Background

We previously reported on increased bone strontium levels in dialysis patients with osteomalacia versus those presenting other types of renal osteodystrophy. A causal role of strontium in the development of osteomalacia was established in a chronic renal failure rat model.

Methods

To further elucidate the latter issue and to find out whether dialysis patients from particular centers/countries are at an increased risk for strontium accumulation, a worldwide multicenter study was established. In total, 834 patients from 34 dialysis centers in 23 countries were included. In each of the patients, a serum sample was taken for strontium determination, and water and dialysate samples were taken at the various steps of the water purification process. For each patient clinical data and for each center dialysis modalities were recorded.

Results

Strontium levels in serum of dialysis patients showed major differences between the various centers, ranging from mean values of 25 8 g/liter in the center with the lowest level up to 466 90 g/liter in the center with the highest concentration. It is of interest that these high levels were mainly found in developing countries. Furthermore, our data point toward a role of the final dialysate in the accumulation of the element, as indicated by the strong correlation (r = 0.74, P < 0.001) between mean serum and dialysate strontium levels. As the high tap water concentration of strontium was adequately reduced during the water purification process, contamination of the final dialysis fluid occurred by the addition of concentrates contaminated with strontium. Besides the dialysate, other factors, such as duration of dialysis, vitamin D supplements, or types of phosphate binders, played a less important role in the accumulation of the element.

Conclusions

Data of this multicenter study indicate patients of particular dialysis centers to be at an increased risk for strontium accumulation, the clinical consequence of which is under current investigation.

Methods

Patients

Subjects with normal renal function

Serum Sr levels were determined in 24 subjects with normal renal function who had a median age of 33 years (range 22 to 40). Thirty-three percent were male subjects.

Patients with stable renal failure

In 75 patients with stable chronic renal failure and who were not yet on dialysis, we also determined serum Sr levels. Forty-eight percent of the patients in this group were male; the median age of these patients was 54 years (range 18 to 86). Based on the creatinine clearance, patients were divided into two groups (creatinine clearance < and> 50 ml/min).

Patients treated by HD

Eight hundred and thirty-four HD patients coming from 17 European centers, 9 South and Central American centers, 1 Middle-Eastern center, 1 Japanese center, and 6 African centers were included in this multicenter study. Serum samples were taken before the start of the HD session. Besides the Sr level in serum, we also determined the element's concentration in tap water, water samples taken at several points of the water-purification process, and the final dialysis fluid. In addition to Sr, the Al concentration in all serum and water samples, as well as in the final dialysis fluid, was determined also.

A questionnaire with personal and clinical data from each dialysis patient was completed, together with information about the dialysis modalities of each center. The dialysis population had a median age of 56 years (range 5 to 96), a median time on dialysis of 36 months (range 1 to 300), and a 55% male predominance. A Sr-rich diet (seafood and grains) was ingested by 65% of the population. Concerning the phosphate binders: 56.6% of the patients took only CaCO₃; 4.9% of the subjects took only Al(OH)₃, and 13.0% of the total population took a combination of Al(OH)₃ and CaCO₃ and/or other phosphate binders. Some of the patients (25.5%) took no phosphate binder or none of the types of phosphate binders we asked about in our questionnaire. Vitamin D supplements were used by 57.2% of the patients. A majority of the patients (57.1%) received erythropoietin (EPO) treatment, and 54.5% underwent transfusions. Only 1.6% of the patients were treated by deferoxamine (DFO) therapy, whereas 2.6% had undergone a parathyroidectomy.

