The impact of shrunken pore syndrome in patient with rheumatic diseases on bone mineral metabolism

Abstract The study aimed to investigate the influence of shrunken pore syndrome (SPS), defined as a cystatin C (CysC)-based estimated glomerular filtration rate (eGFRCysC) <60% of the creatinine (Cr)-based eGFR (eGFRCr), on bone mineral density (BMD) in patients with rheumatic diseases. A total of 831 patients with rheumatic diseases were enrolled in the study. Patients were classified into the SPS group (G-SPS) and non-SPS group (G-nSPS). The correlation between the presence of SPS and BMD of the lumbar spine (BMD_LS), BMD of the femoral neck (BMD_FN), serum parathyroid hormone (PTH) level, chronic kidney dysfunction (CKD), and parameters were evaluated statistically. The prevalence of SPS was 4.0%. Serum PTH level, tartrate-resistant acid phosphatase-5b (TRACP-5b), and eGFRCr in the G-SPS were significantly higher than in the G-nSPS, whereas BMD_LS and BMD_FN in the G-SPS were significantly lower than in the G-nSPS. Serum PTH level was significantly correlated with CysC. BMD_LS had no significant correlation with BMD_FN. The presence of SPS was the only factor that demonstrated significant negative correlation with both BMD_LS and BMD_FN. Relationship between BMD_LS and the presence of SPS was present regardless of CKD stage; however, the negative relationship between BMD_LS and serum PTH was observed only in CKD stage 1 and 2 patients. BMD_FN demonstrated significant negative correlation with serum PTH in the group with progression of CKD. These results suggest that there is a serious potential risk of osteoporosis in patients with SPS and increased PTH, and BMD_LS poses a higher risk in CKD stage 1 and 2.


Introduction
Shrunken pore syndrome (SPS) is a new concept of renal dysfunction first defined by Grubb et al. based on their research on cystatin C (CysC) and glomerular filtration of CysC [1][2][3][4].They suggested the combined use of a cystatin C-and creatinine-based glomerular filtration rate (GFR)estimating equation as the best way of estimating GFR and simultaneously allowing the diagnosis of SPS.
Normal kidney-filtering capacity relies on the function and structural integrity of the glomerular filtration barrier.The glomerular filtration barrier between the vasculature and the urinary space consists of three layers: the fenestrated endothelial cells, the podocytes, and the intervening glomerular basement membrane [5].Normally, a remarkably small fraction of albumin and other large plasma proteins pass across the glomerular capillary wall despite the massive filtration of water and small solutes [6].Glomerular endothelial cell fenestrae are essential for the clearance of lowmolecular-weight products from circulation by filtration [7].However, once a reduction in the pore diameter of the fenestra in the endothelial cells occurs, GFR varies in accordance with the molecular weight of the polypeptides [8].The molecular weight of CysC is 13,000 dalton (Da), whereas that of Cr is 133 Da.Thus, the estimated GFR based on CysC (eGFR cysC ) would change while the eGFR based on Cr (eGFR Cr ) would not change.Cr should penetrate the filtration system; however, some quantity of CysC will not penetrate [1][2][3][4]9].Serum CysC would rise and then eGFR cysC would increase, and therefore, a split between eGFR cysC and eGFR Cr should occur.This is indicated to be pathogenesis of SPS.Grubbs et al. defined SPS as eGFR cysC /eGFR Cr <0.6 [4].
SPS is known to produce increased serum levels of not only cystatin C but also a large number of other 5-30 kDa proteins, many of which have signaling functions [10,11].SPS is suggested to be a risk factor for cardiovascular disorder [12][13][14], preeclampsia [4], higher mortality and morbidity, and of course renal dysfunction [13,[15][16][17][18][19].However, to our knowledge, the influence of SPS on bone and mineral metabolism has not been investigated.Because the molecular weight of parathyroid hormone (PTH) is 9,400 Da [20], some PTH should not penetrate the glomerular filtration system as well as CysC.Thus, when the serum PTH concentration rises, bone metabolism should change.
The aim of this study is to clarify the influence of SPS on bone mineral metabolism and bone mineral density (BMD) in patients with rheumatic diseases.

