Asymmetry in internal oblique muscle thickness: Not the key to a cricket fast bowler's performance

Abstract

Purpose

Previous studies found that trunk muscle asymmetry may play a role in preventing injury in cricket fast bowlers, while the association with bowling performance has not been investigated. This study aims to describe the side-to-side differences in trunk muscle thickness and determine the association between bowling performance and these side-to-side differences in trunk muscle thickness in adolescent fast bowlers.

Methods

In this observational cross-sectional study, bowling performance, namely ball release speed and bowling accuracy, was recorded in adolescent fast bowlers. Ultrasound imaging measured external oblique, internal oblique, transversus abdominis and lumbar multifidus muscle thickness.

Results

Fast bowlers (n = 46) with a mean age of 15.9 (±1.2) years participated. On the non-dominant side, the external oblique and internal oblique at rest were thicker than on the dominant side (external oblique: p = 0.011, effect size = 0.27; internal oblique: p < 0.0001, effect size = 0.40), while the transversus abdominus (p = 0.72, effect size = 0.19) and lumbar multifidus (p = 0.668, effect size = 0.04) were symmetrical. Weak correlations existed between bowling performance and the side-to-side differences in the thickness in all muscles, except for two moderate correlations: 1. The smaller the side-to-side difference in absolute thickness of the external oblique when contracted, the faster the ball release speed (Spearman's (ρ) = −0.455, p = 0.002). 2. Also, a smaller side-to-side difference in external oblique contraction ratio (Spearman's (ρ) = −0.495, p = 0.0001) was associated with faster ball release speed.

Conclusions

No relationship between bowling performance and side-to-side differences in internal oblique muscle thickness could be established, while more symmetrical external oblique muscles may be linked to faster ball release speeds.

Keywords

Abdominal muscles adolescence ball release speed sports technique trunk

Introduction

In cricket, the fast-bowling action is a repetitive, asymmetrical action, during which one arm extends above the head. At the same time, the trunk goes into extension, flexion, lateral flexion and rotation, often near the end of its anatomical range of movement.¹ Ground reaction forces and consequently lumbar spine segment loads are at their highest during the power phase – between front foot contact and ball release – of the bowling action.² The inherent nature of the fast-bowling action predisposes the bowler to injury while allowing for high ball release speeds to be attained.¹

The performance of a cricket fast bowler is judged in terms of the speed with which a ball can be delivered and how accurately the delivery is placed in relation to the wickets behind the batsman on strike. Various intrinsic determinants of ball release speed and accuracy have been explored in the literature, including anthropometrics,^3–6 bowling biomechanics,^7–10 run-up velocity,^5,11,12 lower body strength⁴ and trunk muscle stability.¹³ However, side-to-side differences in the thickness of the abdominal muscles and lumbar multifidi i.e., asymmetry in muscle morphometry, have not been investigated as a determinant of ball release speed or accuracy.

As a result of the asymmetrical bowling technique actioned repeatedly, asymmetry in the morphometry, particularly in terms of the thickness of fast bowlers’ abdominal muscles, is a common phenomenon.^14–16 Amongst injury-free bowlers, the internal oblique (IO) muscle of the non-dominant side was thicker than that of the dominant side.^14–16 On the other hand, bowlers who experienced lower back pain at the time of testing,¹⁵ as well as those who were injury-free at the time of testing but who went on to sustain an injury during the cricket season,¹⁴ showed no side-to-side differences in terms of thickness of the IO muscle. The symmetry in trunk muscle morphometry may play a role in lower back pain in cricket fast bowlers.

In the lumbar multifidi however, asymmetry seems to be a precursor to, rather than protective against, injury. In a study by Olivier et al.¹⁷ injury-free bowlers presented with symmetry of the lumbar multifidi muscles. However, amongst injury-free bowlers who were assessed at the start of the cricket season but who sustained a lower back injury during the season, the cross-sectional area of the non-dominant lumbar multifidi at L3 and L5 was smaller than the dominant side.¹⁷ Toumazou et al.¹⁸ also found no side-to-side differences in the lumbar multifidi thickness at L5/S1 level when measured at rest amongst injury-free fast bowlers. In contrast, Hides et al.¹⁹ found that cricketers with lower back pain had more symmetry in their lumbar multifidi muscles, while in those without lower back pain, lumbar multifidus measured larger on the dominant side. It should be noted that Hides et al.'s¹⁹ participants were not bowlers specifically, but cricketers in general, while the participants in the studies of Olivier et al.¹⁷ and Toumazou et al.¹⁸ were classified as fast bowlers. Therefore, it is suggested that the fast-bowling action is responsible for a unique pattern of muscle adaptation in fast bowlers.

The link between trunk muscle thickness, as measured through ultrasound, and injury has been determined, but no evidence exists of the relationship between trunk muscle thickness and bowling performance. It is crucial to determine this link specifically in the context of the unique sporting demands of the fast bowler. The fast bowler's main aim is to perform well on the field, i.e., to deliver the ball in a way that makes it difficult for the batter to accurately address the delivery while remaining injury-free to optimise bowling performance. Therefore, it is important to explore both the link between side-to-side differences in trunk muscle thickness and injury and the relationship between these same side-to-side differences and performance to get a holistic view. An association (or lack thereof) between bowling performance and trunk muscle thickness can form the basis of future research where cause and effect can be established. Hence, we set out to describe the side-to-side differences in trunk muscle thickness and determine the association between bowling performance (ball release speed and bowling accuracy) and these side-to-side differences in trunk muscle thickness in adolescent fast bowlers. We furthermore investigate the role of age, height and weight as potential confounding factors.

