help button home button Endocrine Society JCEM
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Submit a related Letter to the Editor
Right arrow Purchase Article
Right arrow View Shopping Cart
Right arrow Alert me when this article is cited
Right arrow Alert me when eLetters are posted
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow Request Copyright Permission
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Indridason, O. S.
Right arrow Articles by Quarles, L. D.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Indridason, O. S.
Right arrow Articles by Quarles, L. D.
The Journal of Clinical Endocrinology & Metabolism Vol. 83, No. 11 3860-3866
Copyright © 1998 by The Endocrine Society

Original Studies

Predictors of Short-Term Changes in Serum Intact Parathyroid Hormone Levels in Hemodialysis Patients: Role of Phosphorus, Calcium, and Gender1

Olafur S. Indridason, Carl F. Pieper and L. Darryl Quarles

Department of Medicine, Division of Nephrology, and the Division of Biometry, Department of Community and Family Health, Center for Aging and Human Development (C.F.P.), Duke University Medical Center, Durham, North Carolina 27713

Address all correspondence and requests for reprints to: Olafur S. Indridason, M.D., M.H.S., Department of Medicine, Box 3014, Duke University Medical Center, Durham, North Carolina 27713. E-mail: osi{at}acpub.duke.edu


   Abstract
Top
 Abstract
Introduction
 Materials and Methods
 Results
 Discussion
 References
 
Several factors have been identified as important in the pathogenesis of secondary hyperparathyroidism in end-stage renal disease, including serum calcium, phosphorus, and calcitriol. To examine the independent effects of key factors, we prospectively studied 52 new hemodialysis patients with mild secondary hyperparathyroidism (PTH, 110–670 pg/mL) treated with a standardized regimen of calcium supplements, phosphorus binders, and no vitamin D derivatives. We used simple and multivariable linear regression analysis to examine the relationship between changes in PTH ({Delta}PTH) levels observed over a 4-week period and various biochemical and demographic variables. By simple linear regression we found that changes in serum phosphorus (r2 = 0.31; ß = 41.6; P = 0.0001), initial phosphorus concentration (r2 = 0.15; ß = 33.4; P = 0.005), initial PTH level (r2 = 0.29; ß = 0.58; P = 0.0001), changes in serum calcium (r2 = 0.12; ß = -74.0; P = 0.01), and gender (r2 = 0.07; ß = 76.1; P = 0.05) were significantly associated with {Delta}PTH. However, upon multivariable regression analysis, only the changes in phosphorus (partial r2 = 0.31; ß = 37.0; P = 0.0001), initial PTH level (partial r2 = 0.23; ß = 0.50; P = 0.0001), and gender (partial r2 = 0.05; ß = 63.1; P = 0.02) remained significantly associated with {Delta}PTH. Neither the serum concentration of 1,25-dihydroxyvitamin D3, bicarbonate, aluminum, or albumin nor changes in the serum bicarbonate concentration, the presence of diabetes, KT/V, or age were significantly associated with the {Delta}PTH. Our findings are consistent with independent effects of phosphorus and gender on parathyroid gland function in patients with dialysis-dependent renal failure through mechanisms that remain to be defined.


   Introduction
Top
 Abstract
 Introduction
Materials and Methods
 Results
 Discussion
 References
 
IN ADDITION to the well characterized effects of serum calcium and 1,25-dihydroxyvitamin D3 [1,25-(OH)2D3] on parathyroid gland function, phosphorus has been implicated as an important factor in the pathogenesis of secondary hyperparathyroidism in patients with renal failure (1). It is well established that phosphorus restriction retards the progression of secondary hyperparathyroidism in these patients (2, 3, 4). The mechanisms by which phosphorus controls parathyroid gland function are not entirely clear. Earlier studies suggest that hyperphosphatemia stimulates PTH production indirectly through lowering the serum ionized calcium concentration and/or decreased 1,25-(OH)2D3 synthesis (5, 6, 7, 8, 9). It is also possible that phosphorus acts through novel hormone systems or by modulation of the PTH effect on bone (9, 10). In addition to these indirect effects, recent reports support a possible direct effect of extracellular phosphorus on parathyroid gland function. Increments in serum phosphorus accelerate, whereas phosphorus restriction attenuates, secondary hyperparathyroidism in animal models independent of changes in serum calcium and 1,25-(OH)2D3 (11, 12, 13, 14, 15). Moreover, studies of cultured parathyroid glands show that high extracellular phosphorus directly increases PTH secretion by a posttranscriptional effect independent of the medium calcium concentration (16, 17, 18).

Independent effects of phosphorus on parathyroid gland function are difficult to demonstrate in the clinical setting because of simultaneous changes in serum calcium and other confounding factors. Cross-sectional studies relating PTH levels with other biochemical parameters have yielded variable results, indicating the importance of not only serum calcium, phosphorus and 1,25-(OH)2D3 concentrations, but also other factors, including acidosis, degree of renal failure, and age, in predicting PTH levels (19, 20, 21, 22, 23, 24, 25, 26). These cross-sectional studies are limited by a static look at a dynamic process and cannot predict the response of PTH to changes in biochemical parameters. However, the results of longitudinal studies exploring changes in PTH levels in relationship to changes in phosphorus concentration are also conflicting (4, 9, 27, 28, 29, 30). In addition, these existing studies are limited by small sample size, which precludes detailed analysis of independent predictors of the PTH response.

