation of the pelvis allowing training and testing of the lumbar extensors in isolation. The device has high test-retest reliability values of r = 0.81–0.97 in asymptomatic persons (Graves et al., 1990). An intra-class correlation coefficient (ICC) was conducted to assess pre-test MVC test-retest reliability at 72° between conditions (ICC = 0.931 (95% CI [0.845–0.972])) supporting the use of MVC for fatigue response testing in our laboratory.

Fatigue response testing

Following the familiarisation session, participants were invited to return to the laboratory on three separate occasions (with not less than 72 h between) where they were assigned each testing condition (HL, LL, and CON) in a computer generated randomised order. The FRT required participants to complete the previously described testing procedure (pre-MVC) and then complete one of three conditions; a single set of ILEX exercise with a load equating to 80% of the maximum torque (HL), a single set of ILEX exercise with a load equating to 50% of the maximum torque (LL) or a control condition where the participants performed no exercise. Exercise was performed at 2 s concentric: 4 s eccentric (2:4) repetition duration to ensure standardisation with visual time feedback on a display in front of the participant. The ILEX machine also provides an audible sound at the completion of each phase of the repetition (e.g., at full flexion and full extension) to ensure the full range of motion is performed for each repetition. As fatigue incurred, the repetition duration generally increased but participants were encourage never to move faster than the predetermined repetition duration. The exercise was ceased when, despite their maximum effort, participants could not complete the concentric phase of a repetition (e.g., momentary failure; see Steele et al. (2017a) for a detailed description). Immediately (<10 s) following cessation of the dynamic exercise the participant repeated the above detailed isometric testing procedures (post-MVC) to assess the decrement in force production incurred as a result of the dynamic exercise. FRT testing using the ILEX machine has previously been reported elsewhere (Edinborough, Fisher & Steele, 2016).

As a result of pilot testing, a difference was identified for the time-under-load (TUL) between the HL and LL conditions. As such to ensure parity in time interval (∼3 min) between the pre- and post-MVC testing following the HL, LL, and CON conditions, the dynamic exercise was delayed by 120 and 30 s for the HL and LL conditions, respectively. This rest interval was applied between the pre-MVC and exercise condition so that the post MVC could be performed immediately after the exercise condition. For the CON condition participants remained seated in the ILEX machine for 3 min before completing the post-MVC.

Immediately following each condition (but before the post-MVC), each participant was asked to report a rating of perceived exertion for effort (RPE-E) and discomfort (RPE-D) using 0–10 scales that permitted appropriate differentiation of the 2 perceptions (Fisher, Farrow & Steele, 2017; Fisher, Ironside & Steele, 2017; Steele et al., 2017b).

Statistical analyses

Strength was considered as peak MVC which occurred at 72° for all participants. Fatigue was considered as the decrease in MVC as a result of the training condition in both absolute (post-MVC Nm—pre-MVC Nm) and relative units ([post-MVC Nm/pre-MVC Nm] × 100). The independent variable considered was the exercise condition (HL, LL, and CON), as well as sex (male and female), and the dependent variables included pre-MVC (the MVC prior to each condition), absolute and relative fatigue, time under load (TUL), and RPE-D. Analysis was not performed for RPE-E since all participants reported maximal effort (i.e., 10) at the cessation of dynamic exercise at both HL and LL.

Shapiro–Wilk test was conducted to examine whether data met assumptions of normality of distribution and Mauchly’s test was used to examine assumptions of sphericity for repeated measures. For variables collected across all three conditions (pre-MVC, absolute, and relative fatigue), data met assumptions of normality and sphericity and thus a 3 × 2 repeated measures analysis of variance (ANOVA) was used to compare within participants across the within participant factor ‘condition’ and between participant factor ‘sex’. A 2 × 2 repeated measures ANOVA was used in the case of TUL as it was only measured during HL and LL conditions. Post hoc pairwise comparisons with a Bonferroni correction were used for significant effects by ‘condition’, ‘group’, or ‘condition × group’. RPE-D data did not meet assumptions of normality of distribution and thus a Wilcoxon test was conducted to examine differences between conditions. Mann–Whitney U tests were used to compare dependent variables between sexes.

Further, within participant ESs using Cohen’s d (Cohen, 1992) were calculated for absolute and relative fatigue for each condition (d = μ_change∕σ_change), where an ES of 0.20–0.49 was considered as small, 0.50–0.79 as moderate and ≥0.80 as large. All statistical analyses were performed using IBM SPSS Statistics for Windows (version 23; IBM Corp., Portsmouth, Hampshire, UK) and p < 0.05 set as the limit for statistical significance.