Methods

Study design and setting

The data collection for this observational cross-sectional study took place in the outdoor nets of the respective schools.

Study participants

Study participants consisted of male fast bowlers between the ages of 13 and 18, who were nominated for the study by their A-team coaches based on the accepted condition that a wicketkeeper would stand back from the stumps for a fast bowler.²⁰ Participants with previous upper or lower limb surgery or extensive lower back interventions such as spinal facet infiltrations were excluded from the study. Ethical clearance was obtained by the Human Ethics Research Committee of the associated tertiary institution, and parents and participants respectively signed both consent and assent.

Instrumentation and outcome measures

Muscle thickness was measured using a DP-6600 Digital ultrasonic imaging system® (Shenzhen Mindray Bio-medical Electronics Co., Ltd, China) with a 5 MHz curvilinear transducer and a large footprint (≥60 mm). Ultrasound imaging (USI) is an accurate, reliable and non-invasive way of measuring the abdominal and lumbar multifidi muscles to report on the morphometry thereof.²¹ Koppenhaver, Hebert²² investigated the intra-rater reliability of USI in the measurement of Transversus abdominis (TA) and lumbar multifidus on the same day and found it to be very high or excellent at 0.96 to 0.99 (intraclass correlation coefficient (ICC_3,2)). Hides et al.²³ studied the intra-rater reliability of a novice sonographer with eight hours of training and found that the ICC when assessing muscle thickness on the same image was very high (> 0.97). Ultrasound imaging has been validated as reliable against magnetic resonance imaging (MRI) in the measurement of lumbar multifidus and TA.²⁴ Concerning the calculation of percentage change, the reliability of ultrasound has been validated against electromyography (EMG) during voluntary contractions of TA and IO.^25,26

Ball release speed was measured using a hand-held radar speed gun (Stalker, ATS, Texas).²⁷ The average ball release speed was used in the analysis. Bowling accuracy was measured using a black shade cloth target with scoring zones sewn onto it in white and a horizontal line in red, 50 centimetres off the ground (Figure 1).^6,28 A maximum of 100 points was scored if the ball made contact in the zone of the off-stump, 50 and 25 points in adjacent zones and zero if the target was missed altogether or the ball hit the target below the red line. The median accuracy score was used in the analysis.

Figure 1.

Bowling accuracy target based on the work of Elliott et al.²⁸ and Portus et al.⁶.

Procedures

After having warmed up in their own preferred manner, participants bowled six match pace deliveries (one over) using a white County Premier match ball weighing 156 g while ball release speed and accuracy were recorded. The outdoor nets comprised of a standard pitch length of 20.12 m and an artificial surface.

Dominant and non-dominant side muscle thickness of TA, external oblique (EO), and IO was measured at rest and during the abdominal drawing-in manoeuvre using ultrasound imaging according to the methods used by Gray et al.¹⁵ Ultrasound imaging was measured after the bowling speed and accuracy assessment for all bowlers. The participants were asked to lie supine in a supine position on a plinth, with their legs supported over a pillow. The probe was placed in contact with the skin, using heated ultrasound gel halfway between the inferior border of the ribcage and the iliac crest, in the mid-axillary line. The participants were instructed to breathe in and hold at the end of expiration to measure the abdominal wall at rest The three muscle layers of the abdominal wall were hereafter measured in the contracted state while the participant performed the “draw-in” manoeuvre. This frame was frozen, saved and measured. Additionally, the lumbar multifidus muscle at the L4, L5 facet joint was also measured at rest from the inferior layer of the subcutaneous fat layer to the deepest bony landmark of the facet. Measurements were taken in millimetres with on-screen callipers.

Data reduction

Due to both right- and left-handed bowlers in the sample, all data were presented as dominant and non-dominant instead of left and right. The following parameters were determined and were included as derivatives of absolute muscle thickness to allow for comparison with previous studies:

Percentage change = a percentage of muscle thickness at rest: (muscle activated - muscle at rest) ÷ muscle at rest × 100.¹⁶

Contraction ratio = muscle thickness contracted/muscle thickness at rest²⁹

TA preferential activation ratio (difference in the TA proportion of the total lateral abdominal muscle thickness in going from the relaxed to the contracted state) = (TrA contracted/TrA + OI + OE contracted) – (TrA at rest/TrA + OI + OE at rest).²⁹

Relative thickness = a percentage of total thickness of all three muscles together - EOrest / (EOrest + IOrest + TArest) × 100.²⁹

Relative thickness is expressed as a percentage of total thickness of all three muscles together. EOcont / (EOcont + IOcont + TAcont).²⁹

Percentage difference between sides = [(largest/smallest value × 100) – 100].^17,30

Statistical analyses

Data were analysed using IBM SPSS Statistics for Windows (version 26.0. Armonk, New York, USA). Non-parametric tests were used as the data violated assumptions for normality. Median differences in trunk muscle thickness for dominant versus non-dominant side were assessed using Wilcoxon's signed rank test According to the principles recommended by the American Statistical Association, no arbitrary cut-off was set for the p-value,^31,32 however, effect sizes were used as an additional means of interpreting the findings. The effect size was calculated using the equation $r = z / \sqrt{N}$ , where r is the effect size coefficient, z is the z-statistic from the Wilcoxon's test and N is the number of observations for each pair.^33,34 Effect sizes of 0.2, 0.5 and 0.8 were interpreted as small, medium and large, respectively.