Materials and methods
A total of 831 patients with rheumatic diseases (515 with rheumatoid arthritis, 110 with psoriatic arthritis, 67 with Sjogren syndrome, 66 with systemic lupus erythematosus, 17 with pustulosis palmaris et plantaris, 16 with ankylosing spondylitis, 13 with systemic sclerosis, 11 with ulcerative colitis, 8 with Behc ¸et's disease, 3 with polymyositis/dermatomyositis, 2 with mixed connective tissue disease, 2 with familial Mediterranean fever, and 1 with polyarteritis nodosa) were recruited from the two institutes with which the authors are affiliated.We measured BMD using dualenergy X-ray absorptiometry (DXA) and calculated the levels of CysC, Cr, creatinine phosphokinase (CPK), serum PTH, serum albumin (Alb), serum calcium (Ca) after calculating Alb, serum phosphorus (IP), and tartrate-resistant acid phosphatase-5b (TRACP-5b).Both eGFR cysC and eGFR Cr were also calculated.Dual-energy X-ray absorptiometry measurements were made with the DPXV R Bravo ME9309 Bone Densitometer (GE Health Care, Chicago, IL, USA: Coefficients of variation; CV: 1.1% (LS), 0.9% (FN)) at one institute and with the Horizon Wi V R (Hologic, Bedford, MA, USA: CV: 1.4% (LS), 1.0% (FN)) at the other.The minimum BMD of the first to fourth lumbar spine was calculated as the BMD_LS, and the average BMD of the bilateral femoral neck was calculated as the BMD_FN.We also calculated the T-score by means of standard deviation from the mean value of the BMD of patients aged 30-39 years and for each sex.
A total of 831 participants (77 men, 754 women) were enrolled in the study.Patients were divided into two groups: the SPS group (G-SPS, n ¼ 33) and the non-SPS group (G-nSPS; n ¼ 798).Using the Mann-Whitney U test, we statistically analyzed and compared the following parameters between the two groups: sex, mean age, disease duration, serum CysC, Cr, serum Cr-to-CysC ratio (Cr/CysC), TRACP-5b, CPK, Ca, Alb, IP, eGFR Cr, eGFR CysC , BMD_LS, BMD_FN, T-scores of LS and FN, and body mass index (BMI).We also evaluated patient drug history between the groups, including glucocorticoid steroid (GCS) administration, current mean dosage at test, and cumulative dose of GCS; osteoporosis drug interventions, such as teriparatide, denosumab, bisphosphonate, and selective estrogen receptor modulator (SERM); and vitamin D administration.Parameters found to be significantly different between the groups using the Mann-Whitney U test were evaluated using binary logistic regression analysis.

Correlation between serum PTH level and parameters
The relationship between serum PTH and parameters such as sex distribution, mean age, disease duration, serum CysC, Cr, Cr/CysC, TRACP-5b, CPK, Ca, Alb, IP, eGFR Cr, eGFR CysC , BMD_LS, BMD_FN, BMI, GCS administration, current GCS mean dosage at test, cumulative dose of GCS, teriparatide administration, denosumab administration, bisphosphonate administration, SERM administration, and vitamin D administration were evaluated for each parameter using univariate model linear regression analysis.Linear regression analysis using a multivariate model was performed for the correlation between PTH and the parameters that demonstrated a significant correlation (within 5%) in the univariate model.

Correlation between BMDs and parameters
We used linear regression analysis to evaluate the correlation between BMD_LS and the following parameters: being female, older age, longer disease duration, higher serum CysC concentration, higher serum Cr concentration, higher Cr/CysC, higher serum CPK concentration, higher PTH, higher serum Alb concentration, higher serum Ca concentration, higher serum IP concentration, higher serum TRACP-5b concentration, eGFR Cr , eGFR CysC , presence of SPS, higher BMI, GCS administration, current GCS dosage, cumulative GCS dose, teriparatide administration, denosumab administration, bisphosphonate administration, SERM administration, and vitamin D administration.The correlation between BMD_LS and each parameter was analyzed first using the univariate model, and then the correlation between BMD_LS and parameters that demonstrated a significant correlation (within 5%) was analyzed using the multivariate model.The correlation between BMD_FN and such parameters was also evaluated in a same manner.