Strontium and Al in serum, water, and dialysis fluids were analyzed using a Zeeman 3030 atomic absorption spectrometer equipped with an HGA-600 graphite furnace, an AS-60 autosampler, and an Anadex DP-9500B silent scribe printer, all from Perkin-Elmer (Norwalk, CT, USA). To determine Sr, all samples were diluted fourfold in a solution of 0.5 ml/liter Triton X-100 and 1 ml/liter HNO₃. Aluminum in serum was determined after a 1:3 dilution of the samples in reversed osmosis (RO) water. Interassay coefficient of variance (CV) was <6% (N = 12, N = 10, respectively), and the recovery of added analyte was close to 100% for all of the biological matrices under study. Methods for the determination of Sr and Al have been described by us in detail previously^19,20. In order to reduce the risk for contamination, Sr- and Al-free material for sampling and sample storage was sent to each of the participating centers. All serum, water, and dialysate analyses were performed in a single laboratory, that is, the Department of Nephrology/Hypertension of the University of Antwerp.

Statistics

Data were analyzed using the statistical package SPSS for Windows. Data are expressed as mean SD. Univariate analysis was performed with t-test and chi-square test. The a priori level of significance was set at P 0.05. For multivariate analysis, multiple linear regression and logistic regression were used. The squared correlation coefficient (R²) was calculated for evaluating the explained proportion of variability in a model. Odds ratios and 95% confidence intervals (95% CI) were estimated using the exp (B) of the logistic regression equation.

RESULTS

The mean SD serum Sr level in individuals with normal renal function was 14 7.6 g/liter. In the group of patients with a creatinine clearance of more than 50 ml/min, the mean serum Sr level was 28 15 g/liter. In the group of patients with preterminal renal failure (creatinine clearance <50 ml/min), the mean serum Sr level was 52 21 g/liter. The significant increase (P < 0.0001) of the serum Sr level with decreasing renal function is shown in Figure 1.

Figure 1.

Serum strontium (Sr) levels in subjects with various degrees of renal function. *P < 0.001 vs. normal renal function; °P < 0.001 vs. creatinine clearance> 50.

Full figure and legend (16K)

The overall mean SD serum Sr level of the HD patients under study was 95 103 g/liter. A center-to-center comparison of the mean serum Sr level of all dialysis patients indicated significant differences ranging from a mean SD value of 25 8 g/liter in the center with the lowest value up to 466 90 g/liter (P < 0.0001) in the center with the highest level Figure 2. According to the mean Sr level, the dialysis centers can be categorized into two groups: a first group of centers with Sr levels below 100 g/liter (low Sr group, 42 19 g/liter) and a second group of centers with levels above 100 g/liter (high Sr group, 231 104 g/liter). Of interest was that all dialysis centers in the latter group were located in developing countries. As shown in Figure 3, the distribution of the serum Sr levels in both groups indicates only a very small overlap (0.72%) between groups. These data document that nearly all patients from dialysis centers in the high Sr group will have increased serum Sr levels. Figure 4 shows the mean serum Sr levels in the South and Central American dialysis centers that participated in the multicenter study. Although most dialysis centers in this region have increased Sr levels, we cannot generalize this because levels in both Chile and Argentina are comparable to the concentrations of the element found in Western countries.

Figure 2.

Mean serum strontium (Sr) levels of dialysis patients from the different dialysis centers.

Full figure and legend (29K)

Figure 3.

Distribution of serum strontium (Sr) levels in the low-Sr group versus the high-Sr group. There were 599 low-Sr patients [outliers 100 g/liter: 0.15% (N = 3: 105, 106, 155 g/liter)] and 235 high-Sr patients [outliers 100 g/liter: 1.3% (N = 3: 37, 56, 86 g/liter)]

Full figure and legend (15K)

Figure 4.

Serum Sr levels in the South and Central American dialysis centers.

Full figure and legend (18K)

Another interesting finding was that in the high-Sr group, increased Sr levels were accompanied by increased levels of Al, that is, 72 63 g/liter compared with 19 22 g/liter in the low-Sr group. Serum Ca was significantly decreased in the high-Sr group (4.35 0.57 mEq/liter) in comparison to the low-Sr group (4.65 0.67 mEq/liter; Figure 5). Also in Figure 5, the differences in personal and clinical factors between both groups are shown. Multivariate analysis of all of these data indicates these factors can explain 20% of the differences in serum Sr levels between both groups.

