Clinical Biomechanics 20 (2005) 233241 www.elsevier.

com/locate/clinbiomech

Regional morphology of the transversus abdominis and obliquus internus and externus abdominis muscles
Donna M Urquhart a,*, Priscilla J Barker b, Paul W Hodges Ian H Story a, Christopher A Briggs b
b

c,d

a School of Physiotherapy, The University of Melbourne, Victoria 3010, Australia Department of Anatomy and Cell Biology, The University of Melbourne, Victoria 3010, Australia c Prince of Wales Medical Research Institute, New South Wales 2031, Australia d Division of Physiotherapy, The University of Queensland, Queensland 4072, Australia

Received 14 August 2004; accepted 15 November 2004

Abstract Background. The mechanisms by which the abdominal muscles move and control the lumbosacral spine are not clearly understood. Descriptions of abdominal morphology are also conicting and the regional anatomy of these muscles has not been comprehensively examined. The aim of this study was to investigate the morphology of regions of transversus abdominis and obliquus internus and externus abdominis. Methods. Anterior and posterolateral abdominal walls were dissected bilaterally in 26 embalmed human cadavers. The orientation, thickness and length of the upper, middle and lower fascicles of transversus abdominis and obliquus internus abdominis, and the upper and middle fascicles of obliquus externus abdominis were measured. Findings. Dierences in fascicle orientation, thickness and length were documented between the abdominal muscles and between regions of each muscle. The fascicles of transversus abdominis were horizontal in the upper region, with increasing inferomedial orientation in the middle and lower regions. The upper and middle fascicles of obliquus internus abdominis were oriented superomedially and the lower fascicles inferomedially. The mean vertical dimension of transversus abdominis that attaches to the lumbar spine via the thoracolumbar fascia was 5.2 (SD 2.1) cm. Intramuscular septa were observed between regions of transversus abdominis, and obliquus internus abdominis could be separated into two distinct layers in the lower and middle regions. Interpretation. This study provides quantitative data of morphological dierences between regions of the abdominal muscles, which suggest variation in function between muscle regions. Precise understanding of abdominal muscle anatomy is required for incorporation of these muscles into biomechanical models. Furthermore, regional variation in their morphology may reect dierences in function. 2004 Elsevier Ltd. All rights reserved.
Keywords: Anatomy; Abdominal muscles; Transversus abdominis; Obliquus internus abdominis; Obliquus externus abdominis; Regional morphology; Spinal control

1. Introduction
* Corresponding author. Address: Department of Epidemiology and Preventive Medicine, Central and Eastern Clinical School, Monash University, Alfred Hospital, Commercial Road, Melbourne 3004, Vic., Australia. E-mail address: donna.urquhart@med.monash.edu.au (D.M Urquhart).

There is increasing evidence to indicate the importance of the anterolateral abdominal muscles in control and movement of the lumbar spine and pelvis (Cholewicki et al., 1999; Cresswell et al., 1992; Hodges and Gandevia, 2000). However, few studies have

0268-0033/$ - see front matter 2004 Elsevier Ltd. All rights reserved. doi:10.1016/j.clinbiomech.2004.11.007

234

D.M Urquhart et al. / Clinical Biomechanics 20 (2005) 233241

comprehensively investigated the morphology of these muscles. Transversus abdominis (TrA), obliquus internus abdominis (OI) and obliquus externus abdominis (OE) have a number of primary bro-osseous attachments that include the costal cartilages, the lumbar spine via the thoracolumbar fascia (TLF), and the iliac crest and pubis (Williams et al., 1999). It is hypothesised that the muscle fascicles originating from these dierent structures may have dierent functions. For example, the upper fascicles of TrA arising from the costal cartilages may stabilise the rib cage, the middle fascicles attaching to the TLF may contribute to control of the lumbar spine, and the lower fascicles arising from the iliac crest may support the abdominal contents and generate forces that compress the sacroiliac joints (Richardson et al., 2002 and Snijders et al., 1995). Recent evidence from electromyographic (EMG) studies suggests there is regional variation in abdominal muscle function. During upper limb movements, greater tonic activity of the lower region of TrA was reported compared to middle region (Hodges et al., 1999). The supercial abdominal muscles of the lower abdominal wall were also more active in erect standing than those of the upper abdominal wall (Strohl et al., 1981). Furthermore, dierences in the recruitment of regions of OI and OE have been documented during trunk extension and rotation (Davis and Mirka, 2000; Mirka et al., 1997). While these studies provide preliminary evidence of regional dierences in function, investigation of regional morphology is critical for evaluation of the mechanical eect of each region on the lumbar spine, pelvis and rib cage. Although one study has described the fascicle orientation of regions of TrA, OI and OE (Askar, 1977), there has been no detailed, quantitative investigation of regional morphology of the abdominal muscles. The aims of this study were to compare the morphology (fascicle orientation, thickness and length) of the abdominal muscles and regions of these muscles.