Considering the multitude of factors that may influence parathyroid gland function, we examined the relationship between short term changes in serum intact PTH levels and several important variables, including simultaneous changes in serum phosphorus and calcium concentration, in a large cohort of end-stage renal disease (ESRD) patients beginning dialysis therapy. These patients were not being treated with vitamin D derivatives and had received standardized calcium dialysis, calcium supplementation, and phosphate binders to maintain or lower serum phosphorus to acceptable levels. Multivariable regression analysis was performed to control for possible confounders, including initial serum intact PTH, calcium, and phosphorus concentrations; changes in serum calcium concentration; 1,25-(OH)2D3, bicarbonate, albumin, and aluminum concentrations; age; and gender.


   Materials and Methods
Top
 Abstract
 Introduction
 Materials and Methods
Results
 Discussion
 References
 
Design

The current study is a prospective cohort study involving new hemodialysis patients with mild secondary hyperparathyroidism. The data are derived from a 4- to 5-week run-in period of a randomized clinical trial designed to compare three different therapies in this population: calcium carbonate alone, daily oral calcitriol, and thrice weekly iv calcitriol. The purpose of this run-in period was to standardize diet, calcium intake, and dialysate calcium as well as to control serum phosphorus below 6.6 mg/dL by dietary phosphorus restriction and oral phosphate binders. In the beginning of this period, the dialysate calcium concentration was standardized at 2.5 mEq/L (1.25 mmol/L). All subjects were instructed how to change their diet to contain 500–600 mg elemental calcium and 800–900 mg phosphorus/day. In addition, the existing calcium supplementation was discontinued, and the subjects were started on calcium carbonate therapy (Oscal-500, SmithKline Beecham). The dose of calcium carbonate was determined by the initial serum calcium and phosphorus concentrations in the following manner. Subjects with normal serum calcium (8.7–10.5 mg/dL) and phosphorus (2.3–4.5 mg/dL) levels were given 1500 mg elemental calcium/day in three divided doses (one pill of Oscal-500 three times a day with meals). Those with calcium concentrations below 8.7 mg/dL and/or phosphorus concentrations above 4.5 mg/dL were given 2500 mg elemental calcium/day in divided doses (one pill with breakfast, two with lunch, and two with dinner). Aluminum-containing phosphorus binder was added in doses up to 3 g aluminum hydroxide/day, given with meals, if calcium carbonate in this dose failed to lower serum phosphorus below 6.6 mg/dL. If phosphorus control was not adequate after 4 weeks, an additional week was allowed after enforcement of medication and diet. Dietary recall was used to monitor compliance with the diet 2–3 weeks into the period.

Variability of intervention

The interventions resulted in variable changes in calcium supplementation and dialysate calcium concentration among the subjects. The mean prestudy dose of elemental calcium per day was 1070 ± 96.1 mg, with a range of 0–3000; eight subjects were not taking any calcium supplements. Twelve patients started on 1500 mg elemental calcium/day; the remainder had either hypocalcemia and/or hyperphosphatemia requiring a calcium dose of 2500 mg/day. The average change in elemental calcium supplementation was an increment of 1199 ± 93.4 mg/day, ranging from a 2500-mg increase to a 500-mg decrease. In only two subjects did our intervention result in a decrease in calcium supplementation. The change in dialysate calcium upon entering the study ranged from no change to a decrease of 0.50 mmol/L. In this regard, before entry, 13, 21, and 18 subjects were being dialyzed against a 1.75, 1.50, and 1.25 mmol/L dialysate, respectively, with a mean dialysate calcium concentration of 1.48 ± 0.03 mmol/L. Upon enrollment, none of the subjects was taking aluminum-containing phosphorus binders, and none had been treated previously with vitamin D derivatives. During the study, eight subjects were given aluminum-containing phosphorus binders in doses ranging from 1.5–3.0 g/day to maintain the serum phosphorus concentration below 6.6 mg/dL. No subject started therapy with vitamin D derivatives during the study.

Subjects

Patients with mild secondary hyperparathyroidism who were initiating hemodialysis therapy were recruited from five dialysis units affiliated with Duke University Medical Center and Durham Regional Hospital between November 1994 and July 1996. To identify eligible patients, we examined the initial intact PTH level measured at the respective dialysis unit for every patient starting dialysis therapy during this period. Those with serum intact PTH concentrations between 150–600 pg/mL were eligible for our main study comparing the three therapies for mild secondary hyperparathyroidism and thus were employable for the current study. Excluded were subjects who had been on dialysis for over a year, subjects younger than 18 yr of age, pregnant patients, and those with human immunodeficiency viral disease or serious coexisting illnesses, subjects with aluminum overload (serum aluminum concentration above 80 µg/L), subjects with an iron overload (serum ferritin levels above 1000 ng/mL), those with history of parathyroidectomy or treatment with vitamin D analogues within the past 2 months, and subjects being treated with steroids, phenytoin, or phenobarbital. In addition, we excluded subjects who were unlikely to be able to finish the full 40 weeks of the main study based on serious coexisting diseases, plans for transplantation, or transfer to other dialysis modalities or other dialysis units. One hundred and fifteen of the 295 patients initiating dialysis at the participating dialysis units during this period had mild secondary hyperparathyroidism by our definitions, and 91 met all inclusion and exclusion criteria for the main study. Seventy-four patients consented to participate; however, during the run-in period, 22 were excluded from randomization to calcitriol or calcium therapy based on a priori defined criteria [development of hypercalcemia (serum calcium concentration greater than 10.5 mg/dL), uncontrollable hyperphosphatemia (serum phosphorus greater than 6.5 mg/dL despite addition of 3 g aluminum hydroxide), low serum intact PTH (<150 pg/mL) on our baseline measurement, protocol violation, or death]. None of these 22 subjects had sufficient laboratory evaluation required for assessment of changes in serum intact PTH levels over the run-in period, leaving 52 patients who had laboratory measurements adequate for analysis in this part of the study.

