Predictive factors for the development of scoliosis in Duchenne muscular dystrophy.

OBJECTIVE

Scoliosis is a frequent complication (68-90%) of Duchenne muscular dystrophy (DMD). Prevention of limb deformities, rehabilitation in knee-ankle-foot-orthoses (KAFOs) and glucocorticoids prolong walking and standing, and might reduce scoliosis. We evaluated possible predictive factors for scoliosis development in a large DMD population.

METHODS

Case notes of 123 DMD boys, > or = 17 years, followed at our centre between 1992 and 2002 were reviewed. Univariate analysis was used to relate two outcome measures (age at onset of scoliosis and severity at 17 years) with (i) glucocorticoids treatment; (ii) ages at (a) loss of independent ambulation, (b) rehabilitation into KAFOs, (c) loss of standing, (iii) forced vital capacity (FVC) (%) between 11 and 12 years and (iv) lower limb contractures.

RESULTS

In total, 37/123 boys (30%) received intermittent prednisolone (0.75 mg/kg/day, 10 day/month) for a median 1-year (2 months-9 years), starting between 7.7 and 12.4 years (mean 9.5). About 96/123 (78%) were rehabilitated into KAFOs at 10.2+/-1.6 years. Age at loss of ambulation in KAFOs was 12.3+/-1.9 years and at loss of standing 12.8+/-2.1 years. About 95/123 (77%) boys developed scoliosis (Cobb angle >30 degrees ). Mean age+/-S.D. at scoliosis onset was 12.7+/-1.6 years. Forty-three boys (35%) had scoliosis surgery by 15+/-1.2 years. Later age at loss of ambulation (p<0.0001) and longer duration of prednisolone treatment (p=0.01) related to later scoliosis onset. Ages at loss of ambulation and standing were inversely related to scoliosis severity at 17 years (p<0.005). Hip asymmetry and %FVC at 11-12 years were directly related to scoliosis severity (p=0.02).

CONCLUSIONS

Our data indicate a significant association between prolonged ambulation and a reduced risk of scoliosis development. Glucocorticoid administration, in our series, appear to be associated with a later onset of scoliosis, but did not alter the severity at 17 years, probably reflecting the shorter overall glucocorticoid exposure in this population.