Effects of individualised training programmes based on the force-velocity imbalance on physical performance in rugby players

Abstract

BACKGROUND:

The use of force-velocity imbalance (Fvimb) has been proposed as an effective method for prescribing training loads aiming to enhance physical performance.

OBJECTIVE:

This study aimed to analyse the effects on lower body strength, jump and sprint performance of different individualised resistance training (RT) programmes based on Fvimb in rugby players.

METHODS:

Thirty-four senior rugby players were divided into four groups according to their Fvimb: Not individualised (NI, $n=$ 8), velocity imbalance (Vimb, $n=$ 6), force imbalance (Fimb, $n=$ 11) and well balanced (WB, $n=$ 9). The intervention period consisted of a 7-week resistance, plyometric and sprint training programme performed twice a week. Pre- and Post-training measures included Force-velocity (Fv) profile, countermovement jump (CMJ), 30 m sprint test and estimated one-repetition maximum in squat (1RM-SQ).

RESULTS:

Significant “group” $\times$ “time” interactions ( $P<$ 0.05) were observed for theoretical maximal velocity (V0), theoretical maximal force (F0), and Fvimb. No significant “group” $\times$ “time” interactions were observed for the rest of variables analysed. The Vimb group significantly ( $P<$ 0.05) increased V0, whereas decreased F0 and Fvimb. The Fimb group showed significant ( $P<$ 0.05) decreases in V0, whereas increased F0 and Fvimb. The WB and NI groups did not show significant changes in these parameters. The WB group induced significant enhancements in 10 m, 20 m, and 30 m sprint times, maximum sprint speed and sprint momentum (SM), whereas Fimb attained significant changes in 20 m and 30 m sprint times. The NI group attained significant improvements ( $P<$ 0.05) in SM. No significant changes were observed for 1RM-SQ and jump performance.

CONCLUSION:

Individualised RT programs based on Fvimb induced improvements in sprint performance. Moreover, individualised RT programs based on Fvimb induced better adjustments of the Fv profile to the theoretical “optimal” Fv profile, although any group improved SQ strength, jump height or maximum power.

1. Introduction

An elite rugby player must be competent in the different phases of sprinting speed, including the initial acceleration and maximum speed [1]. Rugby players are required to perform maximal and explosive movements for contact situations and sprint actions [2]. In this regard, due to the nature of the modern game, it requires players to be strong to successfully tackle, agile to quickly evade opposition, have high cardiovascular endurance to sustain performance more than 80 minutes of activity, and have rapid force production capabilities for accelerating and tackling [3].

According to load magnitude employed in resistance training, Mora-Custodio et al. [4] reported that low-loads (40–60% 1RM) training programmes produce similar or more beneficial effects on neuromuscular performance than moderate loads (60–80% 1RM). Likewise, Pareja-Blanco et al. [5] showed that squat training with low/moderate loads combined with light loads resisted sprints may be an effective stimulus for improving leg strength, jumping ability, change of direction (COD), and sprint performance. In rugby players, the combination of two different complex training loading, both enhanced positive adaptations across the entire force-velocity spectrum, while simultaneously improving maximum strength and sprint [6].

Samozino et al. [7] proposed a model, known as Fv profile, based on the measurement of three simple variables, which allows practitioners to accurately assess strength, speed and power output in lower-limbs during a squat jump (SJ) in field conditions. In addition to being simple and low cost, test validation has been reported recently [8, 9]. The comparison of an actual Fv profile with a theoretical optimal Fv profile (i.e. balance between force and velocity) could help identify which component (force or velocity) should be prioritised during training to improve performance [7, 10]. For a given maximum power (Pmax), an imbalance between force and velocity (Fvimb) could lead to a 30% loss of jump performance [7]. The loss of performance, depending on the calculated imbalance, would indicate how far (expressed in %) an athlete is from an optimal Fv profile. In practical terms, individualised training programmes aiming to improve vertical push-off performance, should focus on increasing Pmax and/or reducing the Fvimb [11]. Morin and Samozino [11] suggested that athletes with a significant Fvimb, should follow an individualised RT program that prioritises the development of the unbalanced mechanical variable to optimise jumping performance. According to a recent study [1], both coaches and sports professionals involved in the development of rugby players should consider a sprinting force dominant Fv profile to improve acceleration ability, particularly on the initial acceleration phase. Another study [12] mentioned that to improve sprint speed in rugby players, training strategies should aim to optimise the athlete’s power to weight ratio, and that RT for lower limbs should focus on movement velocity.

In the light of the above, only two studies [13, 14] have examined the effects of an individualised resistance and plyometric programme based on Fvimb. These studies [13, 14] concluded that individualised training programmes based on the Fvimb improve jumping ability in soccer and rugby players. However, they did not examine the effects of these protocols on lower limbs strength (i.e. 1RM) and sprint performance. In contrast, regarding sprint Fv profile, it has been reported that an individualised sprint-training based on the Fvimb was not more effective in improving sprint performance than a generalised sprint-training programme [15].

