Introduction

β-Thalassemia is a group of heterogeneous disorders inherited in the form of autosomal recessive. The disease is classified according to the type of defect in the β-globin chain, characterizing insufficient erythropoiesis and iron overload [1]. Some factors, including the age of receiving a red blood cell (RBC) transfusion, the age of beginning iron chelation therapy, iron siderosis in vital organs (e.g., liver, cardiac, and endocrine glands), tolerance to chelation therapies, and red-cell transfusion requirements have been considered as the disease intensity in these cases [2,3]. According to the continuous glucose monitoring system (CGMS), the highest incidence of diabetes mellitus (DM) has been reported in patients with a β-thalassemia major in their second and third decades [4].

Zinc, as a vital element, in addition to its role in pancreatic cells, also plays a pivotal role in improving the control of blood glucose in peripheral tissues [5]. Serum zinc concentration regulates blood sugar, insulin storage and secretion, and RBC survival [6]. Zinc deficiency (hypozincemia) is found in three to 30% of β-thalassemia patients. An increased zinc requirement [7] and hyperzincuria ensuing from iron chelation therapy, hyperferritinemia, and liver dysfunction can be enumerated as the underlying causes of hypozincemia in thalassemic cases.

Insulin resistance and damaging beta cells caused by an iron overload may play a significant role in emerging impaired glucose homeostasis (IGH) in patients with β-thalassemia [8,9]. Moreover, such determinations as the frequency of red-cell transfusion, and treatment with iron chelators, for instance, deferoxamine, can impact the blood glucose homeostasis in these patients [10]. Therefore, this study planned to evaluate serum zinc levels in β-thalassemia patients with DM and perform a risk assessment for hypozincemia.

Method

Results

Of 64 diabetic β-thalassemia patients, 41 cases had zinc data in their medical file (aged 38 ± 9 years, 48.8% female). 78.05% of patients (n = 32) were TDT, and 21.95% were NTDT (n = 9). The mean zinc level was 110.2 ± 127.6 μg/dL [median 89.0, first quartile 78, third quartile 105 μg/dL]. The prevalence of hypozincemia and hyperzincemia was 9.76%, 95% CI 0.27 to 19.24 (four cases) and 2.44%, 95% CI -2.49 to 7.39 (one case), respectively. The basic and clinical characteristics of diabetic β-thalassemia patients, alongside their laboratory and paraclinical information, have been shown in Tables 1 and 2.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 1. Basic and clinical characteristics of diabetic β-thalassemia patients.

https://doi.org/10.1371/journal.pone.0284267.t001

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 2. Laboratory and paraclinical information of diabetic β-thalassemia patients.

https://doi.org/10.1371/journal.pone.0284267.t002

The correlation between serum zinc level and age was approximated at -0.39; P = 0.01. An increase in the length of DM and using insulin were associated with the risk of hypozincemia (OR 1.94, 95% CI 0.29 to 12.6 and OR 5.18, 95% CI 0.76 to 34.53, correspondingly). A risk for hypozincemia was also found with an increase of serum ALP, OR 2.40, 95% CI 0.36 to 15.68.

A crude odds of 5.17, 95% CI 0.60 to 44.18, was estimated for hypozincemia in the cases under deferasirox (DFX) therapy. After controlling age, the OR of hypozincemia for using DFX was 8.77, 95% CI 0.60 to 127.1. Following converting ORs to RRs index, the modified risk of hypozincemia for patients under treatment with DFX changed to 21% (RR 1.21 95% CI 0.88 to 1.24).

The crude odds of hypozincemia in the cases affected by hepatitis C were obtained at 3.44, 95% CI 0.27 to 44.32. In β-thalassemia patients, the age-adjusted risk of hypozincemia was calculated at 15.85, 95% CI 0.47 to 529.3 for hepatitis C. The adjusted risk of 11% was also found for hypozincemia in cases with hepatitis C, RR 1.11, 95% CI 0.89 to 1.12. A crude odds of 2.36, 95% CI 0.29 to 18.97, was computed for hypozincemia in those who received antacids treatment. The age-adjusted risk of hypozincemia for antacid use based on age was 6.34, 95% CI 0.39 to 102.7. The risk of hypozincemia, 36%, rose in those who took antacids after a bias correction (RR 1.36, 95% CI 0.67 to 1.46).

