Comprehensive Conservative Management Versus Dialysis in Uric Acid Control

Abstract

Background: Hyperuricemia is a well-known problem in end-stage kidney disease. Currently, the end-stage kidney disease patients may be treated with comprehensive conservative management, hemodialysis, or peritoneal dialysis, which impact uric acid levels distinctly. We assessed the impact of these strategies on uric acid control and identified the factors that influence it. Methods: We conducted a preliminary case–control study comparing patients in comprehensive conservative management, hemodialysis and peritoneal dialysis. For each patient, we evaluated demographic characteristics, comorbidities, body mass index, protein intake, urine output and blood test results. Results: In the entire population, uric acid levels were slightly higher in the comprehensive conservative management group. Furthermore, uric acid control was influenced primarily by body mass index (β = −0.005, p = 0.03) and treatment modality (β = −0.0026, p = 0.05). In comprehensive conservative management, body mass index (β = −0.007, p = 0.02) and urine urea excretion (β = 0.014, p = 0.04) were independent predictors of uric acid level. Conversely, only the suggested protein intake (β = 0.16, p = 0.05), potassium levels (β = −0.046, p = 0.04) and allopurinol therapy (β = −0.073, p = 0.03) were independent predictors of uric acid in hemodialysis patients. Finally, only the recommended protein intake (B = −0.005, p = 0.03) was associated with uric acid levels in patients undergoing peritoneal dialysis. Conclusions: In our series, uric acid control correlates with the treatment modality used for end-stage kidney disease and dietary protein intake.

1. Introduction

Uric acid (UA) is the final product of purine metabolism, and its high level depends on impairment of purine metabolism or excretion. Specifically, the UA clearance relies on renal and gut excretion, which account for 66% and 34% of total excretion, respectively [1,2,3]. The presence of chronic kidney disease (CKD) has a multifaceted impact on UA metabolism. UA levels significantly increase in CKD due to impaired renal excretion, which is related to the activity of urate anion transporter 1 (URAT1) and glucose transporter 9 (GLUT9) in the proximal tubule, resulting in increased UA reabsorption and reduced urinary excretion [1,4,5].

Furthermore, the presence of CKD disrupts the intestinal microbiota, which is another crucial factor in UA excretion and balance. Microbiota dysbiosis in CKD reduces fecal excretion of UA and increases its intestinal reabsorption [1,6]. If kidney disease disrupts UA balance, the presence of hyperuricemia also negatively impacts kidney function and the progression of CKD. Higher UA levels harm kidney function by dysregulating the inflammatory response and inducing oxidative stress, thereby promoting glomerular sclerosis and fibrosis [7]. Furthermore, in this vicious cycle, the presence of other comorbidities, such as hypertension, diabetes and obesity, exacerbates the deleterious interaction between UA control and CKD, especially in end-stage chronic disease.

Although it remains under evaluation, UA appears to be an independent cardiovascular risk factor [8]. Some studies have suggested its role in oxidative stress [9,10,11,12], systemic inflammation [11,13], activation of the renin–angiotensin–aldosterone axis [14,15] and endothelial dysfunction [13,16,17], with a significant impact on the development of metabolic syndrome [18], heart failure [19,20] and coronary artery disease [21]. Currently, there is limited evidence that urate-lowering therapies reduce cardiovascular risk in the general population [20]. However, UA can act as an antioxidant in the extracellular environment [21,22,23,24,25] and as a pro-oxidant in the intracellular environment, enhancing endothelial nitric oxide activity and inhibiting cellular proliferation and migration [26,27]. Consequently, it is not surprising that allopurinol and fexustat therapies can improve hypertension, diabetes and kidney function in some clinical and experimental studies [28,29,30,31].

In end-stage chronic kidney disease (ESKD), UA levels may depend on various factors, including residual renal function, the efficiency of renal replacement therapy in removing UA in dialysis patients, treatment with UA-lowering medications and the patient’s diet. Currently, ESKD patients can choose from different treatment strategies [32], namely hemodialysis (HD), peritoneal dialysis (PD) and comprehensive conservative management (CCM), based on their performance status, comorbidities, age, primary nephropathy and patient preference. In elderly and frail patients with preserved urine output, CCM is a viable alternative to dialysis [33], as initiation of renal replacement therapy (RRT) has not shown significant benefits in survival or quality of life [34,35,36,37,38,39,40]. On the contrary, they may experience deterioration in overall condition and autonomy upon initiation of dialysis [41,42,43,44]. In CCM patients, UA control depends solely on dietary restriction and pharmacological treatment.

