Methods

Twenty-eight female floorball teams in Finland participated in the intervention study, which extended from September 2005 to February 2006. We first arranged the baseline test schedule with the participating teams in the summer of 2005, and then in August 2005 the tests were performed at the training venue of each team. The tests were repeated in February 2006. The study was approved by the ethics committee of the Pirkanmaa Hospital District, Tampere, Finland.

Participants and randomisation

All healthy players who performed the baseline tests and were official members of the participating teams were included in the study. Informed consent was obtained from each player for final participation. A total of 347 players from 28 teams were tested during August 2005. Stratified cluster randomisation to the intervention and control teams was performed at three league levels (elite league, 1st division and 2nd division) using a team as the unit of randomisation. The statistician (MP) who ran the computer-based randomisation was not involved in the intervention. Teams allocated to the intervention group were informed about the upcoming warm-up programme for preventing injuries. Teams in the control group were asked to do their usual training during the entire study period.

Intervention programme

The neuromuscular warm-up programme consisted of four different types of exercise: (1) running technique exercises; (2) balance and body control exercises; (3) jumping exercises; and (4) strengthening exercises to the lower limbs and trunk. The neuromuscular training was carried out like a warm-up session just before floorball exercises, with low-to-moderate intensity for each exercise type. One warm-up session lasted 20–30 min, each exercise type taking 5–7 min. The teams were asked to use the programme 1–3 times per week during the study period. An exact description of the intervention programme has been published.7

Test battery

The baseline and follow-up tests were performed during a training session for each team at their own training venue. Six blinded research physiotherapists carried out the tests. Before the tests, the players warmed-up for 5 min by jogging on a floorball court. The test battery included five performance tests performed in the following order.

Static jump and countermovement jump

Two different types of vertical jump tests were performed: a static jump and a countermovement jump.13 14 Both tests measured the maximal vertical jump height (cm) (ie, the muscle power of the extensor muscles of the lower extremities). The electronic apparatus (New Test Powertimer; New Test, Oulu, Finland) including the contact mat computed the height of the jump (cm) by measuring time in the air with a digital timer. For the static jump, the subject was asked to jump as high as possible on the contact mat, starting the jump from a static squatting position with a 90° knee angle. In the countermovement jump, the subject started the jump from standing upright and then making a countermovement (squat) before the vertical jump. For both jumps, the best result of three trials was used in the analyses.

Jumping over a bar

The jumping-over-a-bar test was used to assess maximal jumping speed.15 The subject was asked to perform repeated sideways jumps as quickly as possible over a foamed plastic bar (length 50 cm, width 4 cm and height 4 cm) that had been placed on the ground. The jumping time was 15 s and number of correctly accomplished two-leg jumps was recorded (ie, one-leg steps and jumps touching the bar were excluded). The stopwatch was started simultaneously with the starting signal, and the finishing signal ended the test. The better result from two trials was used in the analyses.

Standing on a bar

The standing-on-a-bar test measured one-leg static balance,16 thus measuring ability to control the stationary one-leg standing position. The subject was asked to stand with her dominant leg on a narrow bar (width 2 cm, height 4 cm and length 50 cm) for 1 min. The stopwatch was stopped every time the subject touched the floor with the free foot and restarted when the balanced position was achieved again. The number of balance losses (and thus restarts) was the studied variable. The subject was allowed to use her unsupported arms for balance. The dominant leg was tested only once.

Figure-of-eight running

The figure-of-eight running test measured running agility,17 thus measuring ability to move, accelerate, decelerate and change direction effectively and quickly in a controlled manner. The subject was asked to run a figure-of-eight course as fast as possible. The course was marked with two vertical cones placed 10 m apart and the start/finish line was next to the first cone. The stopwatch was started simultaneously with the subject’s takeoff and was stopped when the subject completed the course and crossed the finish line. The time was recorded in seconds. The better result of two attempts was recorded.

Data collection

At baseline, players completed a questionnaire to provide background information such as anthropometrics, previous injuries, floorball experience and preseason training volume. During the 6-month study period, each team coach recorded the players’ scheduled practice and game hours in an exercise diary. In addition, the warm-up instructors kept a diary about the scheduled warm-up sessions in the intervention teams. Also possible injuries were recorded. After each follow-up month, the coach and instructor mailed the completed diaries to the UKK Institute.

