Neural, biomechanical, and physiological factors involved in sex-related differences in the maximal rate of isometric torque development

The rate of tension development has received increased attention as a critical aspect of dynamic muscle performance during activities of daily living (e.g., balance maintenance) and sport performance (Aagaard et al. 2002; LaRoche et al. 2010; Paasuke et al. 2001; Pijnapples et al. 2008; Schultz et al. 1997; Tillin et al. 2013). A number of studies have shown that there is an inextricable link between the ultimate strength of the muscle and its rate of tension development (Andersen and Aagaard 2006; Holtermann et al. 2007). However, neural factors can also play an important role, as training-related increases in the maximal rate of tension development are associated with an increase in muscle activation at the onset of contraction (Van Cutsem et al. 1998; Van Cutsem and Duchateau 2005). Inglis et al. (2013) recently showed that neural factors also play a role in sex-differences in the maximal rate of tension development in the upper limb. Maximum strength was used as a covariate and was only able to account for a portion of the sex-differences. The addition of a second variable, the maximum rate of electromyographic (EMG) activity at the onset of contraction (Q₃₀), was able to eliminate statistically significant sex-differences.

Given the inherent relationship between muscle strength and the maximum rate of tension development, few studies have explored additional mechanisms that would explain sex-differences in the maximal rate of tension development (Aagaard et al. 2002; Andersen and Aagaard 2006; Folland et al. 2014; Van Cutsem et al. 1998). However, sex-differences in lower extremity musculoskeletal injury rates (DiStefano et al. 2015) and falls incidences (Hess and Woollacott 2005; Stevens and Sogolow 2005) may be linked back to this critical aspect of muscle contraction (Bento et al. 2010). Hannah et al. (2012) explored potential neural and biomechanical factors involved in sex-differences in the maximal rate of tension development in the quadriceps, in addition to maximum strength. Twitch properties, electromechanical delay, and muscle activation using surface EMG (sEMG) were also assessed. Consistent with the more general findings for the relationship between maximum strength and the rate of tension development, when the peak rate of tension development was normalized with respect to maximum isometric strength of the muscle, the sex-differences were completely eliminated.

The results by Hannah et al. (2012) in the lower limb are contrary to the finding of Inglis et al. (2013) who showed a role for neural factors in the upper limb. It may be hypothesized that since males and females were more comparable in absolute strength in the quadriceps (Δ33%) than the biceps (Δ55.5%; Inglis et al. 2013), strength may entirely explain sex-differences in the rate of tension development in the lower limb. In contrast, the difference in maximum strength between males and females in the upper limb observed by Inglis et al. (2013) was much greater (Δ55.5%), possibly allowing for additional factors, such as the rate of muscle activation, to play a role. Inglis et al. (2013) explored the possibility that the interpretation of the results may differ based on normalization to maximum voluntary strength versus the use of a covariate approach. It was found that normalization to maximum voluntary strength also failed to account for all sex-differences in the maximal rate of tension development.

Since there is a strong relationship between maximum strength and the rate of tension development, sex-differences may be accounted for by maximum strength only when the two groups are ‘more’ comparable with respect to maximal strength as exists in the lower limb compared with the upper limb. For example, it has been shown that males and females are more comparable in maximal isometric dorsiflexion strength (Δ28.9%) and identical with respect to root-mean-square (RMS) sEMG magnitude (Heyward et al. 1986; Hoffman et al. 1979; Lenhardt et al. 2009). Unfortunately, Lenhardt et al. 2009 did not assess the maximum rate of torque development. The 28.9% strength difference between the sexes is consistent with Holmbäck et al. (2003) who concluded that muscle cross-sectional area was the principal determinant of dorsiflexion strength. In general, the muscle cross-sectional area of the TA in males is only 20% larger than that for females (Holmbäck et al. 2003; Jaworowski et al. 2002).

The purpose of this paper was to determine if sex-differences in the maximum rate of isometric dorsiflexion torque development are determined by maximum isometric dorsiflexion torque alone or other factors as observed for the upper limb where the strength differences are more pronounced. Based on the work of Hannah et al. (2012), it was hypothesized that maximum isometric strength would account for sex-differences in the maximum rate of torque development in the TA, because males and females are more comparable in maximum isometric strength in the lower limb. Studying mechanisms behind sex-differences in distal muscles that are responsible for balance and explosive activity can guide specific training interventions or rehabilitation techniques.

The main finding of the current work was that maximum torque was able to account for almost all of the sex-related differences in the maximum rate of torque development, as the percent difference in least-square means was reduced to 1.2% when using maximum torque as a covariate. There also were no significant differences between males and females in sEMG RMS amplitude magnitude as observed by Lenhardt et al. (2009). The absence of sex-differences in the sEMG signal amplitude, Vpp, and CT may reflect a comparable number of hours per week of training for the two groups (Aagaard et al. 2002). Contraction time was used in the study to determine if differences in the maximal rate of torque development were associated with either descending voluntary control or peripheral factors associated with muscle composition (Close 1972). Based on the fact, no sex-related differences in the CT (0.01%) were found suggests that the two groups were also similar with respect to muscle fiber composition, which is consistent with other research (Hicks and McCartney 1996).