Biochemistry

We performed serial measurements of serum biochemistry over the 4- to 5-week period. Upon entry into the study, we measured serum concentrations of total calcium and phosphorus (Hitachi 911 autoanalyzer, Hitachi Scientific Instruments, Hialeah, FL), intact PTH (immunoradiometric assay, Nichols Institute Diagnostics, San Juan Capistrano, CA), and aluminum (graphite furnace, Perkin Elmer, Norwalk, CT). We measured total calcium and phosphorus weekly, and at the end of the study we repeated the measurements of intact PTH, total calcium, and phosphorus. At that time we also measured 1,25-(OH)2D levels (RRA; Incstar Corp., Stillwater, MN). The serum albumin and bicarbonate (Hitachi 747–200 autoanalyzer) were measured at each dialysis unit as part of their monthly biochemical profile, and we used the measurements closest to the beginning and the end of the 4-week period. As an indicator of the dialysis dose, we used the monthly KT/V calculated at the dialysis units in this period (KT/V calculations use pre- and postdialysis serum urea, time dialyzed, fluid removed, and patient weight to estimate dialysis dose). In our analysis and throughout this article, serum calcium concentration refers to albumin-corrected calcium according to the formula: calcium = measured calcium + [0.8 x (4.0 - albumin)].

All blood was drawn before dialysis at the first session of each week. The blood was spun within 2 h of the collection, and the serum was frozen at -20 C until biochemical measurements were performed within 1–4 weeks of the sample collection.

Statistical analysis

We used the paired t test to examine the significance of changes in biochemical measurements between baseline and the end of the study and regression analysis to examine the association between changes in PTH levels and other variables. Based on the known role of serum calcium and the potential effects of phosphorus and PTH levels, our a priori hypothesis was that changes in serum calcium, changes in serum phosphorus, the serum 1,25-(OH)2D3 concentration, and the initial intact PTH level were the most important variables predicting changes in PTH levels. The association of these variables with the changes in intact PTH levels ({Delta}PTH) was therefore specifically tested with both simple and multivariable linear regression analyses including all interaction terms. The interaction terms were tested as a block in a model including all of the main effects and were only considered further if the block was significant in the model. To further examine the relationship between {Delta}PTH and physiologically relevant and demographic variables, we used simple linear (bivariate) regression analysis and subsequently a multivariable linear regression analysis with stepwise addition of the variables that were significant in our a priori analysis and upon bivariate regression. In addition to the changes in serum calcium ({Delta}calcium), the changes in serum phosphorus ({Delta}phosphorus), the serum 1,25-(OH)2D3 concentration, and the initial intact PTH level, the variables tested were initial serum calcium concentration, initial serum phosphorus concentration, initial serum albumin concentration, initial serum bicarbonate concentration, change in serum bicarbonate over the 4 weeks of the study ({Delta}bicarbonate), serum aluminum, time since initiating dialysis treatment, KT/V, primary disease (diabetes = 0; other = 1), and gender (female = 0; male = 1) and age. To take into account differences in baseline calcium levels and its possible effects on subsequent changes in calcium and PTH responses, initial serum calcium was included in all regression models where changes in calcium was one of the independent variables.

With a sample size of 52 patients, multivariable regression analysis should produce reliable results with 5–6 independent variables in a given model. Because of the limitations a relatively small sample size imposes upon our analysis, we used the bootstrapping technique to further explore the association of the independent variables with {Delta}PTH. Bootstrapping is a methodology used in regression analysis to adjust for the effects of outliers and to decrease the effect of variability in a given sample (31). This technique has been validated in regression models where no selection of variables occurs; however, it has been proposed as a method of model validation for variable selection as well (32). This is achieved through the creation of multiple subsamples by random selection of subjects from the original sample. The selected subject is immediately returned to the original sample and thus can be selected more than once into a given subsample (i.e. sampling with replacement). The size of each subsample is the same as the original size (52 observations in our case); therefore, some subjects will be selected into a given subsample more than once, and others not at all. Repeated samples are then used to examine the relationship between the variables of interest. We generated 100 bootstrap samples to which a stepwise multivariable regression analysis was performed using all 15 independent variables (see Table 2Go), with {Delta}PTH as the dependent variable. The frequency of entry for each variable in the final model in these 100 samples is an indicator of the importance of this variable in relation to the dependent variable (see Table 4Go). Data are presented as the mean ± 1 SEM unless otherwise specified.


View this table:
[in this window]
[in a new window]
  		Table 2. Bivariate relationship between changes in serum intact PTH ({Delta}PTH) and other variables

 

View this table:
[in this window]
[in a new window]
  		Table 4. The variables most frequently significantly associated with changes in intact PTH levels using the bootstrapping technique and stepwise multivariable regression analysis

 

   Results
Top
 Abstract
 Introduction
 Materials and Methods
 Results
Discussion
 References
 
Baseline characteristics

The baseline characteristics of the 52 subjects are shown in Table 1Go. Upon enrollment, the mean age was 55 ± 1.9 yr. Sixty percent of the subjects were male, and 92% were African-American, reflecting the demographics of our dialysis population. Forty-six percent of the subjects had diabetes as the cause of ESRD. The mean duration of chronic maintenance hemodialysis was 12 ± 1.4 weeks.


