Introduction

Muscle power is the maximum strength that can be generated over a short period of time (1). During aging, muscle power declines earlier and to a greater extent than other physical performance parameters (e.g., muscle strength) (2–5). In terms of sex-specific changes, it is observed that women experience more significant declines in muscle power compared to men (3, 6). In older adults, muscle power is a strong, independent predictor of mobility disability, poor functional status, hospitalization, and death (3–5, 7). Comparative studies have found that this physical capacity is a better indicator of musculoskeletal health (3), physical performance (4), mobility (5), functional independence (8), and death (7) than muscle strength.

Measurement of muscle power entails the use of complex equipment which impacts its clinical applicability. More recently, equations to estimate lower-limb muscle power measures based on 5-repetition sit-to-stand (5STS) performance and anthropometric parameters were validated (8). Although several studies have found significant associations between these estimations of muscle power and health outcomes (8–10), not all investigations have confirmed their association with physical performance (7, 10). Moreover, some authors argued that 5STS-based muscle power equations should be refined due to the presence of conceptual problems in the operationalization of concentric movements (11).

Lower-limb muscle power impacts physical performance through its contribution to the production of more complex motor gestures. Indeed, muscle power might be associated with the propulsion required to climb stairs and walk fast (4) and the ability to rapidly respond to balance challenges (12, 13). Notably, these motor tasks demand different phases of muscle power, from maximal concentric performance to eccentric actions, which might not be entirely captured by the time taken to perform the 5STS test. As such, it may be proposed that the analysis of specific biomechanical aspects of 5STS performance could represent a valid discriminator of people with different physical function levels.

These premises are partially supported by prior empirical observations (14–17) that found significant associations between muscle power obtained during a single stand movement using a linear encoder and physical performance tests, including gait speed, handgrip strength, and 30s chair STS performance. In addition, this method of muscle power measurement can distinguish fallers from non-fallers (18).

To provide further and more detailed information on this topic, the present study explored and examined the biomechanical aspects of 5STS performance, and compared the capacity of this measure and 5STS muscle power equations to discriminate older women with high and low physical function levels.

Methods

This is an observational cross-sectional study that described the biomechanical aspects of 5STS performance in older women. We also compared the association between 5STS biomechanical aspects, and 5STS muscle power equations, with physical performance measures. This study was approved by the Research Ethics Committee of the University of Mogi das Cruzes (UMC, São Paulo, Brazil). All study procedures were conducted in compliance with the Declaration of Helsinki and the Resolution 196/96 of the National Health Council. The article was prepared according to the Strengthening the Reporting of Observational Studies in Epidemiology criteria (19).

Participant recruitment took place between January 2015 and January 2018. Participants were recruited in a community senior center located in the metropolitan area of São Paulo, Poá, Brazil. The study was advertised through posters placed in public sites (e.g., parks, city hall, public offices, bus stops, and train stations) as well as via local radio and newspapers. People were also invited to participate by direct contact. Candidates were considered eligible if they were 60 years or older, lived independently, and possessed sufficient physical and cognitive abilities to perform all measurements required by the protocol. No other selection criteria were set. Written informed consent was obtained before inclusion from each participant.

5-Time Sit-To-Stand Test (5STS)

5STS was performed twice and the best result was used for analysis. The test involved rising from a chair 5 times as quickly as possible with arms folded across the chest. Timing began when participants raised their buttocks off the chair and was stopped when they were seated at the end of the fifth stand. Time performance was quantified using a stopwatch (Vollo Sports, São Paulo, Brazil). A 50-Hz linear encoder (Peak Power, CEFISE, Brazil) was attached to the wrist of the arm that was at the waist to obtain absolute muscle power (AMP, W). Then, concentric (MPC1-5) and eccentric (MPE1-5) muscle power variables were created from each stand and sit movement. The duration of each concentric (DC1-5) and eccentric (DE1-5) movement was also obtained. The delta (Δ) of variation from the first concentric (ΔMPC2-5) and eccentric (ΔMPE2-5) movements was calculated. Compositive scores summing the duration of concentric (DC1-2, DC1-3, DC1-4) and eccentric variables (DE1-2, DE1-3, DE1-4) were also created.