Spearman's Rank order correlation was conducted to determine the relationship between absolute differences in side-to-side muscle thickness, and ball release speed and accuracy. A value of 1 or −1 for Spearman's (ρ) correlation was considered a perfect association. Positive and negative values above 0.7 were considered strong correlations, those between 0.4 and 0.7 were moderate correlations, and those below 0.4 were weak correlations. Furthermore, the association between age, height and weight with differences in side-to-side muscle thickness was also determined using Spearman's Rank order correlation.

Results

Participants and bowling performance

A total of 46 fast bowlers participated in this study. The mean age (standard deviation (SD)) was 15.9 (1.2) years. Anthropometric measurements revealed a mean (SD) weight of 65.0 (16.8) kg and height of 173.9 (8.0) cm. The majority of fast bowlers (n = 42, 91.3%) were right-hand bowlers.

Side-to-side differences

Table 1 shows the muscle thickness of the IO, EO, TA and multifidus muscles. In terms of side-to-side comparison of absolute muscle thickness, on the non-dominant side, the EO and IO at rest were thicker than on the dominant side (EO: p = 0.011, effect size = 0.27; IO: p < 0.0001, effect size = 0.40), while the transversus abdominus (p = 0.72, effect size = 0.19) and lumbar multifidus (p = 0.668, effect size = 0.04) were symmetrical. Only the IO on the non-dominant side was thicker than on the dominant side while in the contracted state (p < 0.0001, effect size = 0.44).

Table 1.

Muscle thickness on the non-dominant and dominant sides at rest and in the contracted state (n = 46).

	Non-dominant side	Dominant Side	Absolute Difference
	Median (range)	Median (range)	Median (range)	p Value	Effect size
EO at rest (mm)	6.48 (3.87–12.33)	5.83 (3.44–11.73)	0.41 (−4.45–6.65)	0.011	0.27
IO at rest (mm)	10.87 (6.94–17.73)	9.10 (5.57–14.23)	1.81 (−1.94–2.27)	<0.0001	0.40
TA at rest (mm)	4.56 (2.91–7.64)	4.72 (2.78–8.21)	0.22 (−2.69–1.37)	0.072	0.19
Total thickness at rest (mm)	21.86 (15.32–32.41)	19.24 (13.06–34.17)	2.18 (−12.31–12.1)	<0.0001	0.38
EO contracted (mm)	5.98 (3.25–13.63)	5.42 (3.29–13.3)	0.27 (−4.08–5.75)	0.148	0.15
IO contracted (mm)	12.49 (7.87–20.07)	10.65 (3.77–18.3)	1.80 (−4.20–8.42)	<0.0001	0.44
TA contracted (mm)	7.67 (4.77–11.5)	7.62 (3.90–13.4)	0.12 (−5.92–2.24)	0.394	0.09
Total thickness contracted (mm)	27.16 (18.54–37.54)	24.86 (13.6–37.9)	2.60 (−7.25–11.1)	<0.0001	0.43
Multifidus (mm) at rest	31.04 (22.9–41.5)	31.40 (24.57–43.63)	0.05 (−3.96–6.47)	0.668	0.04
IO % change	11.91 (−26.48–78.80)	13.70 (−32.32–116.14)	2.12 (−78.17–75.32)	0.627	0.05
TA % change	72.74 (15.37–136.77)	67.50 (−2.74–130.87)	9.70 (−59.32–70.99)	0.112	0.17
EO contraction ratio	0.95 (0.54–1.91)	0.98 (0.49–1.58)	0.01 (−0.60–0.73)	0.468	0.08
IO contraction ratio	1.12 (0.74–1.79)	1.14 (0.68–2.16)	0.02 (−0.78–0.75)	0.627	0.05
TA contraction ratio	1.73 (1.15–2.37)	1.68 (0.97–2.31)	0.10 (−0.59–0.71)	0.112	0.17
EO-IO contraction ratio	1.05 (0.71–1.77)	1.05 (0.83–1.74)	0.02 (−0.72–0.76)	0.987	0.002
TA preferential activation ratio	0.09 (−0.05–0.16)	0.09 (−0.02–0.17)	0.01 (−0.12–0.09)	0.806	0.03
EO relative thickness at rest (%)	29.54 (18.55–45.25)	29.84 (17.85–38.94)	0.48 (−7.59–16.92)	0.797	0.03
IO relative thickness at rest (%)	49.09 (31.74–61.81)	46.18 (35.50–55.22)	3.09 (−13.96–14.81)	0.002	0.32
TA relative thickness at rest (%)	20.91 (12.23–34.95)	24.20 (15.71–32.48)	3.15 (−14.47–10.92)	0.001	0.36
EO relative thickness contracted (%)	23.12 (11.64–38.46)	24.11 (12.68–43.60)	0.40 (−16.1–16.51)	0.148	0.15
IO relative thickness contracted (%)	46.54 (31.95–62.27)	41.90 (27.72–62.39)	3.57 (−7.36–19.33)	0.001	0.36
TA relative thickness contracted (%)	30.26 (19.93–39.32)	32.54 (24.65–45.29)	2.92 (−20.28–8.72)	0.001	0.36

EO = external oblique; IO = internal oblique; TA = Transversus abdominis.