View this table:
[in this window]
[in a new window]
  		Table 1. Baseline characteristics of patients

 
At baseline, the mean PTH was 302 ± 17.7 pg/mL, with a range of 110–670 pg/mL, indicating mild secondary hyperparathyroidism. Mean serum calcium and phosphorus concentrations were 8.7 ± 0.09 mg/dL (normal range, 8.7–10.5 mg/dL) and 4.9 ± 0.22 mg/dL (normal range, 2.3–4.5 mg/dL), respectively. The mean serum concentration of 1,25-(OH)2D3 was 12.4 ± 1.26 pg/mL, with 83% of the subjects having values below the normal range (16–42 pg/mL). The mean serum aluminum level was 11.2 ± 1.12 µg/L. Mean serum bicarbonate and albumin levels were 20.6 ± 0.49 mEq/L and 3.8 ± 0.06 mg/dL, respectively. Finally, KT/V ranged from 1.00–1.97, with mean of 1.34 ± 0.03.

Changes in biochemistry over the study period

The study was designed to obtain control of serum phosphorus and normalize serum calcium by administering calcium carbonate and aluminum hydroxide. Our interventions resulted in control of the serum phosphorus concentration and mild decreases in mean serum intact PTH levels without significant changes in the mean serum calcium concentration. Mean intact PTH levels decreased from 303 ± 17.7 to 271 ± 17.6 pg/mL (P = 0.10, by paired t test). The mean serum phosphorus level decreased from 4.9 ± 0.22 to 4.5 ± 0.18 mg/dL (P = 0.18), whereas the mean calcium concentration changed minimally from 8.7 ± 0.09 to 8.6 ± 0.09 mg/dL (P = 0.84). There was, however, marked variation in the biochemical response in individual subjects. Changes in intact PTH levels ranged from a 440 pg/mL decrease to a 352 pg/mL increase, changes in calcium ranged from a 1.6 mg/dL increase to a 1.8 mg/dL decrease, and changes in phosphorus ranged from a 4.2 mg/dL increase to a 4.4 mg/dL decrease. This variability permitted examination of variables associated with the observed changes in intact PTH levels.

Bivariate relationship between {Delta}PTH and other variables

By simple linear regression analysis (Table 2Go), we observed a strong relationship between {Delta}phosphorus and {Delta}PTH (r2 = 0.31; P = 0.0001). The regression coefficient (ß = 41.6) indicates that every 1.0 mg/dL decrease in the serum phosphorus concentration was associated with a 41.6 pg/mL drop in the PTH level (Fig. 1Go). In addition, the initial phosphorus concentration was significantly related to {Delta}PTH (r2 = 0.15; ß = 33.4; P = 0.005); the higher the initial phosphorus level, the greater the predicted drop in PTH. The initial intact PTH level was also strongly associated with {Delta}PTH (r2 = 0.29; ß = 0.58; P = 0.0001); the higher the initial PTH concentration, the larger the fall in PTH levels. The change in the serum calcium concentration was significantly related to {Delta}PTH (r2 = 0.12; ß = -74.0; P = 0.01). The regression coefficient indicates that with a 1 U (1.0 mg/dL) increase in the serum calcium concentration, the intact PTH level is predicted to decrease by 74.0 pg/mL. Gender was also significantly related to {Delta}PTH (r2 = 0.07; ß = 76.1; P = 0.05). Males responded more favorably than females; the mean intact PTH decreased by 61 ± 22.5 pg/mL in males, whereas it increased by 14 ± 30.7 pg/mL in females.



View larger version (17K):
[in this window]
[in a new window]
  		Figure 1. Bivariate relationship between {Delta}PTH and {Delta} serum phosphorus in 52 patients with endstage renal disease over the 4- to 5-week observation period. A significant relationship was observed between changes in serum phosphorus concentrations (mg/dL) and changes in PTH (pg/mL) (r2 = 0.31; P = 0.0001).

 
Multivariable regression analysis

To examine our a priori hypothesis, we used the initial PTH level, the change in the calcium concentration, the initial calcium level, the change in the phosphorus concentration, and the 1,25-(OH)2D3 concentration along with the interaction terms in a multivariable regression analysis. The interaction terms were not significant when tested as a block and thus were not included in the model. The model, including the five main variables, indicated that the change in phosphorus concentration and the initial PTH level continued to be significantly related to {Delta}PTH (ß = 28.8; P = 0.0006 and ß = 0.55; P = 0.0001, respectively). Additionally, both initial serum calcium and changes in calcium were significant independent predictors of the changes in PTH (ß = 61.7; P = 0.02 and ß = -55.0; P = 0.03, respectively) in this model. However, the 1,25-(OH)2D3 concentration remained insignificant in the model (P = 0.32).

To further identify factors independently predicting the PTH response, we performed a multivariable regression analysis with stepwise selection of significant variables, using variables found to be significantly associated with {Delta}PTH by our a priori analysis as well as those significant on bivariate regression analysis (Table 2Go). The variables that entered the model and remained statistically significant are the initial PTH level, the changes in phosphorus concentration, and gender. When controlled for these variables, the changes in serum calcium concentration and initial calcium concentration show only borderline significant association with the changes in PTH levels when added to the model. Moreover, the initial phosphorus concentration was not significant when we controlled for the change in phosphorus concentration. The final model therefore includes the {Delta}phosphorus, the initial PTH level, and gender as significant independent predictors of changes in PTH levels (Table 3Go).


