Are reference limits for serum creatine kinase valid in detection of the carrier state for Duchenne muscular dystrophy?

We evaluated serum creatine kinase (CK) as an index to heterozygosity in Duchenne muscular dystrophy. When the 97.5th percentile of the CK normal reference interval was selected as the cutoff point, only 31% of 28 obligate carrier mothers and 24% of 43 mothers of simplex cases (those with only one occurrence of dystrophy in the kindred) exceeded this cutoff value. The outcome depended to some degree on the method used for establishing the reference limit for 379 presumably non-carrier, ambulatory women. The considerable overlap of CK activities between the control and carrier population as well as the non-gaussian distribution of the data permitted no satisfactory approach for differentiating these two populations. Neither the application of likelihood ratios, which evaluates a continuum of results without the dichotomy of a cutoff point, nor the application of predictive value based on sensitivity and specificity, which involves use of a cutoff value for decision making, provided a reliable estimate of carrier status. There was no significant difference (2 alpha = .19) between the median CK activity of obligate carrier mothers and mothers of simplex cases. The serum CK test does not provide data that either support or reject the Haldane hypothesis.