Association of Anemia with Parathyroid Hormone Levels and Other Factors in Patients with End-Stage Renal Disease Undergoing Hemodialysis: A Cross-Sectional, Real-World Data Study in Pakistan

End-stage renal disease (ESRD) patients are mostly managed with maintenance hemodialysis (MHD). ESRD patients on MHD also present with many complications, such as anemia, hyperparathyroidism, and hepatitis prevalence. This study depicts the real-world scenario of anemia among MHD and end-stage renal disease patients in the Pakistani population. A retrospective, multicentric, and real-world data analytical study was conducted at 4 dialysis centers in Pakistan. The study had a sample size of n = 342 patients on maintenance hemodialysis. The data were gathered from the medical records of patients. Data analysis was performed using STATA Version 16. Statistical significance was gauged at a 0.05 level of significance. According to our results, the mean age of the patients was 45 (±15) years. Most of the patients were male (n = 234, 68.4%), whereas 58.1% of the patients were maintained on twice-weekly hemodialysis. The most commonly reported comorbidities were hypertension and diabetes mellitus. The frequency of dialysis (P < 0.01) and comorbidities (P = 0.009) had a significant association with anemia in MHD patients. The majority of the patients had hyperparathyroidism (52%) with anemia. Upon performing binary logistic regression, multivariate analysis displayed a similar odds value for having anemia in patients with every additional month in the duration of hemodialysis (OR 1.01, P = 0.001), the odds of anemic patients having a positive antihepatitis-C antibody (OR 2.22, P = 0.013), and the odds of having anemia in patients in the age category below 45 years (OR 1.93, P = 0.013). In conclusion, the study results depict that every additional month in the duration of hemodialysis, age (<45 years), and positive anti-HCV antibody status, these variables were more likely to have anemia in our study MHD patients. While in our final multivariate model, no statistically significant association was observed between hyperparathyroidism and anemia.


