Does the choice of phosphate binder affect trace element levels in chronic kidney disease patients treated by regular haemodialysis?

BACKGROUND

Ion exchange resins have been reported to bind copper and zinc. As the phosphate binder sevelamer hydrochloride is an ion exchange resin, we audited trace element levels in our haemodialysis cohort to determine whether sevelamer prescription affected trace element levels compared with other phosphate binders.

METHODS

Samples for zinc, copper and selenium were taken in special tubes and measured by atomic absorption spectroscopy or inductively coupled plasma-mass spectrometry from 211 patients attending an inner city university hospital main dialysis centre.

RESULTS

Of the patients, 12.9% were prescribed oral or intravenous trace element supplementation. Of the remainder, 5.5% of patients had low plasma copper, 37.4% low zinc and 45.6% low selenium. There was no difference in copper (16.7 ± 0.2 vs 16.8 ± 0.4 μmol/L, respectively) and zinc (12.0 ± 0.3 vs 11.6 ± 0.2 μmol/L) comparing patients prescribed sevelamer compared with other phosphate binders. Despite a high prevalence of statin prescription, total cholesterol (3.42 ± 0.12 vs 3.89 ± 0.08, P < 0.01), LDL-cholesterol (1.46 ± 0.1 vs 2.00 ± 0.07, P < 0.01) and total cholesterol/HDL-cholesterol ratio (2.82 ± 0.15 vs 3.5216.7 ± 0.2 vs 16.8 ± 0.4 μmol/L, P < 0.01) were lower in the sevelamer group compared with those prescribed other phosphate binders. On logistic regression analysis, serum zinc levels were associated with serum albumin (F 20.36, β 0.174, CL 0.086-0.265, P < 0.001) and dialysis vintage (F 8.1, β 0.008, CL 0.002-0.013, P = 0.005), copper levels with log CRP (F 31.4, β 3.04, CL 0-1.97, P < 0.001) and urine volume (F 5.1, β - 0.01, CL - 0.002-0, P = 0.024), and selenium levels with serum albumin (F 23.2, β 0.016, CL 0.02-0.1, P < 0.001) and race (F 31.4, β 3.62, P = 0.032), with selenium levels being greater in non-Caucasoids (0.9 ± 0.02 vs 0.76 ± 0.02 μmol/L, P < 0.01).

CONCLUSIONS

Trace element and micronutrient deficiencies were relatively common in this inner city population of outpatient haemodialysis patients. However, the prescription of different phosphate binders did not have an observable effect on serum copper and zinc levels, but those prescribed sevelamer did have lower lipid profiles compared with those prescribed other phosphate binders. Trace element concentrations were more associated with albumin, a marker of general nutritional status, with some differences according to ethnicity, most likely due to differences in dietary intake.