It seems that no research has analysed the combination of resistance and plyometric programme based on Fvimb on strength, jump and sprint performance. It was hypothesised that if the Fvimb is reduced, it would increase vertical jump height along with a positive impact on lower limbs strength and sprint. Therefore, the aim of this study was to experimentally test the effects on lower limbs strength, jump and sprint performance after following an individualised RT programme based on Fvimb.

Table 1

Groups according to the Force-velocity imbalance cut-off values and main characteristics of participants

Group	Fvimb (%)	Body mass (kg)	Height (cm)	Age
Vimb ( $n=$ 6)	$>$ 110%	84.05 $\pm$ 11.65	174.33 $\pm$ 6.83	21.50 $\pm$ 3.53
Fimb ( $n=$ 11)	$<$ 90%	89.38 $\pm$ 11.07	177.64 $\pm$ 5.26	23.73 $\pm$ 3.32
WB ( $n=$ 9)	90–110%	93.46 $\pm$ 15.65	177.89 $\pm$ 7.36	22.00 $\pm$ 3.77
NI ( $n=$ 7)	1–199%	84.37 $\pm$ 15.59	179.43 $\pm$ 9.22	21.43 $\pm$ 2.51

${}^{*}$ Vimb: group with velocity imbalance; Fimb: group with force imbalance; WB: well-balanced group or low imbalance; NI: not individualised. Fvimb: Force-velocity imbalance.

Table 2

Loads and exercises selected to target each group force-velocity imbalances

Group	Resistance/loads	Jumps/loads	Sprint
Vimb ( $n=$ 6)	Low 40–60% 1RM	SJ (40–60% BW)	Unresisted
	Exercise: Squat	BW Horizontal and vertical
Fimb ( $n=$ 11)	High 75–85% 1RM	SJ (75–85% BW)	Unresisted
	Exercise: Squat	BW Horizontal and vertical
WB ( $n=$ 9)	Moderate 60–75% 1RM	SJ (60–75% BW)	Unresisted
	Exercise: Squat	BW Horizontal and vertical
NI ( $n=$ 7)	Moderate 60–75% 1RM	SJ (60–75% BW)	Unresisted
	Exercise: Squat	BW Horizontal and vertical

Vimb: velocity imbalance group; Fimb: force imbalance group; WB: well-balanced group; NI: not individualised group. 1RM: one-repetition maximum; SJ: squat jump; BW: body weight.

2. Methods

2.1 Participants

Thirty-four men, highly trained rugby players volunteered to participate in this study. Participants competed in top club competitions and had a minimum of ten years of playing experience. Participants were familiar with the tests performed, and were normally involved in $\sim$ 5 training sessions per week (2 sessions for strength and speed, and 2–3 rugby sessions) and one weekly competition. Participants were recruited on the basis that they were free from any injury or training restriction, as verified by the club physiotherapist and had a minimum of two years of structured training experience under the supervision of a qualified strength and conditioning coach. All athletes rested the day before testing and were asked to attend testing in a fed and hydrated state, similar to their normal practices before training. The study met the ethical standards and was approved by an Institutional Research Ethics Committee and conformed to the recommendations of the Declaration of Helsinki. After being informed of the purpose and experimental procedures, participants signed a written informed consent form prior to participation.

2.2 Procedures

2.2.1 Experimental design

The present study used a longitudinal follow-up with Pre-Post experimental design with testing sessions. The intervention was performed at the middle of the competitive season. Participants were allocated according to their individual Fvimb (%) obtained from Pre-training tests and then randomly assigned to one of four training conditions. Three groups carried out a targeted and individualised resistance, plyometric and sprint programme according to their Fvimb: 1) velocity imbalance (Vimb, $n=$ 6), who had a Fvimb higher than 110%; 2) force imbalance (Fimb, $n=$ 11), who had a Fvimb lower than 90%; and 3) well-balanced (WB, $n=$ 9), who had a Fvimb between 90 and 110%. The fourth group carried out a non-individualised intervention (NI, $n=$ 8), consisting of performing the same training programme as the WB group but their Fvimb was not taken into consideration. The randomisation process was performed by a co-author not directly involved in testing or the training intervention. Participants’ anthropometric characteristics and Fvimb cut-off values are presented in Table 1.

Resistance and plyometric training loads used in the present study (Table 2) were adapted and modified from Jiménez-Reyes et al. [13]. The training programme was performed over a 7 weeks period (two sessions per week). Participants were required to complete 14 out of 14 intervention-training sessions (100%) and all Pre- and Post-training tests in order to be included. The RT program was different regarding RT loads (% 1RM) and plyometric (% body mass, “BM”) but identical in volume and exercises among groups. Sprint training was identical for all groups. The training intervention consisted of sprint training in the first part of the session followed by two lower-limb resistance and plyometric exercises, whose duration did not exceed 40 min. These training sessions were held before the specific rugby training so that, in this way, fatigue did not affect performance. Session rating of perceived exertion (RPE) was registered throughout the intervention period for all the rugby session based on previously published guidelines [16]. The specific rugby sessions were $\sim$ 60 min duration in total and RPE was between 6–8 for all sessions and players. The weekly characteristics of the training programme of each sub-group are presented in Table 3.