Acute fatigue

While similarly to our previous work (Fisher, Farrow & Steele, 2017) LL conditions resulted in both a higher absolute, as well as relative fatigue. Our analyses also revealed that males and females differ in their fatigability in the lumbar extensors. Significantly greater absolute fatigue was incurred by males performing exercise with the same relative load although, whilst still greater compared with females, differences in relative fatigue were not significant; possibly as a result of a type II error as previous research has supported sex differences in relative fatigue (Gentil et al., 2017). In the HL condition males showed a mean decrement in force production of 77.1 Nm (∼21%) compared to 32.6 Nm (∼11%) for females, whereas following the LL condition males showed a mean decrease of 128.8 Nm (∼33%) compared to 70.7 Nm (∼26%) for females. Correspondingly, endurance capacity appears to differ between males and females, as time-under-load (time to task failure) was shorter for males than females in the LL condition (mean = 133.3 vs. 174.0 s, respectively), but not the HL condition (mean = 57.7 vs. 62.4 s, respectively). The present data suggests that, for the lumbar extensors, when exercising to momentary failure using a HL (80% MVC) males and females show similar time to task failure but show a large disparity in their absolute and relative fatigue following exercise. In contrast, at a LL (50% MVC) females showed greater endurance capacity by being able to continue exercising for ∼41 s (∼24%) longer than males, though showed a more similar decrement in force production (males = 33%, females = 26%).

The present findings agree with the body of literature suggesting women show a smaller decrement in maximal force production compared to males, following fatiguing exercise (Hunter, 2016). The greater TUL in females compared to males is also supported by studies showing sex-based discrepancies in endurance capacity during muscular contraction (Maughan et al., 1986; Clark et al., 2003; Gentil et al., 2017). However, where previous data suggested a similar number of repetitions prior to task failure between males and females for dynamic TEX exercise at 50%MVC (Clark et al., 2003), our data suggests that where the lumbar extensors are considered in isolation males show a shorter time to failure, and greater degree of fatigue.

Our data also adds to our understanding of fatigue at HL exercise. Maughan et al. (1986) showed that males and females show similar time to task failure at high-force or HL exercise (80% MVC for isometric knee extension; 80% and 90% 1-repetition maximum for dynamic elbow flexion). Whilst our data supports this regarding time to task failure (57.7 and 62.4 s for males and females, respectively) we have shown that, despite the similar duration of exercise, males incurred a greater absolute and relative degree of fatigue.

Differences between males and females for degree of fatigue and time to failure has been hypothesised to relate to differences in muscle fibre type (Bajek et al., 2000). Type I muscle fibres have a greater oxidative capacity which makes them markedly more fatigue resistant allowing contractions to be sustained over a longer period but of lower force output than type II muscle fibres. Indeed, our results revealed significant between sex differences in pre-MVC—supportive of greater strength in males compared to females. In addition, because the cross-bridge cycling is faster in type II muscle fibres, utilisation of adenosine triphosphate (ATP) occurs more rapidly than in type I muscle fibres (Westerblad, Bruton & Katz, 2010). Any disparity between HL and LL might be explained by the location of the onset of fatigue in relation to the α-motor neuron. Research has suggested that LL activities likely incur elevated central motor output and as such incur higher levels of fatigue and greater reductions in maximal torque. With HL exercise fatigue appears to develop progressively and exercise cessation occurs as a result of centrally mediated factors (a decrease in the number and discharge of motor units). In contrast, fatigue as a result of LL activities appears to occur peripherally as a result of metabolic changes within the muscle preventing transmission of muscle action potentials (Boyas & Guével, 2011; Gandevia, 2001). Research has suggested women have either a greater relative percentage of type I fibres compared to males (Bajek et al., 2000), or have a similar number but show a larger area of type I than type II muscle fibres (70–75% compared to 54–58%) when compared to males (Thorstensson & Carlson, 1987; Mannion et al., 1997). The lumbar extensors are postural muscles and, as such, it would be expected that type I fibres predominate in this group due to a need for sustained or repeated lower force actions. However, this disparity in relative percentage and/or size of type I fibres might explain the greater fatigue resistance in females compared to males observed in the present study, both in strength, the degree of fatigue incurred at both HL and LL, and the time to task failure in LL. However, we should acknowledge that the present study was not intended to assess muscle fibre type disparity or causal factors of fatigue, and as such, we only speculate on potential underpinning mechanisms.

