Letters to the editors

Dear Editors and readers, We thank Dr Verhoef very much for his interesting letter and for raising this important statistical point. We agree that we should have used a method, such as that of Oldham, to examine the relationship. However, using Oldham's method (Tu and Gilthorpe 2007) does not change the conclusion. Patients with higher mean of admission and day 42 α have a greater reduction in alpha in convalescence (Figure 1, y = −67x + 76, R2 = 0.21, P < 0.001). Therefore, in this case statistical analysis using Oldham's method does not change the conclusion and the evidence indicates that the extent of improvement in thiamin status does depend on the degree of admission thiamin deficiency. Relationship between (admission α and convalescent α)/2 and percentage change in α. We are very grateful to Dr Sue Lee for her statistical advice on this response.