Duchenne Muscular Dystrophy (DMD) is a hereditary muscular disease, caused by the deficiency of the dystrophin protein, which is responsible for muscle membrane stability.1 The disease is characterized by progressive muscle weakness that initially affects pelvic girdle muscles and proximal and extensor muscles of the lower extremities and trunk.2,3 The muscle weakness and imbalance lead to muscle shortening, contractures of lower extremity joints and postural deviations as the disease progresses. As a result, compensatory spinal deformities such as hyperlordosis develop in boys with DMD, which may affect balance and gait.2

Increased lumbar lordosis is a compensatory strategy to adapt to initial gluteal weakness. In DMD, the pelvis is unstable due to this weakness and compensatory tension in the iliotibial bands often draws the pelvis into an anterior tilt. The trunk then adopts a lordotic alignment. This altered posture may change the position of the center of mass and result in balance and gait difficulties.1,4

Determining the factors that affect walking performance in DMD is challenging as many factors such as muscle weakness, balance problems, shortness of the tendons, postural changes, or contractures have possible influence. By the nature of the disease, it is obvious that the progenitor factor is muscle weakness. However, in children at the early stages of the disease, muscle tests may not be sensitive enough to detect initial weakness. Although functional decline may have started, the muscle strength tests may not be sensitive enough to show deficits.1 Therefore, present study is aimed to determine the relationship between the factors that may influence ambulatory function, to investigate the effects of lumbar lordosis increment on gait and balance in ambulant boys with DMD, and to determine when the increase in lumbar lordosis begins to affect ambulation.

Out of 39 boys with DMD, 18 were excluded due to non-ambulatory status, and all the remaining ambulatory patients (n = 21) aged between 4 and 13 (8.00 ± 2.4 years) met the inclusion criteria and agreed to participate. Ten healthy boys aged between 6 and 13 (8.9 ± 2.4 years) were included as controls. Twenty of 21 patients’ genetic analysis results were available and a mutation was detected in 17 of 20 patients. Fourteen of 17 patients had distal hot-spot deletions encoding for exons between 44 and 55, three of them had proximal hot-spot deletions encoding for exons between 3 and 19. The clinical characteristics of the participants are summarized in Table 1. The boys with DMD had significantly worse scores in 6MWT, TUG, BBS, TRF, and MMT (both upper and lower limbs), and lumbar lordosis was significantly higher than the healthy controls. Age, BMI, fall index, and thoracic kyphosis were similar in both groups.

The results of repeated inclinometric measurements were statistically similar according to t-test (p = 0.58 for lumbar lordosis, 0.86 for thoracic kyphosis) and displayed excellent reliability (Cronbach's α = 0.95 for lumbar lordosis, 0.92 for thoracic kyphosis).

In the group with DMD, according to Vignos scale, seven boys were at level 1 (33%), seven boys were at level 2 (33%), four boys were at level 3 (19%), and three boys were at level 4 (14%). Twelve boys were wearing night splints (57%). Sixteen (76%) boys were on steroids, 12 of them were on daily, and four of them were on alternate days regimen. Five (24%) children had mild scoliosis with a mean Cobb angle of 12° ± 2.5° (min-max: 10°−15°). One of these boys had thoracic scoliosis (apex: T6), while the others had thoracolumbar scoliosis (apexes varied between T11 and L1).

There was a moderate positive correlation between lumbar lordosis and age (r = 0.532, p = 0.01), lumbar lordosis and BMI (r = 0.522, p = 0.01), and age and BMI (r = 0.633, p = 0.002). The amount of lumbar lordosis correlated strongly and negatively with 6MWT (r=−0.710, p = 0.001), and moderately and positively with TUG (r = 0.666, p = 0.002) and TRF (r = 0.462, p = 0.046), and moderately and negatively with BBS (r=−0.683, p = 0.05). The thoracic kyphosis did not correlate significantly with any of these variables (p > 0.05). There was a moderate and negative correlation between 6MWT and TUG (r=−0.565, p = 0.02), a lesser distance on the 6MWT being correlated with more time to perform the TUG (negative correlation due to worse performance being smaller value of 6MWT and bigger value on TUG). Upper limbs and lower limbs strength did not correlate with any of the variables (p > 0.05).

When the patients were stratified according to the distance covered in the 6MWT (below or above 350 m), the boys who could walk <350 m in 6 min had a significantly higher lumbar lordosis angle and worse scores in BBS and TUG. There were no significant differences in terms of age, BMI, fall index, kyphosis angle, TRF, and MMT between children who could walk less than or more than 350 m in 6 min (Table 2).