2.2.2 Testing procedures

Assessment of CMJ and Fv profile tests were performed on day one. After 48 hours of rest, isoinertial squat loading test was performed to estimate 1RM in the squat (1RM-SQ). Finally, on day three, also ensuring 48 hours of rest, 30 m sprint speed test was performed. Participants had abstained from training on

Table 3

Weekly training characteristics for the participants during the intervention period

		Week 1		Week 2		Week 3		Week 4		Week 5		Week 6		Week 7
		Session 1	Session 2	Session 3	Session 4	Session 5	Session 6	Session 7	Session 8	Session 9	Session 10	Session 11	Session 12	Session 13	Session 14
Vimb	% 1RM	40%	40%	50%	50%	55%	55%	60%	60%	40%	40%	50%	50%	55%	55%
Fimb	% 1RM	75%	75%	75%	75%	80%	80%	85%	85%	75%	75%	85%	85%	80%	80%
WB	% 1RM	60%	60%	65%	65%	70%	70%	75%	75%	60%	60%	65%	65%	60%	60%
NI	% 1RM	60%	60%	65%	65%	70%	70%	75%	75%	60%	60%	65%	65%	60%	60%
	Speed	3 $\times$ 20 m;	3 $\times$ 20 m;	2 $\times$ 20 m;	CMJ;	3 $\times$ 10 m;	3 $\times$ 20 m;	2 $\times$ 20 m;	CMJ;	3 $\times$ 20 m;	2 $\times$ 20 m;	3 $\times$ 20 m;	Hur;	3 $\times$ 20 m;	3 $\times$ 10 m;
	R $\times$ D	6 $\times$ 10 m	6 $\times$ 10 m	2 $\times$ 30 m	4 $\times$ 5	5 $\times$ 5 m	3 $\times$ 10 m	2 $\times$ 10 m;	4 $\times$ 5	3 $\times$ 10 m	4 $\times$ 10 m	3 $\times$ 30 m	4 $\times$ 5	3 $\times$ 10 m	3 $\times$ 20 m
								2 $\times$ 30 m
	SQ S $\times$ R	3 $\times$ 6	3 $\times$ 6	3 $\times$ 5	3 $\times$ 4	3 $\times$ 3	4 $\times$ 4	3 $\times$ 2	3 $\times$ 4	3 $\times$ 5	3 $\times$ 5	2 $\times$ 3	2 $\times$ 3	3 $\times$ 3	3 $\times$ 3
	Jumps	SJ;	Hor;	SJ;	Hor;	SJ;	Hur;	CMJ;	Hor;	Val;	Hor;	SJ;	Hor;	Val;	Hor1P;
	S $\times$ R	3 $\times$ 6	3 $\times$ 6	3 $\times$ 5	3 $\times$ 4	3 $\times$ 3	4 $\times$ 4	3 $\times$ 4	3 $\times$ 4	3 $\times$ 5	3 $\times$ 5	2 $\times$ 3	2 $\times$ 3	3 $\times$ 3	3 $\times$ 3

${}^{*}$ Vimb: group with velocity imbalance; Fimb: group with force imbalance; WB: well-balanced group or low imbalance; NI: not individualised; % 1RM: percentage of one-repetition maximum; R $\times$ D: repetitions $\times$ distance; SQ: Squat; S $\times$ R: sets $\times$ repetitions; Jumps: SJ (Squat Jump); CMJ: countermovement jump; Hur: plyometric hurdles; Hor: Horizontal jump; Hor1P: Horizontal, unilateral jump. ${}^{*}$ Recovery between sets was 2 min for resistance and plyometric training and between 2 to 3 min for sprints.

the day before testing. The Pre- and Post-training tests were conducted in the same rugby facilities. All participants completed the tests in the same order and at the same time of day.

2.2.3 Jump test and force-velocity profile

SJ and CMJ height were assessed using a portable jump mat (Chronojump Boscosystem, Barcelona, Spain) as reported previously [8, 13]. Since take-off and landing positions may affect jump height, participants were instructed to keep their feet extended during the jump until landing. Warm-up consisted of 10 min of non-fatiguing activation and mobilisation exercises, including bodyweight lunges and squats and finally two sets of 4 repetitions of CMJ. To obtain an individual Fvimb, each subject performed a maximal vertical unloaded SJ and against four extra loads (20, 30, 40 and 50 kg) in a randomised order. With regard to Fv profile, mechanical parameters were calculated for all conditions using the method proposed by Samozino et al. [8]. A detailed description of the protocol and testing procedures used in this study has been reported elsewhere [17]. Briefly, it consists of performing SJ, starting from 90 ${}^{\circ }$ in the knee joint, without any counter-movement. The test was performed with hands on the hips for the unloaded SJ and a barbell for the loaded conditions [18], and participants were asked to maintain their individual starting position ( $\sim$ 90 ${}^{\circ }$ knee angle) for about 2 s and then jump for maximum height. Countermovement was verbally forbidden and carefully checked [13]. Two valid trials were performed with each condition with 2 min of recovery between trials and 4–5 min between conditions. SJ conditions were named as follows: SJ (0 kg), SJ20 (20 kg), SJ30 (30 kg), SJ40 (40 kg), SJ50 (50 kg). Intraclass correlation coefficient (ICC) and coefficient of variation (CV) from Pre-training for all jumps were $>$ 0.95 and $<$ 4.9% respectively. Post-training ICC and CV were $>$ 0.97 and $<$ 4.2% respectively. Coefficient R ${}^2$ of Pre- and Post-training were $>$ 0.94.

