Effects of a Short Proprioceptive Neuromuscular Facilitation Stretching Bout on Quadriceps Neuromuscular Function, Flexibility, and Vertical Jump Performance

Abstract

Place, N, Blum, Y, Armand, S, Maffiuletti, NA, and Behm, DG. Effects of a short proprioceptive neuromuscular facilitation stretching bout on quadriceps neuromuscular function, flexibility, and vertical jump performance. J Strength Cond Res 27(2): 463–470, 2013—The inclusion of relatively long bouts of stretching (repeated static stretches of ∼30 seconds) in the warm-up is usually associated with a drop in muscle performance. The purpose of this study was to assess the effect of a novel self-administered proprioceptive neuromuscular facilitation (PNF) paradigm with short periods of stretching and contraction on quadriceps neuromuscular function, vertical jump performance, and articular range of motion (ROM). Twelve healthy men (age: 27.7 ± 7.3 years, height: 178.4 ± 10.4 cm, weight: 73.8 ± 16.9 kg) volunteered to participate in a PNF session and a control session separated by 2–7 days. The PNF stretching lasted 2 minutes and consisted of 4 sets of 5-second isometric hamstring contraction immediately followed by 5 seconds of passive static stretch of the quadriceps immediately followed by 5 seconds isometric quadriceps contraction for each leg. For the control session, the participants were asked to walk at a comfortable speed for 2 minutes. Active ROM of knee flexion, vertical jump performance, and quadriceps neuromuscular function were tested before, immediately after, and 15 minutes after the intervention. The PNF stretching procedure did not affect ROM, squat jump, and countermovement jump performances. Accordingly, we did not observe any change in maximal voluntary contraction force, voluntary activation level, M-wave and twitch contractile properties that could be attributed to PNF stretching. The present self-administered PNF stretching of the quadriceps with short (5-second) stretches is not recommended before sports where flexibility is mandatory for performance.

Introduction

Flexibility is an important physical quality in a number of sports such as gymnastics, figure skating, synchronized swimming, ice hockey (goalie), some martial art postures, dance, and other activities (¹⁶). Nevertheless, the potential benefits of good flexibility on sport performance are still under debate. Despite an improvement in articular range of motion (ROM) induced by stretching, reduced muscle force generating capacity, and sport-related performance are common findings (for a review see [^3,28]). The large majority of previous studies have used relatively long periods of static stretching (generally >90 seconds, e.g., 3 × 30 seconds), which have been shown to acutely impair static (^1,12–14,24) and dynamic (¹⁸) muscle strength, vertical jump (^4,15,30), and sprint performance (²²). Neural (^{2,9,11,13,14,24}) and contractile-mechanical (^8,11,13,14) impairments have been evoked to explain the reduction in muscle performance induced by passive stretching.

An alternative to pure static stretching is the combination of stretching with isometric contractions of the agonist and antagonist muscles, that is, the so-called proprioceptive neuromuscular facilitation (PNF). Compared with conventional stretching, PNF-induced gains in ROM have been proposed to be partially because of Golgi tendon organ-mediated autogenic inhibition (because of prior contraction of the muscle of interest) and Ia-inhibitory interneurons-mediated reciprocal inhibition (because of prior contraction of the antagonist muscle) (for a review see [²⁶]). Similar to pure static stretching, PNF with relatively long stretching bouts (bout usually ≥30 seconds, with several repetitions and stretching exercises for a total stretching duration of several minutes) has been shown to acutely impair dynamic strength of quadriceps muscle (²⁰) and vertical jump performance when quadriceps, hamstrings, and plantar flexors were stretched (⁵). Results seem to differ when PNF uses stretches of short duration; indeed, as opposed to earlier studies using several bouts of static stretching ≥30 seconds, Murphy et al. (²¹) demonstrated that short-duration static stretches of 6 seconds (6 repetitions) increased hamstrings ROM, countermovement jump (CMJ) height, static balance, and movement time. Because short-duration static stretching is less likely to cause functional impairments than longer stretching bouts and could even result in functional improvements (²¹), we hypothesize that the reduced stretching time would not reduce quadriceps force generating capacity and jump height together with an increase in ROM. Thus, the purpose of this study was to examine the acute effect of quadriceps PNF stretching with short (5-second) periods of stretching and contraction on neuromuscular function, vertical jump performance, and ROM. Furthermore, to determine the mechanisms underlying the eventual changes in performance (jump height and ROM), neuromuscular function was investigated with noninvasive techniques based on voluntary and evoked contractions combined with surface electromyography (EMG). The unique aspect of this short PNF paradigm is that subjects could perform the PNF stretching without the assistance of a partner and thus could be time saving and costless for warming up before training and competition. Because athletes may have a limited time period to warm-up (e.g., in the case of substitution in team sports), this study will provide information about the potential usefulness of short-duration PNF stretching before exercise.

