During pregnancy, there is a high risk of developing diastasis recti (Gilleard & Brown 1996; Sperstad et al. 2016) defined as an increased separation of the two rectus abdominis muscles along the linea alba (Akram & Matzen 2014; Gluppe, Engh & Bø 2020). The linea alba is a fibrous structure formed by the fusion of the aponeuroses of the anterior abdominal wall muscles (Axer, Keyserlingk & Prescher 2001). It stretches from the xiphoid process to the symphysis pubis (Boissonnault & Blaschak 1988). According to Gilleard and Brown (1996), in pregnant women, hormonal changes and increased uterine volume decrease the linea alba stiffness and increase its width measured as the interrecti distance (IRD).

Normal values of the IRD have been determined in nulliparous and primiparous women and threshold values have been proposed, above which the diastasis recti is diagnosed (Beer et al. 2009; Mota et al. 2018; Qu et al. 2021). For postpartum women, the 80th percentile (P80) of the IRD measured in 84 primiparous women at 6 months postpartum was used by Mota et al. (2018) as the threshold value for the diastasis recti. They reported a P80 value of 24 mm at 5 cm above the umbilicus and of 28 mm at 2 cm above the umbilicus. The 90th percentile (P90) of IRD values has been proposed for nulliparous women to define this threshold. At 3 cm above the umbilicus, the P90 value of the IRD is 22 mm according to Beer et al. (2009) and 14 mm according to Qu et al. (2021). In these two studies, nulliparous women were very similar regarding the sample size (150 vs. 116), participants’ average age (29 ± 6 vs. 30 ± 5 years old) and body mass index (BMI) (22 ± 4 kg/m² vs. 21 ± 3 kg/m²). However, one study was conducted in Western Europe (Beer et al. 2009) and the other in China (Qu et al. 2021). This suggests that a threshold value measured in one population might not be valid for another one. Studies from Africa on diastasis recti in postpartum women do not report on IRD threshold values. These studies focused on the prevalence of the diastasis recti, the resolution of the diastasis recti and the correlation between IRD and parity, waist and hip circumference (Alamer, Kahsay & Ravichandran 2019; Igwe & Okoye 2020; Igwe, Okoye & Chukwu 2020; Ojukwu et al. 2021).

In the literature, IRD threshold values are all measured at rest in the supine position using ultrasound (Beer et al. 2009; Mota et al. 2018; Qu et al. 2021). In the Beninese clinical environment, palpation methods like finger width or measurements with calipers are more commonly used than ultrasound. Palpation methods require a slight contraction of the rectus abdominis muscle and are therefore performed in a ‘head lift position’ or during curl-up, not at rest. Because rectus abdominis contraction reduces the IRD (Chiarello, McAuley & Hartigan 2016; Gluppe et al. 2020; Mota et al. 2015; Pascoal et al. 2014; Sancho et al. 2015), smaller IRD values are expected in a head lift position or during curl-up than at rest (Gluppe et al. 2020). So far, the IRD threshold values specifically measured in a head lift position or during curl-up are not known. The first two aims of our study were to describe IRD values for men, nulliparous and postpartum women in the Benin Republic, and to define IRD threshold values measured at rest with ultrasound and in a head lift position with calipers. The IRD threshold values measured at rest with ultrasound will be compared with those already published in Caucasian and Chinese samples. Those measured in a head lift position with calipers will be proposed for use in the clinical environment.