Prevalence of G-SPS and comparison of BMDs between G-SPS and G-nSPS based on CKD stage
The patients' grade of chronic renal disorder was classified based on the following eGFR Cr values (in mL/min/1.73m 2 ): stage 1, >90; stage 2, !60 to <90; stage 3a, !45 to <60; stage 3b, !30 to <45; stage 4, !15 to <30; and stage 5, 15 [21,22].Chi-square test was used to evaluate the prevalence of SPS for each CKD stage as calculated using the eGFR Cr .
The mean values of BMD_LS and BMD_FN in the G-SPS and G-nSPS were compared for each CKD stage, CKD group, and in total using the Mann-Whitney U test.

Correlation between CKD stage and parameters
Using the Mann-Whitney U test, the following parameters were compared between patients with CKD stage 1 and 2 and patients with CKD stages 3a, 4, and 5: percentage of male patients, mean age, mean disease duration, mean value of CysC, Cr, CPK, PTH, Alb, Ca, IP, TRACP-5b, BMI, mean BMD_LS, BMD_FN, GCS administration, current GCS dosage at test, cumulative dose of GCS, teriparatide administration, denosumab administration, bisphosphonate administration, SERM administration, and vitamin D administration.The correlation between these parameters and eGFR Cr was evaluated for each CKD stage group using univariate model linear regression analysis.

Correlation between BMD and parameters categorized by CKD stage
As a post hoc test, the correlation between BMD and the parameters based on CKD stage was calculated using univariate and multivariate linear regression analysis for BMD_LS and BMD_FN.This analysis was performed in the same manner as for all total cases.

Sensitivity and specificity for osteoporosis regarding SPS
Finally, the sensitivity and specificity of the osteoporosis index (T-score À2.5) was calculated using the chi-square test, and the odds ratio was determined for SPS.

Statistical procedures and software used
Statistical significance was set to be lower than 5% for all procedures.All statistical analyses were performed using StatPlus:mac V R (AnalystSoft, Inc., Walnut, CA, USA).

Results
Of the 831 patients, 33 ( As compared with the G-nSPS, patients in the G-SPS were more likely to be male and to have older age, higher serum CysC concentration, lower Cr/CysC, lower serum CPK concentration, higher serum PTH concentration, lower Alb, higher eGFR Cr , lower eGFR CysC , higher serum TRACP-5b concentration, lower BMI, lower BMD_LS and BMD_FN, lower T-scores in both the bony parts, higher teriparatide administration rate, and lower SERM administration rate.However, there was no significant between-group difference in disease duration, Cr, Ca, or IP.The mean PTH value in the G-SPS (5.70 pmol/L) was significantly greater than that in the G-nSPS (4.41 pmol/L).The osteoporosis drug intervention rate, GCS administration rate, current GCS dosage, and cumulative GCS dose demonstrated no significant difference between the two groups, with p-values of 0.88, 0.64, 0.74, and 0.40, respectively (Table 1).

Correlation between serum PTH level and parameters
In the univariate model, serum PTH concentration was significantly correlated with CysC, Ca, IP, eGFR CysC , BMD_LS, SERM administration, and vitamin D administration.Among these parameters, serum CysC concentration was significantly positively correlated with serum PTH concentration, whereas serum IP concentration and BMD_LS had a significantly negative correlation in the multivariate model (Table 3).
Parameters that demonstrated a significant correlation with BMD_FN were being female, older age, higher CysC, higher Cr/CysC, higher Alb, higher TRACP-5b, higher eGFR CysC , the presence of SPS, higher BMI, GCS administration, current GCS dosage, cumulative GCS dose, denosumab administration, and vitamin D administration.BMD-FN demonstrated a significant positive correlation with higher Cr/CysC and higher BMI, whereas it demonstrated negative correlations with being female, older age, having SPS, denosumab administration, older age, higher TRACP-5b, denosumab administration, and vitamin D administration (Table 4).
In patients with CKD stage 1, BMD_LS and BMD_FN were significantly lower in the G-SPS than in the G-nSPS.However, there was no significant difference in BMD between the groups in other CKD stages, except for BMD_FN in patients with CKD stage 3a, in whom BMD was significantly lower in the G-SPS as compared with the G-nSPS.In patients with CKD stage 1 to 3a, both BMD_LS and BMD_FN of those in the G-SPS were significantly lower than in those in the G-nSPS (p¼1.5 Â 10 À2 and p¼2.1 Â 10 À4 ), respectively.However, no significant difference in either BMD_LS or BMD_FN was observed between the G-SPS and G-nSPS among patients with CKD stage 3 b to 5 (Supplemental Table 2).

