Nephrol Dial Transplant (2008) 23: 315 320 doi: 10.1093/ndt/gfm639 Advance Access publication 23 October 2007 Original Article Glycaemic control and serum intact parathyroid hormone levels in diabetic patients on haemodialysis therapy Reiichi Murakami 1, Shuichi Murakami 1, Rumi Tsushima 1, Chiyuki Ueda 1,KyokoMikami 1, Takanori Ebina 1, Ryuichiro Kumasaka 2, Norio Nakamura 2 and Ken Okumura 2 1 Department of Clinical Laboratory, Murakami Shinmachi Hospital, Aomori and 2 Department of Nephrology, Hirosaki University Hospital, Hirosaki, Japan Abstract Background. Osteodystrophy is one of the long-term haemodialysis complications, and in diabetic patients, it mainly occurs as an aplastic or low-turnover type due to their low serum intact parathyroid hormone (ipth) levels. In the present study, we investigated the role of glycaemic control to the serum ipth levels in diabetic haemodialysis patients. Methods. A total of 162 patients who had started haemodialysis at our hospital in the last 10 years were enrolled. Among them, 80 patients suffered from diabetic nephropathy as a primary cause of end-stage renal failure, 69 chronic glomerulonephritis, 9 polycystic kidney and 4 from other causes. We examined the serum ipth and HbA 1c levels of all patients at the start of haemodialysis. In 80 diabetic patients, we examined those levels both at the start of haemodialysis and 1 year later and investigated how glycaemic control affected the ipth levels. Results. The serum ipth levels at the start of haemodialysis were significantly lower in patients with diabetes than without diabetes (P = 0.032). The levels were lower in patients with poor glycaemic control than with good control (P = 0.045). In the analysis of diabetic patients 1 year later, the serum ipth levels were significantly reduced in those with poor glycaemic control (P = 0.002). The multiple regression test showed that the serum HbA 1c levels were strongly related to the serum ipth levels (P < 0.001). Conclusions. The status of glycaemic control in diabetic haemodialysis patients affects the serum ipth levels. Good glycaemic control should be required to prevent osteodystrophy. Keywords: diabetes mellitus (DM); intact PTH; renal osteodystrophy Introduction For the last several years, the dialysis population in our country has been increasing by more than 10 000 patients per year and reached a total of 219 183 at the end of 2001 [1]. Among dialysis patients, the predominant primary disease is diabetic nephropathy which accounts for up to 40% of all patients every year [1]. In haemodialysis patients, it has been reported that diabetes mellitus confers a protective effect from the skeletal manifestations of secondary hyperparathyroidism because diabetic patients had significantly lower intact parathyroid hormone (ipth) levels than nondiabetic patients [2 4]. However, the low serum ipth levels suppress the rate of bone formation and result in renal osteodystrophy, which is characterized by an aplastic or low-turnover bone disorder in diabetic haemodialysis patients [5 7]. When considering the increasing number of patients and extended duration of dialysis therapy, it is important to maintain their bone turnover by adjusting the serum ipth level. It has been reported that the high serum concentration of glucose reversibly suppresses PTH secretion from cultured bovine parathyroid cells in vitro [8]. It is also reported that the serum glucose concentration was inversely correlated with the serum PTH levels in diabetic patients with chronic renal failure [9]. However, few reports have been available about the effect of long-term glycaemic control on parathyroid function in diabetic haemodialysis patients who do not receive any active forms of vitamin D. In the present study, we investigated how glycaemic control affected serum ipth level during a 1-year period of haemodialysis therapy. Patients and methods Correspondence and offprint requests to: Reiichi Murakami, MD, PhD, Department of Clinical Laboratory, Murakami Shinmachi Hospital, 13-1-2 Shinmachi, Aomori, 030-0801 Japan. Tel: +81-17-723-1111; Fax: +81-17-723-1118; E-mail: 01-murakami@hkg.odn.ne.jp Patients This retrospective cohort study included 162 consecutive patients (97 males and 65 females with the mean age of 61 ± 13 years) who started haemodialysis therapy between February 1995 and January 2005, and had continued the C The Author [2007]. Published by Oxford University Press on behalf of ERA-EDTA. All rights reserved. For Permissions, please e-mail: journals.permissions@oxfordjournals.org