2.2.4 Isoinertial squat loading test

Strength test consisted of 1RM estimation in the squat exercise, using a linear position transducer (Chronojump Boscosystem, Barcelona, Spain) [19]. Strength assessment began after a specific warm up protocol used previously [20]. Squat was performed using a Squat rack and a barbell, likewise vertical displacement was verified at all times. The participants performed the squat from an upright position, descending at a controlled velocity (0.50–0.70 m $\cdot$ s ${}^{-1}$ ) until the tops of the thighs were below the horizontal plane, then immediately reversed motion and ascended back to the upright position at maximal intended velocity. Participants were not allowed to jump or take off the bar of the shoulders. The initial load was set at 40 kg for all participants and gradually increased by 5–10 kg. When participants reached an average mean propulsive velocity (MPV) near $\sim$ 0.5 m $\cdot$ s ${}^{-1}$ , the test was concluded. Only the best repetition at each load, according to the criteria of fastest MPV, was considered for subsequent analysis. The inter-set recovery time was 3 min. The MPV derived from this test were used to estimate 1RM-SQ, which was calculated from the MPV attained with the heaviest load (kg), as follows: Load (% 1RM) $=$ $-$ 5.961 MPV ${}^2$ $-$ 50.71 MPV $+$ 117.0 [21].

2.2.5 Sprint test

Participants performed three 30 m sprints on a rugby pitch (natural turf) in dry weather conditions. In Pre- and Post-training tests wind speed was less than 2 km $\cdot$ h ${}^{-1}$ , which was measured with a digital anemometer (Perfect-Prime, WD0081, Netherlands). The test started standing, from a 2-point staggered and 0.5 m behind the first timing gate, to prevent any early triggering of the initial gate. Photoelectric cells (Chronojump, Boscosystem, Barcelona, Spain) were placed at the beginning and at 5, 10, 20 and 30 m (1 m above ground level). Participants were encouraged to run the 30 m as quickly as possible. The best 30 m sprint time was considered for subsequent analysis. Recovery time between trials was 3–4 min. Times were recorded in the following distances: 0–5 m (T5); 0–10 m (T10), 0–20 m (T20) and 0–30 m (T30), and in the following split times: 10–20 m (T10–20) and 20–30 m (T20–30). Maximum sprint speed (MSS) was calculated by dividing the best split time by the 10 m interval (m $\cdot$ s ${}^{-1}$ ) and SM was calculated by multiplying the BM of each participant by the MSS of the best measured interval of the 30 m sprint (kg $\cdot$ m ${}^{-1}$ $\cdot$ s ${}^{-1}$ ) [22]. ICC and CV from Pre-training were: T5: 0.84 and 2.4%; T10: 0.94 and 1.3%; T20: 0.97 and 1.0%; T30: 0.98 and 0.9%. Post-training ICC and CV were: T5: 0.84 and 2.4%; T10: 0.91 and 1.9%; T20: 0.97 and 1.0%; T30: 0.98 and 0.7%. All participants completed a 20 min standardised warm-up [15].

Table 4

Changes in sprint and strength performance pre- and post-training (Mean $\pm$ SD)