2.1. Regional anthropometric measures Anthropometric measures were documented to dene regions of the abdominal wall and to determine the extent to which TrA attaches to the costal cartilages (upper region), TLF (middle region) and the pelvis (lower region) (Fig. 1). The upper region was measured with a tape measure (accuracy 0.1 cm) from the 6th costal cartilage or the superior border of the T9 vertebral body to the inferior border of the rib cage. The T9 vertebral body is in close alignment to the notch of the 6th costal cartilage (Snell, 2000) and was selected in specimens where parts of the rib cage had been previously removed. The middle region was dened as the distance between the inferior border of the rib cage and a line connecting the superior borders of the iliac crest, and the lower region was measured from this line to the pubic symphysis. 2.2. Fascicle orientation The orientation of fascicles of TrA, OI and OE was examined in the anteriorly dissected specimens with reference to a line connecting the left and right anterior superior iliac spines (ASIS). Fascicles parallel to this reference line were considered to be horizontal. The upper, middle and lower regional measures of TrA and OI were documented bilaterally at the level of the 11th costal cartilage, halfway between the iliac crest and rib cage, and adjacent to the ASIS. Additional orientation measures were reported for TrA, 2 cm below the ASIS, and for OI, halfway between the ASIS and pubic symphysis. The fascicle orientation of OE was calculated at the level of the 8th costal cartilage (upper region) and halfway between the iliac crest and the rib cage (middle region). The fascicles of OE did not extend below the ASIS and therefore no measurements for a lower region of

2. Methods Twenty-six human cadavers, embalmed in a solution of 4% formaldehyde, were serially dissected to investigate the morphology of TrA, OI and OE. Twenty-four sides of 12 specimens (6 female, 6 male, mean age: 84 (7395) years) were dissected to investigate the anterior abdominal wall, and 28 sides of 14 specimens (6 female, 8 male, mean age: 83 (6096) years) to expose the inner aspect of the posterolateral wall. A midline incision was made from the xiphoid process to the pubic symphysis, and OE, OI and TrA were separated and reected. The abdominal viscera and overlying fascia were subsequently removed to reveal the posterolateral wall.

Fig. 1. Anterior and lateral view of the upper, middle and lower regions of the abdominal wall. Horizontal lines show the borders of the regions. Key landmarks for the measurement of muscle length and fascicle orientation are indicated by A (8th costal cartilage), B (11th costal cartilage), C (halfway between the iliac crest and rib cage), D (ASIS) and E (a line connecting the left and right ASIS). Note the dierent bro-osseous attachments for each region of TrA; the costal cartilages in the upper region, the TLF in the middle region, and the pelvis in the lower region.

D.M Urquhart et al. / Clinical Biomechanics 20 (2005) 233241

235

this muscle were made. The orientation measures were calculated from digital photographic images using an image analysis program, Image J 1.20s (National Institute of Mental Health, Bethesda, Maryland, USA; accuracy 0.001 deg). To standardise the photographic conditions, the distance between the camera and the cadaver was kept constant and the centre of eld of view was maintained halfway between the level of the pubic crest and the manubriosternal joint. 2.3. Muscle thickness Thickness of the upper, middle and lower regions of TrA and OI, and the upper and middle regions of OE was measured using a manual micrometer (Mituyo, Tokyo, Japan; accuracy 0.001 mm). TrA and OI thickness measures were obtained from the same locations as the orientation measures, while the OE thickness measures were recorded adjacent to the 11th costal cartilage and halfway between the rib cage and iliac crest. 2.4. Fascicle length Fascicle length was measured bilaterally with a tape measure (accuracy 0.1 cm). The lengths of TrA and OI fascicles were measured from the anterior edge of the musculotendinous aponeurosis to the osseous or fascial attachment at the 11th costal cartilage (upper), midway between the rib cage and iliac crest (middle), and at the ASIS (lower). The fascicle lengths of OE in the upper region were measured from the aponeurosis to their insertion on the 8th costal cartilage, and those of the middle region (halfway between the rib cage and iliac crest) from the musculotendinous junction to their attachment on the lower rib cage. 2.5. Inferior extent The distal extent of the fascicles of TrA and OI was categorised as terminating at the mid-point of the inguinal ligament or above, or extending below the mid-point of the inguinal ligament. 2.6. Repeatability Three separate measures of each parameter were documented for the middle region of TrA to evaluate measurement consistency. It was predicted that greater error may be associated with these data because determination of the site for measurement required location of several anatomical landmarks. 2.7. Statistics As there was no systematic dierence between measures for left and right sides, the data were pooled for