The crude odds of hypozincemia in the cases with liver MRI T2* lower than 6.3 ms was calculated at 1.47, 95% CI 0.12 to 17.91. The adjusted risk of hypozincemia in patients with reduced liver MRI T2* (<6.3 ms) (< 10 ng/mL) has been valued at 3.07, 95% CI 0.09 to 109.1, after fixing the effects of age and length of using zinc supplementation. A 65% boost in the risk of hypozincemia was computed for patients with reduced liver MRI T2* (<6.3 ms) (RR 1.65, 95% CI 0.14 to 2.37). The potential determinant factors for hypozincemia in β-thalassemia patients have been displayed in Table 3. After resampling by the Jackknife method, all the estimated ORs did not change. The results of resampling (Jackknife method) have been displayed in Table 4.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 3. Potential determinant factors for hypozincemia (< 70 μg/dL) in diabetic β-thalassemia patients (n = 41).

https://doi.org/10.1371/journal.pone.0284267.t003

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 4. The results of resampling (Jackknife method).

https://doi.org/10.1371/journal.pone.0284267.t004

The multivariate logistic regression analysis revealed that the risk of hypozincemia for using DFX surged to 15.81, 95% CI 0.56 to 446.8. An increase in the risk of hypozincemia for hepatitis C was also obtained, OR 7.11, 95% CI 0.18 to 275.1. Moreover, the risk of hypozincemia for using antacids and reduced liver MRI T2* (<6.3 ms) changed to 4.35, 95% CI 0.17 to 108.9 and 2.67, 95% CI 0.06 to 108.9, correspondingly. The multivariate logistic regression model results have been presented in Table 5.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 5. The results of multivariate logistic regression analysis for hypozincemia (< 70 μg/dL) in diabetic β-thalassemia patients (n = 41).

https://doi.org/10.1371/journal.pone.0284267.t005

Discussion

In this study, around 10 percent of diabetic β-thalassemia suffered from zinc deficiency. In our center, administering zinc supplementation led to conserving zinc levels in the optimum ranges. Therefore, prescribing zinc supplementation at the right dose and duration can significantly help correct serum zinc levels, particularly in at-risk cases. Furthermore, the evaluation of the results indicated that using DFX was associated with hypozincemia risk, nearly 21%. In these cases, the risk of hypozincemia in diabetic patients with hepatitis C was 11%, compared to non-hepatitis C cases. The use of antacids was linked to a 36% risk of hypozincemia. Any grade of liver hemosiderosis was accompanied by a 65% risk of zinc deficiency. Although the small sample of our study resulted in non-significant associations between hypozincemia and the potential risk factors in these cases, the resampling approach revealed that the calculated ORs were not affected by the low sample size.

Previously, zincuria was reported as a frequent adverse consequence in patients with DM [15]. Increased urinary excretion of zinc can be the key driver for zincuria in thalassemia patients with DM [16]. A cross-sectional study [17] demonstrated that the rate of hypozincemia was equal to 100% in patients with β-thalassemia cases, which prevailed more in males. In a study [18], the prevalence of hypozincemia was reported to be 33% higher in men. Another study [19] also observed hypozincemia in 24% of 75 cases of β-thalassemia, which was 2-fold in men. The prevalence of hypozincemia in our study was 9.76% and was not predominant in either sex, contrary to the above studies.