Figure 5.

Patient characteristics in the high-Sr and low-Sr groups.

Full figure and legend (44K)

Out of this relatively small percentage, it is obvious that other factors must play a role in the accumulation of Sr in dialysis patients. To further explain this issue, we assessed whether the observed differences between the high and low Sr groups could be due to center-to-center variations in either tap water or dialysate. Assessing the evolution of Sr levels during the various steps of the water purification process, data indicated Sr levels in the tap water to differ greatly from center to center. During the water purification process, however, Sr was removed adequately from the water in both groups Figure 6. In the low-Sr centers, this results in a low concentration of the element in the final dialysis fluid. In the high-Sr group, however, an increase in the Sr concentration between points 4 (after reversed osmosis) and 5 (inlet artificial kidney) of the water treatment process was noted, resulting in an increased concentration of the element in the final dialysis fluid. It can be assumed that this accumulation is due to the addition of concentrates contaminated with Sr. Indeed, analysis of the concentrates used in certain dialysis centers of the high-Sr group revealed these solutions to be contaminated with Sr. Interestingly, the contamination occurred only in acetate-based concentrates (up to 15750 g/liter; undiluted concentrate), whereas Sr levels in bicarbonate-based concentrates were below the analytical detection limit (<3 g/liter). A strong correlation (r = 0.74, P < 0.001) between mean serum and dialysate Sr levels was noted. When the Sr concentration in the final dialysis fluid was assessed in a multivariate model, it explained 50% of the difference in Sr levels between both groups. Interestingly, when considering only the high-Sr group, other factors also seem to play a substantial role in the accumulation of the element, specifically, the serum Ca concentration, the use of vitamin D supplements, and the consumption of seafood contributed to 43.1% of the variation in serum Sr levels between dialysis patients within the high-Sr group. No correlation was present between Sr and Al within the high-Sr group, indicating that in a number of these centers, high Sr levels were associated with low levels of Al.

Figure 6.

Evolution of the strontium levels during the different steps of the water purification process in both the high () and low () Sr groups.

Full figure and legend (21K)

DISCUSSION

As we previously reported, in a histological/chemical survey of 100 bone biopsies from dialysis patients originating from several centers in various countries, increased Sr concentrations as well as Sr/Ca ratios were observed in the bone of patients suffering from osteomalacia as compared with all other types of renal osteodystrophy³. Interestingly, a recent epidemiological study in Turkey demonstrated a significantly higher prevalence of the clinical diagnosis of rickets in children with normal renal function who were living in a region with a high soil Sr content and where nutrition was mainly based on cereals as compared with that in children living in a low soil Sr region²¹. Experimentally, it has been shown that administration of Sr to animals with normal renal function may result in a bone lesion with a histological resemblance to that seen in vitamin D-deficient rickets²². However, the so-called Sr-induced rickets^23,24 differs from Ca- or phosphorus-deficiency rickets in that vitamin D supplements do not correct the lesion^24,25. In an experimental study in chronic renal failure rats loaded with Sr, we could also demonstrate the presence of histological lesions that were highly comparable with the histological features of osteomalacia seen in dialysis patients⁴.

Our results clearly show that serum Sr levels are increased in subjects with decreasing renal function, as compared with individuals with normal renal function. This is not unexpected, as the element is mainly excreted by the kidney.

As early as 1973 using x-ray fluorescence, Rudolph, Alfrey, and Smythe noted distinctly (five- to eightfold) increased Sr concentrations in skeletal muscle of dialyzed uremics compared with those of nondialyzed individuals and controls that they suggested to originate from the use of untreated high Sr tap water in the region of Denver that was used to prepare the dialysis fluid²⁶. At that time, they suggested that in addition to Al, Sr might play a role in the development of renal osteodystrophy in dialysis patients. A contributive role for Sr and Al in the development of dialysis osteomalacia was also suggested by Canavese et al¹⁷. In this respect, epidemiological findings of our bone biopsy study, demonstrating a good correlation between Al and Sr in the osteomalacic patients but not in those presenting the other types of renal osteodystrophy, are of particular interest³. These findings could be confirmed in our recent epidemiological survey, where in the serum of dialysis patients from the high-Sr group, besides the mean Sr levels, mean Al levels were also increased as compared with the low-Sr group.