analysis. Dierences between the abdominal muscles and muscle regions for each variable were compared using one-way repeated measures analysis of variance (ANOVA). Duncans multiple range test was used for the post-hoc analysis and the alpha level was set at 0.05. Intraclass correlation coecients (ICCs), (twoway mixed-model with absolute agreement), were calculated to determine the repeatability.

3. Results 3.1. Regional anthropometric measures The vertical dimensions of each region, expressed as a mean and a proportion of the total vertical dimension of the abdominal wall, are summarised in Table 1. 3.2. Fascicle orientation There were regional dierences in the orientation of TrA and OI fascicles (Table 2). The fascicle orientation of upper TrA was horizontal, however 8 of the 23 sides had fascicles angled greater than 10 deg from the horizontal, both superomedially and inferomedially. The middle and lower fascicles of TrA were inferomedially directed and the angle of the lower fascicles was greater than that of the middle region (P = 0.007) (Fig. 2A). Superior to the iliac crest, the upper and middle fascicles of OI were directed superomedially (Fig. 2B). In contrast, below the iliac crest the OI fascicles were oriented horizontally (P < 0.001), with increasing inferomedial
Table 1 Mean vertical dimensions (cm) and proportions of the abdominal wall and upper, middle and lower TrA (n = 52) Region Vertical dimensions Mean (SD) Upper Middle Lower 15.0 (3.3) 5.2 (2.1) 17.0 (2.6) Proportion (%) 40 14 46

Table 2 Mean (SD) fascicle orientation (deg) for regions of TrA, OI and OE (n = 24) Region Fascicle orientation TrA Upper Middle Lower (1) Lower (2) Lower (3) 2.7 13.3 21.2 20.3 (9.3) (9.8) (10.5) (11.3) OI 48.2 35.3 0.0 8.2 15.5 (12.9) (9.9) (7.2) (9.1) (10.3) OE 49.3 (7.0) 58.6 (10.5)

Lower (1)ASIS level; lower (2)2 cm below ASIS; lower (3) halfway between ASIS and pubic symphysis. Negative values inferomedial orientation, positive valuessuperomedial orientation.

236

D.M Urquhart et al. / Clinical Biomechanics 20 (2005) 233241

Fig. 2. Fascicle orientation of the abdominal muscles. Serial dissections of the right abdominal wall showing right TrA (A), OI (B) and OE (C) for a single specimen. Horizontal lines divide the upper, middle and lower abdominal regions in each panel. Dierences in fascicle orientation can be observed between muscles (compared between panels) and between regions (compared within each panel). In the middle region, fascicles of TrA are the most horizontal. The fascicles of OI radiate superomedially and inferomedially from the iliac crest. Fascicles of OE are inferomedial in all regions. RA-rectus abdominis.

Thickness SD (mm)

angulation below the ASIS. The fascicles of OE were also directed inferomedially, with greater angulation in the middle region (P = 0.009) (Fig. 2C). Standard deviations were large indicating high variability in fascicle orientation between specimens. 3.3. Muscle thickness There were regional dierences in thickness of TrA, OI and OE (Table 3, Fig. 3). The upper region of TrA had a greater mean thickness than the lower and middle regions (P < 0.001), while the lower and middle regions did not dier in their thickness (P = 0.9). Lower and middle regions of OI had a similar thickness (P = 0.4), which diered from the thickness of the upper region (P < 0.001). However, on closer examination of OI, two layers of muscle were evident in the lower and middle regions (refer to Section 3.7). The middle region of OE was thicker than the upper region (P < 0.001). The upper regions of OI and TrA had a similar thickness (P = 0.5), but in the middle region both OI and OE

3 2.5 2

1.5 1

0.5 0 Upper MiddleLower TrA Upper Middle Lower OI Upper Middle OE

Fig. 3. Muscle thickness of upper, middle and lower TrA and OI, and upper and middle OE. Thickness of lower OI is shown in two sections with dierent shading. These represent the division of OI into the true OI and the additional muscle layer (horizontal shading). In the middle region, OI is thicker than OE, and OE thicker than TrA.