A strong correlation was observed between the IRD width and the abdominal muscle strength in patients with diastasis recti (Gunnarsson et al. 2015), and lower abdominal muscle strength and endurance were observed in postpartum women with diastasis recti compared to those without (Gluppe, Engh & Kari 2021; Hills, Graham & McLean 2018; Liaw et al. 2011). According to Beamish et al. (2019), functional impairments associated with the diastasis recti may result from a mechanical deficit related to the stiffness of the linea alba rather than to its increased width. Qualitative structural changes resulting from pregnancy could reduce the linea alba stiffness and negatively affect its function in the management of intra-abdominal pressure, the transfer of force across the midline of the abdomen and the stability of the trunk (Axer et al. 2001; Hodges, Cresswell & Thorstensson 1999; Rath et al. 1996). As a result, postpartum women with an IRD below the threshold value could present a deficiency in abdominal function and may be diagnosed with diastasis recti. Protrusion of the abdomen during contraction of the rectus abdominis muscle has been used to confirm the diastasis recti in postpartum women with a normal IRD (Gluppe et al. 2020; Mahalakshmi et al. 2016). The third aim of our study was to assess the relationships between IRD, linea alba stiffness and abdominal muscle strength and endurance in postpartum women in order to propose recommendations for the diagnosis of diastasis recti and to identify women who would need abdominal rehabilitation.

Thus, in our study we (1) describe the IRD values in three samples of the Beninese population (men and nulliparous and postpartum women); (2) define the IRD threshold values measured at rest using ultrasound and in head lift position using calipers; (3) assess the relationships between IRD, linea alba stiffness, strength and endurance of abdominal muscles; and (4) propose recommendations for diagnosing diastasis recti and for the management of abdominal rehabilitation.

Our observational cross-sectional study was conducted between March and August 2020, and was carried out at the South of Benin University Hospital called Centre National Hospitalier Universitaire (CNHU) at the Clinique Universitaire de Médecine Physique et Réadaptation (CUMPR). Three categories of participants were enrolled: men and nulliparous and postpartum women. Healthy postpartum women who had given birth more than 6 weeks and less than 6 months previously were recruited while waiting for a consultation in the Physical Medicine and Rehabilitation Department of the University Clinic. Healthy men and nulliparous women were recruited from clinic attendees and from students of the Health Sciences Faculty through an informative note to class delegates.

To be included in our study, participants had to be between 18 and 40 years old. Participants who had undergone abdominal surgery, except caesarean section, were excluded.

The IRD values were measured 5 cm above the umbilicus. Postpartum women had greater IRD values than men and nulliparous women. In men and nulliparous women, the P90 of the IRD values was, respectively, 20 mm and 18 mm measured at rest using ultrasound, and 17 mm and 15 mm measured in the head lift position using the calipers. In primiparous women, the P80 of the IRD values was 23 mm at rest and 18 mm in the head lift position. With this threshold (P80), the diastasis recti prevalence among all postpartum women was 46%. Using the linea alba stiffness, a new threshold value of 19 mm at rest and 15 mm in the head lift position was defined in postpartum women. A slack linea alba was associated with lower abdominal muscle function, even in women with a normal IRD value.

The median IRD values measured at rest using ultrasound were 12 ± 5 mm in men, 14 ± 6 mm in nulliparous women and 21 ± 12 mm in postpartum women. In men, Chiarello et al. (2016) measured 16.2 ± 10.4 mm close to our median (IQR) value, while Lee and Hodges (2016) and Qu et al. (2021) measured much lower values, respectively, 8.0 ± 3.8 mm and 6.2 ± 5.6 mm. In nulliparous women, Beer et al. (2009) measured 13 ± 7 mm similar to our median value, while Chiarello et al. (2016) and Lee and Hodges (2016) measured much lower values, respectively, 7.5 ± 4.3 mm and 7.7 ± 3.4 mm. In postpartum women, Mota et al. (2018) measured 23 ± 9.3 mm at 6 weeks and 18.7 ± 8.4 mm at 6 months postpartum, Qu et al. (2021) measured 23.6 ± 8.6 mm and Chiarello et al. (2016) measured 20.3 ± 10.5 mm. Our median values were close to the mean values measured in these three studies but much lower than the 42.2 ± 9.7 mm measured by Ojukwu et al. (2021). The large disparity between the IRD values measured in these studies can be partially explained by the level of the measurement along the linea alba. For instance, the very high values observed by Ojukwu et al. (2021) were measured at the level of the umbilicus, where the IRD is known to be the largest (Rath et al. 1996). Nevertheless, this factor cannot explain most of the observed differences, and it would be worthwhile to investigate other potential predictors of IRD values in different populations.

