More restrictive dietary phosphate prescriptions were not associated with longer survival in prevalent hemodialysis patients, reported an observational analysis published early online by the Clinical Journal of the American Society of Nephrology. There also was a suggestion of higher mortality with a more restrictive prescription in some patient subgroups.
“There's an increasing realization that dialysis patients are very subject to untoward effects of malnutrition,” said senior author Steven M. Brunelli, MD, MSCE, Associate Physician and Director of Dialysis Services at Brigham and Women's Hospital and Assistant Professor at Harvard Medical School, in a phone interview.
“Therefore, it seems paradoxical that so many of our treatments involve dietary restriction. At the same time, there's been this longstanding practice of restricting dietary phosphate that's never really been studied with respect to patient outcomes; it's been studied primarily with respect to short-term effects on serum phosphorus or parathyroid hormone levels.
“So we hypothesized that the benefits that may be anticipated on the basis of reducing phosphate levels through dietary restriction may be offset to greater or lesser degree by unintended compromises in nutritional status, particularly since foods that are intrinsically rich in phosphate tend to be high in other macronutrients, particularly protein.”
Katherine E. Lynch, MD, of Beth Israel Deaconess Medical Center and Harvard Medical School, was the lead author of the study.
The study design is strong, but there are some caveats, noted Myles Wolf, MD, MMSc, Associate Professor of Medicine and Assistant Dean for Translational and Clinical Research at the University of Miami Miller School of Medicine, in a phone interview.
“I think it's extremely well done by an extremely strong group of investigators,” he said. “Being able to base this analysis within a randomized controlled trial means that the quality of the data is quite high.
“Another important point is that it's a first-in-class study. Nobody has done an analysis like this in a dialysis population, and certainly these types of studies are needed. So I think it's a very good paper.”
There are limitations, he added.
“The fundamental underlying question is: Should we be restricting dietary phosphorus intake in dialysis patients? And this study could not directly address whether actually successfully restricting phosphorus intake alters mortality.
“This studied the prescription of dietary phosphorus restriction and whether that altered mortality. They're very different things. What the team taking care of the patients tried to do may not coincide with what the patients actually were doing.
“While that is a limitation, it's also a strength of the study because the study is asking: As we currently know how to do it, does attempting to implement dietary phosphorus restriction make a difference?
“And the answer is that at best it was neutral, and in certain analyses it could have been harmful. What remains an open question after doing this study is: Is dietary phosphorus restriction inadvisable, or are we currently doing it in an incorrect fashion, and if we were to do it differently, would it be successful?”
Plant vs Animal vs Additive
This analysis of prescribed dietary phosphate (PDP) used data from the Hemodialysis (HEMO) Study, a randomized controlled trial of the effects of dialysis dose and dialysis membrane flux on clinical outcomes (Eknoyan G et al: N Engl J Med 2002;347:2010-2019). Patients were enrolled in the trial between March 1995 and October 2000 and followed through Dec. 31, 2001.
Of the 1,846 hemodialysis patients randomized to the HEMO study, 1,751 had enough data for inclusion in this post hoc analysis. At baseline, the mean age of these individuals was 57.7, 56.5% were female, 63.0% were black, and 44.7% had diabetes. Mean serum phosphate level was 5.8 mg/dL, and 22.2% were using nutritional supplements.
Prescribed dietary phosphate was categorized by quartile: 870 mg or less (300 participants), 871 mg to 999 mg (314 participants), 1,000 mg (307 participants), and 1,001 mg to 2,000 mg (297 participants). Participants with no prescribed restriction in dietary phosphate comprised a separate category (533 participants).
There were several reasons behind the decision to examine prescribed phosphate restriction instead of measured phosphate intake, Dr. Brunelli said.
“The first major thing is that as patients get sick, they'll decrease their global nutrient intake. It's well known that patients, particularly patients with preterminal disease, are less hungry, and they eat less.
