Contents

Peritoneal dialysis (PD) is a technique that exposes the patient to glucose solutions and may cause metabolic complications, such as central obesity, hypertriglyceridaemia and hyperglycaemia. Glucose absorbed from the peritoneal cavity may lead to the development of insulin resistance (IR) and de novo diabetes1. Furthermore, exposure to glucose degradation products (GDP) leads to structural and functional damage of the peritoneal membrane2. In a study by Fortes et al., PD patients had higher fasting glucose, glycated haemoglobin (HbA1C) and estimated IR rate using the HOMA-IR index3 than haemodialysis patients. In addition, patients who receive dialysis with glucose-free dialysate show a lower absorption of glucose, lower weight gain and fat accumulation, and improved IR and dyslipidaemia4; furthermore, the use of icodextrin leads to increased adipocytokines in the plasma of PD patients, without changes in cholesterol levels, but with a reduction in triglycerides5. In a study aimed at observing differences in the lipid profile of 22 non-diabetic patients on PD between 6 and 48 months compared to a control group of a similar age, there were significantly higher levels of very low-density lipoproteins, cholesterol bound to low-density lipoproteins and triglycerides, and significantly lower levels of cholesterol bound to high-density lipoproteins compared to the control group6; using a 72-hour continuous glucose-monitoring system, we studied the effect of PD glucose solutions on patient glucose levels and we observed that the percentage of glucose levels above 90mg/dl was influenced by high glucose concentrations in the fluids and the high transporter state. However,  a Spanish study recently published that non-diabetic PD patients did not have a significant increase in HOMA-IR levels, or modifications to these values after one year of treatment on PD, or statistically significant changes in the lipid profile7.

We carried out a retrospective observational study with 39 non-diabetic PD patients of the Hospital Clínico San Carlos de Madrid, 26 on continuous ambulatory PD and 13 on automated PD, of 61+14 years of age, in which we analysed baseline glucose and lipids (total cholesterol and triglycerides) before beginning PD and after 1, 3, 6, 12, 18, 24, 30 and 36 months using the technique, and a prospective 12-month study in 18 of the patients, also analysing HbA1C. We studied time on PD, the type of PD, the type of transporter and use of solutions with a high glucose load (two or more exchanges at 2.3%) or a low load (fewer than two at 2.3% and/or icodextrin). We only used fluids with a high glucose load in 6 patients and we did not use 3.86%-4.25% solutions in any patient. The type of transporter was high (medium-high, high) in 16 and low (medium-low, low) in 23.

We did not find significant differences between the pre-PD glucose means and those found over the 36 months of follow-up (Table 1), which remained at normal levels throughout the study. Cholesterol levels rose suddenly in the sixth month with respect to baseline values (171±45 vs. 193.5±46mg/dl; p=.008), without changes in the triglyceride figures and with normal levels being maintained in both factors throughout follow-up. In the prospective study with 18 patients, we did not observe significant differences in glycaemia evolution: baseline 103±14 vs. 105±17 after 1 month, 112±14 after 3 months, 108±20 after 6 months and 104±14mg/dl after 12 months. No significant differences were observed in HbA1C: baseline 5.5±0.5 vs. 5.5±0.5 after 1 month, 5.4±0.6 after 3 months, 5.7±0.8 after 6 months and 5.4±0.6% after 12 months. Glycaemia and HbA1C do not seem to change in accordance with the glucose load. There is a good correlation between glucose and HbA1C. High transporters have higher glucose values after one month on PD (P=.039), but not of HbA1C.

During the first years in which PD has been reported, and on the basis of the glucose load that was contributed to obtain sufficient ultrafiltration, it was considered to be a dialysis technique with a potential diabetogenic effect. It is possible that in these first few years, due to a lack of knowledge about the deleterious effect that glucose contribution has on the peritoneum with the development of GDP2, the relatively common use of very hypertonic solutions, which furthermore did not use bicarbonate as a buffer, may have caused some cases of diabetes. In the last decade since the introduction of solutions in dual chambers with a mixture of lactate and bicarbonate or bicarbonate alone, with which the formation of GDP is minimal and use of 3.86%-4.25% glucose PD dialysate is practically nil, the induction of diabetes and even the development of moderate hyperglycaemia, as our study shows, have become anecdotal. The increase in lipids reported in some articles6 is not relevant in our study in terms of its maintenance over time and it has not been confirmed by other authors7.

In conclusion, our non-diabetic PD patients treated with glucose solutions did not show changes in their glucose levels throughout the 36 months on dialysis. HbA1C was unchanged after a year on the technique. The potential development of diabetes in PD was not confirmed by our results.

Conflicts of interest

The authors declare that they have no conflicts of interest related to the contents of this article.

12394_19157_60665_en_w4777147514ref.1239430222_12394_19115_53936_es_12394_tabla1_copy1.doc

Table 1. Evolution of glycaemia in peritoneal dialysis.