View this table:
[in this window]
[in a new window]
  		Table 3. Results of a stepwise multivariable regression analysis

 
Finally, Table 4Go shows the results of the bootstrapping analysis (see Materials and Methods) and lists the frequency of the 5 variables most often significantly associated with {Delta}PTH on stepwise multivariable regression analysis of 100 bootstrapped samples of our cohort. As before, the most important variables are the changes in serum phosphorus and initial intact PTH levels, appearing as significantly related to {Delta}PTH in 96% and 99% of the samples, respectively. Gender also continues to be significant in 76% of the samples. The other variables are seen less frequently (initial serum calcium concentration in 73% and {Delta}calcium in 55%). All other variables appeared as significant in less than 50% of the samples.

In our cohort, the 1,25-(OH)2D3 concentration was subnormal as expected in patients with ESRD (Table 1Go). Despite theses low levels, we found no association between the PTH response and the 1,25-(OH)2D3 status. In addition, neither serum aluminum, serum bicarbonate, nor changes in the serum bicarbonate concentration were significantly associated with changes in PTH levels in our study. Moreover, other variables tested, including age, KT/V, serum albumin concentration, primary disease, and time since initiation of dialysis therapy, were not significantly associated with the changes in PTH levels on simple linear or multivariable regression. Due to the high proportion of African Americans in our sample, we were not able to examine possible racial differences in PTH responses.


   Discussion
Top
 Abstract
 Introduction
 Materials and Methods
 Results
 Discussion
References
 
The current study examines factors associated with changes in serum intact PTH levels in a large cohort of ESRD patients beginning hemodialysis therapy. The study design provided an opportunity to examine the association between short term changes in serum intact PTH levels and key factors involved in the pathogenesis of secondary hyperparathyroidism in a background of 1,25-(OH)2D3 deficiency and mild secondary hyperparathyroidism. We found that the change in serum phosphorus, the initial PTH level, the change in serum calcium, and the initial calcium concentration were independently associated with the change in PTH, whereas the 1,25-(OH)2D3 concentration was not. Additional analysis also identified gender as a possible important independent predictor of changes in PTH.

Prior investigations have implicated changes in serum phosphorus as an important predictor of simultaneous changes in PTH levels, but these studies lacked sufficient sample size to perform multivariable analysis (4, 9, 27, 28, 30). Therefore, the relative importance of our study is that due to the large sample, we could assess the independent effect of changes in serum phosphorus, controlling for several potential confounding factors, including the concomitant change in serum calcium and initial PTH levels as well as concurrent 1,25-(OH)2D3 and aluminum concentrations, acid-base status, age, gender, and primary disease. Although it is also important to note that our demonstration of a significant independent association between the changes in serum phosphorus and the changes in PTH levels may not imply a cause-effect relationship, compelling in vitro and animal data (16, 17, 18) as well as the results of cross-sectional (24, 25, 26) and small longitudinal studies are consistent with an independent physiological effect of phosphorus on parathyroid gland function. In addition, phosphorus restriction has beneficial effects on the development of secondary hyperparathyroidism in predialysis patients (3, 4), and hyperphosphatemia is a predictor of poor response to calcitriol therapy in dialysis patients (33, 34, 35).

Our study does not establish whether this independent phosphorus effect is mediated by direct or indirect mechanisms. A direct molecular target for phosphorus in the parathyroid gland has not yet been identified; however, recent in vitro studies have demonstrated that high extracellular phosphorus may increase PTH secretion independent of the calcium concentration, possibly through a posttranscriptional effect (16, 17, 18). Animal studies also show that phosphorus restriction can prevent the development of secondary hyperparathyroidism and/or control serum PTH levels even in the absence of concurrent elevations in the serum calcium concentration (11, 12, 13, 14, 15).

Indirect effects are also possible, including through changes in the 1,25-(OH)2D3 concentration (6). We did not measure changes in the serum 1,25-(OH)2D3 concentration over the 4 weeks and therefore cannot exclude the possibility that such changes might in part be responsible for the observed changes in PTH levels. However, in the ESRD population, several studies have demonstrated that in contrast to what is seen in patients with mild to moderate renal failure, the serum 1,25-(OH)2D3 concentration does not increase in response to lowering the serum phosphorus concentration (4, 8, 36). None of our patients received calcitriol during the study; thus, it is unlikely that major changes in the serum 1,25-(OH)2D3 concentration occurred to confound our results. Indirect effects of phosphorus through other hormone systems or by alterations in bone responsiveness to PTH (10) also cannot be excluded in our study; however, the role of phosphorus in PTH resistance of bone has not been well characterized in humans (37).

We also found that the initial PTH level was significantly related to subsequent changes in PTH, such that the higher the initial PTH level, the more decrease in PTH was predicted. One might argue that the opposite would have been expected, i.e. higher PTH levels signifying larger and more refractory glands, resulting in fewer decreases in PTH levels. This, however, is not substantiated by our observations, and in this population of patients with only mild elevations of PTH, gland size and refractoriness may not be significant problems; rather, a "floor effect" may be responsible for our observations. This floor effect is well known in regression analysis (in our study, when the higer the PTH level, the more it can decrease, whereas low PTH levels have a shorter distance to fall). An alternative explanation is that the PTH level may be a marker of parathyroid gland sensitivity to various stimuli. In that regard, patients with lower PTH may have less responsive parathyroid glands. Indeed, not all patients with renal failure have the same propensity to develop secondary hyperparathyroidism, and there is heterogeneity in the PTH response to manipulations of the serum calcium concentration (38, 39, 40).