316 R. Murakami et al. Fig. 1. Categorization of diabetic patients based on their glycaemic control during the follow-up period. Eight patients were excluded during the study because of the administration of an active form of vitamin D. therapy for 1 year with outpatient status at Murakami Shinmachi Hospital, Aomori, Japan. Eighty patients suffered from diabetic nephropathy induced by type 2 diabetes as a primary disease causing end-stage renal failure, and the remaining 82 patients had diseases other than diabetes. The primary diseases in the patients without diabetes included chronic glomerulonephritis in 69 patients, polycystic kidney in 9 and the other diseases in 4. At the time of starting haemodialysis, none of the patients had received any active forms of vitamin D, and in addition, none of them had had a history of percutaneous ethanol injection therapy (PEIT) or parathyroidectomy as treatment for primary or secondary hyperparathyroidism. Blood sample collection Blood samples were taken from all of the 162 patients at the time of starting haemodialysis therapy to measure serum ipth, calcium, phosphorus, alkaline phosphatase, creatinine, urea nitrogen and albumin levels. The samples were drawn from either the arteriovenous fistula or catheter inserted into the femoral vein. We used a cut point of 6.5% for HbA 1c according to the guideline of the Japan Diabetes Society. Of the 80 patients with diabetic nephropathy, 60 had a serum haemoglobin A 1c (HbA 1c )level< 6.5% (good glycaemic control), while the other 20 had an HbA 1c 6.5% (poor glycaemic control) at the start of the haemodialysis therapy. All of the 80 diabetic patients were followed at our hospital while undergoing haemodialysis therapy. They were maintained with a dialysis fluid of 3.0 meq/l of calcium concentration, and there was no difference in dialysis frequency and efficacy between diabetic and non-diabetic patients. Blood samples were taken again from each patient 1 year after the start of haemodialysis. At that time, 74 patients underwent haemodialysis three times a week, 5 two times a week and the remaining patient once a week. In the patients dialyzed three times a week, blood sampling was performed at the mid session (on Wednesday or on Thursday) exactly 44 h after the previous session. In the patients dialyzed two times a week, it was performed exactly 68 h after the previous session. In the patient dialyzed once a week, it was done at the beginning of the session. Eight diabetic patients were excluded from the analysis because they were administered some active forms of vitamin D, which influences serum ipth levels, during the study period. None of the patients experienced PEIT or parathyroidectomy for the treatment of hyperparathyroidism during the period. Blood samples were taken from the other 72 diabetic patients at 1 year after starting haemodialysis. The patients were divided into four groups according to the status of glycaemic control at the start of the haemodialysis therapy and at 1 year later: Group A consisted of 35 patients with good glycaemic control (HbA 1c < 6.5%) both at the start of the therapy and 1 year later; Group B: 19 patients with good glycaemic control at the start of the therapy but poor control (HbA 1c 6.5%) 1 year later; Group C: 11 patients with poor glycaemic control at the start of the therapy and good control 1 year later; Group D: 7 patients with poor glycaemic control both at the start and 1 year later (Figure 1). In each group, the change in the serum ipth level during a 1-year period of haemodialysis therapy was examined. ipth assay and measurements of biochemical parameters The serum ipth was measured by immunoradiometric assay (IRMA; Allegro Intact PTH; Nichols Institute, San Juan, Capistrano, CA) [10]. The normal range for healthy subjects determined in our laboratory was 10.0 65.0 pg/ml. Biochemical parameters described above were measured by means of an autoanalyzer (Hitachi, Ltd, Japan). Statistical analysis As the serum ipth is not normally distributed, it was expressed as median and range, and analysed using Mann Whitney s U-test (Tables 1 and 2, Figure 2). Other values were expressed as mean ± one standard deviation unless otherwise indicated, and differences in the mean values between two groups were analysed using Student s t-test (Tables 1 and 2). Chi-square test and Fisher s exact probability test were used for comparison of categorical parameters (Tables 1 and 2). Differences of ipth in the