were thicker than TrA (P < 0.001). In the lower region, OI was 24% thicker than TrA (P < 0.002).
Table 3 Mean (SD) muscle thickness (mm) for regions of TrA, OI and OE (n = 24) Region Muscle thickness TrA Upper Middle Lower
a b

3.4. Fascicle length Dierences in fascicle length were evident between TrA, OI and OE, and between regions of each muscle (Table 4). The fascicles of all muscles were longest in the middle region and shortest in the lower region. The middle fascicles of OE were the longest of all muscle regions, measuring approximately 7 cm greater than TrA and OI in the middle region (P < 0.001). In con-

OI 1.3 (0.5) 1.8 (0.9) 1.9 (0.7)b

OE 1.4 (0.5) 1.7 (0.8)a

1.2 (0.4) 0.7 (0.2) 0.7 (0.3)

Posterior aspect of OE in the middle region. Includes the additional muscle layer.

D.M Urquhart et al. / Clinical Biomechanics 20 (2005) 233241 Table 4 Mean (SD) fascicle lengths (cm) for regions of TrA, OI and OE (n = 24) Region Fascicle length TrA Upper Middle Lower
a

237

71% of these the fascicles reached the pubic crest (n = 24). In contrast, the fascicles of TrA terminated more superiorly, with only 4% extending below the mid-point of the inguinal ligament (n = 24). 3.6. Intramuscular septa of TrA Septa between muscle fascicles of TrA were observed just below the rib cage and at the level of the iliac crest (Fig. 4A). The upper septa, with a mean width of 0.5 cm, were present in 42% of specimens (n = 52). In contrast, the lower septa had a greater mean width (0.8 cm) and were observed more frequently (77% of cases). 3.7. Subdivision of OI In all of the anteriorly dissected cadavers (n = 24), OI could be separated into two muscle layers in the lower

OI 8.8 (2.3) 10.8 (2.4) 5.7 (1.1)

OE 10.6 (2.4) 18.4 (3.0)a

9.0 (1.2) 11.3 (1.5) 3.6 (1.1)

Posterior aspect of OE in the middle region.

trast, the lower fascicles of TrA were the shortest, reaching only 3.6 (SD 1.1) cm, and were approximately 2 cm shorter than those of OI (P < 0.001). 3.5. Inferior extent The muscle fascicles of OI extended below the midpoint of the inguinal ligament in all specimens and in

Fig. 4. Features of the deep abdominal muscles. (A) Intramuscular septa separating the fascicles of TrA in the middle and upper regions. The right panel shows a line drawing highlighting the extent of the septa, from the lateral raphe around the lateral abdominal wall to three quarters along the length of the TrA fascicles. ISintramuscular septa; RCrib cage; ICiliac crest. (B) Anterior fascial attachment of TrA and the two distinct muscle layers of OI in the lower and middle abdominal region. From left to right; attachment of supercial fascicles of OI, deeper fascicles of OI (additional muscle layer), and TrA can be observed medial to the ASIS. The additional muscle layer is shaded in the right panel. Note the tiered arrangement of the muscle layers as they become aponeurotic, with the fascicles of TrA being the shortest and deepest. ASISanterior superior iliac spine; RArectus abdominis.

238

D.M Urquhart et al. / Clinical Biomechanics 20 (2005) 233241

and middle regions (with the exception of one damaged side). The muscle layer was located between the umbilicus and the pubic symphysis, becoming aponeurotic medial to the insertion of TrA, and 1.0 (SD 0.6) cm lateral to the insertion of OI (Fig. 4B). The thickness of this layer was 0.98 (SD 0.39) mm at the level of the ASIS (n = 24), and its fascicle orientation at the ASIS level and halfway between the ASIS and pubic symphysis was 5.9 (SD 15.3) deg and 18.9 (SD 10.7) deg inferomedial respectively (n = 12). The thickness and fascicle orientation of this layer were similar to those of the more supercial component of OI at the level of the ASIS (thickness: P = 0.7, orientation: P = 0.1). 3.8. Anatomical variations Five variations in the morphology of TrA were identied in separate specimens. These were; complete and partial detachment of the muscle from the iliac crest and an abrupt change in fascicle orientation between the lower and middle regions, absence of fascicles below the iliac crest, passage of the iliohypogastric and ilioinguinal nerves through the septa in the muscle, and fusion of the lower fascicles with OI. 3.9. Repeatability ICCs for repeated measures of the middle region of TrA were high for the vertical dimension (ICC = 0.99), fascicle orientation (ICC = 0.96), thickness (ICC = 0.82) and length data (ICC = 0.95).