Correlation between CKD stage groups and parameters
Patients with CKD stage 1 and 2 were relatively young, and there was a higher proportion of men than women as compared with patients with CKD stage 3a to 5. The serum concentration of CysC, Cr, and Ca as well as the BMI were significantly lower in patients with CKD stage 1 and 2 than in patients with CKD stage 3a to 5. BMD_LS was significantly lower in those with CKD stage 1 and 2 whereas BMD_FN was significantly greater in patients with CKD stage 1 and 2 than in the CKD stage 3a to group (Table 4).The cumulative dose of GCS, bisphosphonate administration rate, SERM administration rate, osteoporosis drug administration rate, and vitamin D administration rate in the CKD stage 3a to 5 group were significantly lower than in the CKD stage 1 and 2 group (Table 5).

Correlation between BMD and parameters based on CKD stage
In  Abbreviations: CKD: chronic kidney diseases; eGFR Cr : estimated glomerular filtration rate based on creatinine; CysC: cystatin C; Cr: creatinine; CPK: creatinine phosphokinase; PTH: parathyroid hormone; Alb: albumin; Ca: calcium; IP: phosphorus; TRACP-5b: tartrate-resistant acid phosphatase-5b; BMI: body mass index; SPS: shrunken pore syndrome; BMD_LS: bone mineral density in the lumbar spine; BMD_FN: bone mineral density in the femoral neck; GCS: glucocorticoid steroids (prednisolone); SERM: selective estrogen-receptor modulator.In p-value columns, real number shows no statistically significant difference within 5%, and exponential number with bold style shows statistically significant difference within 5%.In each column demonstrates p-value, whereas real number expresses no statistically significant within 5%, and power number expresses statistically significant within 5%.Bold style shows significant correlation within 5%.Plain style demonstrates positive correlation.Italic style demonstrates negative correlation.
to 5, BMD_FN demonstrated a significant correlation with being female, age, serum PTH, BMI, eGFR CysC , and cumulative GCS dose; a significant positive correlation with BMI; and a significant negative correlation with being female and serum PTH (Supplemental Table 3).

Sensitivity and specificity for osteoporosis regarding SPS
The sensitivity and specificity of a T-score -2.5 in the LS for determining SPS were 48.3% and 74.0%(p ¼ 7.9 Â 10 À3 ), whereas in the FN, they were 73.3% and 62.0% (p ¼ 1.1 Â 10 À4 ), respectively.The odds ratio for each bone was 2.66 and 4.48 for BMD_LS and BMS_FN, respectively.

