This study investigated BLD in voluntary and electrically-evoked explosive contractions of the knee extensors and considered the contribution of agonist neuromuscular activation and measurement issues to any BLD. We observed a BLD in voluntary explosive force/RFD but not MVF. The BLD in explosive force occurred at 100 ms only and reflected a BLD specific to RFD50100. BLD measurement issues made only a minor contribution to the observed BLD and thus these results support an underlying physiological mechanism explaining BLD. However, the fact that we observed a BLD in evoked force production and no change in EMG during explosive voluntary efforts suggests the BLD was not solely attributable to reduced agonist or antagonist neural drive.
The finding of no BLD in MVF is consistent with numerous reports (e.g. 
), but in contrast to an equal number that have shown a BLD in knee extensor MVF (e.g. 
). As there was no BLD in MVF, it is unsurprising that there was no difference in agonist or antagonist activation, evoked peak force measures with high force values, or influence of methodological factors. This is in accordance with previous findings of no BLD or mechanistic differences between BL and UL MVCs 
Despite no BLD for MVF, we observed a BLD in explosive force of 11.2% during these single joint voluntary contractions. The BLD was specific to F100
, but there was a tendency for a BLD in F150
. Furthermore, there was a 14.9% BLD for RFD50100
, with no BLD for RFD050
. This is the first study to investigate the possibility of a BLD in explosive strength by analysing force/RFD throughout the rising force-time curve. Previously, only pRFD had been assessed in this context, with BLD reported to range from 0–20% 
. The mechanisms for the observed BLD in explosive force could have been due to measurement issues in the comparison of UL and BL performance, neuromuscular activation of agonist, antagonist muscles that were assessed in this study, or even activation of stabiliser muscles that we did not assess.
The assessment of single limb performance during BL contractions allowed for the delineation of measurement artefacts that may have contributed to any observed BLD. Although, the BLUL measure reported only a tendency for a difference to UL for F100, there was a difference for RFD50100, confirming a BLD due to a physiological effect exclusive of measurement issues. There were also no differences in explosive or maximal force/RFD between the two BL measures, indicating measurement artefacts played only a minor role in the observed BLD. Surprisingly, the onset of force discrepancy between the two limbs during BL contractions was relatively small (3.2 ms), which suggests that neuromuscular system is capable of near simultaneous activation of the knee extensor muscles of both legs during BL actions.
The current study found no differences in agonist EMG between UL and BL explosive contractions. This is despite the widely suggested mechanism for BLD being a reduction in neural drive to the agonist muscles. Our findings support previous research demonstrating a BLD in RFD in the absence of a change in agonist EMG 
. It is important to note that the sensitivity of EMG for assessing BLD has been questioned 
. However, in the present study we normalised the EMG amplitude to Mmax
, which would be expected to increase the effect size and power of statistical comparisons between the conditions. Additionally, we averaged across three quadriceps muscles and across the best three contractions during the explosive efforts. These methods would be expected to improve the reliability and sensitivity of the EMG measurements. Furthermore, we also measured neural efficacy, which assesses agonist neuromuscular activation during the initial phase of the contractions (50 ms), and provided further evidence that agonist activation was not different during the early phase of UL and BL explosive contractions. These findings suggest that the observed BLD in RFD was not attributable to agonist activation, and indicates a role for an alternative mechanism.
Agonist and antagonist activation contribute simultaneously to net joint torque and thus the level of co-activation could account for any BLD. This is the first study to assess if antagonist activation influenced the BLD during explosive force production, and found that the observed BLD in RFD was not attributable to antagonist activation. A possible remaining explanation concerns stabiliser activation.
BL evoked contractions were utilised within the present study to help establish if the BLD was influenced by a physiological mechanism(s) exclusive of neural drive to the agonist muscles. Interestingly, there was a BLD in evoked force production, which occurred in both twitch and octet F50
(8.7 and 6.3%, respectively), and twitch PF (9.0%) and was of a similar magnitude to the observed declines in explosive voluntary force/RFD (8.6–14.9%). This is the first study to investigate a potential BLD in evoked force production and provides further support to the notion that the BLD in voluntary explosive force production was due to mechanisms other than agonist neural drive. A possible explanation for the BLD in both evoked and voluntary force is a difference in postural stability/stabiliser activation requirements during UL and BL actions. Stabiliser activation was not measured within the present study, but is thought to be important for optimal force expression 
. For instance, Nozaki et al. 
demonstrated that even during a relatively simple task such as an isometric knee extension used within the current study, that there was a large variation, both between and within-participants in the ability to stabilise the adjacent joint torque through effective inter-muscular coordination. The greater postural requirement for BL than UL strength tasks has been proposed as the mechanism accounting for the BLD in MVF 
. In support of this suggestion, the BLD has been observed to be higher in an action requiring greater activation of postural stabilising muscles (leg press versus hand grip, 
). In the current study insufficient stabilisation during BL explosive contractions may have afforded greater movement of adjacent joints, particularly the hips, increasing biological compliance and reducing explosive force production. Whilst the BLD in evoked explosive force we have observed might appear to contradict this possibility (as only the agonists are activated by the stimulation), there is undoubtedly stabiliser activation in anticipation of, and/or in response to, the stimulation, and this could be similarly less effective in the BL compared to UL situation. The similarity of MVF across BL and UL contractions might also argue against a role of stabiliser activation in the BLD we have observed, however, during these longer contractions force production is unlikely to be influenced by compliance and hence stabilisation. Future research should consider the role of stabiliser muscle activation in the BLD. The observed 15% deficit in RFD, despite no influence of BL actions on MVF has important implications for sport and exercise training science and suggests specific training to offset this deficit should be performed in order to maximise the performance of BL explosive sporting tasks. The observed deficit may have been explained by reduced inter-muscular coordination (lower stabiliser activation) during BL efforts and suggests that specific practice of coordinated explosive BL tasks and improved core/joint stability could be expected to improve the expression of BL explosive sporting tasks through reducing this explosive force/RFD BLD.
In summary, there was a BLD in explosive but not MVF of the knee extensors, which was specific to RFD50100. Measurement artefacts not previously considered were shown to play only a minor role on the observed BLD confirming a BLD due to a physiological effect. The novel finding of a BLD in evoked force production and no change in agonist or antagonist EMG during explosive voluntary efforts suggest the BLD in voluntary explosive force may be attributable to changes in stabiliser activation.