In the quality control samples sent to each center every quarter, measurements were quite similar across centers. In particular, results from the laboratory at Minnesota and Missouri were in close agreement. We concluded, therefore, that differences in laboratory methods did not explain our findings.