Statistical analysis
The analyses were performed in SPSS Statistics for Windows version 23 (IBM Corp., Armonk, NY, USA). The average concentrations of Hb, ferritin and 25OHD over one year period prior to the study were used for the analysis. Data distribution was assessed before performing the analyses. Descriptive statistics are presented as number (percentage), mean ± SD or median (Q1-Q3), as appropriate. The χ2 or Fisher’s exact test was used for the comparison of proportions. Participants were stratified according to tertiles of Dkk-1 concentration [low: range <197 pg/mL, mid: 197-523 pg/mL, high: >523 pg/mL]. For the group comparison, a t-test, one-way ANOVA or one-way analysis of covariate (ANCOVA) was used for normally distributed data, and a Mann-Whitney U test or Kruskal-Wallis test was used for skewed data.
Spearman’s method was used for the correlation analysis. Multivariable regression models were fit to evaluate the associations between Dkk-1 and BMD z-scores. Participant BA, sex, body mass index (BMI) z-score, average pre-transfusion Hb and 25OHD concentrations, the history of delayed puberty, subtype of non-deletional Hb H disease, type of iron chelator, and prednisolone use were considered as covariates based on previous data supporting their influences on BMD in thalassemia.10,25,26 Our analysis was considered exploratory and hypothesis-generating, therefore, we did not perform multiple comparisons adjustment. P-value <0.05 was considered statistically significant.