Serum Levels of Phosphorus, Parathyroid Hormone and Calcium and Risk of Death and Cardiovascular Disease in Individuals with Chronic Kidney Disease - Article Review - Dr. Ross Morton
This article appraisal is part of the EMiNEM Bone and Mineral Metabolism Series. Click here to reach the EMiNEM homepage on UKidney
The quality of the evidence on which clinical practice guidelines for mineral metabolism control in CKD is based is poor.
Study design and study population:
This is a systematic review and meta-analysis of published cohort studies examining the relationship between phosphorus (sic), parathyroid hormone, or calcium and death or cardiovascular events. By nature, the study is dealing with retrospective data for the most part. The population is derived largely from USRDS and DOPPS datasets, but studies involving patients on peritoneal dialysis, patients with renal transplants and non-dialysis CKD patients were also included as available.
Outcome of Interest:
The outcome of interest was the quality of the evidence linking abnormalities in phosphorus, parathyroid hormone or calcium to all cause mortality and cardiovascular disease.
Although the authors were required to exclude a large number of patients for methodological reasons, they were left with a dataset of > 100 000 patients for analysis.
Their conclusion from adequately adjusted studies was that there was no relationship between mortality and cardiovascular outcomes for parathyroid hormone or calcium. The authors found a relationship between elevated phosphorus and mortality with an increasing risk of death when phosphate levels exceed 5.5 mg/dL (1.78 mmol/L). (However, visual inspection of Figure 2 in the paper suggests that this relationship is driven mainly by one study.)
The relative risk of death was considered to be 1.18 (95% CI 1.12 – 1.25) for every 1 mg/dL (0.323 mmol/L) increase in phosphate above 3.5 mg/dL (1.13 mmol/L).
The authors found that the quality of evidence to permit strong Clinical Practice Guidelines to be promulgated was poor and did not justify their creation.
This study is well worth reading. It is an important attempt to synthesis data from multiple sources and includes peritoneal dialysis, transplant and CKD patients without renal replacement (although hampered by numbers in some of these categories). While much of the data is derived from USRDS (country specific) and DOPPS (more globally representative) datasets, it does represent the acquired knowledge to date.
The study has a number of weaknesses. In a polite way, the authors paraphrase the “garbage in, garbage out” nature of their analysis. They note that they were required to assume linearity in any relationship between abnormal mineral metabolism parameters and outcome, when a “U”- or “S”-shaped relationship is equally (if not more) likely. They then provide some self-criticism based on their lack of inclusion of associative data on vitamin D and its analogues and alkaline phosphatase and outcome. It is unlikely that the data from numerous studies on either of these associations is any stronger that that for phosphorus, calcium and vitamin D.
Overall it may be naïve to assume that any single marker of mineral metabolism can be examined in isolation. Certainly clinical measures to change any one of calcium, phosphorus or parathyroid hormone impact significantly on each of the others.
The authors conclude, as most Nephrologists accept, that the evidence on which mineral metabolism guidelines are based is poor.
Impact on practice:
It is now “open season” on Clinical Practice Guidelines for mineral metabolism management in CKD. (See Effect of Bone Mineral Target Achievement on Mortality in Incident Dialysis Patients: An Analysis of the United Kingdom Registry. Am J Kidney Dis 2011;57:415)
This is one of a number of papers reminding us that we do not have solid evidence on which to base our mineral metabolism targets. There is a strong caution in the paper reminding the reader that the medical community has been wrong before in its adoption of apparently rational treatment strategies – with notable examples given (although no mention that the community has been correct before in many instances). This paper may have the effect of allowing a sense of security in having a laissez-faire attitude towards abnormal mineral metabolism for the general Nephrology practitioner.
The essential, if repetitive, call for placebo controlled randomized studies (for which funding is unlikely to come from either industry or payers) provides little help for clinical practice.
Reviewed by Reviewed by Dr. Ross Morton