Glycaemic control and ipth in diabetic patients with HD 317 Table 1. Comparison of the parameters measured at the start of the haemodialysis therapy between the patient groups with and without diabetic nephropathy Diabetic nephropathy Other diseases P No. of patients 80 82 Age (y) 62 ± 10 58 ± 15 0.082 ipth (pg/ml) 152(4.8 477) 187.5(3.7 860) 0.032 Ca (mg/dl) 7.4 ± 1.0 7.7 ± 1.5 0.136 Pi (mg/dl) 5.1 ± 1.8 5.5 ± 1.7 0.148 Alkaline phosphatase (IU/L) 247.0 ± 102.4 207.5 ± 95.3 0.012 Creatinine (mg/dl) 7.6 ± 2.3 8.1 ± 2.7 0.207 Urea nitrogen (mg/dl) 69.3 ± 22.7 76.8 ± 27.4 0.060 Total protein (g/dl) 6.2 ± 0.8 6.4 ± 0.7 0.096 Albumin (g/dl) 3.2 ± 0.7 3.4 ± 0.7 0.067 Administration of Ca-based phosphate binders 6/80 5/82 0.648 Statistical significance (P < 0.05). Table 2. Comparison of the parameters measured at the start of the haemodialysis therapy between the groups with good and poor glycaemic control in diabetic patients Good control Poor control (HbA 1c < 6.5%) (HbA 1c 6.5%) P No. of patients 60 20 Age (y) 63 ± 9 62± 10 0.599 HbA 1c (%) 5.6 ± 0.6 7.3 ± 0.7 <0.001 ipth (pg/ml) 155.5(4.8 477) 101(25.4 269) 0.045 Ca (mg/dl) 7.4 ± 1.0 7.3 ± 1.2 0.708 Pi (mg/dl) 5.2 ± 1.8 5.0 ± 2.0 0.677 Alkaline phosphatase (IU/L) 237.6 ± 104.3 275.4 ± 93.2 0.154 Creatinine (mg/dl) 7.5 ± 2.5 7.6 ± 1.7 0.868 Urea nitrogen (mg/dl) 66.4 ± 22.6 77.5 ± 22.0 0.058 Total protein (g/dl) 6.3 ± 0.7 6.1 ± 1.0 0.320 Albumin (g/dl) 3.2 ± 0.6 3.1 ± 0.7 0.523 Haemoglobin (g/dl) 8.3 ± 1.2 8.3 ± 0.8 0.770 Haematocrit (%) 25.6 ± 3.6 26.0 ± 2.0 0.766 Administration of Ca-based phosphate binders 4/60 2/20 0.637 Therapy of diabetes:insulin( )/(+) 38/22 7/13 0.027 Statistical significance (P < 0.05). Fig. 2. Comparison of the serum ipth levels at the start of haemodialysis among patients without diabetes and with diabetes with good and poor glycaemic control.
318 R. Murakami et al. Fig. 3. Changes in the serum ipth levels in four groups during 1-year period of haemodialysis therapy. Table 3. Multiple regression analysis for the contribution of the biochemical parameters to the serum ipth level after 1-year period of haemodialysis in diabetic patients (n = 72) β t P Age (y) 0.054 0.500 0.619 HbA 1c (%) 0.515 4.762 <0.001 Ca (mg/dl) 0.238 0.726 0.471 Pi (mg/dl) 0.058 0.048 0.962 Creatinine (mg/dl) 0.183 1.318 0.192 Urea nitrogen (mg/dl) 0.124 1.000 0.321 Total protein (g/dl) 0.026 0.196 0.845 Statistical significance (P < 0.05). mean values between before and after 1-year period of haemodialysis in each group were analysed using Wilcoxon signed-ranks test (Figure 3). The contribution of each biochemical parameter to the serum ipth level at 1 year after the therapy was assessed using a multiple regression test (Table 3). Results Profiles at the start of the haemodialysis therapy Table 1 shows the baseline parameters for the patients with diabetic nephropathy and those with the other diseases. At the start of haemodialysis, the mean serum ipth level in patients with diabetic nephropathy was 152 (4.8 477) pg/ml, which was significantly lower than that in patients without diabetes (188(3.7 860) pg/ml, P = 0.032). The serum alkaline phosphatase levels in patients with diabetic nephropathy were significantly higher than those in patients without diabetes. There were no significant differences in age and serum levels of calcium, phosphorus, creatinine, urea nitrogen and albumin between the two groups. The rate of the administration of calcium-based phosphate binders was not different either. Table 2 shows the comparison of the parameters measured at the start of the haemodialysis therapy between the groups with good and poor glycaemic control in diabetic patients. The mean serum ipth level in patients with poor glycaemic control was 101(25 269) pg/ml and was significantly lower than that in patients with good glycaemic control (156(4.8 477) pg/ml, P = 0.045). It was also significantly lower than that in the patients without diabetic nephropathy (Figure 2). The serum alkaline phosphatase level in patients with poor glycaemic control was higher than that in patients with good control but the difference was not significant (P = 0.154). The usage of insulin was significantly higher in patients with poor glycaemic control than in those with good glycaemic control. There were no significant differences in the other parameters between the two groups. Changes in the serum ipth and other biochemical parameters during 1-year study period in diabetic patients As shown in Figure 3, in Groups A and D in which no change in the state of glycaemic control was noted during 1-year period, no marked change in the serum ipth level was noted 1 year later. In Group B, with glycaemic control having worsened during the 1-year period, the serum ipth level was decreased from 110 (48 304) pg/ml at the start of the haemodialysis therapy to 76 (14 174) pg/ml at 1 year after the therapy (P = 0.002). In Group C, with glycaemic control having improved during the 1-year period, the serum ipth level was slightly increased but the change was not significant (P = 0.266). The other parameters including haemoglobin, calcium, phosphorus and