attachment to the lumbar spine via the TLF (Cresswell et al., 1992; De Troyer et al., 1990; Dumas et al., 1991; Hodges and Richardson, 1996). It has been proposed that one mechanism by which TrA may increase spinal control is by tensioning the TLF. The mean vertical distance between the rib cage and iliac crest was found to be 5.2 (SD 2.1) cm, suggesting that only a limited region of TrA or OI may attach to the middle layer of lumbar fascia and inuence vertebral motion of predominately the mid-lumbar levels. However, due to the inferomedial and superomedial orientation of the laminae of the posterior layer (Barker and Briggs, 1999; Bogduk and Macintosh, 1984; Tesh et al., 1987; Vleeming et al., 1995), TrA and OI may attach to the spinous processes of the lumbar vertebrae over a larger area. 4.2. Fascicle orientation An assumption in many anatomical texts is that the fascicles of TrA are consistently directed horizontally (with the exception of the lower inferomedial bres) (Moore and Dalley, 1999; Snell, 2000; Williams et al., 1999). In contrast, this study found the fascicles of TrA varied in their orientation between regions. The fascicles were horizontal in the upper region, and became increasingly inferomedial in the middle and lower regions. Similarly, Askar (1977) observed the upper fascicles of TrA to be superomedially angulated, the middle fascicles almost transverse, and the lower fascicles inferomedially directed. Although 40 cadavers were examined in that study, quantitative measures were not reported. Fascicle orientation of the abdominal muscles has been measured previously, however, comparison of results is dicult as only reference to the semilunaris and the rib cage was reported (McGill, 1996). Regional dierences in fascicle orientation were also evident for OI. The upper and middle fascicles of OI that attach to the costal cartilages and iliac crest were oriented superomedially. This is consistent with the role of this muscle in trunk exion and rotation (Carman et al., 1972; Cresswell et al., 1992; Goldman et al., 1987). In agreement with previous reports, the lower fascicles of OI were horizontal at the ASIS level (0.0 (SD 7.2) deg) and were directed inferomedially 2 cm below the ASIS (8.2 (SD 9.1) deg) (Ng et al., 1998). These ndings support proposals that lower OI may be involved in compression of the sacroiliac joint (Richardson et al., 2002; Snijders et al., 1995). However, considerable variation in fascicle orientation between specimens has been reported in this and other studies (Ng et al., 1998), with measures varying from 13 deg inferomedial to 20 deg superomedial at the ASIS level. Such variation needs to be carefully considered when determining the function of the lower fascicles of both OI and TrA.

4. Discussion Regional dierentiation in the morphology of TrA, OI and OE may suggest functionally distinct zones of the abdominal wall. While all the regions of TrA may contribute to increasing intra-abdominal pressure and support of the abdominal contents, individual regions may be recruited independently to perform specic functions. For example, the upper region may stabilise the rib cage, the middle region may tension the TLF, and suggest that the lower regions may compress the sacroiliac joints Richardson et al., 2002 and Snijders et al., 1995. The reported variation in fascicle orientation, thickness and length between regions of TrA, along with the presence of septa dividing the fascicles of TrA and the separation of OI into two distinct muscles in the lower and middle regions, provide structural evidence to support these hypotheses. 4.1. Regional anthropometric measures The middle region of TrA has been primarily investigated in EMG and biomechanical studies because of its