Data of this multicenter study indicate that patients of particular dialysis centers are at an increased risk for Sr accumulation. We found Sr levels in dialysis patients to differ significantly from center to center, as well as from country to country. These great variations in Sr levels appeared to be mainly due to the use of Sr-contaminated dialysis fluids. High Sr levels in the dialysate seemed to originate from the addition of Sr-containing concentrates, particularly acetate-based concentrates, which are added to prepare the final dialysis fluids. Other factors, such as time on dialysis, vitamin D supplements, or type of phosphate binder used, seem to play a less important role in the accumulation of the element. With regard to the latter, Canavese et al suggested the oral intake of Al-containing phosphate binders, particularly Maalox® tablets, which may contain up to 217 g/g of the element, to be an important source of Sr within the dialysis population¹⁷.

Serum Ca levels were decreased in the high-Sr group as compared with the low-Sr group. This may point to an effect of Sr on Ca absorption and, as such, confirm previous observations of other groups, indicating that high Sr concentrations negatively affect Ca absorption, either directly or indirectly, possibly through the element's interference with vitamin D synthesis^16,25.

Although the prevalence of osteomalacia has decreased in the overall dialysis population, the disease is still regularly observed in developing countries. In this context, it is of interest that we only found increased serum Sr concentrations in patients living in these countries. Hence, for these individuals, Sr accumulation may present an increased risk for the element's toxic effects. One of the centers in this multicenter study also participated in the previous bone biopsy study³. Out of the five patients in whom a bone biopsy was taken, four had the histological features of osteomalacia in the presence of increased bone Sr levels. As noted in this study, all patients of that particular center also have increased serum Sr levels, which appeared to be due to the use of a high Sr dialysate (as high as 243 g/liter). Hence, these data indicate that dialysis patients treated with Sr contaminated dialysis fluids are at an increased risk for Sr accumulation and perhaps the development of osteomalacia. Further studies are necessary, however, to clarify the relationship between Sr and osteomalacia.

Data of this multicenter study allow us to conclude that patients of particular dialysis centers are at an increased risk for Sr accumulation, which appears to be mainly due to Sr contamination of the dialysis fluid. To what extent this Sr accumulation has clinical consequences remains to be elucidated.