Total thickness = EO + IO + TA.

With reference to the side-to-side differences in the derivatives of absolute muscle thickness, no side-to-side difference was found in percentage change and contraction ratio for any of the muscles. The relative thickness at rest as well as in the contracted state of the IO was thicker on the non-dominant side than on the dominant side (p = 0.001, effect size = 0.36), while the opposite was found for the TA where the dominant side was thicker (p = 0.001, effect size = 0.36).

The percentage difference between the dominant and non-dominant sides are shown in Table 2. The IO percentage difference is the largest at rest (median = 24.76%) as well as the contracted state (median = 22.03%) compared to the other two abdominal muscles (EO at rest median = 14.31%, contracted median = 11.57%; TA at rest median = 11.37%, contracted median = 7.73%). A relatively small percentage difference is noted for the lumbar multifidus (median = 2.69%)

Table 2.

Percentage difference between the non-dominant and dominant sides (n = 46).

	Median (range) Mm	95% CI Mm
EO % difference at rest	14.31 (0.42–143.94)	7.23;21.96
IO % difference at rest	24.76 (0.59–133.93)	18.31;38.85
TA% difference at rest	11.37 (0–51.34)	6.26;14,70
Total thickness % difference at rest	15.45 (1.12– 65.07)	9.51;22.39
EO % difference contracted	11.57 (0–127.21)	7.70;21.69
IO % difference contracted	22.03 (0.32–154.11)	15.04;33.80
TA % difference contracted	7.73 (0–79.14)	4.13;10.88
Total thickness % difference contracted	12.73 (0.52–49.71)	8.20;17.37
Multifidus % difference	2.69 (0–25.54)	1.87;4.39

SD = standard deviation; CI = confidence interval; mm = millimetre; EO = external oblique; IO = internal oblique; TA = Transversus abdominis.

Total thickness = EO + IO + TA.

The relationship between ball release speed and accuracy, and side-to-side differences in muscle thickness

Table 3 shows the association between ball release speed and side-to-side differences in trunk muscle thickness. Two moderate correlations were identified. The smaller the side-to-side difference in absolute thickness of the EO when contracted the faster the ball release speed (Spearman's (ρ) = −0.455, p = 0.002). Also, a smaller side-to-side difference in EO contraction ratio (Spearman's (ρ) = −0.495, p = 0.0001) is associated with faster ball release speed. The side-to-side difference in IO was not associated with ball release speed. No association between bowling accuracy and side-to-side differences in muscle thickness could be determined.

Table 3.

The relationship between ball release speed and accuracy, and side-to-side differences in trunk muscle thickness (n = 46).

	Ball release speed		Accuracy
	Spearman's (ρ)	p-value	Spearman's (ρ)	p Value
EO at rest (mm)	.049	0.742	<.0001	1.00
IO at rest (mm)	−.058	0.700	−.016	0.914
TA at rest (mm)	.148	0.327	−.115	0.445
Total thickness at rest (mm)	−.024	0.876	−.091	0.547
EO contracted (mm)	−.455	0.002	.073	0.632
IO contracted (mm)	−.130	0.389	−.158	0.294
TA contracted (mm)	−.015	0.922	−.118	0.435
Total thickness contracted (mm)	−.346	0.019	−.039	0.798
Multifidus at rest (mm)	−.199	0.185	.294	0.048
IO % change	.041	0.789	−.109	0.470
TA % change	−.110	0.468	.127	0.400
EO contraction ratio	−.495	0.0001	.151	0.317
IO contraction ratio	.041	0.789	−.109	0.470
TA contraction ratio	−.110	0.468	.127	0.400
EO-IO contraction ratio	−.266	0.074	.069	0.650
TA preferential activation ratio	.176	0.243	.092	0.543
EO relative thickness at rest (%)	.133	0.377	.035	0.817
IO relative thickness at rest (%)	−.112	0.457	.117	0.440
TA relative thickness at rest (%)	.094	0.536	−.123	0.414
EO relative thickness contracted (%)	−.280	0.059	.124	0.413
IO relative thickness contracted (%)	.082	0.589	−.185	0.219
TA relative thickness contracted (%)	.311	0.035	−.007	0.961
EO % difference at rest	−.094	0.531	−.107	0.481
IO % difference at rest	.227	0.129	−.078	0.604
TA% difference at rest	−.204	0.174	.097	0.519
Total thickness % difference at rest	.091	0.548	−.067	0.657
EO % difference contracted	.154	0.306	−.068	0.655
IO % difference contracted	.192	0.202	.056	0.712
TA % difference contracted	−.146	0.334	−.101	0.506
Total thickness % difference contracted	.191	0.203	.082	0.589
Multifidus % difference	.250	0.093	.086	0.568

Correlations in this table refers to the side-to-side differences, and ball release speed/accuracy, and not to the correlation between non-dominant/dominant values and ball release speed/accuracy. EO = external oblique; IO = internal oblique; TA = transversus abdominis; Total thickness = EO + IO + TA.