Discussion
Renal function is primarily characterized by the filtration function of the glomerulus.Glomerular filtration function is estimated using GFR, which is GFR evaluated based on the clearance of inulin or iohexol; however, these procedures are complicated to apply to a population.eGFR using serum concentration of creatinine or cystatin C is usually used for convenience [3,21,22].SPS is a renal disorder that involves the glomerular filtration system and induces a selective impairment of glomerular filtration of 12-to 29-kDa molecules.The representative plasma peptide is CysC.CysC is a protein encoded by the CST3 gene [23], a basic nonglycosylated 13.3-kDa protein comprising a single polypeptide chain of 120 amino acid residues.CysC is produced at a stable rate from a housekeeping gene in all nucleated cells.It is eliminated from the circulation by free filtration through the glomerular membranes and is widely used as a biomarker of renal function [24].CysC is well known to correlate with age, sex, weight, and serum Cr levels, and it is more closely correlated with age than with the other factors [25].Even in elderly people aged 64-100 years, CysC level has been shown to increase with age [26].It may provoke metabolic disorder in various organs [4,10,[12][13][14].SPS is closely associated with cardiovascular disorders and increases the risk of mortality and morbidity [11,17].However, the influence of SPS on mineral metabolism is still unclear.PTH has a molecular weight similar to that of CysC (about 9.5 kDa), and it contains 84 amino acids [27].If the penetration of CysC through the endothelial cells in glomerulus is disturbed, the serum concentration of PTH increases, which affects both renal and bone mineral metabolism.
PTH promotes binding to the PTH receptor on the osteoblast membrane.The induced RANKL binds to the RANKL receptor RANK on the osteoclast precursor membrane and stimulates calcium mobilization from the bone into the blood [28].At the same time, the activated transient receptor potential cation channel subfamily V member 5 (TRPV5) stimulates calcium reabsorption of the allantoic membrane calcium channels in the distal renal tubule [29].On the other hand, PTH suppresses the action of the Na-Pi co-transporter in the proximal tubule to reduce phosphorus reabsorption [30].
Our hypothesis is that serum PTH concentration will increase regardless of the serum Ca concentration in patients with SPS and that an increased PTH level suppresses resorption in the renal tubule or activates osteoclasts and promotes Ca release from the bone.As a result, serum TRACP-5b is increased and BMD is reduced.However, when renal dysfunction occurs, Ca resorption is inhibited, and the serum Ca concentration decreases, serum PTH further increases, and BMD decrease progresses even more.
In this study, the serum PTH concentration was correlated positively with the concentration of serum CysC as expected.We found a negative correlation between the serum concentration of PTH and IP, which might be explained by the suppressive action of PTH on the Na-Pi co-transporter.
The serum concentration of PTH in the G-SPS was greater than that in the G-nSPS in this study, and it was significantly correlated with the levels of serum CysC and eGFR CysC , despite the fact that there was no correlation between the serum concentration of PTH and Cr.Normally, PTH secretion is tightly controlled according to serum Ca concentration [27]; however, in the patients with SPS, PTH is retained in the serum and its level remains high, regardless of the serum concentration of Ca.Thus, bone resorption is more likely to occur in patients with SPS.
PTH is controlled by serum calcium and vitamin D levels.Unfortunately, serum 25-OH-vitamin D was not measured in the present study; however, 25(OH)-vitamin D administration rate was compared between the G-SPS and G-nSPS groups.As shown in Table 1, administration rate demonstrated no significant difference between the two groups.Further, as shown in Table 3, serum PTH level was not correlated significantly with vitamin D administration, which was found using multivariate linear regression model.Thus, the influence of 25-OH-vitamin D on serum PTH may be diminished.
In this study, the serum TRACP-5b level, a representative biomarker of bone resorption that is independent of renal dysfunction [31], was significantly higher in the G-SPS than in the G-nSPS.However, the serum concentration of Ca and IP did not differ significantly between the two groups.Serum Ca concentration is strictly controlled not only by PTH but also by other bioactive molecules such as vitamin D, FGF23, and Klotho [32].
The accuracy of BMD is a concern while comparing BMD using a dual DXA-testing machine.A previous report discussed cross calibration for the lumbar spine DXA and compared Hologic QDR with GE Lunar DPX; the results demonstrated that the difference was <1% over the range of BMD covered [33].Therefore, we estimated that the accuracy of the dual DXA-testing machine in the present study is comparable with that of the previous study for estimating BMDs.
BMD should decrease in patients with SPS as a result of bone resorption caused by the elevated level of serum PTH.As expected, both BMD_LS and BMD_FN were significantly lower in the G-SPS than in the G-nSPS.It should be noted that various confounding factors such as sex, age, use of anti-osteoporosis drugs, and GCS administration affect BMD.Among these factors, there was no significant difference in the use of anti-osteoporosis drugs or GCS administration between the two groups.Sex and age are major factors that affect the serum concentration of CysC [25], so that the calculation of eGFR CysC includes sex and age as correction terms.The ratio of males to females was significantly higher in the G-SPS than in the G-nSPS (Table 1); nonetheless, male sex was a positive predictive factor for increasing BMD at both the LS and FN (Table 3).After adjusting for confounding factors, the presence of SPS still remained a negative predictive factor for decreasing BMD (Table 3).