Separate diagnostic performances of lumbar lordosis, BBS, and TUG tests were evaluated to assess their ability to discriminate the ambulatory function assessed with the 6MWT. According to the ROC analysis performed, AUC values were excellent for BBS and TUG (0.948, 0.906, respectively) and good for lumbar lordosis (0.850). Based on the maximum value of Youden Index, the optimal threshold was determined to be 36° for lumbar lordosis (Table 3).

The boys whose lumbar lordosis was higher than 36° had significantly worse scores in BBS, TUG, and MMT of the upper and lower limbs and were more likely to have poorer functional levels. The amount of lumbar lordosis had no significant effects on the other variables (Table 4).

The main result of this study is that ambulation and dynamic balance are negatively associated with the increment of lumbar lordosis with a cut-off point of 36° in boys with DMD. It is known that an increase in lumbar lordosis is a compensatory mechanism that occurs secondary to muscle weakness.23 However, our results suggest that balance and gait are affected by the increased lumbar lordosis before obvious strength loss occurs.

This study has several important limitations. The digital inclinometer, which was used to measure lumbar lordosis is reliable in healthy adults, but, although it has been used in pediatric studies,8,9 its reliability was not assessed among children. But, to address this limitation, we checked the intrarater reliability and found that it was very reliable as used in this study. Also, using a pediatric version of the BBS [Pediatric Balance Scale (PBS)] for balance evaluation would have helped us achieve more reliable results. However, a validity and reliability study of the PBS had not been performed for the language spoken by the patients on the date we conducted the study. Limitations related to muscle strength testing include lack of testing of the trunk musculature as well as the use of a method that is highly operator dependent. Normative values of sagittal spinal alignment of children are lacking in the literature, thus it is difficult to determine what values are normal. We have tried to overcome this issue by comparing boys with DMD with healthy control participants. As the functional tests were a part of a routine visit, the intrarater reliability of these tests in our setting has not been determined.

Muscle strength, which is a part of the efferent system of the central nervous system, is an important factor contributing to postural control and balance.24–26 An intact muscular system is required to maintain upright balance, during both stance and ambulation.25 Kids with DMD have poor pelvic control and have difficulty maintaining balance due to proximal weakness.25 However, MMT using the MRC is not a reliable tool to assess the strength of only mildly affected muscles.16 In children who are in an early stage of DMD, clinical or functional tests may not be sensitive enough to detect early compensatory strategies. Moreover, muscle strength has previously been found to be less sensitive than timed function tests to changes in disease status in ambulatory boys with DMD.10 McDonald et al.10 reported that, although there were significant declines in ambulatory function, relatively small changes occurred in muscle strength of patients with DMD. Our findings are consistent with these earlier findings because no significant relationship was detected between muscle strength and the 6MWT and balance tests. In addition, according to the current results, although lumbar lordosis angle was found to be increased in boys with DMD compared to controls, there seems to be no correlation between muscle strength and lumbar lordosis, which contradicts the knowledge that the increase in lumbar lordosis is caused by a decrease in muscle strength.2 These discrepancies may be explained by the muscle strength approaching the floor effect.

When a child with DMD becomes hyperlordotic due to weakness and/or shortness of pelvic girdle muscles, the center of mass demonstrates greater displacements which may lead to balance and gait disturbances.2 Similarly, in the current study, the increase in lumbar lordosis was found to be highly correlated with the ambulation function measured by the 6MWT. Also, significant relationships were detected between lumber lordosis and dynamic balance tests in terms of TUG and TRF, consistent with previous studies.1–3 The main question of the present study was at what degree of lumbar lordosis these alterations start. Based on the ROC analysis, 36° has been determined as a threshold of gait deterioration. And the patients with lumbar lordosis over 36° were shown to have worse scores in dynamic balance tests, muscle strength, and functional level.

Horlings et al.25 and Kaya et al.26 investigated the effects of muscle weakness on balance and found that patients with DMD with proximal weakness had worse scores in static balance than the distal weakness group. Our data showed that dynamic balance was better correlated with ambulation and disease severity than static balance evaluated by posturography. It should be noted, however, that all the patients in this study were at high-risk category according to posturography (median fall index: 100), which might have resulted in a ceiling effect. But, although Kohen-Raz15 reported that the reliability for posturography scores collected among 43 healthy children in second and third grades was between 0.68 and 0.95, because this test was not capable of discriminating between kids with DMD and healthy kids in the present study, a suspicion arises about the validity of the test among children. So, we cannot simply claim that static balance has no relationship with sagittal alignment or functional performance.

An increase in lumbar lordosis has negative effects, especially on dynamic balance and ambulation, in children with DMD. Therefore, to maintain balance and ambulation, it is recommended to monitor these patients in terms of sagittal curvatures such as lumbar lordosis, and tailor physical therapy programs accordingly.