Body mass index and peritoneal dialysis: "exceptions to the exception" in reverse epidemiology?

Abstract

Until recently, it was not apparent whether apparent paradoxical associations of body mass index (BMI), lipids, and blood pressure with survival observed in hemodialysis (HD) patients, which contradict observations from the general population, also applied to peritoneal dialysis (PD) patients. Studies of survival in PD patients must account for differences in adjusted survival relative to HD patients, namely, early equivalent to superior survival, but after about 1-2 years, inferior survival. Several recent observational studies have analyzed the association between BMI and survival in PD patients from different perspectives and using different patient populations. In general, these studies found that any survival advantage associated with obesity is significantly less likely in PD than HD patients. Among PD patients, those who are obese can be said to have equivalent survival to PD patients with normal BMI. Studies of lipids and blood pressure in PD patients also yield conflicting associations with survival. Obese patients, especially if diabetic, may have increased risk of death after starting on PD compared to HD, although firm conclusions are premature given the limitations of current evidence. At present, the levels of lipids and blood pressure which are best associated with survival in PD patients are not well-defined.