As expected, we found a significant association between the changes in PTH levels and the simultaneous changes in calcium. This finding is in concordance with the known role of extracellular calcium in the acute control of PTH secretion (41). However, we had expected the changes in calcium to be the dominating predictor, rather than changes in serum phosphorus or initial PTH level. Under our study conditions, where the control of serum phosphorus was the primary treatment goal, and changes in calcium were limited, a calcium effect might be dampened. A larger sample would increase the power to detect a significant association between {Delta}calcium and {Delta}PTH in these circumstances; indeed, the results of the bootstrapping method identify calcium as an important predictor of the PTH response. Assessment of ionized calcium might also have uncovered a stronger relationship between calcium and PTH in our analysis. By using the albumin-corrected calcium concentration and incorporating the initial serum calcium level into the regression, we were still unable to show a stronger predictive role for calcium changes. Even though other studies indicate a relationship between persistent elevations in serum calcium and PTH suppression (4, 33, 42), simultaneous calcitriol administration and/or changes in serum phosphorus may be partially responsible for the observed changes in PTH in these studies. Indeed, the results of a large cross-sectional study of hemodialysis patients did not demonstrate a significant association between the serum calcium concentration and PTH levels once controlled for serum phosphorus and age (26). Similarly, in a recent study in predialysis patients with different levels of renal insufficiency, PTH was significantly related to serum calcium only in advanced renal failure, not in mild to moderated disease, whereas the opposite was true for phosphorus (24). Other cross-sectional studies have yielded variable results, and the relationship may be confounded by PTH-driven increases in serum calcium in patients with severe disease (19, 20, 21, 22, 23, 25). Our study has the advantage of examining changes in biochemical parameters over time in a relatively uniform population of patients. Although our results are not inconsistent with a role of calcium in long term control of PTH levels, they indicate that other factors are important as well, including serum phosphorus.

Another interesting aspect of our study is the association between gender and PTH response. We found that PTH suppression was less in women than in men. Although the explanation for this effect is not clear, it is important to note that in our sample there were statistically significant differences between the genders with respect to the primary disease and dialysis dose. In this regard, females had more diabetes, and the dialysis dose, KT/V, was higher in women. In addition, women showed slightly fewer increases in calcium concentration over the 4 weeks than men (data not shown). However, when controlling for these factors in a multivariable regression, the effect of gender remained significant. Thus, gender seems to be an independent predictor of changes in PTH over time. In fact, some studies support a sexual difference in bone metabolism in dialysis patients (43), and in the general population, there are gender differences in bone metabolism as well as in propensity to develop parathyroid abnormalities (32). A possible mechanism for this gender effect includes effects of estrogens and progestins on the secretion of PTH (44) or modulation of the PTH response to various stimuli by these hormones (45). In any event, this interesting relationship needs further investigation and confirmation in larger studies.

Other variables tested, including serum aluminum concentration at study entry, were not associated with changes in serum intact PTH. We did not measure serum aluminum levels at the end of the study and therefore could not examine the relationship between changes in aluminum concentration and changes in PTH. Even if such a relationship exists, we do not believe that there were major changes in serum aluminum over this 4-week period to confound our results. None of the patients had changes in aluminum therapy upon study entry, and when we excluded from analysis the eight patients initiating Al(OH)3 treatment during the study and who might have had an increase in serum aluminum over the 4-week period, the findings are identical to the overall results. Thus, our results suggest that the aluminum concentration is not related to short term changes in PTH levels, at least not in patients without evidence of aluminum overload.

In conclusion, our study demonstrates that changes in the serum phosphorus concentration are strongly associated with simultaneous changes in intact PTH levels in patients with mild secondary hyperparathyroidism who are not being treated with vitamin D derivatives. It thus provides additional evidence for an independent role of phosphorus in the pathogenesis of secondary hyperparathyroidism. Changes in serum calcium levels were also associated with changes in PTH levels, but this effect was relatively small compared to the effect of phosphorus, possibly as a consequence of our study design. In addition, males responded more favorably than females, indicating a possible gender difference in parathyroid gland responsiveness. Although we do not know the basic mechanisms underlying the apparent effects of phosphorus and gender on intact PTH levels, our results support the fundamental importance of phosphorus control in the management of secondary hyperparathyroidism in patients receiving hemodialysis.


   Footnotes
 
1 This work was supported in part by a grant from Hoffman-La Roche, Inc. Back

Received March 19, 1998.

Revised June 29, 1998.

Accepted July 29, 1998.