Conversely, HD patients can rely on the dialytic removal of UA, which depends on dialysis membrane characteristics and results in a UA reduction of 5.5–7.5% [45,46,47]. As well, PD patients can rely on UA removal during peritoneal exchange, with a clearance of approximately 7 L/24 h [48,49]. Although dialysis helps control UA, dialysis procedures require an increase in protein intake at the beginning of renal replacement therapy, which could nullify the potential advantage of UA dialysis removal. Both HD and PD patients lose a significant amount of protein during each treatment session. In HD, it is estimated that over 15 g of essential amino acids are consumed per session [50,51,52]. Whereas in PD, it is estimated that 5–15 g of protein is lost during 24 h and even more during a peritonitis episode [53,54]. Finally, the kinetics of urate-lowering drugs may differ in RRT patients. Recent studies have shown a 25–35% reduction in exposure to allopurinol metabolites after a hemodialysis session [55] and a 50–60% re-duction in the allopurinol dose in PD patients to achieve urate control [56].

In this context, where ESKD patients exhibit a distinct profile in controlling UA levels, it is essential to evaluate the determinants of UA control. This preliminary study aims to assess potential differences in UA levels across treatment modalities (CCM, HD and PD) and investigate how diet and pharmacological interventions affect UA control. Specifically, we focus on protein intake, patient characteristics and the impact of diuretics and urate-lowering drugs.

2. Materials and Methods

We conducted a preliminary case–control study comparing patients undergoing dialysis (HD or PD) with those in CCM at the Nephrology, Dialysis, and Transplantation Unit of Padua University Hospital, Italy.

Specifically, we enrolled all adult patients who had a CCM or RRT (i.e., hemodialysis and peritoneal dialysis) for at least six months. CCM patients had no change in their low-protein diet over the previous three months, and HD and PD patients had no changes in their dialysis prescriptions in the three months preceding the study. Specifically, to make the conditions comparable between CCM and RRT patients, we decided to enroll patients aged 75 years or older, considering the marginal role of CCM in younger patients.

The exclusion criteria were as follows:

-

A known diagnosis of active oncologic disease;

-

Diagnosis of cachexia;

-

Chronic inflammatory bowel disease;

-

The presence of a colostomy;

-

An unexpected weight loss in the last 6 months;

-

A history of a gout episode in the last 6 months;

-

Incremental dialysis;

-

A failure in dialysis adequacy (kt/V < 1.2 in HD patients, weekly kt/V < 2 in PD patients) in RRT patients;

-

Septic episode in the last three months.

For each patient, the following parameters were collected:

-

Demographic data: gender, age and dialytic or CCM vintage. Specifically, the CCM vintage was defined as achieving a clearance creatinine of less than 12 mL/min.

-

Comorbidities such as hypertension, diabetes, previous acute myocardial ischemia, history of cancer, peripheral vascular disease, previous stroke and chronic obstructive pulmonary disease (COPD).

-

Blood examinations: UA, hemoglobin, creatinine, urea, sodium, potassium, calcium, phosphorus, magnesium, parathyroid hormone (PTH), vitamin D 25-OH, albumin. All blood fasting samples were taken between 7:00 and 8:00 am on weekdays. Specifically, in HD, samples were collected at the end of a prolonged interval between dialysis sessions.

-

Urine output in 24 h;

-

Body mass Index (BMI);

-

Home therapy. Specifically, we assessed each patient’s allopurinol dose and the presence of potassium binders, phosphate binders and diuretics, including loop, thiazide and antialdosterone diuretics. Finally, we would like to emphasize that CCM patients were treated with sodium bicarbonate to correct metabolic acidosis, with the target of a venous bicarbonate level of 24–26 mmol/L, as per the KDIGO guidelines [57].

-

Protein intake. Specifically, CCM patients followed a personalized diet with a recommended daily protein intake of 0.6 g/kg/day in a low-protein diet, or 0.4–0.5 g/kg/day in a very low-protein diet, which was supplemented with keto-analogs. Furthermore, in the CCM group, protein intake was estimated from measured urine urea excretion using the Maroni equation [58]. In PD and HD patients, protein intake was based on dietary recommendations, with an adequate intake of 1.2 g of protein per kilogram of body weight, because the Maroni equation is not applicable for assessing dietary protein intake due to the critical influence of dialysis on urea balance. Specifically, RRT patients received a diet plan focused on the recommended protein intake, based on their body weight, in accordance with KDIGO guidelines [32,57].