References

1. D'Haese, PC, Couttenye, MM, Goodman, WG, Lemoniatou, E, Digenis, P, Sotornik, I, Fagalde, A, Barsoum, RS, Lamberts, LV, De Broe, ME: Use of the low-dose desferrioxamine test to diagnose and differentiate between patients with aluminum-related bone disease, increased risk for aluminum toxicity or aluminum overload. Nephrol Dial Transplant 1995 10: 1874–1884,  | PubMed | ChemPort |
2. Couttenye, MM, D'Haese, PC, Van Hoof, VO, Lemoniatou, E, Goodman, WG, Verpooten, GA, De Broe, ME: Low serum levels of alkaline phosphatase of bone origin: A good marker of adynamic bone disease in haemodialysis patients. Nephrol Dial Transplant 1996 11: 1065–1072,  | PubMed | ISI | ChemPort |
3. D'haese, PC, Schrooten, I, Goodman, WG, Cabrera, W, Lamberts, LV, Elseviers, MM, Couttenye, MM, Debroe, ME: Increased bone strontium levels in dialysis patients with osteomalacia. (in press)
4. Schrooten, I, Cabrera, W, Dauwe, S, Lamberts, LV, Marynissen, R, Dorriné, W, Goodman, WG, De Broe, ME, D'Haese, PC: Strontium causes osteomalacia in chronic renal failure rats. Kidney Int 1998 54: 448–456,  | Article | PubMed | ISI | ChemPort |
5. Schroeder, HA, Tipton, JH, Nason, AP: Trace metals in man: Strontium and barium. J Chronic Dis 1972 25: 491–517,  | Article | PubMed | ISI | ChemPort |
6. Sips, AJAM, Van Der Vijgh, WJF: Strontium, in Handbook on Metals in Clinical and Analytical Chemistry, 1994 edited by Seiler H, Sigel A, Sigel H, New York, Marcel Dekker, pp 577–585
7. Rosenthal, HL, Cochran, OA, Eves, MM: Strontium content of mammalian bone, diet and excreta. Environ Res 1972 5: 182–191,  | Article | PubMed | ISI | ChemPort |
8. Underwood, EJ: Trace Elements in Humans and Animal Nutrition. 1977 New York, Academic Press,
9. Skoryna, S: Metabolic aspects of the pharmacologic uses of the trace elements in human subjects with specific reference to stable Sr. Trace Subst Environ Health 1984 18: 3–23,  | ChemPort |
10. Marie, JP, Garba, MT, Hott, M, Miravet, L: Effect of low doses of stable strontium on bone metabolism in rats. Miner Electrolyte Metab 1985 11: 5–13,  | PubMed | ISI |
11. Grynpas, M, Marie, JP: Effects of low doses of strontium on bone quality and quantity in rats. Bone 1990 11: 313–319,  | Article | PubMed | ISI | ChemPort |
12. Marie, JP, Hott, M, Modrowski, D, De Pollak, C, Guillemain, J, Deloffre, P, Tsouderos, Y: An uncoupling agent containing strontium prevents bone loss, depressing bone resorption and maintaining bone formation in estrogen-deficient rats. J Bone Min Res 1993 8: 607–615,  | ISI |
13. Matsumoto, A: Effect of strontium on the epiphyseal cartilage plate of rat tibiae—Histological and radiographic studies. Jpn J Pharmacol 1976 26: 675–681,  | PubMed | ChemPort |
14. Cabrera, WE, Schrooten, I, De Broe, ME, D'Haese, PC: Strontium and bone. J Bone Miner Res 1999 14(5):661–668,  | PubMed | ISI | ChemPort |
15. Omdahl, JL, DeLuca, HF: Strontium induced rickets: Metabolic basis. Science 1971 174: 949–951,  | PubMed | ISI | ChemPort |
16. Corradino, R, Wasserman, R: Strontium inhibition of vitamin D3-induced calcium binding protein (Ca BP) and calcium absorption in chick intestine. Proc Soc Exp Biol Med 1970 133: 960–963,  | PubMed | ISI | ChemPort |
17. Canavese, C, Pacitti, M, Salomone, M, Santoro, A, Strata, P, Mangiarotti, G, Talarico, S, Sabbioni, E, Pietra, R, Vercellone, A: Strontium overload in uremic patients on regular dialytic treatment. Trans Am Soc Artif Intern Organs 1986 32: 120–122,  | ChemPort |
18. Mauras, Y, Ang, K, Simon, P, Tessier, B, Cartier, F, Allain, P: Increase in blood plasma levels of boron and strontium in hemodialyzed patients. Clin Chim Acta 1986 156: 315–320,  | Article | PubMed | ISI | ChemPort |
19. D'Haese, PC, Van Landeghem, GF, Lamberts, LV, Bekaert, VA, Schrooten, I, De Broe, ME: Measurement of strontium in serum, urine, bone and soft tissues by Zeeman atomic absorption spectrometry. Clin Chem 1997 43: 121–128,  | PubMed | ChemPort |
20. D'Haese, PC, Van de Vyver, FL, De Wolff, FA, De Broe, ME: Measurement of aluminum in serum, blood, urine and tissues of chronic hemodialyzed patients by use of electrothermal atomic absorption spectrometry. Clin Chem 1985 31: 24–29,  | PubMed | ChemPort |
21. Özgür, S, Sümer, H, Koçoglu, G: Rickets and soil strontium. Arch Dis Child 1996 75: 524–526,  | PubMed |
22. Engfeldt, B, Hjertquist, SO: The effect of strontium administration on bones and teeth maintained on diets with different calcium contents. Virchows Arch Pathol Anat 1969 346: 330–344,  | ISI | ChemPort |
23. Shipley, PG, Park, EA, McCollum, EV, Simmonds, N, Kinney, EM: Studies on experimental rickets. XX. The effects of strontium administration on the histological structure of growing bones. Bull J Hopkins Hosp 1922 33: 216–221,  | ChemPort |
24. Sobel, A, Goldfarb, A, Kramer, B: Studies of incurable rickets. I. Respective role of the local factor and vitamin D in healing. Proc Soc Exp Biol Med 1934 31: 869–870,
25. Omdahl, JL, DeLuca, HF: Rachitogenic activity of dietary strontium. I. Inhibition of intestinal calcium absorption and 1,25-dihydroxycholecalciferol synthesis. J Biol Chem 1972 247: 5520–5526,  | PubMed | ISI | ChemPort |
26. Rudolph, H, Alfrey, AC, Smythe, WR: Muscle and serum trace element profile in uremia. Trans Am Soc Artif Int Organ 1973 21: 456–461,