From an applied perspective it is important to recognise the differences in performance and fatigue decrement between males and females. Our data suggests that females do not incur the same degree of fatigue as males following similar exercise protocols, and indeed that females might be able to sustain longer exercise duration at the same relative loads when using low, but not high loads. As such females should manipulate training variables accordingly when training with high loads, performing either greater repetitions at the same relative load, or using heavier relative loads than males. Furthermore, if females do not show the same performance decrement then it might be feasible that females can train at a higher frequency than males to optimise adaptation. However, it should be noted that in the present paper we have only examined the acute fatigue response. Previous research using a larger training volume (four sets of 10-repetitions @10-repetition maximum, followed by four sets of 10 repetitions @ 80% of 10-repetition maximum) resulted in a similar reduction in peak torque of the elbow flexors between males and females. However, data suggested a longer time to return to baseline torque for females compared to males (Flores et al., 2011). Whilst the relevance to the present study is limited since there is considerable disparity in exercise volume between the studies, future research might consider recovery from fatiguing exercise between males and females using different exercise volumes and across different muscle groups.

Effort and discomfort

No analyses were performed for RPE-E since all participants reported maximal values as a result of training to momentary failure (Steele et al., 2017a). This is supportive of previous research which has shown that effort and discomfort can be distinguished independently (Fisher, Ironside & Steele, 2017; Steele et al., 2017b). Performing repetitions to momentary failure in the LL condition induced greater discomfort compared to the HL condition. This is also in accordance with research considering the knee extensors where effort and discomfort have been assessed independently (Fisher, Ironside & Steele, 2017). Our analyses also showed no sex-based differences for RPE-D (p > 0.05). Previous research has reported varying but submaximal values for effort when assessed alone (Shimano et al., 2006). However, it has previously been discussed that where exercise is continued to the point of momentary failure effort should be maximal irrespective of load, time-under-load, repetitions completed prior to task failure, etc., and that variation in perceptual responses are likely a product of discomfort (Smirnaul, 2012; Steele et al., 2017b). In this sense it is important to be able to anchor effort using the RPE-E and independently assess discomfort (RPE-D), and vice-versa, as we have done so in the present article.

As stated, when performing LL resistance exercise momentary failure appears to occur as a result of peripheral fatigue (Boyas & Guével, 2011; Gandevia, 2001). This prolonged ATP production incurs greater metabolite accumulation in the form of increased inorganic phosphate (P_i) accumulation of hydrogen ions (H⁺) and thus decreased intramuscular pH (Schott, McCully & Rutherford, 1995; MacDougall et al., 1999; Takada et al., 2012). Perceptions of discomfort appear to be linked to afferent feedback and thus a reason why increased metabolic stress may promote greater discomfort during LL resistance exercise (Marcora, 2009). Contrastingly, perceptions of effort are likely related to central motor output (De Morree, Klein & Marcora, 2012) which may explain the similarly maximal RPE-E that participants gave for both LL and HL when performed to momentary failure. Indeed, it has been argued that both LL and HL likely result in similar numbers of motor units being recruited, albeit with differing recruitment strategies (Fisher, Steele & Smith, 2017), and thus likely a similar central motor command being required upon achieving momentary failure.

However, although higher discomfort levels were reported following the LL condition, we can again only speculate upon the potential influences of central or peripheral response. The results from our study may however be useful in a practical sense. Since greater discomfort was incurred in alignment with greater decrements in MVC as a result of the LL condition, and knowing that similar adaptations in muscular strength can occur using both heavier- and lighter-loads of resistance (Assunção et al., 2016; Schoenfeld et al., 2017; Fisher, Ironside & Steele, 2017; Fisher, Steele & Smith, 2017), individuals may be more inclined to exercise using a heavier load since it appears to incur a lower level of discomfort.

Limitations

The present study adds to the dearth of literature considering female participants and comparing sex-based fatigue. However, it is important to recognise any potential limitations within the present study that might be considered for future research. Whilst a power analysis confirmed adequate sample size for between condition comparisons, only 17 participants completed the study and as such between sex comparisons may have been underpowered (possibly explaining the lack of significant interaction effects for relative fatigue; indeed, post hoc observed β for this was 0.318). Future work might consider the typical effect size differences between sexes in determining sample estimates, and also might consider what constitutes a clinically meaningful difference. Further, our results may be limited in application to persons with the characteristics that conformed to our inclusion criteria. Whilst we have discussed the efficacy of strengthening the lumbar extensors to reduce CLBP there were no participants involved in the study with the condition. As such it would be unreasonable to assume these findings would be equated with this particular population, therefore it may be beneficial to replicate this study using both CLBP symptomatic and asymptomatic individuals.