Statistical analysis

The differences in follow-up test means (static jump, countermovement jump, jumping over a bar, standing on a bar and figure-of-eight running) between the two study groups (control and intervention) were analysed by multilevel regression models, taking into account the hierarchical structure of the data due to cluster randomisation. Adjustments were performed according to individual level (baseline test result, age, floorball experience, playing position and number of orthopaedic operations) and team level (league level). Analyses were performed according to the intention-to-treat principle. In addition to the intention-to-treat analyses, efficacy analyses were conducted to evaluate the potential benefits of high training compliance and adherence (high indicating players who carried out the warm-up exercises at least once a week during the 6-month follow-up). p<0.05 was considered significant. The MLwiN (version 2.02) software package was used for statistical analyses.

Results

Study population

Figure 1 gives details of the flow of teams and players through the study. Altogether 345 players and 28 teams were randomised. Of these, 123 players (36%) dropped out of the study, leaving 222 players and 27 teams for current analysis. Table 1 shows the characteristics of the players in the two groups. No significant differences in baseline characteristics were found between the groups.

• Download figure
• Open in new tab
• Download powerpoint

Figure 1

Flow of teams and players through the study.

Training activity

Concerning the training compliance of the measured 13 intervention teams, five teams carried out the warm-up programme according to the plan, six teams had some irregularities in training, and two teams interrupted training during the follow-up. Seventy-one players (60%) from the intervention group participated in structured warm-up sessions at least once a week during the study season, 28 players (23%) trained irregularly, and 20 players (17%) stopped training before the midpoint of the follow-up. No injuries occurred in the intervention group during the warm-up sessions. Concerning the control teams, 40 players (39%) actually carried out some of the intervention exercises weekly as a part of their usual training.

Performance tests

Table 2 presents the data on the five outcome measures and results of multilevel analysis. At 6 months, both groups showed improvements in all performance tests. Statistically significant between-group differences were found in two of the outcome measures, jumping over a bar (number of jumps in 15 s) and standing on a bar (number of balance losses in 60 s): 2.3 jumps (95% CI 0.8 to 3.8, p = 0.003), favouring the intervention group; −0.4 balance losses (95% CI −0.8 to 0.0, p = 0.050), again in the favour of the intervention group. Improvements in the other outcomes did not differ significantly between the two groups.

In efficacy analysis between high-compliance players (n = 71) and the control group (n = 103), we found parallel results to the main analysis (table 3). In jumping speed, the adjusted mean difference between high-compliance players and the control group was 1.1 jumps (95% CI −0.1 to 2.3, p = 0.08), and in the static balance test the adjusted mean difference was −0.4 balance losses (95% CI −0.9 to 0.1, p = 0.10).

Discussion

Participation in a structured 6-month neuromuscular warm-up programme designed to enhance motor skills and prepare the body for upcoming floorball training was found to improve static balance and sideways jumping speed in female floorball players. Concerning changes in vertical jumps and running speed and agility, there were no differences between the intervention and control groups.

Neuromuscular training

Studies have shown that neuromuscular training is likely to enhance athletic performance9 10 11 12 and thereby improve lower extremity biomechanics and reduce injurious forces. The programmes used differ in type, intensity, frequency and duration, and this probably has a marked effect on measured outcomes. Hewett and colleagues10 investigated the effects of 6 weeks of an intensive and progressive jump and weight training programme on landing mechanics and lower limb strength in female athletes. The training session lasted ∼2 h and was repeated three times a week. After the training period, landing forces from jumps decreased and knee control increased among the trained female athletes. In addition, vertical jump height and hamstring-to-quadriceps muscle torque ratios had increased in the trained group.

Chimera and coworkers11 evaluated the effects of jump training on muscle-activation strategies and lower extremity performance during a 6-week intensive plyometric training period. The experimental group of female athletes performed jump exercises twice a week, and one training session took 20–30 min. They found a significant increase in preparatory adductor muscle firing, adductor-to-abductor muscle coactivation and quadriceps-to-hamstring muscle coactivation in the intervention group. Findings supported the importance of hip musculature activation strategies for lower extremity control, which interact in biomechanics and reduce harmful forces.

Emery et al12 studied the effectiveness of a home-based balance-training programme in male high-school students. The training programme included two-leg and one-leg balance exercises with wobble board and trunk stabilisation exercises. Students were advised to use the balance-training programme daily for 6 weeks; one training session lasted ∼20 min. The balance test battery included a static balance test (one-leg standing on the floor) and a dynamic balance test (one-leg standing on a balance pad). Improvements in static and dynamic balance during the follow-up were significantly greater in the intervention group than in the control group.