Zinc deficiency may occur in β-thalassemia cases for various reasons, including constant red-cell transfusion, hyperferritinemia, liver hemosiderosis and dysfunction, urinary overexcretion, and higher demand for zinc—as cofactors of antioxidant enzymes—in response to raised oxidative stress attributable to hemolysis and iron overload [18,20]. In addition, hypozincemia tends to be presented as a repercussion of excessive zinc losses that stem from chronic hemolysis in this population [20]. The rigorous iron burden may damage hepatocytes under the surged oxidative stress levels by overproducing free radicals, which may more widely release zinc into the plasma [21]. As such, iron chelators can cause excessive zinc excretion following an increase in the extracellular zinc pool [22]. In β-thalassemia patients, the heme production process is where zinc element partakes to compensate for ineffective hematopoiesis, taking place based on anemia severity [23,24]. Furthermore, to regulate iron overload, some nutritional considerations restricting dietary iron-enriched sources, such as meats and cereals, are most often recommended for β-thalassemia cases, putting them at risk of trace elements deficiency, e.g., hypozincemia [24,25]. Moreover, to adhere to the transferrin binding sites competitively, the iron surplus cuts zinc absorption down in the gastrointestinal tract [26].

Thalassemia patients treated with iron chelators suffer from the chelation of zinc minerals, which is the most plausible mechanism of hypozincemia in these patients [26]. Other roles of iron chelators in the development of hypozincemia are related to the interaction of these medications with double bonds of zinc ions, decreasing serum zinc levels [27]. Moreover, iron chelators have a high affinity to zinc elements [28].

Deferasirox (Exjade, ICL670) is a potential factor in developing hypozincemia, increasing the chance of hypozincemia. Several studies have also pointed out that serum zinc levels in thalassemic cases under DFX therapy were lower than those treated with deferiprone (DFP) [22,27,29]. Based on the results of an investigation on 59 β-thalassemia major patients, although urinary zinc excretion after DFX therapy was significantly low compared to deferoxamine) DFO (plus DFP, the level of serum zinc was slightly less amid cases treated with DFX (64.8±14.8 μg/dL in DFX group versus 66.5±15.1 μg/dL in DFO plus DFP group). The urine levels of zinc in DFX group were 662.2±428.2 μg/day in comparison with 1182.3±980.3μg/day in DFO plus DFP group [26]. Consequently, zinc is lost in non-urinary ways, such as the gastrointestinal system, resulting in hypozincemia in this population.

The most excretion of zinc elements following DFX treatment occurs through biliary elimination via the gastrointestinal tract and feces [30]. A pharmacokinetics study [30] exhibited that the excretion of DFX in urine is minor. Only 6 to 11% of the administrated dose of DFX was excreted by urine in rats (mean: 101% of the dose, within seven days post-dose). 6.3, 69.3, and 21.5% of the administrated dose of DFX were recovered in urine, bile, and feces, respectively [30]. Besides, a disturbance in the zinc metabolism and homeostasis might tend to exist after DFX therapy, causing a lifted risk of hypozincemia in β-thalassemia patients, especially if the serum level of DFX peaks at the higher level of the therapeutic level [31,32].

Hepatitis C is another prominent factor in the development of hypozincemia. Nosheen Aslam et al. [33] conducted a case-control study involving patients with hepatitis C (HCV) (N = 25) and healthy control (N = 25), who examined the levels of heavy metals in the two groups. Their evaluation showed that the heavy metal levels were higher in hepatitis C cases. Serum zinc level in the HCV group was lower than healthy controls, indicating hypozincemia (42.9 ± 17.09 μg/dL versus 81.8 ± 13.52 μg/dL; P< 0.001) [33]. Chih-Hung Guo et al. [34] designed a study made up of three groups of patients, HCV (n = 30), HCV plus non-alcoholic fatty liver disease (NAFLD) (n = 32), and healthy controls (n = 30). They reported that serum zinc levels in the HCV group were lower compared to the healthy group (median 0.68 mg/L first quartile 0.65, third quartile 0.77 versus median 0.90 mg/L first quartile 0.77, third quartile 1.03), while the values were greater than HCV+NAFLD (median 0.58 mg/L first quartile 0.52, third quartile 0.63 in HCV+NAFLD group). The correlation estimated for serum ferritin levels and zinc concentrations was -0.60, presenting a moderate strength of association between them [34].