Introduction
Anemia is one of the most common complications associated with maintenance hemodialysis (HD) patients. Tere are several reasons for developing anemia in end-stage renal disease (ESRD) patients. Te reduction in erythropoietin (EPO) production is one of the main reasons for anemia in ESRD patients [1].
Anemia is considered to be one of the predictors of poor quality of life (pQOL), numerous morbidities, and mortality for maintenance HD (MHD) patients [2]. Some researchers reported that the mortality rate increases by 1.58 times with every 1 g/dL reduction in hemoglobin (Hb) levels [3]. Severe cardiac complications and alteration in cardiac output are the signifcant consequences of anemia that intensely deteriorate the quality of life of HD patients. Anemia can be defned as the hemoglobin (Hb) level becoming <12 g/dL in women and <13 g/dL in men [4]. Anemia is of multifactorial origin [5], and many researchers suggested the possible mechanisms for anemia in ESRD. Te mechanisms suggested by researchers are as follows: iron defciency, infammation, and the raised parathyroid hormone (PTH) level also contribute to causing anemia [6].
Renal anemia is a common complication with raised serum PTH levels in ESRD patients, and its pathogenesis ( Figure 1) is complex and remains incompletely understood. Intact parathyroid hormone (iPTH) is mainly responsible for the regulation of calcium and phosphorus in blood, while calcium and phosphorus homeostasis is disturbed; that is, when calcium levels decrease in blood, phosphorus levels increase. Tis decreased calcium level due to defciency of vitamin D (vit-D) or decreased mineral metabolism in ESRD patients eventually results in hypocalcemia and hyperphosphatemia.
Hyperphosphatemia stimulates fbroblast growth factor 23 (FGF-23), a protein that is secreted by bones. First, FGF-23 increases the phosphorus excretion from the kidneys; later on, it further enhances hyperphosphatemia by retention of phosphorus from the kidneys [7]. Consequently, increased serum phosphorus binds with bioavailable calcium to produce calcium hydrogen phosphate (CHPO 4 ), leading to hypocalcemia that stimulates PTH production and secretion. FGF-23 also decreases the conversion of the inactive form of vit-D to active form of vit-D by reduction of 1-alpha hydroxylase activity [7]. Decreased availability of the vitamin D active form leads to further hypocalcemia; as a result, it produces drastic increments in the iPTH levels and contributes to secondary hyperparathyroidism in ESRD patients [8].
Secondary hyperparathyroidism or renal hyperparathyroidism (rHPT) decreased the serum Hb level in ESRD patients by two mechanisms. During the frst mechanism, the increased serum iPTH level leads to a signifcant decrease in erythropoietin production that causes an inhibitory efect on erythroid progenitors by downregulating erythropoietin receptors and increasing the insensitivity of the erythroid progenitor's cell to EPO. Consequently, reduced erythropoiesis and decreased serum Hb levels were observed in ESRD patients [9]. Te increased serum iPTH level leads to a signifcant decrease in erythropoietin production. Many researchers have proved this hypothesis by demonstrating that parathyroidectomy reduces the EPO requirement in ESRD patients [4]. Several researchers reported that numerous EPO receptors on erythroid progenitor cells might be regulated by various factors such as calcitriol. Te defciency of calcitriol leads to downregulation of EPO receptors on the erythroid progenitor, and as a result, the erythroid progenitor becomes insensitive to the erythropoietin action. Te second mechanism, increased iPTH, leads to a hemolytic efect due to calcium disturbance and enhances osmotic pressure and the fragility of red blood cells (RBCs). Both mechanisms lead to renal anemia in ESRD patients [9].
According to the "Kidney Disease Outcomes Quality Initiatives" (KDOQI) [10,11] and the "Kidney Disease Improving Global Outcomes" (KDIGO) guidelines 2013 [12], chronic kidney disease (CKD) is defned as the incidence of kidney damage and deterioration of kidney function, irrespective of the cause, for ≥3 months. Secondary hyperparathyroidism is an insidious disease, and it is commonly associated with the early course of CKD and worsens with the decrease in the glomerular fltration rate (GFR) [13]. KIOQI recommendations for targeted serum iPTH levels according to CKD stages are as follows: in stage 3, the serum iPTH level should be within the range of 35 to 70 pg/mL, in stage 4, serum iPTH lies between 70 and 110 pg/mL, and in stage 5 or ESRD, serum iPTH should be between 150 and 300 pg/mL [11,14]. In Pakistan, CKD afects 21.2% of the population [15], and a potential predictor >100 per million population of Pakistan annually undergoes ESRD [16]. Published studies have reported the association of anemia with hyperparathyroidism in MHD patients. Some studies have reported a negative correlation between Hb and iPTH [1,17,18], and another study has reported a weak negative correlation between Hb and iPTH levels [13]. However, their results regarding the correlation between PTH and serum Hb were not consistent. Till today, no research study has used a multivariate model to evaluate the association of Hb and iPTH levels and validate their associations in ESRD Pakistani patients. Also, all previously reported studies were conducted on a small pool of hemodialysis patients [1,4,13,17,18]. Most of the studies did not mention which recommended cutof levels of hyperparathyroidism (iPTH > 300 pg/mL) were used for the determination of anemia associated with raised iPTH levels [1,18]. Te existing literature is insufcient for healthcare professionals in Pakistan to devise local recommendations for the clinical prognosis and treatment modalities in patients with raised iPTH undergoing maintenance hemodialysis.
Besides this, there was a need to conduct a study on a large group of maintenance hemodialysis patients to validate the association between hyperparathyroidism and anemia. Terefore, the rationale of our study is to evaluate the prevalence of anemia in ESRD-undergoing MHD patients and the correlation between serum Hb levels with raised intact PTH (iPTH > 300 pg/mL) levels and other factors. Te attained information would be benefcial to healthcare professionals and health education programs for better management of renal anemia with secondary hyperparathyroidism in ESRD patients undergoing MHD.

Study Design and Clinical Settings.
A retrospective, multicentric, and real-world data analytical study was conducted from January 2021 to May 2022 at the dialysis centers of Lahore, Faisalabad, and Peshawar in Pakistan. Te data were collected from patient fles of institutes that include District Headquarter (DHQ) Hospital, Faisalabad; Bahria Town International Hospital (BTIH), Lahore; Khyber Teaching Hospital (KTH), Peshawar; and the Institute of Kidney Diseases (IKD), Peshawar. DHQ, Faisalabad, KTH, Peshawar, and IKD, Peshawar, are public sector tertiary care teaching hospitals, while BTIH, Lahore, is a private sector tertiary care hospital.