Group	Variable	Pre		Post		Intra-group $P$ -value		ES		95% CI
Vimb ( $n=$ 6)	T5 (s)	0.90	$\pm$ 0.04	0.88	$\pm$ 0.05	0.	55	$-$ 0.	24	( $-$ 0.90; 1.37)
	T10 (s)	1.66	$\pm$ 0.06	1.62	$\pm$ 0.05	0.	08	$-$ 0.	61	( $-$ 0.56; 1.76)
	T20 (s)	2.96	$\pm$ 0.10	2.92	$\pm$ 0.08	0.	08	$-$ 0.	45	( $-$ 0.70; 1.59)
	T30 (s)	4.16	$\pm$ 0.16	4.12	$\pm$ 0.14	0.	15	$-$ 0.	28	( $-$ 0.86; 1.40)
	MSS (m $\cdot$ s ${}^{-1}$ )	8.31	$\pm$ 0.45	8.31	$\pm$ 0.43	1.	00	0.	00	( $-$ 1.13; 1.13)
	SM (kg $\cdot$ m $\cdot$ s ${}^{-1}$ )	682.00	$\pm$ 109.23	684.00	$\pm$ 103.42	0.	81	0.	02	( $-$ 1.14; 1.11)
	1RM-SQ (kg)	129.15	$\pm$ 23.26	136.01	$\pm$ 12.21	0.	11	0.	36	( $-$ 1.5; 0.78)
Fimb ( $n=$ 11)	T5 (s)	0.93	$\pm$ 0.03	0.92	$\pm$ 0.04	0.	37	$-$ 0.	30	( $-$ 0.54; 1.13)
	T10 (s)	1.70	$\pm$ 0.06	1.68	$\pm$ 0.06	0.	28	$-$ 0.	25	( $-$ 0.59; 1.08)
	T20 (s)	3.04	$\pm$ 0.11	3.00	$\pm$ 0.10	0.	05	$-$ 0.	31	( $-$ 0.52; 1.15)
	T30 (s)	4.28	$\pm$ 0.17	4.23	$\pm$ 0.16	0.	03	$-$ 0.	29	( $-$ 0.55; 1.12)
	MSS (m $\cdot$ s ${}^{-1}$ )	8.04	$\pm$ 0.36	8.11	$\pm$ 0.40	0.	18	0.	18	( $-$ 1.02; 0.65)
	SM (kg $\cdot$ m $\cdot$ s ${}^{-1}$ )	715.50	$\pm$ 63.15	725.61	$\pm$ 69.25	0.	11	0.	15	( $-$ 0.98; 0.68)
	1RM-SQ (kg)	139.67	$\pm$ 19.15	145.48	$\pm$ 15.66	0.	07	0.	33	( $-$ 1.16; 0.51)
WB ( $n=$ 9)	T5 (s)	0.91	$\pm$ 0.06	0.89	$\pm$ 0.04	0.	10	$-$ 0.	48	( $-$ 0.46; 1.41)
	T10 (s)	1.67	$\pm$ 0.09	1.62	$\pm$ 0.06	0.	002	$-$ 0.	66	( $-$ 0.30; 1.60)
	T20 (s)	2.97	$\pm$ 0.16	2.91	$\pm$ 0.11	0.	007	$-$ 0.	38	( $-$ 0.55; 1.31)
	T30 (s)	4.23	$\pm$ 0.19	4.17	$\pm$ 0.14	0.	02	$-$ 0.	34	( $-$ 0.58; 1.27)
	MSS (m $\cdot$ s ${}^{-1}$ )	8.15	$\pm$ 0.34	8.28	$\pm$ 0.35	0.	04	0.	35	( $-$ 1.28; 0.57)
	SM (kg $\cdot$ m $\cdot$ s ${}^{-1}$ )	759.54	$\pm$ 120.03	775.24	$\pm$ 124.51	0.	03	0.	12	( $-$ 1.05; 0.79)
	1RM-SQ (kg)	152.54	$\pm$ 20.04	155.35	$\pm$ 19.66	0.	41	0.	14	( $-$ 1.06; 0.78)
NI ( $n=$ 8)	T5 (s)	0.91	$\pm$ 0.04	0.89	$\pm$ 0.06	0.	33	$-$ 0.	28	( $-$ 0.70; 1.26)
	T10 (s)	1.68	$\pm$ 0.07	1.69	$\pm$ 0.08	0.	69	0.	09	( $-$ 1.13; 0.95)
	T20 (s)	3.03	$\pm$ 0.14	3.02	$\pm$ 0.15	0.	84	$-$ 0.	02	( $-$ 1.01; 1.07)
	T30 (s)	4.21	$\pm$ 0.24	4.18	$\pm$ 0.26	0.	20	$-$ 0.	12	( $-$ 0.85; 1.10)
	MSS (m $\cdot$ s ${}^{-1}$ )	8.12	$\pm$ 0.52	8.23	$\pm$ 0.61	0.	10	0.	18	( $-$ 1.16; 0.79)
	SM (kg $\cdot$ m $\cdot$ s ${}^{-1}$ )	681.54	$\pm$ 99.20	698.09	$\pm$ 91.41	0.	03	0.	17	( $-$ 1.15; 0.81)
	1RM-SQ (kg)	124.15	$\pm$ 23.34	129.52	$\pm$ 19.15 ${}^{\text{WB}}$	0.	15	0.	25	( $-$ 1.23; 0.73)

Groups: velocity imbalance (Vimb); force imbalance (Fimb); well-balanced (WB); Not individualised (NI); ES: effect size of intra-group effects; 95% CI: confidence interval; T5: sprint time in 5 m; T10: sprint time in 10 m; T20: sprint time in 20 m; T30: sprint time in 30 m; MSS: maximum sprint speed; SM: sprint momentum; 1RM-SQ: estimated one-repetition maximum in the squat exercise. Inter-group significant differences at the corresponding time-point: ${}^{\text{Vimb}}$ indicates significant differences with Vimb group; ${}^{\text{Fimb}}$ indicates significant differences with Fimb group; ${}^{\text{WB}}$ indicates significant differences with WB group. Time effects ( $P$ -values): T5: 0.05; T10: 0.007; T20: 0.002; T30: 0.001; MSS: 0.002; SM: 0.004; 1RM-SQ: 0.007. Group $\times$ time interactions ( $P$ -values): T5: 0.91; T10: 0.08; T20: 0.38; T30: 0.88; MSS: 0.58; SM: 0.53; 1RM-SQ: 0.88.