Glycaemic control and ipth in diabetic patients with HD 319 alkaline phosphatase levels were all unchanged during the 1-year period in all groups. Moreover, the efficacy of dialysis (Kt/V) in the four groups 1 year later was 0.91±0.16 in group A, 0.89±0.26 in group B, 0.89±0.14 in group C and 0.86 ± 0.26 in group D, respectively. There was not any statistical difference in dialysis efficacy among four groups. To determine which parameter was significantly and independently related to the serum ipth level after a 1-year period of haemodialysis, we performed multiple regression analysis (Table 3). The serum HbA 1c and alkaline phosphatase levels were related to the serum ipth level independently (P < 0.001 and P = 0.012, respectively). Discussion While confirming the low levels of the serum ipth in diabetic haemodialysis patients compared with those in nondiabetic patients, this study showed that poor glycaemic control further reduced the serum ipth level without administration of any active forms of vitamin D, whereas improved glycaemic control resulted in an increase in an serum ipth level. The multiple regression analysis showed that the serum HbA 1c level was closely related to the serum ipth level after haemodialysis therapy. The serum calcium and phosphorus levels, which were known to influence the serum ipth level, did not show any significant correlation with the ipth level. This may be because that the calcium and phosphorus levels had already been adjusted and maintained by using phosphate binders during the haemodialysis therapy. In this study, a significant elevation of the serum alkaline phosphatase level was noted at the start of haemodialysis, also suggesting low bone turnover in diabetic patients, especially in those with poor glycaemic control. This is probably because the serum alkaline phosphatase levels are elevated in both types of osteodystrophy with low bone turnover and high bone turnover. More sensitive markers for bone formation, such as intact osteocalcin or bone-specific alkaline phosphatase [11,12], should have been measured to assess more accurately the effect of glycaemic control on bone turnover. In diabetic patients without renal insufficiency, it was reported that their bone turnover is suppressed because the bone formation was decreased with low intact osteocalcin levels according to their glycaemic control [13 15]. In diabetic patients without renal insufficiency, poor glycaemic control is associated with the increase in the urine calcium secretion, and therefore PTH secretion is stimulated by the decreased serum calcium level [16,17]. This is the reason why the serum ipth levels do not show significant differences in diabetic patients without renal insufficiency although bone formation is decreased. On the other hand, when the diabetic nephropathy becomes so severe that the patients need haemodialysis therapy, their serum levels of ipth are significantly decreased compared with those without diabetes in inverse proportion to the increased serum calcium level followed by the reduction in urine calcium secretion [2,3]. Thus, renal osteodystrophy showing the aplastic or low-turnover bone disorders, which is most characteristic in diabetic haemodialysis patients, is caused by both impaired secretion of ipth and decreased bone formation [5 7]. It has been reported that advanced glycation endproducts (AGEs) play an important role in the pathogenesis of both impaired secretion of ipth and decreased bone formation [18 20]. The serum levels of AGEs are elevated in patients with diabetic nephropathy, correlating directly with serum creatinine levels, and were reported to be higher by about 5-fold in diabetic haemodialysis patients compared to normal subjects [18]. It was demonstrated that the increase in PTH secretion in response to the low serum concentration of calcium is inhibited by AGEs [20]. In addition, it was also shown that AGEs inhibit osteocalcin synthesis in response to 1,25(OH) 2 vitamin D 3 in cultured human osteoblasts, which were proven to have receptors to AGEs [19,20]. Therefore, diabetic haemodialysis patients with high serum levels of AGEs are indicated