“Inasmuch as phosphorus is one component of overall nutrition, if we just studied what patients were actually taking in, there would be differential phosphorus intake among patients who were going to die versus patients who weren't going to die in a proximate sense, and so that leads to confounding. We might observe associations that are themselves true but don't reflect any influence of phosphate intake on mortality itself.
“At the same time, and in the other direction, many patients are given advice on what to do with their phosphorus intake, and some patients will heed that advice, and other patients won't. Of course it's not dichotomous; there are different degrees of adherence.
“In general, patients who tend to heed one form of medical advice tend to heed other forms of medical advice, so inasmuch as being an adherent patient and adhering to other forms of one's therapy or other therapeutic recommendations might impact a patient's mortality, it becomes difficult to say what's the direct impact of the dietary phosphorus manipulation.
“Studying the prescribed dietary phosphate takes a step back. The clinician can only make a determination as to whether or not to provide a certain piece of advice, and they really don't have much of a direct role as to whether or not that advice is followed. It really is studying what should the clinician practice be.”
Patients with more restrictive prescribed dietary phosphate were more likely to be female, to be black, and to have a graft for dialysis access. They also tended to show poorer nutritional status, even though they took in more calories and protein. The poorer nutritional status was evidenced by lower serum albumin, creatinine, body weight, mid-arm muscle circumference, and triceps skinfold thicknesses; greater use of nutritional supplements; and poorer appetite.
“Observed phosphate intake tended to track with PDP (except for the group with no specified phosphate prescription), but differences across groups were modest,” the researchers reported.
Over a median follow-up of 2.3 years, 817 patients died. In an unadjusted baseline analysis, prescribed dietary phosphate was not associated with mortality. In a multivariable analysis that adjusted for baseline differences between the groups, there was a tendency toward longer survival in the unrestricted group, but the difference was not statistically significant. Further adjustment for protein and caloric intake had little effect on the results.
Since 29.1% of patients had a dietary phosphate prescription change after baseline, the researchers used a marginal structural analysis to minimize misclassification and account for potential time-dependent confounding. In this analysis, there was a step-wise trend toward greater survival with less restrictive prescription, particularly in patients who were not black, had a serum phosphate below 5.5 mg/dL, and were not taking vitamin D.
“I think the implications are twofold,” Dr. Brunelli said. “One, I think it points to the fact that dietary phosphate counseling as it's been done needs to be revisited. Now this is an observational analysis of a randomized trial that was done for other purposes, so it should be validated before anything is directly influenced, but it should give us pause.”
In his opinion, a dedicated randomized trial would be the preferred design for such a subsequent study, Dr. Brunelli noted.
“I definitely would like to emphasize that these data really pertain to recommendations made about foods that are intrinsically high in phosphate.” They do not apply to foods that are high in phosphate because of additives, he added.
“Increasingly so, there's been addition of phosphate to many food products in the form of food additives. Of course, the risk-benefit ratio of restricting those latter foods may be very different because they're not necessarily as nutrient rich as foods that are intrinsically high in phosphate.”
Even in terms of foods that are intrinsically high in phosphate, there may be differences. Drs. Brunelli and Wolf both pointed to a recently published study that compared vegetarian and meat dietary protein sources, reporting that one week of the vegetarian diet led to lower serum phosphorus levels (Moe SM et al: Clin J Am Soc Nephrol 2010 Dec 23;Epub ahead of print).
Referring again to the study by Dr. Brunelli and colleagues, Dr. Wolf said, “I think what should be taken from this study is a reevaluation of how we approach phosphorus restriction from purely looking at the total reported phosphorus content of foods to a model where we think not only about the total amount of phosphate in food but also the bioavailability of the phosphate in specific foods and figure out a way to keep foods in the diet that have a low phosphorus bioavailability but high protein content, and restrict foods that have high amounts of bioavailable phosphate but are otherwise of poor nutritive value, like processed foods that are enriched with phosphorus additives. Then we can begin to ask if implementing those kinds of changes would have a beneficial effect on clinical outcomes.”