Confounders

No association between age and weight, and side-to-side differences in trunk muscle thickness were present, respectively, as shown in Table 4. However, a weak negative correlation between the side-to-side difference in EO in the contracted state and height was found (Spearman's (ρ) = −0.324, p = 0.033). Also, weak positive correlations between the side-to-side difference in IO percentage change (Spearman's (ρ) = 0.336, p = 0.022) and IO contraction ratio (Spearman's (ρ0.336, p = 0.022), and height were found. No relationship between age, height or weight and percentage difference in any of the muscles were found.

Table 4.

The relationship between age, height and weight, and side-to side differences in trunk muscle thickness (n = 46).

	Age		Height		Weight
	Spearman's (ρ)	p Value	Spearman's (ρ)	p Value	Spearman's (ρ)	p value
EO at rest (mm)	−.005	0.972	−.135	0.373	−.110	0.467
IO at rest (mm)	.003	0.984	−.201	0.181	−.109	0.470
TA at rest (mm)	.005	0.947	−.045	0.767	−.101	0.502
Total thickness at rest (mm)	−.026	0.862	−.230	0.124	−.170	0.259
EO contracted (mm)	−.045	0.766	−.314	0.033	−.204	0.173
IO contracted (mm)	.019	0.898	.034	0.823	.055	0.719
TA contracted (mm)	−.019	0.205	.151	0.317	−.191	0.203
Total thickness contracted (mm)	−.127	0.401	−.183	0.223	−.136	0.368
Multifidus at rest (mm)	.059	0.699	−.238	0.111	−.210	0.161
IO % change	.086	0.571	.336	0.022	.162	0.282
TA % change	−.124	0.412	−.043	0.776	−.020	0.894
EO contraction ratio	−.106	0.483	−.181	0.229	−.152	0.314
IO contraction ratio	.086	0.571	.336	0.022	.162	0.282
TA contraction ratio	−.124	0.412	−.043	0.776	−.020	0.894
EO-IO contraction ratio	−.015	0.921	.118	0.435	.058	0.700
TA preferential activation ratio	−.159	0.291	−.104	0.492	−.053	0.729
EO relative thickness at rest (%)	.050	0.742	.137	0.364	.084	0.577
IO relative thickness at rest (%)	−.028	0.856	−.184	0.222	−.022	0.885
TA relative thickness at rest (%)	.072	0.635	.196	0.193	.045	0.764
EO relative thickness contracted (%)	.044	0.770	−.228	0.127	−.194	0.196
IO relative thickness contracted (%)	.108	0.474	.183	0.223	.184	0.221
TA relative thickness contracted (%)	−.029	0.850	.105	0.489	.012	0.939
EO % difference at rest	−.036	0.815	−.078	0.667	−.021	0.817
IO % difference at rest	.044	0.773	.044	0.772	.029	846
TA% difference at rest	−.095	0.531	−.167	0.267	−.070	.0644
Total thickness % difference at rest	.017	0.909	.001	0.993	.025	0.867
EO % difference contracted	−.036	0.811	.044	0.772	−.045	0.769
IO % difference contracted	.184	0.220	−.017	0.909	.028	0.854
TA % difference contracted	−.087	0.564	−.117	0.438	−.004	0.981
Total thickness % difference contracted	.160	0.287	−.108	0.476	−.050	0.739
Multifidus % difference	.233	0.119	.027	0.861	−.071	0.640

All variables are presented as a correlation between the difference between the dominant and non-dominant sides, and ball release speed and accuracy, respectively; Spearman's Rank order correlation test performed for non-parametric data;.

EO = external oblique; IO = internal oblique; TA = transversus abdominis; total thickness = EO + IO + TA.

Discussion

This is the first study to explore the link between bowling performance, specifically ball release speed and accuracy, and the side-to-side differences in morphometry of the abdominal and lumbar multifidus muscles. While the link between trunk muscle thickness and injury has been established, its link with performance is as important considering that the fast bowler plays a key role in the cricket team's performance. Previous studies investigating trunk muscle thickness in cricket fast bowlers have found similar patterns of morphometry on both dominant and non-dominant sides.^14–17 These investigations, however, were focused on muscle adaptation due to the repetitive nature of the bowling action¹⁶ and injury prevention^14,15,17 not bowling performance per se. This paper makes a unique contribution to the literature as it focuses specifically on the side-to-side differences in muscle thickness of the IO, EO, TA and multifidus muscles and their relationship with ball release speed and accuracy.