In comparing the LS and FN, the influence of aging is more evident in BMD_FN than in BMD_LS.On the other hand, the effect of PTH on bone resorption was more evident in BMD_LS.These differences may be explained by the difference in Cr/CysC.Cr/CysC is a newly advocated biomarker that may express muscular volume [34][35][36][37][38].Both CysC and Cr are popular biomarkers of kidney function.However, Cr differs from CysC, as it is produced from only cross-striated muscle tissue.Therefore, serum Cr is affected by muscle volume, whereas CysC is produced in every nucleated cell.Cr/CysC may increase if the muscular volume relative to the whole body increases.Cr/CysC has received attention as a biomarker of muscular volume, the so-called sarcopenia index.Cr/CysC has a more powerful influence on BMD at the FN than at the LS [39], and sarcopenia is usually more evident in elderly patients.Therefore, age and the sarcopenia index were predictive factors of BMD_FN (Table 3).Cr/CysC decreases as CysC increases; thus, SPS reduces Cr/CysC.In this study, the concentration of serum Cr in the G-SPS tended to be low compared with the G-nSPS, and eGFR Cr was significantly higher in the G-SPS than in the G-nSPS.We previously demonstrated that Cr was significantly correlated with CPK (p¼.039),CysC (p < 1.0 Â 10 À12 ), and eGFR CysC (p < 1.0 Â 10 À12 ).Serum Cr levels were relatively low in the patients with SPS, along with increasing concentration of CysC.Thus, the sarcopenia index (i.e.Cr/CysC) decreased in patients with SPS (Table 1).One concern that arises is that SPS is a result of severe sarcopenia, and therefore, osteoporotic pressure is strong if Cr/CysC reflects body muscle volume.This concern may be valid.The presence of SPS and sarcopenia may partly overlap.However, the presence of sarcopenia cannot explain the relationship between the presence of SPS and elevated serum PTH level.Thus, we believe that the pathology of SPS should exist independent of sarcopenia.
Having SPS was significantly negatively correlated with BMD_LS independent of CKD stage.At the same time, serum PTH level was significantly negatively correlated with BMD_LS only in patients with CKD stage 1 and 2. In contrast, there was no significant correlation between SPS and BMD_FN independent of the CKD stage, and the serum level of PTH was a predictive factor of BMD_FN only in the group of patients with CKD stage 3a to 5 (Table 5).
These results suggest that the effect of the presence of SPS on BMD may be more evident at the LS than at the FN.
The effect of SPS on reducing BMD was more evident in patients with CKD stage 1 and 2 than in those with CKD stage 3a to 5. One of the reasons for this may be that the small number of patients with stage 3a to 5 affected the ability to detect a difference.Another reason could be that the serum concentrations of both Cr and CysC increase in moderate to severe stages of CKD and that the serum PTH level may elevate and bone resorption would increase independent from SPS.Previous reports have already described that SPS is not related to renal dysfunction but rather to the split of the filtration ratio between freely fenestrated small molecules and blocked plasma polypeptides [7][8][9].These results suggest that the decrease of BMD in SPS patients is caused by SPS itself, and those with hyperactive PTH status may be especially prone to BMD loss.
The prevalence of SPS was higher in patients with CKD stage 1 and 2 as compared with patients in stage 3a to 5. Accordingly, BMD_LS among those with CKD stage 1 and 2 was significantly lower than in the stage 3a to 5 group.On the contrary, BMD_FN in patients with CKD stage 1 and 2 was significantly higher than in patients with CKD stage 3a to 5. The reason for this difference is believed to be due to the drug intervention rate for osteoporosis, including vitamins.One more reason may be the higher cumulative dose of GCS administration, despite the fact that there was no significant difference in the GCS administration rate.Patients with CKD stage 3a to 5 may have a long history of GCS administration because of rheumatic disease treatment; therefore, the cumulative dose should increase.In addition, the drug intervention for steroid-induced osteoporosis including vitamin D would increase.Drug response in BMD_LS was more sensitive than that in BMD_FN.Therefore, BMD_LS in patients with CKD stage 3a to 5 was significantly higher than that in patients with CKD stage 1 and 2.
Serum Ca and IP demonstrated a significant negative correlation with eGFR Cr in the CKD stage 3a to 5 group.These results are paradoxical because eGFR Cr represents renal dysfunction as it decreases.When eGFR Cr is reduced, renal function is deteriorated, and serum Ca should decrease.This paradoxical phenomenon may be explained by a higher rate of vitamin D administration in the CKD stage 3a to 5 group.Vitamin D administration leads to an increase in serum Ca concentration, which in turn may result in hyposecretion of PTH.However, when the serum Ca concentration is maintained within the normal range, the change in the secretion of PTH is only minimal.
The reported prevalence of SPS in patients with rheumatic diseases is similar among studies [12,13,16,17].Rheumatic diseases may not be related to the occurrence of SPS, even though rheumatic disease is a chronic inflammatory disease.However, the glomerulus may not be influenced.There was no significant difference in the mean disease duration between the G-SPS and G-nSPS in this study, which suggests that disease duration did not affect the prevalence of SPS.
As reported in the study, SPS is associated with a high decrease in BMD, probably caused by elevated serum PTH levels.However, there are some confounding underlying factors, such as sex, age, BMI, and drug administration.Furthermore, as a posthoc analysis, comparison between the G-SPS and G-nSPS groups was performed after propensity score matching for eliminating such parameters' influence using Mann-Whitney U test.The results demonstrated that the serum PTH levels in the G-SPS group were higher than that in the G-nSPS group, and BMDs and T-scores were significantly lower in the G-SPS group than in the G-nSPS group for both LS and FN (Supplemental Table 4).These results suggest that there is a serious potential risk of osteoporosis in patients with SPS, and increased PTH and BMD_LS lead to a higher risk of CKD stage 1 and 2.
BMI demonstrated no significant difference in between the G-SPS and G-nSPS groups, although higher BMI contributed gaining BMD significantly in both of the LS and the FN.These results suggested that BMI and presenting SPS have no correlation with each other.
There are some limitations to this study.First, we did not measure other biomarker GFR estimation, such as beta-2 microglobulin.Secondly, the influence of serum 25-OHvitamin D level, comorbidities and ethnic considerations were not considered.Third, a basic problem arose with this study's design.This study was cross-sectional, so no longitudinal results were presented.However, a rough outline of the influence of SPS on bone metabolism was provided.These results suggest that SPS is associated with a potential risk of osteoporosis.The findings of this study might be helpful for treating rheumatic and osteoporotic diseases.