to suffer from low-turnover bone disorders caused by both impaired ipth secretion and decreased bone formation. It is difficult to interpret the HbA 1c levels in diabetic haemodialysis patients as a measure of glycaemia because of their reduced red blood cell survival, erythropoietin and iron therapy. That is the limitation of this study, however, in the present study, we demonstrated that keeping good glycaemic control has a beneficial effect on the serum ipth levels even in diabetic haemodialysis patients on high levels of AGEs. In face of the increasing population of diabetic patients on haemodialysis and their expanding duration of haemodialysis therapy, glycaemic control is exceedingly important to prevent aplastic or low-turnover osteodystrophy, one of the most severe complications of long-term haemodialysis therapy. Conflict of interest statement. None declared. (See related article by Luigi Gnudi. Serum intact parathyroid hormone in diabetic patients on haemodialysis: what is the treatment goal? Nephrol Dial Transplant 2008; 23: 24 26.) References 1. Nakai S, Shinzato T, Nagura Y et al. An overview of regular dialysis treatment in Japan (as of 31 December 2001). Ther Apher Dial 2004; 8: 3 32 2. Vincenti F, Hattner R, Amend WJ Jr et al. Decreased secondary hyperparathyroidism in diabetic patients receiving hemodialysis. JAMA 1981; 245: 930 933 3. Inaba M, Okuno S, Nagasue K et al. Impaired secretion of parathyroid hormone is coherent to diabetic hemodialyzed patients. Am J Kidney Dis 2001; 38: S139 S142 4. Inaba M, Nagasue K, Okuno S et al. Impaired secretion of parathyroid hormone, but not refractoriness of osteoblast, is a major mechanism of low bone turnover in hemodialyzed patients with diabetes mellitus. Am J Kidney Dis 2002; 39: 1261 1269 5. Nishitani H, Miki T, Morii H et al. Decreased bone mineral density in diabetic patients on hemodialysis. Contrib Nephrol 1991;90: 223 227 6. Pei Y, Hercz G, Greenwood C et al. Renal osteodystrophy in diabetic patients. Kidney Int 1993; 44: 159 164 7. Holgado R, Haire H, Ross D et al. Effect of a low calcium dialysate on parathyroid hormone secretion in diabetic patients on maintenance hemodialysis. J Bone Miner Res 2000; 15: 927 935 8. Sugimoto T, Ritter C, Morrissey J et al. Effects of high concentrations of glucose on PTH secretion in parathyroid cells. Kidney Int 1990;37: 1522 1527
320 R. Murakami et al. 9. Martinez I, Saracho R, Moina I et al. Is there a lesser hyperparathyroidism in diabetic patients with chronic renal failure? Nephrol Dial Transplant 1998; 13: S9 S11 10. Goodman WG, Juppner H, Salusky IB et al. Parathyroid hormone (PTH), PTH-derived peptides, and new PTH assays in renal osteodystrophy. Kidney Int 2003; 63: 1 11 11. Urena P, Hruby M, Ferreira A et al. Plasma total versus bone alkaline phosphatase as marker of bone turnover in hemodialysis patients. J Am Soc Nephrol 1996; 7: 506 512 12. Morishita T, Nomura M, Hanaoka M et al. A new assay method that detects only intact osteocalcin. Two-step non-invasive diagnosis to predict adynamic bone disease in haemodialized patients. Nephrol Dial Transplant 2000; 15: 659 667 13. Krakauer JC, McKenna MJ, Buderer NF et al. Bone loss and bone turnover in diabetes. Diabetes 1995; 44: 775 782 14. Sayinalp S, Gedik O, Koray Z. Increasing serum osteocalcin after glycemic control in diabetic men. Calcif Tissue Int 1995;57:422 425 15. Inaba M, Nishizawa Y, Mita K et al. Poor glycemic control impairs the response of biochemical parameters of bone formation and resorption to exogenous 1,25-dihydroxyvitamin D3 in patients with type 2 diabetes. Osteoporos Int 1999; 9: 525 531 16. Thalassinos NC, Hadjiyanni P, Tzanela M et al. Calcium metabolism in diabetes mellitus: effect of improved blood glucose control. Diabet Med 1992; 10: 341 344 17. Nagasaka S, Murakami T, Uchikawa T et al. Effect of glycemic control on calcium and phosphorus handling and parathyroid hormone level in patients with non-insulin-dependent diabetes mellitus. Endocr J 1995; 42: 377 383 18. Makita Z, Radoff S, Rayfield EJ et al. Advanced glycosylation end products in patients with diabetic nephropathy. N Engl J Med 1991; 325: 836 842 19. Takagi M, Kasayama S, Yamamoto T et al. Advanced glycation endproducts stimulates interleukin-6 production by human bone-derived cells. J Bone Miner Res 1997; 12: 439 446 20. Yamamoto T, Ozono K, Miyauchi A et al. Role of advanced glycation endproducts in adynamic bone disease in patients with diabetic nephropathy. Am J Kidney Dis 2001; 38: S161 S164 Received for publication: 24.6.07 Accepted in revised form: 21.8.07