In our study, both external and IO measured thicker at rest, as well as in the contracted state, on the non-dominant side of these pain-free adolescent fast bowlers. In addition, the percentage difference [(largest/smallest value × 100) - 100],^17,30 was also the largest for IO compared to EO, TA and multifidus muscles (refer to Table 1). When IO is presented in relation to the total thickness of all three muscles together [(IOrest / (EOrest + IOrest + TArest) × 100)],²⁹ this value was also higher on the non-dominant side than the dominant side in the contract and at rest states. Studies by Olivier et al.¹⁶ Martin et al.¹⁴ Gray et al.¹⁵ and Hides et al.³⁵ also found IO thicker on the non-dominant side in fast bowlers without pain. The asymmetry in muscle thickness can be explained by sport-specific muscle adaptation associated with the repetitive rotational demands of the bowling action. The fast-bowling action involving trunk flexion, rotation and side flexion towards the non-dominant side during the delivery stride, seems to induce hypertrophy due to high loading of the musculature on the non-dominant side.¹⁶ It is possible that the type of bowling action utilised plays a role in the asymmetry detected in the IO muscle. In other words, the non-dominant IO may hypertrophy to protect the bowler against potentially harmful kinetic and kinematic movement components such as contralateral lumbopelvic side flexion,³⁶ without compromising ball release speeds. However, further research is needed to explore this hypothesis considering that a more symmetrical IO is associated with lower back and lower limb injury.^14,15

In our study, TA were symmetrical as measured at rest and in a contracted state, as shown by the absolute differences and the percentage change calculations in Table 2. In addition, multifidus was also symmetrical. Similar findings are shown in Mannion et al.²⁹ and Springer et al.³⁷ where no difference was found in the thickness of TA between sides in healthy individuals. These findings are also present in asymptomatic cricketers where the TA measured symmetrical in both the relaxed as well as contracted states.²⁴ In addition, symmetry of lumbar multifidus as identified in our study is supported by Hides et al.³⁸ in a study on healthy young adults, and Olivier et al.¹⁷ in a study on injury-free fast bowlers. In our study, the percentage difference of multifidus (median = 2.69%) was the smallest of all muscles although slightly higher than values in Hides et al.,³⁵ which was 2.62% (calculated from the results in Table 1 presented in Hides et al..³⁵⁾ A potential explanation for this finding is that both the TA and multifidus muscles are stabilising muscles. The bilateral activity of TA was reported to have produced optimal stability of the spine.²⁵

The relative thickness of the IO muscle,²⁹ expressed as a percentage of the total thickness of all three abdominal muscles together [IOrest / (EOrest + IOrest + TArest) × 100], at rest, is larger on the non-dominant side than on the dominant side as established in our study. This finding fits in with the fact that the absolute thickness of the IO is more on the non-dominant side, and it means that the IO on the non-dominant side is much thicker in relation to all three muscles together. Interestingly, the relative thickness of the TA is less on the non-dominant side than on the dominant side, even though the absolute thickness of the TA seemed symmetrical. It gives us a clue that the large relative thickness of the IO on the non-dominant side may be closely associated with the smaller relative thickness of the TA on that side. The same happened during the contracted state. These unexpected findings need to be further investigated, although the effect sizes for these differences were small.

In this paper, many derivatives of absolute muscle thickness have been presented to compare the descriptive values in adolescent fast bowlers to those in the existing literature. Percentage change (a percentage of muscle thickness at rest),²² was calculated for TA and IO and did not show any side-to-side differences. Percentage change was found to be a surrogate measure of muscle activity for TA³⁹ and IO,^26,39 but not for EO³⁹ and therefore this calculation was not shown for EO. The percentage change, preferential activation ratio, and contraction ratio were very similar to that of Mannion et al.,²⁹ who studied healthy men and women. In our sample of fast bowlers, a comparison of the derivatives of trunk muscle thickness, namely percentage change, preferential activation ratio and contraction ratio, did not show any side-to-side differences. This finding was surprising because the absolute thickness values did display side-to-side differences. These calculations are based on the absolute muscle thickness values and produced derived values such as percentages or ratios where the ranges were large in most cases. It is possible that the processing of the absolute data through these calculations reduces the sensitivity of the data.

Very few moderate and no strong correlations were identified between ball release speed and the side-to-side differences in muscle thickness, as shown in Table 3. At the same time, only weak correlations were identified between accuracy and side-to-side differences in muscle thickness. The side-to-side difference in IO was not associated with ball release speed, while this muscle has been implicated in earlier work for its potential role in injury, specifically lower back or lower quarter injury.^14,15 Future research needs to explore the effect of an intervention programme to enhance asymmetry of the IO muscle for injury prevention purposes, while we do not foresee an influence on ball release speed or accuracy. It is important to consider this potential trade-off as it is present in other domains of fast bowling such as where a technique-related factor involving the knee angle⁴⁰ is associated with injury and at the same time with higher ball release speed.

Both moderate correlations with ball release speed involved the EO muscle, where a smaller side-to-side difference (a more symmetrical muscle) is associated with higher ball release speeds. Similarly, smaller side-to-side difference EO contraction ratios, are associated with higher ball release speeds. Symmetry in the EO may play a role in ball release speed and the mechanism behind this relationship need to be explored in future research. This link between a more symmetrical EO muscle thickness and higher ball release speed is especially important seeing that Gray et al.¹⁵ found that the EO was symmetrical in bowlers with lower back pain. The simultaneous relationships where a symmetrical EO is associated with enhanced performance while, at the same time, may be implicated in injury is an important finding which warrants further exploration. It is important to note that these correlations only consider the differences between the dominant and non-dominant sides and ball release speed, and not the actual muscle thickness and its relationship to ball release speed, as has been stated elsewhere.⁴¹

Accuracy of a delivery could not be related to side-to-side differences in muscle thickness of any of the trunk muscles, where the highest correlation was interpreted as weak where a thicker multifidus muscle was weakly correlated with a higher accuracy score. Seeing that the trunk muscles contribute to proximal stability,²⁵ the authors hypothesised that some relationship will exist However, it is possible that a more involved measure of accuracy is required to adequately explore this relationship, such as a combination between an accuracy target behind the stumps and one on the pitch. Also, it may be that mechanisms other than side-to-side differences in trunk muscle thickness are determinants of bowling accuracy. No literature exists relating accuracy to trunk muscle morphometry, and future research to explore this interaction is warranted.