2.3 Statistical analysis

Standard statistical methods were used for the calculation of means, standard deviations (SD). Shapiro-Wilk test were used to analyse if the values were nor- mally distributed. Data were analysed using a 4 $\times$ 2 factorial ANOVA with Bonferroni’s post-hoc comparisons using one between-group factor (Vimb vs. Fimb vs. WB vs. NI) and one within-group factor (Pre- vs. Post-training). Additionally, these differences were further analysed based on Cohen’s “d” Effect Size (ES) (95% confidence interval). Significance was accepted at the $p⩽$ 0.05 level. Statistical analyses were performed using SPSS software version 20.0 (SPSS Inc., Chicago, IL, USA).

3. Results

3.1 Running sprint performance

No significant “group” $\times$ “time” interactions were observed for any running sprint performance (Table 4). Significant “time” effects ( $P<$ 0.05) were found for all parameters examined (T5, T10, T20, T30, MSS, and SM). The WB group induced significant enhancements in T10, T20, T30, MSS and SM, whereas Fimb attained significant gains in T20 and T30 performance (Table 4). Moreover, the Vimb group showed almost significant improvements ( $P=$ 0.08) in T10 and T20 performance. The NI group attained significant improvements ( $P<$ 0.05) in SM. The ES values are reported in Table 4.

Table 5

Changes in force-velocity mechanical variables and jump height pre- and post-training (Mean $\pm$ SD)

Group	Variable	Pre		Post		Intra-group $P$ -value		ES		95% CI
Vimb ( $n=$ 6)	V0 (m $\cdot$ s ${}^{-1}$ )	2.14	$\pm$ 0.21	2.57	$\pm$ 0.36	0.	04	1.	43	( $-$ 2.68; $-$ 0.11)
	F0 (N $\cdot$ kg ${}^{-1}$ )	46.45	$\pm$ 7.46	40.07	$\pm$ 5.86	0.	002	$-$ 0.	95	( $-$ 0.27; 2.13)
	Pmax (W $\cdot$ kg ${}^{-1}$ )	24.75	$\pm$ 3.25	25.73	$\pm$ 5.15	0.	31	0.	23	( $-$ 1.35; 0.91)
	Fvimb (%)	133.50	$\pm$ 28.32	95.17	$\pm$ 17.27	0.	001	$-$ 1.	63	(0.27; 2.93)
	CMJ (cm)	34.92	$\pm$ 4.80	34.43	$\pm$ 4.98	0.	62	$-$ 0.	10	( $-$ 1.03; 1.23)
	SJ (cm)	29.61	$\pm$ 2.50	31.15	$\pm$ 4.86	0.	15	0.	40	( $-$ 1.53; 0.75)
Fimb ( $n=$ 11)	V0 (m $\cdot$ s ${}^{-1}$ )	3.04	$\pm$ 0.44 ${}^{\text{Vimb}}$	2.59	$\pm$ 0.28	0.	005	$-$ 1.	20	(0.28; 2.11)
	F0 (N $\cdot$ kg ${}^{-1}$ )	32.28	$\pm$ 3.48 ${}^{\text{Vimb}}$	35.80	$\pm$ 3.52	0.	02	1.	00	( $-$ 1.88; $-$ 0.10)
	Pmax (W $\cdot$ kg ${}^{-1}$ )	24.46	$\pm$ 3.79	23.13	$\pm$ 3.15	0.	07	$-$ 0.	38	( $-$ 0.46; 1.22)
	Fvimb (%)	68.82	$\pm$ 8.52 ${}^{\text{Vimb}}$	87.55	$\pm$ 12.99	0.	02	1.	70	( $-$ 2.67; $-$ 0.70)
	CMJ (cm)	32.70	$\pm$ 4.25	33.26	$\pm$ 4.06	0.	44	0.	13	( $-$ 0.97; 0.70)
	SJ (cm)	29.47	$\pm$ 4.66	28.96	$\pm$ 3.44	0.	50	$-$ 0.	12	( $-$ 0.71; 0.96)
WB ( $n=$ 9)	V0 (m $\cdot$ s ${}^{-1}$ )	2.63	$\pm$ 0.26 ${}^{\text{Vimb, Fimb}}$	2.49	$\pm$ 0.32	0.	41	$-$ 0.	46	( $-$ 0.47; 1.39)
	F0 (N $\cdot$ kg ${}^{-1}$ )	37.04	$\pm$ 2.63 ${}^{\text{Vimb}}$	38.33	$\pm$ 4.75	0.	474	0.	33	( $-$ 1.26; 0.60)
	Pmax (W $\cdot$ kg ${}^{-1}$ )	24.37	$\pm$ 3.03	23.62	$\pm$ 2.45	0.	34	$-$ 0.	27	( $-$ 0.65; 1.19)
	Fvimb (%)	97.33	$\pm$ 6.04 ${}^{\text{Vimb, Fimb}}$	99.44	$\pm$ 23.39	0.	81	0.	26	( $-$ 1.21; 0.69)
	CMJ (cm)	35.78	$\pm$ 6.40	34.99	$\pm$ 5.27	0.	34	$-$ 0.	13	( $-$ 0.79; 1.05)
	SJ (cm)	30.60	$\pm$ 4.88	29.69	$\pm$ 3.84	0.	29	$-$ 0.	20	( $-$ 0.72; 1.13)
NI ( $n=$ 8)	V0 (m $\cdot$ s ${}^{-1}$ )	2.68	$\pm$ 0.21 ${}^{\text{Vimb}}$	2.72	$\pm$ 0.52	0.	82	0.	09	( $-$ 1.07; 0.88)
	F0 (N $\cdot$ kg ${}^{-1}$ )	33.74	$\pm$ 3.82 ${}^{\text{Vimb}}$	33.55	$\pm$ 4.72	0.	919	$-$ 0.	04	( $-$ 0.93, 1.02)
	Pmax (W $\cdot$ kg ${}^{-1}$ )	22.65	$\pm$ 3.29	22.57	$\pm$ 3.49	0.	92	$-$ 0.	02	( $-$ 0.95; 1.00)
	Fvimb (%)	81.38	$\pm$ 14.76 ${}^{\text{Vimb}}$	81.38	$\pm$ 20.01	1.	00	0.	00	( $-$ 0.97; 0.97)
	CMJ (cm)	34.24	$\pm$ 5.33	32.26	$\pm$ 5.02	0.	03	$-$ 0.	38	( $-$ 0.63; 1.36)
	SJ (cm)	28.68	$\pm$ 4.36	28.61	$\pm$ 4.34	0.	34	$-$ 0.	01	( $-$ 0.96; 0.99)