Absolute (AMP_EQUA), relative (adjusted by body weight) (RMP), and allometric (adjusted by height) muscle power (ALMP) values were estimated according to the equations proposed by Alcazar et al. (11):

$$\matrix{{(1){\rm{Absolute\ muscle\ power (W)}}} \hfill = \hfill {{{{\rm{Body}}\,{\rm{weight}}\,{\rm{(kg)}} \times {\rm{0}}{\rm{.9}} \times g \times [{\rm{height}}\,{\rm{(m)}}\, \times \,0.5\, - \,{\rm{chair}}\,{\rm{height}}\,{\rm{(m)}}]} \over {[{{{\rm{SSTS}}\,{\rm{test time}}\,{\rm{(s)}}\,} \over {{\rm{no}}\,{\rm{of}}\,{\rm{STS}}\,{\rm{repretitions}}}}] \times 0.5}}} \hfill \cr {(2)\,{\rm{Relative}}\,{\rm{muscle}}\,{\rm{power}}\,{\rm{(W/kg)}}} \hfill = \hfill {{{{\rm{Absolute}}\,{\rm{muscle power (W)}}} \over {{\rm{Body}}\,{\rm{weight (kg)}}}}} \hfill \cr {(3)\,{\rm{Allometric}}\,{\rm{muscle}}\,{\rm{power}}\,{\rm{(W/}}{{\rm{m}}^2})} \hfill = \hfill {{{{\rm{Absolute}}\,{\rm{muscle power (W)}}} \over {{\rm{Height (}}{{\rm{m}}^2}{\rm{)}}}}} \hfill \cr}$$

Physical Performance Tests

Maximum isometric strength of metacarpophalangeal joints was measured by isometric handgrip strength (IHG) testing. Participants remained seated comfortably on a chair with their shoulders in a neutral position. The arm being assessed (dominant arm) had the elbow flexed at 90° near the torso, and the hand neutral with thumb up. A maximal contraction was performed over four seconds using a handheld hydraulic dynamometer (Jamar®, USA). The Timed “Up and Go!” (TUG) test involved getting up from a chair without the help of the arms, walking a distance of three meters around a marker placed on the floor, coming back to the same position, and sitting back on the chair. The test began when a researcher shouted a “go!” command. The stopwatch was activated when participants got up from the chair and was stopped when they were seated again. To measure walking speed (WS), a three-meter WS test was performed. Participants were required to walk a distance of five meters at their usual cadence. Before testing, both feet remained on the starting line. Measurement was initiated when a foot reached the one-meter line and was stopped when a foot reached the four-meter line. The one-meter intervals at the beginning and end were used to avoid early acceleration and/or deceleration.

Anthropometry

Body height and weight were measured through a stadiometer and an analog medical scale, respectively. The body mass index (BMI) was then calculated as the ratio between body weight (kg) and the square of height (m2).

Statistical Analysis

Normality of data was ascertained using the Kolmogorov-Smirnov test. Data are presented as mean (standard deviation). One-way repeated measures within-subject analysis (ANOVA) was conducted to examine differences in 5STS, AMP, and duration parameters among the repetitions and test the existence of a regression linear model. Participants were divided into high and low function performance groups according to the median results of TUG (13 s), WS (1.3 m/s), and IHG (28 kg) tests. Differences in 5STS muscle power equations and biomechanical aspects were examined through independent t-tests. Pearson’s correlation was conducted to examine the correlations between 5STS parameters and physical performance tests. Significance was set at 5% (P<0.05) for all tests. All analyses were performed using the SPSS software (version 23.0, SPSS Inc., Chicago, IL).

Results

The main characteristics of study participants are shown in Table 1. Older women were relatively young (67.4 ± 6.5 years) and had a mean BMI (30.1 ± 5.4 kg/m2) indicative of overweight/obesity. 5STS performance ranged from 4 to 22 s, with an average time of 13.2 s (± 3.2), suggesting that most participants had preserved physical performance. Results of the TUG (10.0 ± 14.2 s) and WS (1.1 ± 0.5 m/s) tests further supported the good physical status of study participants. Average IHG performance (26.9 ± 9.2 kg) was borderline to sarcopenia.

Table 1 Main characteristics of study participants (n= 78)
Full size table

AMP and contraction duration during concentric and eccentric phases across the 5STS test recorded using the linear encoder are shown in Figure 1. The peak of AMP during the concentric phase was reached in the first repetition and significantly declined thereafter. A linear trend of decline was observed (P= 0.006). No other significant differences were identified.