We compared the IRD values among the three samples of participants. We did not observe a significant difference between men and nulliparous women, consistent with the observations of Lee and Hodges (2016). In contrast, Chiarello et al. (2016) observed a significantly larger IRD value in men compared to nulliparous women. In their study, the men were older and had a greater BMI than the nulliparous women, which suggests that age and BMI could influence IRD values.

Significantly larger IRD values were reported in postpartum women than in nulliparous women. The same observation was made by Chiarello and McAuley (2013) and Lee and Hodges (2016). The high IRD values measured at the end of pregnancy progressively decrease during the first few months after childbirth but never return to pre-pregnancy values (Coldron et al. 2008; Liaw et al. 2011; Mota et al. 2018), which suggests that different thresholds should be used to define IRD for nulliparous and postpartum women.

The IRD threshold values that we defined using the calipers measurement can serve as a reference for physiotherapists in Benin to diagnose the diastasis recti in nulliparous and postpartum women and in men. To date the issue of the diastasis recti in men has been sparsely studied, but it could deserve more attention in the future (Nienhuijs et al. 2021).

The IRD threshold values measured at rest with ultrasound can be compared to already published values in other population samples. In nulliparous women, the IRD was normal if it was less than 18 mm. In two similar populations of young (under 30 years old) nulliparous women with a BMI under 25, the P90 threshold value was 22 mm (Beer et al. 2009) and 14 mm (Qu et al. 2021). In these two studies the measurements were taken 3 cm above the umbilicus rather than 5 cm in our study. Larger IRD values are usually observed closer to the umbilicus (Axer et al. 2001; Beer et al. 2009; Lee & Hodges 2016; Mota et al. 2015), which may explain the difference between the values measured by Beer et al. (2009) and our results, but cannot explain the smaller values measured by Qu et al. (2021). Possible ethnicity-related differences should be investigated as one study was conducted in Europe (Beer et al. 2020), one other in China (Qu et al. 2021) and ours in Africa. These differences could be related to the role played by collagen fibres on the appearance of diastasis recti and hernias (Blotta et al. 2018; Mosanya et al. 2020). Different proportions of collagen fibres have been reported in the anterior rectus sheath of different ethnic populations, namely, an African Nigerian population (Mosanya et al. 2020), South-American Brazilian population (Casanova, Trindade & Trindade 2009; Gonçalves, De Moraes E Silva & Lopes Filho 2014) and a European Portuguese population (Blotta et al. 2018). To move forward on that issue, a study should be designed to compare different ethnic populations regarding histological characteristics of the abdominal wall in postpartum women with and without diastasis recti.

In postpartum women, the IRD was normal when less than 23 mm, a value close to the 24 mm measured in similar conditions in a European population sample by Mota et al. (2018). A subtle difference can nevertheless be noticed related to the shorter time elapsed since delivery in our study (3–6 months) compared to the study of Mota et al. (2018) (6 months). As the IRD spontaneously decreases during the first 6 months after pregnancy (Liaw et al. 2011; Mota et al. 2018), it could be hypothesised that at 6 months after delivery the P80 IRD value would have been smaller in our study than the value measured by Mota et al. (2018).

Regarding the prevalence of diastasis recti in postpartum women, our results collected on women of African ethnicity can be compared to those collected on women of European ethnicity in the study by Sperstad et al. (2016). In both studies, the IRD was measured about 5 cm above the umbilicus during rectus abdominis muscle contraction and the reported prevalence was about 45%. However, the 18 mm threshold value that we used to diagnose diastasis recti was much lower than the 2-finger breadth threshold used by Sperstad et al. (2016) as a finger corresponds to 15 mm according to Turan et al. (2011) and Rett et al. (2009). Applied to our sample of postpartum women, this 2-finger breadth threshold would have resulted in a much lower prevalence. Moreover, our sample had several differences with the sample examined in Sperstad et al. (2016); in our sample, the average parity was higher (we examined both primiparous and multiltiparous women, while Sperstad et al. (2016) studied only primiparous women), and the average time since delivery in our study was shorter (our sample ranged from 6 weeks to 6 months, while Sperstad et al. (2016) examined only the 6 months timepoint). As a result, a higher prevalence would have been expected in our population sample compared to that of Sperstad et al. (2016).