D.M Urquhart et al. / Clinical Biomechanics 20 (2005) 233241

239

Several methodological issues require consideration in the measurement of fascicle orientation. The posture of the cadaver, photographic technique and clarity of muscle fascicles may have inuenced the measures. These eects were minimised by using macrophotography and a digital camera, standardising the photographic and measurement conditions, and comparing all angles to an internal reference. 4.3. Muscle thickness Dierences in thickness were evident between muscles and between regions. In agreement with ultrasound studies, the upper abdominal wall was thicker than the lower abdominal wall (Strohl et al., 1981) and the middle region of OI was thicker than OE, and OE thicker than TrA (Cresswell et al., 1992; De Troyer et al., 1990; McGill et al., 1996). In contrast, one study reported OE to be thinner than both OI and TrA at functional residual capacity (Misuri et al., 1997). With respect to regional measures, McGill et al. (1996) also found the lower region (2 cm superior to the ASIS) of TrA to have a similar thickness to the middle region (midway between iliac crest and rib cage). While the upper region of OI was previously reported to be thicker than the lower region (Carman et al., 1972), similar results were only obtained in this study if a single layer of lower OI was measured. Although the relative thicknesses reported in this study are consistent with previous studies, the absolute thickness values are considerably smaller (Cresswell et al., 1992; De Troyer et al., 1990; McGill et al., 1996; Misuri et al., 1997). This may be explained by muscle atrophy in the elderly specimens dissected and tissue changes that can occur with preparation and embalming processes. Nonetheless, the cross sectional area of a muscle is associated with force production and these data may provide a relative indication of force generating capacity of dierent regions (Williams et al., 1999). 4.4. Fascicle length The fascicle length of each abdominal muscle and region has important functional implications. The lower fascicles of TrA and OI were the shortest and OE fascicles the longest. Similar results were reported in a previous study that measured between the attachments of OI and OE, including the length of the aponeuroses (McGill, 1991). Although the relative change in length for short and long fascicles is the same, there is less absolute change in length of short fascicles (Gans and de Vree, 1987; Williams et al., 1999). This suggests dierences in function of these muscle regions. The fascicles of OE may therefore have a greater role in torque production and movement, whereas the lower fascicles of TrA

and OI may sustain isometric tension (Hodges et al., 1999; Snijders et al., 1995). These hypotheses are supported by EMG studies that have reported recruitment of OE during trunk movements (Carman et al., 1972; Cresswell et al., 1992; Goldman et al., 1987) and tonic activation of TrA with repetitive movement tasks (Hodges et al., 1999). While biomechanical studies have represented the abdominal musculature with multiple-force vectors (Davis and Mirka, 2000; Dumas et al., 1991; Stokes and Gardner-Morse, 1999), the results of this study emphasise the need to consider the lower, middle and upper regions of the abdominal muscles. 4.5. Inferior extent Although OE fascicles have been reported to consistently become aponeurotic above the iliac crest, considerable variability in the termination of the TrA and OI fascicles has been reported (Anson et al., 1960; Chandler and Schadewald, 1944; Zimmerman et al., 1944). Absence of TrA has even been observed below the iliac crest (McGill et al., 1996; Strohl et al., 1981). While TrA and OI were found to be variable in their distal extent, TrA was consistently aponeurotic proximal to OI and extended below the ASIS in 23 of the 24 specimens. These ndings provide anatomical evidence for the contribution of TrA and OI to sacroiliac joint control and support of the lower abdominal contents (Richardson et al., 2002; Snijders et al., 1995). There is little consensus regarding the role of TrA and OI in the aetiology and management of inguinal hernia, however, the variability in the lower extent of these muscles suggests their contribution may dier between individuals. 4.6. Intramuscular septa of TrA Although absent from most current anatomical texts, septa between the fascicles of TrA have been previously described as aponeurotic intersections or cracks in early French and German texts (Eisler, 1912; Poirier and Charpy, 1901) and more recently have been referred to as defects or banding (Williams et al., 1999; Zimmerman et al., 1944). They have been reported to occur in TrA where the fascicles originating from the TLF do not connect with those from the iliac crest (Eisler, 1912). Septa have also been observed between fascicles of OI (Geller and Machado Da Costa, 1970; Williams et al., 1999; Zimmerman et al., 1944). However, in the current study septa between the fascicles of TrA just below the rib cage were reported. There are a number of possible explanations for these intersections. Firstly, they may prevent the lateral transmission of force (see Sheard (2000) for review) at the interface between adjacent muscle regions allowing independent function. Secondly, they may serve as pathways for the iliohypogastric and ilioinguinal nerves. Thirdly,

