A statistical analysis of the loss of muscle strength in Duchenne's muscular dystrophy

We have confirmed that the loss of muscle strength in 12 boys with Duchenne's muscular dystrophy, as measured by manual muscle testing, approximates a linear decay model, but we have also found that it fits as well a first order decay model. The strength in the same eight muscles was measured over time. Results of an analysis of 111 examinations are reported here. An arbitrary numerical scale for grading muscle strength was used, such that normal was 13 units, and no movement was zero. The maximal sum, if all 8 muscles were normal, would be 104 units. Pooling all measurements, the linear decay rate in this sum was -0.189 ± 0.023 (estimate ± standard error) arbitrary muscle strength units.month^-1. The corresponding first order fractional decay rate was -0.0034 ± 0.0004 month^-1. However, a more detailed statistical analysis indicated that decay rates in muscle strength were not homogeneous, i.e. muscle strength decayed faster in some patients than others. The decay constants in 11 of the 12 subjects spanned a 10-fold range, and in one subject increased the spread to 40-fold. The distribution frequency of decay rates appears to be bimodal. In these assessments, the muscle strength at time zero was not known. Therefore, an estimate of muscle strength at 10 years was made. This varied from 34 to 71 units. The group mean was 53.0 units using the exponential model, i.e. on average, only 51% (= 53.0/104 X 100) of the normal muscle strength remained in the 8 muscle groups assessed at age 10 years. In conclusion, a quantitative characterization of muscle strength deterioration is reported, which emphasizes the heterogeneity in this disease. This approach may eventually allow quantitative distinctions between Duchenne's and Becker's varieties of muscular dystrophy.