In 2007, Spitznagle, Leong and Van Dillen observed the prevalence of diastasis recti in a urogynaecological population consisting of 539 patients of different ethnic groups. The prevalence was significantly lower among African-American women (33%) than among white women (58%). This observation together with our comparison with the study of Sperstad et al. (2016) (see the paragraph above) supports the hypothesis that the prevalence of diastasis recti could be ethnicity related. However, the small sample of primiparous women in our study and the small percentage (14%) of African-American women in the study conducted in North America (Spitznagle et al. 2007) do not allow us to confirm this hypothesis right now. It should be investigated in future studies with a larger number of participants.

Postpartum women diagnosed with diastasis recti (IRD above the P80 threshold) had less strength than those without, as previously observed by several researchers (Gluppe et al. 2021; Gunnarsson et al. 2015; Hills et al. 2018). We used palpation to measure the linea alba stiffness in a binary ‘stiff’ versus ‘slack’ evaluation, and we observed that a slack linea alba was associated with less abdominal strength and endurance, even in women without diastasis recti. Among the postpartum women with a normal IRD, 60% had a slack linea alba. This justifies considering the linea alba stiffness for the diagnosis of diastasis recti and for the decision to offer abdominal rehabilitation. Based on linea alba stiffness, a new IRD threshold value, which was lower than the P80 threshold, was defined through the ROC curve with both a high sensitivity and a high specificity.

We propose to consider this new threshold in the clinical evaluation of postpartum women and to pay close attention to women with an IRD below the P80 threshold but above this new stiffness-based threshold. Linea alba stiffness and abdominal strength and endurance should be evaluated in these women to identify their need for abdominal rehabilitation. Some of them with very poor abdominal function could even be diagnosed with diastasis recti. Considering the IRD increase with parity (Gitta et al. 2017) and the greater IRD reduction observed after a rehabilitation programme in women with a parity of 2 compared to those with a parity of 4 or 5 (Igwe et al. 2020), abdominal rehabilitation should be recommended after each delivery, especially in the Beninese population whose average birth rate per woman is 5 (Leridon 2015).

To diagnose diastasis recti, parameters such as parity, linea alba stiffness, abdominal muscle strength and endurance must be considered besides IRD. In some cases, women may need abdominal rehabilitation even if their IRD is below the diastasis recti threshold value. Hereafter we propose a simple approach allowing a physiotherapist to diagnose diastasis recti and to identify the need for abdominal care through a two-step analysis. Firstly, the decision to treat is based on the comparison of the woman IRD value with a threshold value specific to her obstetrical status, parous or nulliparous. Secondly, for parous women with specific IRD values as defined in the first step, the evaluation of their abdominal function contributes to treatment decisions. This reasoning is proposed as a clinical approach, IRD values will be those measured with calipers.

First step: parous or nulliparous?

Second step: how well do the woman’s abdominal muscles function?

Interrecti distance median and threshold values specific to Beninese men and nulliparous and postpartum women were defined using ultrasound and calipers. Based on a threshold value of 23 mm at rest and 18 mm in a head lift position, a diastasis recti prevalence of 46% was observed in postpartum women. Differences in IRD values or in prevalence with other studies can be related to differences in the methodology of measurement or in population samples. Possible ethnicity-related differences cannot be excluded and should be investigated further.

In postpartum women, a slack linea alba was associated with lower abdominal strength and endurance even in women with a normal IRD value. Based on this observation, a decision-tree approach was built that combines the measurement of IRD, linea alba stiffness and abdominal strength and endurance to help physiotherapists in evaluating a woman’s need for abdominal rehabilitation. Future randomised controlled trials should investigate the effects of specific rehabilitation protocols designed to reduce IRD, restore linea alba stiffness and increase abdominal muscle strength and endurance.