Acknowledgments

The authors are grateful to the participating nephrologists and their staff: Dr. Billiouw (OLV Ziekenhuis Aalst, Belgium), Dr. Schurgers (AZ St. Jan Brugge, Belgium), Dr. Bosteels (Kliniek Maria's Voorzienigheid, Kortrijk, Belgium), Dr. Segaert (Kliniek Heilig Hart, Roeselare, Belgium), Dr. Nenov (Clinic of Nephrology, Sofia, Bulgaria), Dr. Sotornik (Institute for Clinical and Experimental Medicine, Prague, Czech Republic), Dr. Drüeke (Hôpital Necker, Paris, France), Dr. Ittel (Rheinisch-Westfälische TH, Aachen, Germany), Dr. Digenis (four centers, Athens, Greece), Dr. Di Ioro (two centers, Lauria, Italy), Dr. Polenakovic, and Dr. Spasovski (three centers, Skopje, Macedonia), Dr. Barata (Sociedade Portuguesa de Dialize, Lissabon, Portugal), Dr. Dhlopolçek (FD Roosevelt Hospital, Banska Bystrica, Slovak Republic), Dr. Djukanovic (two centers, Belgrade, Yugoslavia), Dr. Fagalde (Instituto Privado de Dialisis, Cordoba, Argentina), Dr. Schor and Dr. Carvalho, and Dr. Draibe (five centers, São Paulo, Brazil), Dr. Ortiz (Universidad de Valpariso, Vina del Mar, Chile), Dr. Herra-Sanchez (Hospital San Juan de Dios, San José, Costa Rica), Dr. Santacruz and Dr. Cabrera (two centers, Asuncion, Paraguay), Dr. Gnionsahe (Abidjan, Côte d'Ivoire), Dr. Barsoum and Dr. Soliman (Cairo Kidney Center, Cairo, Egypt), Dr. Ramzy (Misr Kidney Center, Cairo, Egypt), Dr. Sobh (Mansoura, Egypt), Dr. Huraib (King Fahad Hosp., Riyadh, Saudi Arabia), Dr. Swanepoel and Dr. Halkett (Grote Schuur, Kaapstad, South Africa), Dr. Akiba (Tokyo, Japan) and Dr. Akoglu (Istanbul, Turkey). Iris Schrooten is a recipient of a postgraduate research grant of the Flemish Institute for the promotion of Scientific Technological Research in Industry (IWT). The authors are indebted to D. De Weerdt for his excellent drawings.