Groups: velocity imbalance (Vimb); force imbalance (Fimb); well-balanced (WB); Not individualised (NI); ES: effect size of intra-group effects; 95% CI: confidence interval; V0: theoretical maximum velocity; F0: theoretical maximum force; PMax: maximum power; Fvimb: imbalance of the force-velocity profile; CMJ: countermovement jump; SJ: squat jump; Fvimb: imbalance of the force-velocity profile. Inter-group significant differences at the corresponding time-point: ${}^{\text{Vimb}}$ indicates significant differences with Vimb group; ${}^{\text{Fimb}}$ indicates significant differences with Fimb group. Time effects ( $P$ -values): V0: 0.72; F0: 0.60; Pmax: 0.48; Fvimb: 0.33; CMJ: 0.12; SJ: 0.98. Group $\times$ time interactions ( $P$ -values): V0: 0.01; F0: 0.003; Pmax: 0.26; Fvimb: 0.002; CMJ: 0.18; SJ: 0.32.

3.2 Squat strength and jump height

No significant “group” $\times$ “time” interactions were observed for squat and jump performance. Significant “time” effects ( $P<$ 0.05) were found for 1RM-SQ, whereas no “time” effects were observed for jump height (CMJ and SJ). Although no significant intra-groups improvements were observed for 1RM-SQ, the Fimb group attained almost significant gains ( $P=$ 0.07) in this parameter. The NI group showed a significant decrease ( $P<$ 0.05) in SJ performance.

3.3 Force-velocity profile

Significant “group” $\times$ “time” interactions ( $P<$ 0.05) were observed for V0, F0 and Fvimb (Table 5). Since groups were divided according their Fvimb, significant differences between groups ( $P<$ 0.005, Table 5) were observed in V0, F0 and Fvimb at Pre-training. However, these differences disappeared at Post-training. The Vimb group significantly ( $P<$ 0.05) increased V0, whereas decreased F0 and Fvimb. The Fimb group showed significant ( $P<$ 0.05) decreases in V0, whereas increased F0 and Fvimb. The WB and NI groups did not show any significant change in these parameters. The ES values are presented in Table 5.

4. Discussion

The aim of this study was to experimentally examine the effects on lower limbs strength, jump and sprint performance after following an individualised RT programme based on Fvimb. In the present study, results showed that individualised RT programme based on Fvimb induce improvements in sprint performance, whereas the NI athletes showed improvements only in SM. Moreover, individualised RT programmes based on Fvimb induced better adjustments of the Fv profile to the theoretical “optimal” Fv profile, although none of the groups improved 1RM-SQ strength, jump height or Pmax.

In rugby, sprint speed, BM and SM discriminate between levels of play [23]. Additionally, WB athletes improved not only sprint times in all intervals but MSS and SM. Moreover, the other “individualised” groups (Vimb and Fimb) also improved sprint performance. Young et al. [24] mentioned that sprints from a static start in a relatively short distance, such as 10 m, is a reflection of the acceleration ability, and therefore it could be expected an improvement in their initial sprint momentum. Baker and Newton [25] suggested that rugby players should focus on increasing strength and power along with muscle hypertrophy, improving on this way sprint momentum, which would be beneficial to cope with the contacts and tackles demands during games [26].