Age, height, and weight did not influence the findings, as shown by the weak effect sizes, which means that the older, taller and heavier fast bowlers did not display larger or smaller side-to-side differences in their trunk muscles (refer to Table 4). These findings do not talk to the relationship between age, height, weight and absolute trunk muscle thickness-related variables as was found in Olivier et al.⁴¹ In their⁴¹ study, there were moderate to weak correlations between age, height and trunk muscle thickness-related variables, with strong, moderate and weak correlations between weight and absolute trunk muscle thickness-related variables. It seems like although the older, taller and heavier fast bowlers may have thicker trunk muscles, they don't have more or less asymmetry based on their age, height or weight.

A limitation of this study is that the multicollinearity between variables did not allow the authors to perform a regression analysis. Due to the cross-sectional nature of this study, no causality should be assumed. Future studies can explore the role and mechanism behind EO on bowling performance and the effectiveness of interventions to use side-to-side differences in trunk muscle thickness to improve bowling performance.

Conclusion

Injury-free adolescent cricket fast bowlers present with asymmetrical trunk muscle morphometry, i.e., thickness, as measured through ultrasound. Previous studies have found that trunk muscle asymmetry in fast bowlers may be protective against injury. This study could establish no definitive relationship between bowling performance (specifically ball release speed and accuracy) and side-to-side differences in IO muscle thickness. However, a more symmetrical EO seems to be linked with faster ball release speeds, whereas previous research on injury found that a more symmetrical EO was found in fast bowlers who experienced lower back pain. Some factors known to prevent injury often influence performance, and it is important to consider these factors before implementing any injury prevention measures.

Footnotes

Acknowledgements

Thank you to the headmasters, directors of cricket, parents and participants for taking part in this study.

Declaration of conflicting interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship and/or publication of this article: The authors would like to thank the Faculty Research Committee of the University of the Witwatersrand and National Research Foundation (NRF) for financial assistance towards this research.

ORCID iD

Oluchukwu Loveth Obiora

Benita Olivier

References

Bartlett

Stockill

Elliott

, et al. The biomechanics of fast bowling in men’s cricket: a review. J Sports Sci 1996; 14: 403–424.

Ferdinands

Kersting

Marshall

. Three-dimensional lumbar segment kinetics of fast bowling in cricket. J Biomech 2009; 42: 1616–1621.

Foster

John

Elliott

, et al. Back injuries to fast bowlers in cricket: a prospective study. Br J Sports Med 1989; 23: 50.

Pyne

Duthie

Saunders

, et al. Anthropometric and strength correlates of fast bowling speed in junior and senior cricketers. J Strength Cond Res 2006; 20: 620–626.

Glazier

Paradisis

Cooper

. Anthropometric and kinematic influences on release speed in men’s fast-medium bowling. J Sports Sci 2000; 18: 1013–1021.

Portus

Sinclair

Burke

, et al. Cricket fast bowling performance and technique and the influence of selected physical factors during an 8-over spell. J Sports Sci 2000; 18: 999–1011.

Lillee

. The art of fast bowling. Sydney, NSW: William Collins, 1977.

Portus

Mason

Elliott

, et al. Technique factors related to ball release speed and trunk injuries in high performance cricket fast bowlers. Sports Biomech 2004; 3: 263–283.

Crewe

Campbell

Elliott

, et al. Lumbo-pelvic biomechanics and quadratus lumborum asymmetry in cricket fast bowlers. Med Sci Sports Exercise 2013; 45: 778–783.

10.

Loram

McKinon

Wormgoor

, et al. Determinants of ball release speed in schoolboy fast-medium bowlers in cricket. J Sports Med Phys Fitness 2005; 45: 483–490.

11.

Salter

Sinclair

Portus

. The associations between fast bowling technique and ball release speed: a pilot study of the within-bowler and between-bowler approaches. J Sports Sci 2007; 25: 1279–1285.

12.

Duffield

Carney

Karppinen

. Physiological responses and bowling performance during repeated spells of medium-fast bowling. J Sports Sci 2009; 27: 27–35.

13.

Hilligan

. The relationship between core stability and bowling speed in asymptomatic male indoor action cricket bowlers. Durban University of Technology, 2008, Unpublished.

14.

Martin

Olivier

Benjamin

. Asymmetrical abdominal muscle morphometry is present in injury free adolescent cricket pace bowlers: a prospective observational study. Phys Ther Sport 2017; 28: 34–42.

15.