   References
Top
 Abstract
 Introduction
 Materials and Methods
 Results
 Discussion
 References
 

  1. Delmez JA, Slatopolsky E. 1992 Hyperphosphatemia: its consequences and treatment in patients with chronic renal disease. Am J Kidney Dis. 19:303–317.[Medline]
  2. Lindenau K, Kokot F, Frohling PT. 1986 Suppression of parathyroid hormone by therapy with a mixture of keto analogues/amino acids in hemodialysis patients. Nephron. 43:84–86.[Medline]
  3. Combe C, Aparicio M. 1994 Phosphorus and protein restriction and parathyroid function in chronic renal failure. Kidney Int. 46:1381–1386.[Medline]
  4. Combe C, Morel D, de Precigout V, et al. 1995 Long-term control of hyperparathyroidism in advanced renal failure by low-phosphorus low-protein diet supplemented with calcium (without changes in plasma calcitriol). Nephron. 70:287–295.[Medline]
  5. Reiss E, Canterbury JM, Bercovitz MA, Kaplan EL. 1970 The role of phosphate in the secretion of parathyroid hormone in man. J Clin Invest. 49:2146–2149.
  6. Portale AA, Booth BE, Halloran BP, Morris Jr RC. 1984 Effect of dietary phosphorus on circulating concentrations of. 1,25-dihydroxyvitamin D and immunoreactive parathyroid hormone in children with moderate renal insufficiency. J Clin Invest. 73:1580–1589.
  7. Turner C, Compston J, Mak RHK, et al. 1988 Bone turnover and 1,25-dihydroxycholecalciferal during treatment with phosphate binders. Kidney Int. 33:989–995.[Medline]
  8. Tessitore N, Venturi A, Adami S, et al. 1987 Relationship between serum vitamin D metabolites and dietary intake of phosphate in patients with early renal failure. Miner Electrolyte Metab. 13:38–44.[Medline]
  9. Kaye M. 1995 Hypocalcemia after an acute phosphate load is secondary to reduced calcium efflux from bone: studies in patients with minimal renal function and varying parathyroid activity. J Am Soc Nephrol. 6:273–280.[Abstract]
  10. Rodriguez M, Martin-Malo A, Martinez ME, Torres A, Felsenfeld AJ, Llach F. 1991 Calcemic response to parathyroid hormone in renal failure: role of phosphorus and its effect on calcitriol. Kidney Int. 40:1055–1062.[Medline]
  11. Slatopolsky E, Caglar S, Pennell JP, et al. 1971 On the pathogenesis of hyperparathyroidism in chronic experimental renal insufficiency in the dog. J Clin Invest. 50:492–499.
  12. Denda M, Finch J, Slatopolsky E. 1996 Phosphorus accelerates the development of parathyroid hyperplasia and secondary hyperparathyroidism in rats with renal failure. Am J Kidney Dis. 28:596–602.[Medline]
  13. Hernandez A, Concepcion MT, Rodriguez M, Salido E, Torres A. 1996 High phosphorus diet increases preproPTH mRNA independent of calcium and calcitriol in normal rats. Kidney Int. 50:1872–1878.[Medline]
  14. Lopez-Hilker S, Dusso AS, Rapp NS, Martin KJ, Slatopolsky E. 1990 Phosphorus restriction reverses hyperparathyroidism in uremia independent of changes in calcium and calcitriol. Am J Physiol. 28:F432–F437.
  15. Kilav R, Silver J, Naveh-Many T. 1995 Parathyroid hormone gene expression in hypophosphatemic rats. J Clin Invest. 96:327–333.
  16. Nielsen PK, Feldt-Rasmussen U, Olgaard K. 1996 A direct effect in vitro of phosphate on PTH release from bovine parathyroid tissue slices but not from dispersed parathyroid cells. Nephrol Dial Transplant. 11:1762–1768.[Abstract/Free Full Text]
  17. Slatopolsky E, Finch J, Denda M, et al. 1997 Phosphorus restriction prevents parathyroid gland growth. J Clin Invest. 97:2534–2540.[Medline]
  18. Almaden Y, Canalejo A, Hernandez A, et al. 1996 Direct effect of phosphorus on PTH secretion from whole rat parathyroid glands in vitro. J Bone Miner Res. 11:970–976.[Medline]
  19. Fournier AE, Arnaud CD, Johnson WJ, Taylor WF, Goldsmith RS. 1971 Etiology of hyperparathyroidism and bone disease during chronic hemodialysis. II. Factors affecting serum immunoreactive parathyroid hormone. J Clin Invest. 50:599–605.
  20. Fuss M, De Backer M, Brauman J, et al. 1976 Parathyroid hormone plasma levels in untreated chronic renal failure and in hemodialyzed patients. Nephron. 17:144–154.[Medline]
  21. Wills MR, Fairney A, Varghese Z, Tatler GLV, Baillod RA, Moorhead JF. 1974 Serum parathyroid hormone concentration in chronic renal failure patients on maintenance haemodialysis. Clin Chim Acta. 57:83–89.[CrossRef][Medline]
  22. St John A, Thomas MB, Davies CP, et al. 1992 Determinants of intact parathyroid hormone and free 1,25-dihydroxyvitamin D levels in mild and moderate renal failure. Nephron. 61:422–427.[Medline]
  23. Pitts TO, Piraino BH, Mitro R, et al. 1988 Hyperparathyroidism and 1,25-dihydroxyvitamin D deficiency in mild, moderate, and several renal failure. J Clin Endocrinol Metab. 67:876–881.[Abstract]
  24. Kates DM, Sherrard DJ, Andress DL. 1995 Serum phosphorus (PO4) strongly correlates with PTH in mild chronic renal failure. J Am Soc Nephrol. 6:965.
  25. Rudnicki M, McNair P, Transbol I, Nielsen B. 1991 Lack of relationship between parathyroid hormone and 1,25-dihydroxyvitamin D in chronic renal failure. Nephron. 52:144–149.