Likewise, and despite the substantial changes found in Fvimb across all “individualised” groups, no improvements were observed in jumping ability or Pmax after the intervention programme in highly trained rugby players. These findings do not support previous findings by Jimenez-Reyes et al. [13], who reported that reducing Fvimb without even increasing Pmax lead to clearly beneficial jump performance changes. Our findings do not support this suggestion, since none of the intervention groups increased jump performance despite obtaining a more “optimal” Fv profile. Regardless of this lack of “positive” results, Fvimb could be considered as a potentially useful variable for prescribing optimal RT loads to improve sprint and lower limbs strength. It is plausible to suggest that the lack of effectiveness in jumping ability in all groups could be due, in part, to the presence of neuromuscular fatigue that is accumulating throughout the season, as reported recently with high level athletes [27]. Moreover, another possible explanation could be that rugby players participating in the present study possibly needed more weeks to show positive effects on jumping ability and Pmax.

Regarding RT, it has been shown that higher levels of strength are associated with improvements in Fv characteristics, general and specific sports skills [28]. The results reported in the current study confirm that when imbalances are not individually addressed, as for NI athletes, the training programme was not effective in reducing Fvimb or improving jumping ability. In line with the findings of Jiménez-Reyes et al. [13] athletes participating in training programmes based on their Fvimb will get better results (at least they did not decrease jump performance) than those who participated in traditional training. It is likely, therefore, that an individualised RT based on Fvimb will improve strength and sprint performance in lower limbs, although a question arises as improvements in jumping ability did not occur. In this regard, Jiménez-Reyes et al. [13] only analysed changes in jumping ability. In addition, although no significant intra-groups improvements were observed for 1RM-SQ, the Fimb group was the only one that attained nearly significant gains ( $P=$ 0.07) in 1RM-SQ. In this regard, it seems that carrying out an individualised RT programme based on Fvimb induced increases in 1RM-SQ strength. However, it should be recognised that improvements in maximal strength through RT is not unexpected. Previous studies have clearly shown the effectiveness of strength training aiming at specifically increasing maximal force [29, 30, 31, 32].

The results in the mechanical variables (Table 5), showed positive changes in the area of the curve that was prioritised in their training and, in turn, reduced the opposite one. These findings are in agreement with Jiménez-Reyes et al. [13], who reported that when training was focused on increasing V0, a parallel decrease in F0 was observed, and vice-versa. Therefore, caution should be taken when only is prioritised an extreme of the Fv spectrum aiming to attain the “optimal” profile, because it can induce decreases in the opposite extreme, being this a plausible explanation of the lack of positive adaptations in squat strength and jump performance. Therefore, the interpretation of Fv profile should be considered taking into account the changes in F0 and V0, since a decrease in both variables (F0 and V0) could result in lower Fvimb but could induce a lower athletic performance. Results in Vimb players, are consistent with previous findings in youth soccer players, in which after a RT program with low loads and low volume combined with jumps and sprints, induced important improvements in performance [20]. Regarding the use of heavy loads in RT, a study reported improvements in muscle performance in the upper portion of the force-velocity curve [33]. In Fimb athletes, results showed almost significant ( $P=$ 0.07) improvements in 1RM-SQ, which is in line with the increases in F0.

The aim of a RT programme with different loads is to target all areas of the Fv curve as an attempt to increase adaptations across the whole Fv spectrum [33]. Although no changes were found in jump height and 1RM-SQ, WB players improved sprint performance across distances from T10 to T30. Therefore, these findings suggest that strategies combining low and high loads could be useful for the development of sprinting speed in rugby players with a well-balanced Fv profile. An interesting finding of the current study was the fact that WB had significant changes in some testing variables, whereas the NI athletes did not, despite an identical loading regime was carried out. Consequently, it is plausible that more significant improvements in athletic performance may occur following an individualised RT programme when taking into consideration individual Fvimb. However, this hypothesis needs to be further investigated. Present findings are limited to the specific population analysed, thus caution must be made, as when training is applied to high-level team sports athletes, the effects of periodised RT programmes on strength and power seem to follow the law of diminishing returns, while training exposure increases, adaptation rates are reduced [34].

We invite further research to address some limitations of this study and analyse; a) the effect of individualised loading prescription using individual Fvimb, without modifications on the approach used by Jimenez-Reyes [13, 14]; b) the effect of training programmes with more training sessions and weeks, since only 14 training sessions during 7 weeks were performed in this study. Additionally, a larger sample size would be needed since recruitment led to a Vimb group of only 6 participants while Fimb, WB and NI groups included 11, 9 and 8 participants, respectively.

In practical terms, rugby player’s assessment should include Fvimb to individualise training loads according to the needs of each athlete, aiming to identify in which area of the Fv spectrum an athlete needs to prioritize their RT. However, it seems that the entire Fv spectrum should be targeted (although one zone is prioritised) to avoid decreases in the untrained zone. Therefore, strength and conditioning coaches should also assess 1RM strength (to determine exactly the relative and absolute loads to be used) and sprint speed. The combination of these assessments together with well-designed training programmes can lead to improvements in lower limbs absolute strength and sprint performance, which are key qualities in rugby union players.

5. Conclusion

In conclusion, this study presents data of the effects on key physical characteristics of rugby union players when addressing force-velocity imbalances. The individualisation of a training programme based on Fvimb could be an effective method to improve sprint and strength performance. Reducing the actual force-velocity imbalances, may have a positive impact in other maximal actions such as acceleration ability, speed and strength. However, no positive changes were observed in jump performance for any of the intervention groups.