Figure 1
figure 1

Concentric and eccentric muscle power (1A and 1C) and duration (1B and 1D) during 5STS

*P<0.05 vs 1st repetition

Full size image

Differences in 5STS-based biomechanical aspects according to the performance on WS, TUG, and IHG tests are shown in Table 2 and supplementary material (SM) 1 and 2. No significant differences in AMP, RMP, or ALMP estimated according to 5STS equations were observed between participants with high and low physical performance. In contrast, participants with high or low WS had significant differences in MPC, MPC1, MPC2, MPC3, MPC4, MPC5, MPC1-2, MPC1-3, MPC1-4, DC1, DC3, DC1-2, DC1-3, DC1-4, MPE, MPE1, MPE5, MPE1-2, MPE1-3, and MPE1-4 (Table 2). When participants were compared according to their performance on TUG (SM1) and IHG (SM2), only differences in MPE3 and MPC5 reached significance.

Table 2 Main characteristics of study participants according to walking speed performance (n= 78)
Full size table

Results of Pearson’s correlations between 5STS-based biomechanical aspects and physical performance tests of sarcopenia are shown in Table 3. Aspects of muscle power recorded with the linear encoder and the duration of concentric and eccentric contractions were significantly correlated with IHG, TUG, and WS. Specifically, MPC, MPC2, MPC3, MPC4, MPC5, DC5, MPE, MPE2, and MPE5 were significantly associated with all physical performance tests. MPC1, DC1, DC2, DC3, DC4, DC5, DC1-2, DC1-3, DC1-4, MPE1, MPE3, MPE4, DE1, DE2, DE3, DE4, DE5, DE1-2, DE1-3, and DE1-4 were significantly associated with both TUG and WS, but not IHG. No parameters were exclusively associated with IHG, TUG or WS. Moreover, no significant correlations were observed between 5STS muscle power equations and physical performance measures.

Table 3 Pearson’s correlation between muscle power parameters and physical performance
Full size table

When results were adjusted according to BMI (SM3), all MPC, DC, and MPE (except for EC2) variables become significantly associated with physical performance tests. Furthermore, DE1, DE2, DE3, DE4, DE5, DE1-2, DE1-3, and DE1 remained significantly associated with TUG and WS. In contrast, a positive and significant associations was found between AMP and the time to perform the TUG test.

Discussion

The assessment of lower limb muscle power is an important aspect of the physical evaluation of older adults by providing relevant insights on physical status. This information, in turn, may be used by, contributing to health professionals in the identification of individuals at high risk of experiencing negative events, and to devise the best therapeutic approach. Yet, the clinical evaluation of muscle power is hampered by the lack of an unequivocal assessment tool. Our findings indicate that peak lower limb muscle power is reached in the first concentric contraction and declines significantly thereafter (Figure 1). Furthermore, we observed that 5STS-based biomechanical aspects demonstrate a good capacity to discriminate individuals according to WS performance (Table 2). We also noted that muscle power during specific contractions might contribute to differentiating older women with low and high performance on TUG (SM1) and IHG tests (SM2). Notably, such associations were not observed with muscle power estimated according to 5STS equations. Pearson’s correlation analysis suggested similar results (Table 3 and SM3).

The fact that peak lower-limb muscle power was reached during the first stand movement might suggest that one sit-to-stand action could be a sufficient representation of lower-limb muscle power in older adults. The capacity to produce power decreases in tasks requiring high velocities and is exacerbated during repeated contractions, including cyclic STS movements (20–22). Muscle power is primarily produced by the activity of type II muscle fibers, those more sensitive to fatigue (23–25). As muscles are required to continue producing strength at high velocities, the availability of ATP and phosphocreatine of type II muscle fibers becomes insufficient (24, 26, 27), leading fatigue-resistant fibers to become protagonists of muscle actions (24), causing significant reductions in power output (24, 25).

In the present study, AMP during the second concentric contraction (stand movement) was significantly reduced by 12.5% in comparison to the first muscle action. These abrupt and meaningful losses might represent structural and metabolic changes commonly observed in old muscles, mainly in type II muscle fibers, including relative reduction of these fiber types, reduced phosphorylation of the fast myosin regulatory light chain and rates of myosin force production, and longer myosin attachment times (25).

No significant differences were found in the duration or velocity of sit movements among eccentric contractions (Figure 1). However, visual inspection revealed a “W-shaped” curve during eccentric contractions, in which oscillations are noted between the 2nd-4th and 1st-3rd-5th repetitions. These observations suggest that older adults might have problems controlling eccentric actions performed after explosive moments. These premises are partially supported by clinical observations, given that frail individuals, mainly those institutionalized and hospitalized, have two major motor behaviors when they have to sit: i) extremely controlled movement and ii) fast movement due to incapacity to sustain body weight.