Sample Size and Selection
Criteria. Te sample size (n � 342) was calculated by using the OpenEpi version 3, sample size calculator, keeping power at 80%, level of signifcance at 5%, and confdence interval at 95% [4,18]. Male and female patients ≥18 years of age who had been on HD maintenance for at least 3 months were included in the study. Nonnative origin HD patients, pregnant HD patients, patients who had three months of bleeding history before initiation of the study, and patients with transplant workups were excluded from the study.

Ethical Approval.
Te study was conducted from January 2021 to May 2022 in multiple dialysis centers in Pakistan. Te study was approved by the Ethical Review Committee (ERC) of Faisalabad Medical University, Faisalabad, with approval reference number (F.48-ERC/2021-22/PHRC/FMU/260) presented as (SF-1: appendix I).

Data Collection and Screening.
Te data were collected from the medical records of hemodialysis patients in the abovementioned sites, and screening was conducted based on the inclusion/exclusion criteria of the study. Upon successful screening, the data were flled in the predesigned proforma (SF-2: appendix II), which was later entered into a computer. Te study included the natives of urban and rural areas of Pakistan who were on MHD. Te data gathered in the study include demographics, date of 1 st dialysis or month of dialysis, number of HD per week, vascular access, vaccination for hepatitis B (HBV), cause of renal failure, any other comorbid disease, antihepatitis B surface agent (HBsAg), antihepatitis C (HCV) antibodies, systolic and diastolic blood pressure, and lab profles (if available). For our study of ESRD patients, a cutof value for serum iPTH categorized as <150 pg/mL represented below target serum iPTH levels and 150 to 300 pg/mL represented a targeted serum iPTH level, and for above target serum iPTH levels or hyperparathyroidism, we took a cutof value (hyper-iPTH; >300 pg/mL); the specifc ranges were referred from KIDGO guidelines [11,13,14]. Similarly, a cutof value for hemoglobin in terms of anemia was referred from the "KDIGO Clinical Practice Guideline for Anemia in CKD" [10,12,19], below 12 g/dL represented anemia and 12 g/dL to 16 g/dL represented without anemia. Te overall workfow of data collection and analysis is shown in Figure 2.

Statistical
Analysis. Data analysis was accomplished using STATA version 16. For continuous variables, summary statistics (including the number of patient profle data (n), mean, standard deviation, median, minimum, and maximum values, as well as frequencies and percentages for categorical variables) were reported. Fisher's exact chisquare test was performed for categorical variables to gauge the statistical signifcance of the association between diferent factors and anemia. Cramer's test was performed to identify the correlation of all variables with serum Hb levels and serum iPTH levels. Statistical signifcance was gauged at α � 0.05 level of signifcance. Binary logistic regression was performed (n � 342), at both univariate and multivariate levels using the level of signifcance � 0.05. Te fnal model was synthesized using statistically signifcant covariates from the univariate analysis with a forward approach; the odds ratio and P values were reported for the fnal model as well. Table 1 demonstrates the results; a total of 342 hemodialysis patients were recruited in the study, of which n � 234 (68.4%) were male and n � 108 (31.6%) were female. Te mean age of the study patients was 45 ± 15 years. Our results showed that, out of n � 342 study patients, the frequency of dialysis was found to be more distributed as twice-weekly among n � 198 (58.1%) patients as compared to once or thrice weekly. A graphical representation of the month of duration of dialysis among our study HD patients is presented in Figure 3. Of the total serum Hb levels of 342 patients recorded, the analysis depicted that prevalence of anemia was higher in patients (n � 267, 78%) than that in patients (n � 75; 22%) having normal serum Hb levels.

