Discussion
Adolescents with non-deletional Hb H disease have many risk factors for compromised bone health similar to their peers with β-TI whose disease severity is comparable. In the present study, we demonstrated high prevalence of low BMD in adolescents with non-deletional Hb H disease. Notably, the majority of them receiving frequent blood transfusion. The prevalence of low BMD was higher at the lumbar spine as compared to the total body site. After adjusting for confounding factors, TBBMD was inversely associated with serum Dkk-1 concentration, suggesting that alteration in the Wnt signaling pathway might play an important role in decreased bone mass and might be a potential bone biomarker in non-deletional α-thalassemia.
Our study found that the prevalence of low LSBMD in adolescents with non-deletional Hb H disease is slightly lower (42% vs. 50%) than the report from the United States Thalassemia Clinical Research Network (USTCRN) study.12 Moreover, when compared with USTCRN study, the prevalence of low LSBMD in adolescents with deletional Hb H disease was similar to their peers with β-TM (44%) and β-TI (44%), lower than those with both transfusion-dependent and nontransfusion-dependent forms of Hb E/β-thalassemia (58%), and higher than those with Hb H disease (13%).12 Notably, the high prevalence of low LSBMD in Hb E/β-thalassemia vs. other forms in this study might be related to the clinical severity as demonstrated by shorter height, more delayed BA, lower pre-transfusion Hb, and higher serum transferrin receptor concentration. In addition, sex, race, vitamin D status, blood transfusion and iron chelation protocols as well as different methods used for calculating BMD z-score (i.e., height adjusted vs. unadjusted BMD z-score) may also contributed to the differences observed. Recently, Nakavachara et al. measured the height-adjusted BMD z-score in Thai adolescents with nontransfusion-dependent Hb E/β-thalassemia and found lower prevalence of low LSBMD (23%) and TBBMD (9%)27 as compared to peers with non-deletional Hb H disease or Hb E/β-thalassemia in our study and USTCRN. Again, participants in the Nakavachara’s study seemed to have less severity of anemia and lower ferritin concentration. As expected, the prevalence of low BMD was higher at the trabecular lumbar spine, where bone marrow expansion occurs in response to anemia, compared to the total body site. Taken together, other studies and ours confirm the high prevalence of low LSBMD in non-deletional α-thalassemia, comparable to β-thalassemia of similar anemia severity and blood transfusion requirement.
Our results are in line with other thalassemia studies that age, BMI z-score (i.e., nutritional status) and pubertal status5,12,28 are associated with BMD. Furthermore, male adolescents had lower LSBMD z-score and a trend toward lower TBBMD z-score compared to females despite having comparable blood transfusion frequency, iron overload, and rates of endocrine complication, and even higher Hb concentration. However, the latter is considered to be normal physiology as healthy male vs. female adolescents have ~12% higher Hb concentration due to testosterone effect.29 Previous literature in adults with thalassemia major supports our findings by demonstrating that male vs. female patients had an increased prevalence of low BMD particularly at the LS site.30 Interestingly, a recent meta-analysis showed men vs. women with anemia and otherwise healthy had higher rates of fracture31 suggesting that anemia per se may have different impacts on male vs. female skeletons. Moreover, there is empirical evidence showing that males with β-thalassemia are more susceptible to iron toxicity in many organs than females32,33 which is linked to sex-related expression of various pathways such as oxidative stress and erythropoietin.33 In α-thalassemia, this information remains unknown, thus, further research is needed.
Wnt/β-catenin pathway is crucial for the gain and maintenance of bone mass. Several agonizing and antagonizing molecules involve in the regulation of this intricate pathway.34 To date, many studies have assessed sclerostin,16,17 another Wnt inhibitor, in various causes of osteoporosis including thalassemia, but only a few studies have evaluated serum Dkk-1 concentration in thalassemia.18,35 Voskaridou et al. found that adults with transfusion-dependent β-thalassemia vs. healthy controls had elevated concentrations of Dkk-1 and sclerostin, both of which were associated with lower BMD z-score at the lumbar spine, forearm and femoral neck.17,18 They also found that among various bone metabolism markers (including receptor activator of NF-kB (RANK) ligand and osteoprotegerin), only Dkk-1 was an independent predictor for severe low bone mass. In addition, a reduction in Dkk-1 but not sclerostin was observed in their participants following bisphosphonate therapy.18 In adolescents with non-deletional Hb H disease, we demonstrated similarly that Dkk-1 was inversely correlated with LSBMD and TBBMD z-scores. However, when adjusting for confounders, Dkk-1 remained an independent negative predictor only for TBBMD z-score. We also examined factors associated with Dkk-1 concentration and found that interval of blood transfusion, a surrogate marker of anemia severity, had a negative correlation, and the history of delayed puberty had a positive correlation with Dkk-1. Multiple studies also found people with hypogonadism had increased Dkk-1 and sclerostin concentrations.36-39 However, the mechanism linking sex hormone and Dkk-1 is incompletely understood. Whether circulating Dkk-1 involves in the pathogenesis of decreased BMD in thalassemia needs future research.
Our study had some limitations. First, our study design was cross-sectional, thus, the association results must be interpreted with caution. Further, our study also had a small sample size that may lead to type II error. Finally, we did not measure sex hormone concentrations, therefore, we could not determine their associations with Dkk-1.
In conclusion, adolescents with non-deletional Hb H disease had a high prevalence of low BMD with the lumbar spine being the prominent site. Our reported prevalence is akin to that in β-thalassemia of similar severity. Furthermore, Dkk-1 was independently associated with TBBMD suggesting its potential bone biomarker in thalassemia. Whether Dkk-1 involves in the pathogenesis of low bone mass in non-deletional Hb H disease requires further research.