240

D.M Urquhart et al. / Clinical Biomechanics 20 (2005) 233241

the septa may be the result of incomplete fusion of myotomes that are separated by connective tissue septa during development (Keith, 1921; Schafer et al., 1923). Finally, as suggested by Eisler (1912) they may be a sign of atrophy or pathology. Additional studies are required to determine their aetiology and function. 4.7. Subdivision of OI The subdivision of OI into two separate muscle layers has been previously reported in early dissection studies (Anson and McVay, 1938; Chouke, 1935). However, its frequency and dimensions have been variably reported and it has not been documented in recent investigations. Chouke (1935) observed the muscle division to be present bilaterally in 63% of 136 specimens and unilaterally in 35%, while Anson and McVay reported its presence in 40% of 125 cadavers. In addition, a previous report indicated this additional muscle layer to be 23 in. wide, to fuse with TrA before inserting into the linea alba and to extend inferiorly as far as the linea semicircularis (Chouke, 1935). The remainder of OI was reported to be separated from this additional layer by the iliohypogastric nerve, a branch of the deep circumex iliac artery and one or two layers of fascia (Chouke, 1935). However, in the present study the muscle was observed to vary in width, extend from the umbilicus to at least the mid-point of the inguinal ligament, and to insert into the anterior fascias medial to TrA and lateral to OI. While the similarities in fascicle orientation with the lower region of OI may suggest a synergistic function, the muscle layer was distinctly separate from the more supercial portion of OI with a dierent site of insertion. 4.8. Future investigations This study indicates that there are morphological differences between regions of TrA, OI and OE. The results provide important foundations and normative data for future biomechanical, EMG and clinical studies, and highlight the need to consider the role of dierent regions of TrA and OI in control of the lumbopelvic region and in therapeutic exercise approaches. Future EMG investigations are required to determine if there are dierences in function between these muscle regions and to examine the regional activation of the abdominal wall in subjects with and without spinal pain.

sity of Melbourne), and Steve Martin for his imaging support (School of Physiotherapy, The University of Melbourne). Paul Hodges was supported by the NHMRC.

References
Anson, B.J., McVay, C.B., 1938. Inguinal Hernia. I. The anatomy of the region. Surg. Gynecol. Obstet. 66, 186191. Anson, B.J., Morgan, E.H., McVay, C.B., 1960. Surgical anatomy of the inguinal region based upon a study of 500 body-halves. Surg. Gynecol. Obstet. 111, 707725. Askar, O.M., 1977. Surgical anatomy of the aponeurotic expansions of the anterior abdominal wall. Ann. R. Coll. Surg. Engl. 59, 313321. Barker, P.J., Briggs, C.A., 1999. Attachments of the posterior layer of lumbar fascia. Spine 24, 17571764. Bogduk, N., Macintosh, J.E., 1984. The applied anatomy of the thoracolumbar fascia. Spine 9, 164170. Carman, D.J., Blanton, P.L., Biggs, N.L., 1972. Electromyographic study of the anterolateral abdominal musculature utilising indwelling electrodes. Am. J. Phys. Med. 51, 113129. Chandler, S.B., Schadewald, M., 1944. Studies on the inguinal region. I. The conjoined aponeurosis versus the conjoined tendon. Anat. Rec. 89, 339343. Cholewicki, J., Juluru, K., McGill, S.M., 1999. Intra-abdominal pressure mechanism for stabilizing the lumbar spine. J. Biomech. 32, 1317. Chouke, K.S., 1935. The constitution of the sheath of the rectus abdominis muscle. Anat. Rec. 61, 341349. Cresswell, A.G., Grundstrom, H., Thorstensson, A., 1992. Observations on intra-abdominal pressure and patterns of abdominal intramuscular activity in man. Acta Physiol. Scand. 144, 409418. Davis, J.R., Mirka, G.A., 2000. Transverse-contour modeling of trunk muscle-distributed forces and spinal loads during lifting and twisting. Spine 25, 180189. De Troyer, A., Estenne, M., Ninane, V., Van Gansbeke, D., Gorini, M., 1990. Transversus abdominis muscle function in humans. J. Appl. Physiol. 68, 10101016. Dumas, G.A., Poulin, M.J., Roy, B., Gagnon, M., Jovanovic, M., 1991. Orientation and moment arms of some trunk muscles. Spine 16, 293303. Eisler, P., 1912. Die muskeln des stammes. Verlag von Gustav Fischer, Jena, Germany. Gans, C., de Vree, F., 1987. Functional bases of ber length and angulation in muscle. J. Morphol. 192, 6385. Geller, M., Machado Da Costa, R., 1970. Anatomic considerations of a variable tendinous intersection in the obliquus abdominis internus muscle. Hospital 77, 10051016. Goldman, J.M., Lehr, R.P., Millar, A.B., Silver, J.R., 1987. An electromyographic study of the abdominal muscles during postural and respiratory manoeuvres. J. Neurol. Neurosurg. Psychiatry. 50, 866869. Hodges, P.W., Richardson, C.A., 1996. Inecient muscular stabilization of the lumbar spine associated with low back pain. A motor control evaluation of transversus abdominis. Spine 21, 2640 2650. Hodges, P., Cresswell, A., Thorstensson, A., 1999. Preparatory trunk motion accompanies rapid upper limb movement. Exp. Brain Res. 124, 6979. Hodges, P.W., Gandevia, S.C., 2000. Changes in intra-abdominal pressure during postural and respiratory activation of the human diaphragm. J. Appl. Physiol. 89, 967976. Keith, A., 1921. Human Embryology and Morphology. Edward Arnold, London, pp. 6673.