Methods

Experimental Approach to the Problem

This randomized controlled crossover study consisted of 3 testing sessions. The first session (∼45 minutes) was used for participant familiarization with the testing procedures (PNF stretching, vertical jump techniques, voluntary and electrically evoked contractions). The 2 experimental sessions consisted of testing quadriceps neuromuscular function, flexibility, and vertical jump performance after a quadriceps PNF stretching or a walking period of similar duration (control condition). The following outcomes were considered as the dependent variables: active ROM of knee flexion and hip extension, maximal voluntary contraction (MVC) force of quadriceps and hamstring muscles, quadriceps voluntary activation level, EMG activity of quadriceps and hamstring muscles, M-wave amplitude and peak twitch of the quadriceps, CMJ, and drop jump (DJ) performance. The 2 experimental sessions (duration ∼90 minutes) were separated by 2–7 days. Figure 1 gives a schematic representation of the experimental protocol.

Figure 1: Schematic representation of the protocol. ROM = range of motion; MVC = maximal voluntary contraction; MVC ITT = maximal voluntary contraction with the interpolated twitch technique; CMJ = countermovement jump; DJ = drop jump.

The right and left knee extensors were given a sequence of movements which are based on PNF schemes, and consisted of: a 5-second isometric hamstring contraction (antagonist) immediately followed by 5 seconds of passive static stretch of the quadriceps immediately followed by a 5-second isometric quadriceps contraction (agonist). The antagonist contraction was performed to facilitate the quadriceps stretching and the sequence was terminated with a quadriceps contraction to potentiate the muscle, with the expectation that this procedure would not induce any decrement in jump height. Passive static stretch of the quadriceps was performed as follows: while the subject was standing on one leg (fully extended), the leg to be stretched was flexed at the knee and the subject pushed the ankle back toward the buttocks until the level of discomfort for 5 seconds. Hamstring and quadriceps contractions were performed while the subject was standing on one leg, contracting muscles of the other leg while pushing against a wall with the heel (for the hamstring) or the toes (for the quadriceps) reaching the level of discomfort (vigorous contraction). The subjects were asked to keep both legs extended during these contractions. After each 15-second PNF stretching bout, the subjects were instructed to stretch the other leg, and the procedure was repeated four times for each leg. Hence, for this modified PNF procedure, each leg was stretched for a total duration of only 20 seconds with 20-second contractions of the agonist and antagonist respectively. An illustration of the PNF procedure is provided in Figure 2. The intervention during the control session consisted in a 2-minute walking period at a self-selected comfortable speed.

Figure 2: Pictures of the procedure used for proprioceptive neuromuscular facilitation (PNF) intervention. (A) Hamstring (antagonist) contraction followed by (B) passive static stretch of the quadriceps followed by (C) quadriceps (agonist) contraction. Each step was sustained for 5 seconds once the level of discomfort was reached.

Subjects

Twelve healthy men (age: 27.7 ± 7.3 years, height: 178.4 ± 10.4 cm, weight: 73.8 ± 16.9 kg) volunteered to participate to this randomized crossover study. All the subjects were physically active (i.e., at least 2 hours of physical activity per week). They were verbally informed of the procedures and read and signed an informed consent form before participation. The protocol of the study was approved by the Ethics Committee of the University Hospital of Geneva (protocol 10-051). All the procedures were conducted according to the Declaration of Helsinki. All the subjects received standard testing instructions (nutrition, hydration, sleep) verbally and written and were asked not to take part in any demanding physical activity for 24 hours before each testing session.