Results
Among the patients, 42% were found to be anti-HCV antibodies positive. Our results showed a hepatitis B vaccination rate (n � 185, 66%) in the included study patients, while n � 12 patients (3.5%) were found to have HBV surface antigens (HBsAg). According to our results, the most common comorbidities observed among 230 HD patients were hypertension (n � 185, 80.4%) and diabetes mellitus (n � 23, 10%). Serum iPTH levels were reported for only 179 patients; however, out of those 179 patients, n � 126 patients (70%) were found to have hyperparathyroidism in our study MHD patients. Table 2 presents the study correlation results for covariates of hemodialysis patient profles with hemoglobin levels. Both age categories have a positive correlation (0.147) with serum Hb levels among studied MHD patients. In these age categories, they were also found to have a signifcant (P � 0.009) association with anemia. Both male and female hemodialysis patients had a nonsignifcant (P � 0.237) association with anemia. Furthermore, the association of serum iPTH levels with serum Hb levels was also evaluated. Based on the statistical evaluation, raised serum iPTH levels were found to have a signifcant (P � 0.005) association with anemia as the majority of patients (n � 93, 52%) with hyperparathyroidism had anemia. Likewise, hepatitis C had a signifcant (P < 0.01) association with anemia as the majority of patients (n � 94, 27.5%) with anti-HCV antibodies positive had anemia. However, hepatitis B had a nonsignifcant (P � 0.076) association with anemia, and only n � 12 (3.5%) patients were anemic with positive HBV surface antigens. As per the results, both comorbid diseases and the frequency of dialysis per week have a positive correlation with anemia among hemodialysis patients. Te frequency of comorbid diseases with anemia in HD patients observed in this study was that n � 137 (59%) patients were anemic with hypertension and n � 20 (8.7%) patients were anemic with diabetes mellitus. In addition, the statistical test result shows (Table 2) a signifcant association between anemia with comorbid disease (P � 0.009) and the frequency of dialysis per week (P < 0.01). Table 3 depicts the Fisher's exact statistic and chi-square test results that show a statistically insignifcant (P � 0.106) association between anemia with serum iPTH levels based on stratifcation of age categories 18-44 and above 45 years in HD patients. In our univariate model, serum Hb levels were kept as the dependent variable and the level of signifcance was at 0.250; the covariates such as frequency of dialysis per week, anti-HCV antibodies, age category, and months on hemodialysis were all statistically signifcant when univariate binary logistic regression was performed. All other variables except gender were excluded from the fnal model for the association of covariates with serum Hb levels, as refected in Table 4.
In our fnal model for binary logistic regression, keeping hemoglobin levels as our dependent variable, we found that the odds of having anemia were similar to the odds of having normal hemoglobin levels, for every additional month on hemodialysis (P � 0.011).
Among our patients, the odds of having anemia were 2.2 times higher than the odds of having normal hemoglobin levels for those having a positive anti-HCV antibody status 4 International Journal of Clinical Practice (P � 0.013), whereas when stratifed based on the age group, the odds of having anemia were 1.93 times higher than the odds of having normal serum Hb levels, for patients aged between 18 and 44 years (P � 0.031), as refected in Table 5.
No signifcant interaction and confounding were found in the multivariate model. For the goodness of ft of a model, the Hosmer-Lemeshow test was performed, yielding a chisquare value of 8.95 (P � 0.346), signifying that the fnal selected model was not a poor ft.