Eccentric actions involve a lower activation of motoneurons, when compared to concentric actions, and a specific recruitment of type II muscle fibers (25). It has been seen that eccentric muscle strength presents a virtually lesser decline with age in comparison to concentric and isometric muscle actions, leading many authors to argue that this physical capacity is preserved in advanced age (28). However, these conclusions are obtained after testing exclusively eccentric actions (28) and no studies have examined the impact of explosive movements in a subsequent capacity to generate eccentric strength. In one of the few studies examining this subject, it was found that fatigue induced significant decreases in eccentric muscle strength, mainly when it is required during high velocities (29).

Although this interpretation is only speculative, these results emphasize the need for more specific investigations to understand if the patterns of eccentric actions noted during 5STS are a result of type II muscle fibers’ overload, in which movements in the boundaries, with extreme caution and/or inability to sustain body weight, reveal the lack of strength to maintain muscle actions. Slow (controlled) eccentric movements produce worse performance on 5STS and are captured in test results. But then, a question that arises from this view is whether fast eccentric actions, and consequent better 5STS results, represent good performance or neuromuscular weakness.

Results indicated that 5STS-based biomechanical aspects were better discriminators and more associated with IHG, TUG, and WS performances than 5STS muscle power equations (Table 2, SM1, SM2, SM3). These findings are partially supported by other investigations that found no significant associations between muscle power estimated according to 5STS performance and WS (10). This scenario is surprising given that TUG and WS performances are dependent on muscle power (13, 30–32).

5STS-based muscle power equations have faced criticism due to their assumption that only a portion of the body’s mass accelerates during the concentric phase of movement, whereas the production of mechanical power in similar conditions should account for alterations in the kinetic and potential energy of the entire body (11). The authors also highlighted conceptual problems with the definition of the concentric phase (11). Notably, both 5STS-based muscle power equations aspects criticized by Fabrica & Biancardi (11) are associated with the concentric phase of the movement, in which more balance is likely required (33), affecting TUG performance, and is associated with the propulsion required to move fast in mobility tasks (4).

On the other hand, the analysis of individual biomechanical aspects captures specific moments of physical tasks (e.g., propulsion, dynamic balance), which explain the association among these parameters. The fact, for example, that significant differences in MPE3 were found between participants with high and low TUG performance suggests that eccentric actions might be a better discriminator of mobility than concentric and combined contractions. Moreover, Pearson’s correlation results indicate that numerous biomechanical 5STS parameters (e.g., MPC, MCP2, MCP3, MCP4, MCP5, DC5, MPE2) were significantly associated with IHG, TUG, and WS, suggesting that all these variables might contribute to identifying people with low physical function.

The fact that MPC, MPE, and DC variables becamesignificantly associated with physical performance tests when results were adjusted according to BMI, suggests an important role of body mass distribution in this scenario. Individuals with higher BMI need to produce more strength than those with lower BMI values to move the body rapidly during sit and stand movements. The inability to generate strength properly in these circumstances reflects in lower power production and longer contraction durations and might explain the associations observed in the adjusted analysis.

The present study is not free of limitations. First, specific muscle power, adjusted according to muscle mass (e.g., computer tomography, magnetic resonance imaging, or dual X-ray Absorptiometry), was not estimated in the present study. Second, participants were not screened for sarcopenia or frailty. Third, important information associated with the presence of chronic conditions, such as pharmacological therapy and disease status, was not recorded. Fourth, only community-dwelling women were examined, and extrapolations to men or people in other contexts (e.g., institutionalized) should be made with caution. Fifth, correlation analysis was not corrected according to numerous covariables that might influence muscle power, including physical activity levels, diet quality, and sleep. Sixth, the linear encoder only captures movements in one plane, indicating that disturbances during sit and stand actions were not recorded. Finally, the results shown in this work are derived from cross-sectional observations. The possibility cannot be ruled out that differences in birth cohorts may have influenced some of the assessed parameters. A deeper understanding of age-dependent differences in muscle power requires an analysis of prospective data that are unavailable at this stage for our study.

Conclusion and Implications

The operationalization of our findings is complex. However, results of the present study indicate that examining the biomechanical aspects of 5STS performance might contribute to distinguishing older adults with different physical performance levels. Moreover, if confirmed, our results indicate that a simpler physical evaluation might be conducted to assess lower-limb muscle power in people with difficulties in performing many repetitions at high velocities, such as older adults suffering from chronic pain, mobility limitations, fatigue, respiratory diseases, cancer, and cognitive impairment. More studies are necessary to further explore the potential of our results.