Procedures

Warm-Up

The participants performed a 5-minute submaximal warm-up on a cycle ergometer (Fleisch, Metabo, Epalinges, Switzerland). They were instructed to cycle at 70 rpm and to maintain a power output of approximately 1 W·kg⁻¹

Pretest and Posttest

Except for vertical jumps where both legs were used, all the tests were performed on the right (dominant) leg. Pretest and posttest, which were performed before, immediately after, and 15 minutes after the intervention, consisted in the assessment of (a) active ROM, (b) neuromuscular function, and (c) vertical jump performance (Figure 1).

Data Collection

Active Range of Motion

For knee flexion active ROM measurement, the subject laid prone and actively flexed the leg as far as possible without assistance from the investigator (²⁰). Knee flexion angle was recorded at a sampling frequency of 1 kHz with an electronic goniometer (SG150, Biometrics, Cwmfelinfach, United Kingdom) connected to an acquisition system (MP150; Biopac, Goleta, CA, USA). The goniometer was positioned on the lateral aspect of the right knee as already described (²⁰). Briefly, the proximal arm was taped along the femur using the greater trochanter as a reference, and the distal arm was aligned with the shank using the lateral malleolus as a reference. Hip extension active ROM was also investigated because the biarticular rectus femoris (RF) is both a knee extensor and a hip flexor muscle. For hip extension ROM assessment, the subjects were lying prone on a table. With their anterior superior iliac spine in contact with the table, they attempted to lift their leg as far as possible off the table. The vertical distance from the table to the anterior aspect of the patella was measured using a tape measure (²⁴). Two trials of each stretch were performed at both pretest and posttest.

Evoked Contractions

Electrically evoked contractions of the quadriceps were induced with a high-voltage (maximal voltage 400 V) constant-current stimulator (model DS7AH, Digitimer, Hertfordshire, United Kingdom). The femoral nerve was stimulated transcutaneously using a circular (diameter: 5.1 cm) self-adhesive electrode (Dermatrode; American Imex, Irvine, CA, USA) positioned in the femoral triangle, 3–5 cm below the inguinal ligament. A large (5 × 10 cm) rectangular electrode (Compex, Ecublens, Switzerland) was fixed over the gluteal fold to close the stimulation current loop. Current intensity was progressively increased from 0 mA to full motor unit recruitment, as verified by M-wave and twitch force recordings, and this was further increased by 10% to provide supramaximal stimuli. Once this stimulation intensity was found, it was kept constant throughout the session for each subject. The site of stimulation was marked on the skin so that it could be repeated between the sessions. The stimulus duration was 1 millisecond, and the interstimulus interval in the doublet was 10 milliseconds.

Force Recordings

Isometric MVC and electrically evoked force of the quadriceps and hamstring muscles were recorded using a custom-built ergometer that comprised a chair connected to a strain gauge (STS 2,500 N, sensitivity 2 mV·V⁻¹ and 0.0017 V·N⁻¹, SWJ, Shenzhen, China). The subjects were comfortably seated with a knee angle of 90° and a trunk-thigh angle of 100° (180° = full extension). The strain gauge was securely strapped between the ankle and the chair. Extraneous movements of the upper body were limited by 2 crossover shoulder harnesses and a belt across the abdomen. Two quadriceps MVC and 2 hamstring MVC were performed at pretest; if the difference in MVC force between the 2 trials was >5%, a third attempt was requested. Only 1 MVC per muscle group was performed at posttest to minimize fatigue.