  26. Salem MM. 1997 Hyperparathyroidism in the hemodialysis population: a survey of 612 patients. Am J Kidney Dis. 29:862–865.[Medline]
  27. Martinez I, Saracho R, Montenegro J, Llach F. 1994 The role of acute and chronic phosphate loading in the secondary hyperparathyroidism of early renal failure. J Am Soc Nephrol. 5:869.
  28. Seidel A, Stein G, Schneider S, Schmidt-Gayk H, Ritz E. 1991 Do limited changes in phosphate intake modulate 1,25-(OH)2D3 levels in early renal failure? Clin Nephrol. 36:274–280.[Medline]
  29. Mizumoto D, Watanabe Y, Fukuzawa Y, Yuzawa Y, Yamazaki C. 1994 Identification of risk factors on secondary hyperparathyroidism undergoing long-term haemodialysis with vitamin D3. Nephrol Dial Transplant. 9:1751–1758.[Abstract/Free Full Text]
  30. Fine A, Cox D, Fontaine B. 1993 Elevation of serum phosphate affects parathyroid hormone levels in only 50% of hemodialysis patients, which is unrelated to changes in serum calcium1,2. J Am Soc Nephrol. 3:1947–1953.[Abstract]
  31. Efron B, Gong G. 1983 A leisurely look at the bootstrap, the jackknife, and cross-validation. Am Statistician. 37:36–48.
  32. Harrell FE, Lee KL, Mark DB. 1996 Multivarable prognostic models: issues in developing models, evaluating assumptions and adequacy, and measuring and reducing errors. Stat Med. 15:361–387.[CrossRef][Medline]
  33. Quarles LD, Yohay DA, Carroll BA, et al. 1994 Prospective trial of pulse oral vs. intravenous calcitriol treatment of hyperparathyroidism in ESRD. Kidney Int. 45:1710–1721.[Medline]
  34. Coggins MO, Rao DS, Bhat S, Zasuwa G, Narins RG. 1994 Hyperphosphatemia predicts a poor response to I.V. calcitriol (IVC) in hemodialysis (HD) patients with secondary hyperparathyroidism (HPT). J Am Soc Nephrol. 5:879.
  35. Shoji S, Nishizawa Y, Tabata T, et al. 1995 Influence of serum phosphate on the effecacy of oral 1,25-dihydroxyvitamin D3 pulse therapy. Miner Electrolyte Metab. 21:223–228.[Medline]
  36. Lucas PA, Brown RC, Woodhead JS, Coles GA. 1986 1,25-Dihydroxycholecalciferol and parathyroid hormone in advanced chronic renal failure: effects of simultaneous protein and phosphorus restriction. Clin Nephrol. 25:7–10.[Medline]
  37. Massry SG, Coburn JW, Lee DBN, Jowsey J, Kleeman CR. 1973 Skeletal resistance to parathyroid hormone in renal failure: studies in 105 human subjects. Ann Intern Med. 78:357–364.
  38. Sherrard DJ, Hercz G, Maloney NA, et al. 1993 The spectrum of bone disease in end-stage renal failure–an evolving disorder. Kidney Int. 43:436–442.[Medline]
  39. Felsenfeld AJ, Rodriguez M, Dunlay R, Llach F. 1991 A comparison of parathyroid-gland function in haemodialysis patients with different forms of renal osteodystrophy. Nephrol Dial Transplant. 6:244–251.
  40. Kurz P, Monier-Faugere M-C, Bognar B, et al. 1994 Evidence for abnormal calcium homeostasis in patients with adynamic bone disease. Kidney Int. 46:855–861.[Medline]
  41. Brown EM. 1983 Four-parameter model of the sigmoidal relationship between parathyroid hormone release and extracellular calcium concentration in normal and abnormal parathyroid tissue. J Clin Endocrinol Metabol. 56:572–581.[Abstract]
  42. Tsukamoto Y, Moriya R, Nagaba Y, Morishita T, Izumida I, Okubo M. 1995 Effect of administering calcium carbonate to treat secondary hyperparathyroidism in nondialyzed patients with chronic renal failure. Am J Kidney Dis. 25:879–886.[Medline]
  43. Luisetto G, Bertoli M. 1994 Sexual influences on bone metabolism in uremic patients on regular dialytic treatment. Nephron. 67:150–157.[Medline]
  44. Naveh-Many T, Almogi G, Livni N, Silver J. 1992 Estrogen receptors and biologic response in rat parathyroid tissue and C cells. J Clin Invest. 90:2434–2438.
  45. Cosman F, Nieves J, Horton J, Shen V, Lindsay R. 1994 Effects of estrogen on response to edetic acid infusion in postmenopausal osteoporotic women. J Clin Endocrinol Metab. 78:939–943.[Abstract]



This article has been cited by other articles:


Home page
 Nephrol Dial TransplantHome page
C. Basile, C. Lomonte, L. Vernaglione, F. Casucci, D. Chimienti, A. Bruno, S. Cocola, E. A. Verrelli, and F. Cazzato
A high body mass index and female gender are associated with an increased risk of nodular hyperplasia of parathyroid glands in chronic uraemia
Nephrol. Dial. Transplant., April 1, 2006; 21(4): 968 - 974.
[Abstract] [Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Submit a related Letter to the Editor
Right arrow Purchase Article
Right arrow View Shopping Cart
Right arrow Alert me when this article is cited
Right arrow Alert me when eLetters are posted
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow Request Copyright Permission
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Indridason, O. S.
Right arrow Articles by Quarles, L. D.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Indridason, O. S.
Right arrow Articles by Quarles, L. D.

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
Endocrinology Endocrine Reviews J. Clin. End. & Metab.
Molecular Endocrinology Recent Prog. Horm. Res. All Endocrine Journals