Based on the work of Lenhardt et al. (2009), it was expected that males and females would be more comparable in maximum strength in the lower limb than the upper limb, which is consistent with other studies on sex-differences in maximum strength (Heyward et al. 1986; Hoffman et al. 1979). Instead, a 50.6% difference was observed, which was the same order of magnitude as previously observed for the upper limb (Inglis et al. 2013). One reason may be related to differences in the cross-sectional area (CSA) of the TA in the present sample versus that of Lenhardt et al. (2009). Although CSA was not directly measured, the males and females in this study had a larger difference in lower leg girth (8.4%) than the 5.2% observed by Lenhardt et al. (2009), which may suggest a larger difference in TA CSA.

The large difference in maximum strength between males and females was associated with a comparably large difference in the maximum rate of isometric dorsiflexion torque development (44.6%). Greater co-activation of the antagonist muscle group in females could contribute to both lower maximum strength and rate of torque development. Macaluso et al. (2002) showed that females may use greater antagonist co-activation to stabilize the joint due to greater joint laxity, smaller agonist musculature, and potentially lower quality muscle mass (torque/CSA), as has been seen in older versus younger adults (Solomonow et al. 1988; Thelen et al. 1996). However, antagonist co-activation was not measured in this study as it is markedly lower during isometric contractions compared with dynamic contractions and particularly lower when the muscle is placed in a shortened position as in this study (Pasquet et al. 2006).

There were several novel findings in this study. First, there were no significant differences between males and females with respect to MFCV. This result may be due to comparable muscle fiber diameters between the sexes, which is a large determinant of MFCV (Lange et al. 2002; Lindstrom and Magnusson 1977; Merletti et al. 1995; Nishihara et al. 2005; Zwarts 1989). It has been shown that type I (slow twitch) fibers are situated predominately in the anterior portion of the TA (Henriksson-Larsén et al. 1983). Moreover, it is a general result that, while males have larger fiber diameters than females, the type I muscle fiber diameters in females are larger than their type II muscle fibers. The small interelectrode distance (5 mm) will record from a small pick-up volume that encapsulates these superficial type I fibers of the TA, which are comparable in muscle fiber diameter between the sexes.

A second novel finding is that there were pronounced sex-differences in Q₃₀ that were not evident in either the voluntary (RMS) or evoked (Vpp) sEMG magnitude. Females were significantly greater than males with respect to the rate of muscle activation as assessed by Q₃₀. The difference increased further (4%) with repeated testing. This change was also associated with a significant training-related reduction in voluntary EMD associated with repeated testing. This reduction in voluntary EMG but “not” evoked EMG, and highlights the change in neural control (Gabriel and Boucher 1998). We suggest that the greater Q₃₀ and reduction in voluntary EMD for females were associated with a different motor unit activity pattern at the onset of muscle contraction. However, it is not possible to distinguish exact motor unit behavior that may be responsible for the sex-related differences in the rate of increase in surface EMG. Van Cutsem et al. (1998) established a link between the maximum rate of isometric dorsiflexion torque, the rate of increase in sEMG, and the incidence of doublets associated with progressive resistive training. Similarly, Gabriel et al. (2001)found that only three training sessions were sufficient to produce an increase in the maximum rate of isometric elbow extension torque development and mean spike frequency of the sEMG signal during the torque development phase of the contraction in female participants.

Bojsen-Moller et al. (2005) reported a positive correlation between the rate of torque development and tendon structure stiffness, indicating that 30% of the variation in torque development can be accounted for by the tendons mechanical properties. Winter and Brookes (1991) further hypothesized that joint laxity might also play a role in the tension development phase of the contraction as reflected in the EMD. We believe that the greater rate of increase in sEMG activation for females reflects a compensatory mechanism to transmit force to the tendon more effectively, as evident in the observed decrease in voluntary EMD (Kubo et al. 2003; Wilkerson and Mason 2000; Winter and Brookes 1991). In support of this idea, Rozzi et al. (1999)have shown that females exhibited greater integrated sEMG activity upon landing from a jump as a compensatory mechanism for greater knee joint laxity, which includes both the musculotendinous unit and ligamentous restraint.

sEMG can only provide an indirect measurement of differences and underlying changes in motor unit activity patterns. Similarly, the use of CT to look at fiber composition differences only allows the association of either similarities or differences. Although the possible influence of different muscle structures and sizes in males versus females was discussed in the paper, not directly measuring it with ultrasound is a limitation. It was also assumed that there were differences in TA tendon stiffness without having actually measured it. Moreover, caution must be applied when extrapolating the results to older adults. The use of a healthy college aged population may not account for differences between the sexes with aging, such as the loss of type II fibers, which may render the sexes more similar in the 6th and 7th decades.

The results support the work of Hannah et al. (2012) as maximal strength of the TA as a covariate accounted for nearly all of the sex-differences in the maximum rate of torque development. This was true, even though the difference in maximal strength was quite pronounced (50.6%). Thus, the hypothesis that maximum strength accounts for sex-differences in the maximum rate of tension development in the lower limb due solely to the fact that males and females are more comparable in maximum strength than in the upper limb was not supported. Rather, the maximum strength accounted for sex-difference in the maximum rate of tension development despite a large discrepancy in strength between males and females. Furthermore, females had a greater rate of increase in sEMG activation, and exhibited a significant reduction in EMD with repeated testing, suggesting that they might utilize a different motor unit activity pattern at the onset of contraction. Future research may focus on training modalities and rehabilitation techniques that could optimize RTD rather than only focusing on maximum strength to assist an older population in balance maintenance and recovery from imbalance.