Gray

Aginsky

Derman

, et al. Symmetry, not asymmetry, of abdominal muscle morphology is associated with low back pain in cricket fast bowlers. J Sci Med Sport 2016; 19: 222–226.

16.

Olivier

Stewart

Mckinon

. Side-to-side asymmetry in absolute and relative muscle thickness of the lateral abdominal wall in cricket pace bowlers. S Afr J Sports Med 2013; 25: 81–86.

17.

Olivier

Gillion

Stewart

, et al. Reduced nondominant lumbar multifidi cross-sectional area is a precursor of low back injury in cricket fast bowlers. Scand J Med Sci Sports 2017; 27: 1927–1933.

18.

Toumazou

Carol

Gordon

. Measurement of lumbar multifidus asymmetry in amateur cricket pace bowlers using real-time ultrasound. Int J Exerc Sci 2018; 11: 875.

19.

Hides

Gilmore

Stanton

, et al. Multifidus size and symmetry among chronic LBP and healthy asymptomatic subjects. Man Ther 2008; 13: 43–49.

20.

Orchard

Ranson

Olivier

, et al. International consensus statement on injury surveillance in cricket: a 2016 update. Br J Sports Med 2016; 50: 1245–1251.

21.

Koppenhaver

Hebert

Parent

, et al. Rehabilitative ultrasound imaging is a valid measure of trunk muscle size and activation during most isometric sub-maximal contractions: a systematic review. Aust J Physiother 2009; 55: 153–169.

22.

Koppenhaver

Hebert

Fritz

, et al. Reliability of rehabilitative ultrasound imaging of the transversus abdominis and lumbar multifidus muscles. Arch Phys Med Rehabil 2009; 90: 87–94.

23.

Hides

Miokovic

Belavy

, et al. Ultrasound imaging assessment of abdominal muscle function during drawing-in of the abdominal wall: an intrarater reliability study. J Orthop Sports Phys Ther 2007; 37: 480–486.

24.

Hides

Wilson

Stanton

, et al. An MRI investigation into the function of the transversus abdominis muscle during “drawing-in” of the abdominal wall. Spine (Phila Pa 1976) 2006; 31: E175–E178.

25.

Hodges

Pengel

Herbert

, et al. Measurement of muscle contraction with ultrasound imaging. Muscle Nerve 2003; 27: 682–692.

26.

McMeeken

Beith

Newham

, et al. The relationship between EMG and change in thickness of transversus abdominis. Clin Biomech (Bristol, Avon) 2004; 19: 337–342.

27.

Smith

NDW

Burke

. Influence of angular position on radar gun peak cricket ball speed measurements. Sports Biomech 2021: 1–14. https://doi.org/10.1080/14763141.2021.1887343

28.

Elliott

Plunkett

Alderson

. The effect of altered pitch length on performance and technique in junior fast bowlers. J Sports Sci 2005; 23: 661–667.

29.

Mannion

Pulkovski

Toma

, et al. Abdominal muscle size and symmetry at rest and during abdominal hollowing exercises in healthy control subjects. J Anat 2008; 213: 173–182.

30.

Rankin

Stokes

Newham

. Abdominal muscle size and symmetry in normal subjects. Muscle Nerve 2006; 34: 320–326.

31.

Wasserstein

Lazar

. The ASA statement on p-values: context, process, and purpose. Taylor & Francis, 2016. https://doi.org/10.1080/00031305.2016.1154108

32.

Wasserstein

Schirm

Lazar

. Moving to a world beyond “p < 0.05”. Taylor & Francis, 2019. https://doi.org/10.1080/00031305.2019.1583913

33.

Pautz

Olivier

Steyn

. The use of nonparametric effect sizes in single study musculoskeletal physiotherapy research: a practical primer. Phys Ther Sport 2018; 33: 117–124.

34.

Rosenthal

Rubin

. Requivalent: a simple effect size indicator. Psychol Methods 2003; 8: 492.

35.

Hides

Stanton

Freke

, et al. MRI Study of the size, symmetry and function of the trunk muscles among elite cricketers with and without low back pain. Br J Sports Med 2008; 42: 809–813.

36.

Alway

Felton

Brooke-Wavell

, et al. Cricket fast bowling technique and lumbar bone stress injury. Med Sci Sports Exerc 2021; 53: 581–589.

37.

Springer

Mielcarek

Nesfield

, et al. Relationships among lateral abdominal muscles, gender, body mass index, and hand dominance. J Orthop Sports Phys Ther 2006; 36: 289–297.

38.

Hides

Cooper

Stokes

. Diagnostic ultrasound imaging for measurement of the lumbar multifidus muscle in normal young adults. Physiother Theory Pract 1992; 8: 19–26.

39.

Ferreira

Nascimento

, et al. Discriminative and reliability analyses of ultrasound measurement of abdominal muscles recruitment. Man Ther 2011; 16: 463–469.

40.

Olivier

Stewart

Green

, et al. Cricket pace bowling: the trade-off between optimising knee angle for performance advantages versus injury prevention. S Afr J Sports Med 2015; 27: 79–81.

41.

Olivier

Franso-Mari

Nkazimulo

. The “core” of performance adolescent cricket pace bowlers: trunk muscle stability, maybe, but not strength-endurance and thickness, 2021, Unpublished.