Electromyography Recordings

The EMG activity of the knee extensors vastus lateralis (VL), vastus medialis (VM), and RF and that of the knee flexor biceps femoris (BF) were recorded with pairs of silver chloride circular surface electrodes (Kendall Meditrace 100, Tyco, Canada), with a recording diameter of 1 cm. Electrodes were positioned lengthwise over the middle of the muscle belly with an interelectrode (center-to-center) distance of 2 cm. The electrodes were placed at a distance of two-thirds between the anterior spina iliaca superior and the lateral femoral condyle for the VL, at four-fifths of the distance between the anterior spina iliaca superior and the medial femoral condyle for the VM, midway between the anterior spina iliaca superior an upper border of the patella for the RF, and midway between the ischial tuberosity and the lateral tibial condyle for the BF (²³). Electrode positions were marked to ensure similar placement during the second experimental session. The reference electrode was placed over the patella. Low resistance between the 2 electrodes (<10 kΩ) was obtained by abrading and cleaning the skin with alcohol. The EMG signals were amplified with a bandwidth frequency ranging from 10 to 500 Hz (gain: 1,000), digitized online at a sampling frequency of 2 kHz, and recorded (MP150; Biopac, Goleta, CA, USA).

Vertical Jump Performance

All the jumps were performed on a force plate (9281B, Kistler, Winterthur, Switzerland) connected to an amplifier (9865B, Kistler), and vertical ground reaction force (R_z) data were recorded at a sampling frequency of 1 kHz (MP150; Biopac, Goleta, CA, USA). Jump height was measured for the CMJ and DJ. Two CMJ and DJ were performed at pretest; if the difference in jump height between the 2 trials was >5%, a third jump was requested. Only 1 jump per modality was performed at posttest to minimize fatigue. The interval between 2 consecutive jumps was 10 seconds. For the CMJ, the subjects were instructed to keep their hands on the hips. After a rapid knee flexion to approximately 90°, they left the force plate with knees and ankles fully extended and were asked to land in a similarly extended position to ensure valid test conditions. For the DJ, the subjects started from a 30-cm-high step (²⁴) and were instructed to keep their hands on the hips and step off the platform with the leading leg straight to avoid any initial upward propulsion (thus ensuring a drop height of 30 cm). They were instructed to jump for maximal height and minimum ground contact time. As for CMJ, the participants were instructed to leave the force plate with knees and ankles fully extended and to land in a similarly extended position. Vertical jumps were performed to assess the influence of PNF stretching on slow (CMJ) and fast (DJ) stretch-shortening cycle.

Data Analysis

Active Range of Motion

For pretest and posttest, only the best attempt for knee flexion and hip extension active ROMs was considered for further analysis.

Vertical Jump Performance

Vertical jump height was calculated as h = g × flight time²/8. For each pretest and posttest, the jump achieving the greatest height was used for analysis (⁴).

Force Recordings

The MVC force was considered as the peak force attained during the contraction, and voluntary activation level (interpolated twitch technique) was estimated according to the following formula: voluntary activation level (%) = [1 − (superimposed doublet amplitude × voluntary force level just before the superimposed doublet/MVC force)/potentiated doublet amplitude] × 100 (²⁹). Peak twitch was calculated from the average of 3 trials and peak doublet was analyzed to calculate voluntary activation level.

Electromyography Activity

The M-wave peak-to-peak amplitude was analyzed for VL, VM, and RF muscles with the average of the 3 trials being used for analysis. The EMG activity was quantified as the root mean square estimate for a 500-millisecond interval around MVC force (250-millisecond interval either side of the peak force). Knee extensors EMG activity measured during MVC was averaged for VL, VM, and RF muscles. The EMG activity of BF muscle was determined from hamstring MVC, and the level of coactivation was calculated as the EMG value recorded from BF during quadriceps MVC divided by the corresponding EMG value recorded during hamstring MVC (i.e., when BF acted as an agonist) and then expressed as a percentage.

Statistical Analysis

For the dependent variables used in this study, intraclass correlation coefficients were >0.80. Data were analyzed with separate 2-way (treatment × time) analyses of variance with repeated measures. Post hoc analyses (Tukey) were used to test for differences among pairs of means when appropriate. A significance level of p ≤ 0.05 was used to identify statistical significance. The statistical analyses were performed using Sigmaplot software (version 11.0, Systat, Chicago, IL, USA). Data are presented as mean ± SD within text and table and as mean ± SEM within figures.