Our neuromuscular training programme included running technique, balance, jumping and strengthening exercises with several variations, and it was designed to replace the traditional warm-up before structured floorball training. The intensity in each exercise was low to moderate. Therefore it was obvious that this training might not improve all measured outcomes. The programme did not, for example, include exercises that aim to improve maximal vertical jump height. Systematic strength and power training, such as that in the study of Hewett et al,10 are needed to improve muscle power. On the other hand, it was logical that warm-up exercises would enhance static balance and sideways jumping performance because every warm-up session included different variations of one-leg standing and a rebound jump series.

Although these warm-up exercises did not improve all of the measured outcomes, we feel that they should be practised before sports-specific training because they have been shown to be effective in injury prevention.3 4 5 6 7 8 In the primary analysis of our large injury prevention trial, the greatest reduction was found in ankle ligament injuries.7 Thus improvements in jumping speed and static balance are likely to indicate improvements in ankle control and muscle function of the lower legs.

In an efficacy analysis, we did not find greater improvements among the high-compliance players. However, it is noteworthy that the high-compliance players (n = 71) had slightly better baseline test results than the rest of the intervention group, which may indicate that these regularly trained players may have higher training volume, better condition and neuromuscular performance in general. At the same time, players who had lower baseline results improved their neuromuscular performance by quite a small amount of training. This may indicate that improvements in jumping speed and static balance are very easy to attain at the initial stage of training or among novice players, even with irregular training.

The main point in diverse warm-up exercises is to activate proprioception and motor control, and thereby prepare the neuromuscular system for upcoming sports training. If upcoming sports training or playing includes, for example, one-leg movements, upper body rotations and running in different directions, it would be reasonable to practise those manoeuvres during the warm-up session. Systematic warm-up exercises are also an excellent way to learn and maintain motor skills for each sport. Because of this, in the present intervention the main point in each exercise was to focus on proper technique, such as good playing posture, neutral zone of lumbar spine, trunk stability, and position and function of the hip, knee and foot (especially the “knee-over-toe” position) during the sports-specific manoeuvres.

Test battery

The selected performance tests were carried out during the teams’ training session at their own training venue. We chose widely used field tests, which were feasible and easy to perform. In further intervention studies, however, it would be important to explore more precisely the effects of neuromuscular training on players’ body part movements and muscle activation in sports-specific manoeuvres, because insufficient joint control and side-to-side differences in lower extremity performance are associated with increased risk of sports injury.18

Limitations and strengths of the study

Our study had some limitations. Firstly, although the randomisation phase, data collection and data analysis were blinded, for obvious reasons neither the coaches nor players could be blinded. Secondly, although the initial rate of participation in the tests was quite high, the general rate of participation in the follow-up tests could have been better (fig 1). The drop-out rate in our study was 36%, and this may have influenced the results, although the proportion of drop-outs was similar in both groups. In fact, it was a great challenge to arrange suitable times for testing for all players in this amateur sport. Thirdly, it became obvious that some players in the control group used similar neuromuscular exercises to those in the intervention groups, because these exercises are commonly used in sports training. However, it is likely that the training volume and quality of the controls did not reach the level of the intervention group, and, if this partial contamination of the controls biased the results of the study, it erred on the side of underestimating rather than overestimating the effect of the neuromuscular programme.

Our study also had many strengths. Firstly, a randomised study design was robust and reduced the potential biases and thus increased the reliability of the results. Secondly, the intervention and control groups were similar in baseline characteristics, drop-out rate, and training and playing exposure during the 6-month follow-up. Thirdly, the neuromuscular warm-up activity in the intervention group was effective.

In conclusion, the neuromuscular warm-up programme used was effective in enhancing floorball players’ sideways jumping speed and static balance. Furthermore, the neuromuscular warm-up exercises were safe to perform and can thus be recommended for inclusion in weekly training for this sport.

Perspectives

Floorball players have an increased risk for non-contact ligament injuries of the ankle and knee. Fortunately, regular and structured neuromuscular exercises have been proven to be effective in reducing the risk of such injuries, and therefore they are also widely recommended and used in various sports. However, only a few studies have analysed whether and how these exercises affect musculoskeletal performance. The findings of the present study reveal that the neuromuscular warm-up programme improved floorball players’ static balance and sideways jumping speed, and above all, these exercises are safe to perform. However, further studies are needed to clarify the effects of different components of the neuromuscular training used. Such studies should focus on sports-specific changes in motor control, skills and technique.