Discussion
Anemia is one of the most common complications associated with HD maintenance patients. We were evaluated the anemia association with all levels of iPTH (Above target levels iPTH >300 pg/mL, below target levels iPTH <150 pg/mL & target levels iPTH 150-300 pg/mL) on 179 ESRD patients' data.
n=126 Hyperparathyroidism (iPTH >300 pg/mL) Secondary Analysis: Association of iPTH with Anemia in ESRD patients. International Journal of Clinical Practice frequency of anemia as 38.1% and 24.5% among 45-70 and 20-44 years in HD patients, respectively. However, according to the multivariate model in our study, the age category of 18-44 years has a signifcant (P � 0.031) association with anemia. In our results, young and adult-aged HD patients have more anemia than older HD patients in our study. It is an alarming situation for the Pakistani people that most young and adult-aged people who were to undergo chronic kidney disease have also faced severe anemia during MHD, according to our fndings.
As per our study results, 52% of male patients and 25% of female patients were anemic, indicating a positive but nonsignifcant correlation with anemia. Similar results are observed in the literature, as Lizardi Gómez et al. [6] report that the male gender is an independent risk factor for anemia.
Te present study clearly emphasizes the occurrence of secondary hyperparathyroidism as one of the signifcant factors for persistent anemia among maintenance HD patients. It is evident from the literature that, among ESRD patients, the overproduction of PTH leads to changes in mineral and bone composition, eventually resulting in anemia [17]. Parathyroidectomy in uremic patients has improved the hematocrit levels due to an increase in serum concentration of immunoreactive erythropoietin which is an indirect piece of evidence of the correlation of PTH with anemia in ESRD patients [22]. As per our study, the association of raised serum iPTH levels with serum Hb levels was signifcant; however, the univariate regression analysis showed a non-signifcant association between hyperparathyroidism (serum iPTH level > 300 pg/mL) and anemia.   International Journal of Clinical Practice   Our results were comparable with a study by Adhikary et al. [13] that depicted a nonsignifcant association between serum PTH and Hb levels in HD patients.
In the present study, the overall prevalence of anti-HCV antibodies positive was 42% among HD patients, which is comparable to previously published data by Akhtar et al. in 2021 [23] that found the prevalence of HCV infection in the Pakistani population as 32.33% among HD patients. Another researcher Lodhi et al. [24] also reported that the prevalence of HCV was 43.2% among HD patients in Quetta, Pakistan. As per our results, positive anti-HCV antibodies have a positive correlation with anemia; likewise, HCV is also a signifcant risk factor for anemia in our ESRD study population. Te possible mechanism for a positive association of anemia with HCV in maintenance HD may be due to the negative impact of antiviral therapy, aplastic anemia caused by bone marrow suppression, and liver failure leading to kidney disease, which further causes anemia in HCV-positive patients [25,26]. Our results are in agreement with Elnaggar et al.'s [27] study, in which they demonstrated that slightly higher mean Hb levels were found in HCVpositive (Hb � 8.02 ± 1.08 g/dL) HD patients than in HCVnegative (Hb � 7.64 ± 2.15 g/dL) HD patients, but anemia persisted in their results. Furthermore, our results showed an overall prevalence of Positive-HBV surface antigen among study HD patients, which was found to be comparable with Jeele et al. [28] study.According to our study results, hepatitis B has a nonsignifcant association with anemia, which is also comparable with the data published by Lodhi et al. [24] and Elnaggar et al. [27].
Both comorbid diseases, hypertension and diabetes mellitus, have a signifcant association and positive correlation with anemia among hemodialysis patients, which is comparable with the data published by Junejo et al. [18] and Alshogran et al. [29] study results.
Most of our study patients were on twice-per-week dialysis, depicting a signifcant association with anemia, which is also comparable with the published data by Sabir et al. [20], in which twice-per-week dialysis was observed in 75% of study patients. Te possible cause of anemia with the increased frequency of dialysis per week may be blood loss, which usually gets retained in the dialyzer machine [30]. Furthermore, our study results have represented the duration of hemodialysis and evaluated the association of anemia with every additional month in the duration of hemodialysis in ESRD patients. However, previously reported studies did not identify the association of anemia with every additional month of hemodialysis in ESRD patients in Pakistan.
Anemia is a signifcant contributing factor for the mortality rate among ESRD patients. Terefore, we identifed the correctable or uncorrectable factors of anemia in ESRD patients in Pakistan and also validated our results from multivariate analysis, whereas most researchers identifed only correctable factors and did not validate their results by a multivariate analysis model [4,18].

Study Limitations
Being a real-world data gathering study, this study has some limitations: (i) it is a small sample size study, (ii) there are a smaller number of variables incorporated in the study because of a lack of data availability, (iii) it is not quite a multicentric study (it is suggested that more multicentric studies should be performed on the Pakistani population with large data to have generalizable results), and (iv) fnally, fnancial constraints have also had an impact on availability of expensive lab test (such as ferritin, albumin, interleukin 1, and interleukin 6) data in whole.

Conclusions
In our results, we found a higher prevalence of anemia among ESRD patients with maintenance HD. Based on our fnal multivariate model, we conclude that there was no statistically signifcant association found between anemia and hyperparathyroidism. Moreover, the study results depict that every additional month in the duration of hemodialysis, having age (<45 years), and positive anti-HCV antibody status, these variables were more likely to have anemia in our study MHD patients.

Recommendations
(i) Our results highlighted the clinical relevance of anemia with a positive anti-HCV antibody status and every additional month in the duration of hemodialysis in ESRD patients. So anemia in ESRD patients could be improved by correcting hepatitis-C status and dialyzer machine fuid composition. (ii) Because this study is real world and retrospective in nature, further prospective study and genomic study are required for further investigations of unexplained reasons for less than forty-fve years of age ESRD patients being more anemic. (iii) Further studies are necessary for assessing the correlation of medications (such as erythropoietin stimulating agents, calcium, and vit-D supplementation) with anemia status in ESRD patients.

Data Availability
All relevant data are within the paper and their supporting information fles. Te statistical analysis excel sheet data used to support the fndings of this study are available from the corresponding author upon request.