Results

Active Range of Motion

Knee flexion (p = 0.23) and hip extension active ROM (p = 0.66) were not modified after PNF stretching or control condition (Figure 3).

Figure 3: Active ROM for (A) knee flexion and (B) hip extension. Values (mean ± SEM, N = 12) were collected before (Pre), immediately after (Post), and 15 minutes after 2 minutes of PNF stretching or walking (control). KE = knee extensors; PNF = proprioceptive neuromuscular facilitation; ROM = range of motion; 0° = leg fully extended.

Neuromuscular Function

Quadriceps MVC force was similar before the PNF and control sessions (p = 0.42). Table 1 shows that quadriceps MVC force was slightly though significantly depressed after both PNF stretching and the control walking period (∼5%, p = 0.004), but no difference was found between the two conditions. Quadriceps voluntary activation level was not affected by PNF and control (p = 0.47), and no difference was noted between the 2 sessions (p = 0.68, Table 1).

Table 1: Neuromuscular properties of the knee extensor muscles.^{* †}

Knee extensors EMG activity did not change during the course of both experimental sessions; the normalized EMG activity varied from 100 ± 0 to 94 ± 19% after PNF stretching and to 99 ± 13% after the control condition (p = 0.47). A significant reduction in BF coactivation level was found after PNF stretching (from 10.1 ± 3.8 to 7.5 ± 3.6%, p = 0.043), and also 15 minutes after PNF stretching (7.4 ± 4.5%, p = 0.035). No significant changes were noted after the 2-minute walking period (from 8.0 ± 3.6 to 8.6 ± 3.3%, p = 0.85). Hamstring MVC force (from 195 ± 30 to 194 ± 40 N for the PNF session and from 195 ± 42 to 189 ± 44N for the control session) and EMG of BF muscle remained unchanged after both experimental sessions (p = 0.55 and p = 0.83, respectively). Peak twitch was not different between the two experimental sessions (p = 0.23), and it decreased similarly (∼5%) 15 min after PNF stretching and the control walking period (p = 0.001, Table 1). M-wave amplitude for VL (p = 0.79), VM (p = 0.45), and RF (p = 0.61) muscles was unmodified after the 2 experimental sessions and did not differ between the conditions (Table 1).

Vertical Jump Performance

Vertical jump performance for both CMJ (p = 0.40) and DJ (p = 0.35) was not influenced by PNF stretching and remained also unchanged after the control walking period (Figure 4).

Figure 4: Vertical jump performance for (A) countermovement jump (CMJ) and (B) drop jump (DJ). Values (mean ± SEM, N = 12) were collected before (Pre), immediately after (Post) and 15 minutes after 2 minutes of proprioceptive neuromuscular facilitation (PNF) stretching or walking (control).

Discussion

The aim of this study was to assess the influence of self-administered PNF stretching including short contraction and stretching phases on active ROM, neuromuscular function, and jump performance. Our results showed no change in knee flexion and hip extension active ROM and jumping performance after PNF stretching. Accordingly, most neuromuscular adjustments were comparable between the PNF and the control sessions.