Conclusions
The influence of SPS on bone mineral metabolism was investigated in this cross-sectional practical study.We found that SPS promotes osteolytic processes and that the mechanism might be due to the increased serum PTH concentration in SPS.

Ethical approval
IY: This study was approved by our ethics committee (approval number: Y-RA-2020-1) in accordance with the ethical standards laid down in 1964 Declaration of Helsinki and its later amendments.In addition, anonymity was ensured for all patients and their families who participated in this study, and no names, and/or addresses were issued that could help identify these individuals.Moreover, all patients were informed of the purpose and possible consequences of this study i.e. that this was a retrospective study in clinical practice without any experimentation on subjects.
SN: This study was approved by our ethics committee (approval number: 841) in accordance with the ethical standards laid down in 1964 Declaration of Helsinki and its later amendments.In addition, anonymity was ensured for all patients and their families who participated in this study, and no names, and/or addresses were issued that could help identify these individuals.Moreover, all patients were informed of the purpose and possible consequences of this study i.e. that this was a retrospective study in clinical practice without any experimentation on subjects.

Table 1 .
Demographic characteristics for each group.

Table 2 .
Results of binary logistic regression analysis regarding the presence of SPS for parameters.

Table 4 .
Results of linear regression analysis regarding bone mineral density in each.
Cr, CPK, Alb.TRACP-5b, BMI, eGFR cysC , having SPS, GCS administration, current GCS dosage, drug intervention for osteoporosis, and vitamin D supplementation in the univariate model.In the multivariate model, BMD_FN had a significant positive correlation with Cr, BMI, and GSC administration and a significant negative correlation with being female, age, drug intervention for osteoporosis, and vitamin D supplementation.In patients with CKD stage 3a variate model.In patients with CKD stage 3a to 5, BMD_LS demonstrated a significant correlation with serum CysC, Cr, CPK, Alb, BMI, eGFR Cr , eGFR CysC , and having SPS.It was also significantly positively correlated with CPK and BMI, and had a significant negative correlation with the presence of SPS.In patients with CKD 1 and 2, BMD_FN demonstrated a significant correlation with being female, age, serum CysC,

Table 5 .
Mean values of parameters for CKD Stage groups and correlation of parameters with eGFR Cr .

Table 6 .
Correlation with eGFR Cr with multivariate model linear regression analysis.