In HCV, due to liver damage caused by chronic inflammation and oxidative stress, heavy metal homeostasis is out of regulation, associated with increased accumulation [35]. The accumulation of metals such as copper (Cu) and lead (Pb) in these patients will be alongside the disruption of zinc homeostasis and hypozincemia [36]. Iron overload is of paramount importance to liver injury following HCV, which is mediated by cytotoxicity, immune-related processes, inflammation, and chronic oxidative stress [37]. Hepatotoxicity induced by iron overload alters zinc metabolism and causes zinc deficiency [38]. Another mechanism for the progression of hypozincemia in HCV patients would be the poor absorption of zinc, despite the zinc deficiency [39].

Prescription of gastric acid secretion inhibitors increases the risk of zinc deficiency. Serum zinc levels were assessed in a study (n = 46), including patients with gastroesophageal reflux disease who had a history of more than six months of use of PPIs [40]. The results demonstrated that the patients treated with PPIs had significant hypozincemia compared to non-users. In conformity with our study, their results indicated an almost 4-fold ramp up in serum zinc levels in non-users than the case group [40].

Using PPI might reduce the reabsorption of divalent ions, such as calcium, zinc, and magnesium by ascending the potential of hydrogen (PH) levels and descending the chloride content of the duodenum [41]. PPIs irreversibly inhibit the K-H pump of parietal cells, which deters the production of hydrochloric acid (HCL). The absence of gastric acid can cause significant problems in the absorption of nutrients and pathogens’ elimination [42,43].

Physiologically, zinc trace element plays a crucial role in multiple functions concerning glucose hemostasis, not only in the pancreas but also in insulin performance. Some evidence supports the connection between hypozincemia and the presence of insulin resistance in β-thalassemia patients [18,20]. To be precise, in β-thalassemia cases, the plasma zinc concentrations seem moderately correlated with β-cell function (r = 0.51) [44]. A pilot study on diabetic β-thalassemia patients (n = 9) unveiled that using zinc supplements for three months can bring an improvement in FPG to these cases. Accurately, a 3.7% fall in FPG was aligned with a 17% rise in plasma zinc concentration [44], representing the desired role of zinc element in improving glycemic metabolism. Intriguingly, the relationship between zinc intake and FPG was also shown across a meta-analysis carried out on non-thalassemic patients with type 2 DM (weighted mean difference [WMD] −20.34 mg/dL, CI −29.04 to −11.64), albeit the evidence level was not high [45]. Additionally, several studies have discovered the deleterious impact of inflammation and oxidative stress in emerging hyperinsulinemia, insulin resistance, and overt diabetes in β-thalassemia cases [46–48]. As well as anti-inflammatory properties [49] and stabilizing effects on RBCs [50], zinc can contribute to maintaining the structure of antioxidant enzymes, such as peroxidase, oxide dismutase, and catalase [49,51]. As such, the antioxidant effect of zinc would be positive in making IGH improve in β-thalassemia patients [6,52].

Conclusion

As well as hepatitis C, using DFX and antacids is associated with a high risk of hypozincemia amid diabetic β-thalassemia cases, although overestimation bias should be viewed in this study. Meanwhile, a reduced myocardial and liver MRI T2* as cardiac and liver hemosiderosis can raise the risk of zinc deficiency in this population. The heterogenous β-thalassemia population in transfusion dependency was one of the limitations of the current study. Although the risk assessment revealed several risk factors according to the estimated effect sizes, the relationships were not significant due to the small sample size. Therefore, more studies should be carried out in the future to confirm the results. Ultimately, the results can have an overestimation bias, and readers should be cautiously aware once they interpret the results.

Supporting information

Acknowledgments

This study was derived from Amir Mosanejad-Galchali’s medical doctorate thesis. We are grateful to the nurses and employees at the thalassemia ward at Bu Ali Sina Hospital for providing all the necessary facilities.