Acknowledgments We thank the Department of Anatomy and Cell Biology for provision of cadaveric specimens, Stuart Thyer for photographic support, Ivica Grkovic for his valued advice, Matthew Jackson for technical assistance (Department of Anatomy and Cell Biology, The Univer-

D.M Urquhart et al. / Clinical Biomechanics 20 (2005) 233241 McGill, S.M., 1991. Kinetic potential of the lumbar trunk musculature about three orthogonal orthopaedic axes in extreme postures. Spine 16, 809815. McGill, S.M., 1996. A revised anatomical model of the abdominal musculature for torso exion eorts. J. Biomech. 29, 973977. McGill, S.M., Juker, D., Axler, C., 1996. Correcting trunk muscle geometry obtained from MRI and CT scans of supine postures for use in standing postures. J. Biomech. 29, 643646. Mirka, G., Kelaher, D., Baker, A., Harrison, A., Davis, J., 1997. Selective activation of the external oblique musculature during axial torque production. Clin. Biomech. 12, 172180. Misuri, G., Colagrande, S., Gorini, M., Iandelli, I., Mancini, M., Duranti, R., Scano, G., 1997. In vivo ultrasound assessment of respiratory function of abdominal muscles in normal subjects. Eur. Respir. J. 10, 28612867. Moore, K.L., Dalley, A.F., 1999. Clinically Oriented Anatomy. Lippincott Williams and Wilkins, Philadelphia. Ng, J.K.-F., Kippers, V., Richardson, C.A., 1998. Muscle bre orientation of abdominal muscles and suggested surface EMG electrode positions. Electromyogr. Clin. Neurophysiol. 38, 51 58. danatomie humaine. Masson, Poirier, P., Charpy, A., 1901. Traite Paris. Richardson, C.A., Snijders, C.J., Hides, J.A., Damen, L., Pas, M.S., Storm, J., 2002. The relation between the tranvsersus abdominis muscles, sacroiliac joint mechanics, and low back pain. Spine 27, 399405.

241

Schafer, E.S., Symington, J., Bryce, T.H. (Eds.), 1923. Quains Elements of Anatomy. Longmans Green and Co, London. Sheard, P.W., 2000. Tension delivery from short bres in long muscles. Exerc. Sport Sci. Rev. 28, 5156. Snell, R.S., 2000. Clinical Anatomy for Medical Students. Lippincott Williams and Wilkins, Baltimore. Snijders, C.J., Vleeming, A., Stoeckart, R., Mens, J.M.A., Kleinrensink, G.J., 1995. In: Dorman, T.A. (Ed.), State of the Art Reviews. Hanley and Belfus, Philadelphia, pp. 419432. Stokes, I.A.F., Gardner-Morse, M., 1999. Quantitative anatomy of the lumbar musculature. J. Biomech. 32, 311316. Strohl, K.P., Mead, J., Banzett, R.B., Loring, S.H., Kosch, P.C., 1981. Regional dierences in abdominal muscle activity during various maneuvers in humans. J. Appl. Physiol. 51, 14711476. Tesh, K.M., Shaw Dunn, S.J., Evans, J.H., 1987. The abdominal muscles and vertebral stability. Spine 12, 501508. Vleeming, A., Pool-Goudzwaard, A.L., Stoeckart, R., van Wingerden, J.-P., Snijders, C.J., 1995. The posterior layer of the thoracolumbar fascia. Its function in load transfer from spine to legs. Spine 20, 753758. Williams, P.L., Bannister, L.H., Berry, M.M., Collins, P., Dyson, M., Dussek, J.E., Fergusson, M.W.J. (Eds.), 1999. Grays Anatomy. The Anatomical Basis of Medicine and Surgery. Churchill Livingstone, London. Zimmerman, L.M., Anson, B.J., Morgan, E.H., McVay, C.B., 1944. Ventral hernia due to normal banding of the abdominal muscles. Surg. Gynecol. Obstet. 78, 535540.