The PNF stretching of the quadriceps muscle consisting of several minutes of passive stretching has been shown to acutely improve ROM (e.g., [²⁰]). This increased flexibility was originally explained by neuromuscular facilitation, first evidenced by Sherrington (²⁷), that is, by the effects of autogenic inhibition provided by a prior contraction of the quadriceps and reciprocal inhibition provided by a prior contraction of the hamstrings. Other factors have since been proposed to explain the stretching-induced increase in ROM, such as mechanical changes within the muscle-tendon unit (^8,12,13) and altered perception resulting in an increased tolerance to the stretch (^7,19,26). Our finding of an unchanged active ROM after PNF stretching may be caused either by the short duration of the total (20 seconds per leg) or single stretching sequences (5 seconds). The 5-second passive static stretching is close to the 3 seconds recommended in a recent review (²⁶). However, Murphy et al. (²¹) recently showed that 6 repetitions of 6-second static stretches of the hamstrings were sufficient to increase ROM by about 12%, which suggests that not all muscles behave in the same manner and the quadriceps may need longer stretching durations to achieve increases in ROM. But there may also be methodological difficulties. Behm and Kibele (⁴) reported no significant changes in hip extension ROM with 4 repetitions of 30 seconds of static stretching. They suggested that measuring hip extension ROM can be more difficult than other commonly measured joints such as with supine hip flexion. According to Hubley-Kozey (¹⁷), there is considerable difficulty in locating the true joint center with this method and aligning the limbs. Similarly, it is important but also difficult to maintain contact of the hip with the mat when performing a hip extension ROM test (⁶). Similar difficulties may have increased the variability of these measurements in this study. Moreover, the stretching procedure used here was unique compared with previous PNF studies. In contrast to the classical partner-assisted PNF stretching, the agonist and antagonist contractions used in this study (Figure 2) were performed by the subjects themselves (as is often the case in sports activities) while the quadriceps was not at the end point of the ROM. Finally, it may also be that ROM was transiently changed for a very short period (<1 minute) with this PNF scheme; even if that was the case, the athlete would not take advantage of this procedure during competition, as already suggested (²⁵).

Our results of similar MVC force and jump performance after both PNF and control sessions are in accordance with our active ROM measurements. Indeed, it is unlikely that unchanged ROM (which can be explained by unaltered proprioceptive information and/or musculotendinous viscoelastic properties and/or stretch perception) induced a significant alteration in force-power generating capacity. Again, the low stretching duration might explain the absence of MVC reduction as compared with the control session, because the reduction in MVC force after passive stretching has been suggested to be governed by a dose-response relationship (²⁵). Our results confirm those of Young and Elliott (³¹), who reported no change in vertical jump performance after a relatively short PNF stretching procedure on quadriceps, gluteus, and plantar flexors (including 15-second static stretching bouts for a total stretching duration of 45 seconds for each muscle group). Furthermore, the unchanged quadriceps voluntary activation and EMG activity indicate that neural drive was preserved after our PNF procedure. In earlier studies, impaired neural drive has been hypothesized as a cause of poststretch force decrements (^2,13,14,24). It is also interesting to note that the reduced antagonist muscle coactivation induced by PNF stretching did not affect quadriceps MVC force nor jumping performance, suggesting that a small change in coactivation has a negligible impact on maximal voluntary efforts.

In this study, we used electrically induced contractions to gain insights into the peripheral component of neuromuscular function after PNF stretching. The results clearly indicate that muscle excitability was preserved, as reflected by M-wave characteristics of the knee extensor muscles; this finding was also reported after longer stretching duration on the plantar flexors (⁸). In addition, the unchanged peak twitch in our study indicates that the final steps of excitation-contraction coupling (involving Ca²⁺) were not affected by the PNF procedure. This is in line with the preserved jump and MVC performances and confirms the findings obtained in previous studies in which longer stretching durations were used, for both plantar flexor (⁴) and knee extensor (²⁴) muscles. As our subjects were in a complete resting state for approximately 6–7 minutes from the end of the posttests to the start of the post 15-minute tests, the reduced peak twitch amplitude found 15 minutes after both PNF and control sessions may be attributed to a slight decrease in muscle temperature (¹⁰), although not measured in this study.

Practical Applications

Self-administered PNF stretching including 5-second static stretches of the quadriceps did not improve active ROM of knee flexion. Thus, although this PNF procedure did not have any detrimental effect on quadriceps muscle function, we do not recommend this stretching modality (5-second isometric hamstring contraction immediately followed by 5 seconds of passive static stretch of the quadriceps immediately followed by 5-second isometric quadriceps contraction) for sports activities in which flexibility is a key factor. Nevertheless, stretching is routinely used by many athletes as an important component of warm-up procedures and, although physiological benefits are contestable, it is certainly psychologically beneficial.

Acknowledgments

The authors thank Marc Buclin for the design and conception of the ergometer and all the subjects who volunteered to participate. This study was supported by De Reuter foundation, Geneva Academic Society and Ernest Boninchi foundation. The results of this study do not constitute endorsement of the product by the authors or the National Strength and Conditioning Association.