Sample size: No previous studies reported UA levels in CCM and RRT patients. We adopted the rule of thumb suggested by Hertzog M.A. [59], who recommended a range of 10 to 40 per arm. Therefore, to avoid selection bias in our cohort, we enrolled all patients in our department who met the study’s inclusion criteria.

Statistical analysis: Continuous variables were reported as median and interquartile range (IQR) or mean ± standard deviation (SD), depending on their distribution. The categorical variables were described in terms of absolute numbers and percentages. We tested the normality of the variable distribution using the Shapiro–Wilk test. Student’s t-test, the Mann–Whitney U and the chi-square test were used to compare continuous and categorical variables as appropriate. Univariable and multivariable linear regression were used to evaluate the impact of covariates on UA levels after the normalization of variables through Log transformation. Specifically, in multivariable linear regression, we evaluated collinearities among covariates through the Variance Inflation Factor (VIF), considering a VIF of 2.5 as the threshold limit [60].

Statistical significance was assessed by a two-tailed test with a p ≤ 0.05.

Statistical analysis was performed with the SPSS software version 28.

3. Results

We enrolled 98 patients, with a median age of 80 years (IQR, 77–83), of whom 63 (64.3%) were male.

Table 1. Characteristics of patients.

Variable
Age (years)	80 (77–80)
Treatment vintage (months)	12 (8–15)
BMI (kg/m2)	24.3 (22.1–27)
Male	63 (64.3%)
Diabetes	48 (49%)
Hypertension	97 (99%)
Previous acute myocardial ischemia	42 (42.9%)
History of cancer	33 (33.7%)
Peripheral vascular disease	37 (37.8%)
Previous stroke	17 (17.3%)
COPD	9 (9.2%)
Potassium Binder	28 (28.6%)
Phosphate Binder	62 (63.3%)
Loop diuretics	87 (88.8%)
Thiazide diuretics	6 (6.1%)
Antialdosterone diuretics	12 (12.2%)
Allopurinol	62 (63.3%)
Allopurinol dose (mg/day)	75 (0–100)
Uric acid (mmol/L)	0.35 (±0.1)
Creatinine (µmol/L)	545 (±164)
Urea (mmol/L)	23.6 (±6.8)
Albumin (g/dL)	3.6 (3.3–4)
Sodium (mmol/L)	138 (136–140)
Potassium (mmol/L)	4.4 (4–4.9)
Hemoglobin (g/L)	10.9 (10.1–11.8)
Calcium (mmol/L)	2.3 (2.2–2.4)
Phosphate (mmol/L)	1.37 (1.26–1.74)
PTH (ng/L)	200 (135–331)
Bicarbonate (mmol/L)	24.2 (±4)
Urine output (cc/day)	1400 (950–1700)

Footnotes: Categorical variable reported as number (perceptual value), normally distributed variable reported as mean (standard deviation), non-normally distributed variable reported as median (Interquartile range), BMI body mass index, COPD chronic obstructive pulmonary disease, PTH parathyroid hormone.

reports all clinical and laboratory results for the entire population.

3.1. Comparison According to the Type of ESKD Treatment

In our cohort of patients, 45 (45.9%) received CCM, 33 (33.7%) were HD patients and 20 (20.4%) were PD patients. CCM patients were slightly older in age, with a higher prevalence of female gender and a history of cancer disease compared to dialysis patients. In contrast, HD and PD had higher creatinine and phosphate levels and lower urine output.

Specifically, among CCM group patients, 10 (22.2%) followed a low-protein diet (0.6 g/kg/day) and 35 followed a very low-protein diet (0.4–0.5 g/kg/day) supplemented with keto-analogs. In contrast, all HD and PD patients followed a free diet with suggested protein intake of 1.2 g/kg/day.

UA levels were slightly higher in CCM patients, with a mean of 0.37 mmol/L, compared with those in the dialysis group: HD patients had a mean UA of 0.33 mmol/L (p = 0.08), and PD patients had a mean UA of 0.33 mmol/L (p = 0.093). Allopurinol use did not differ significantly among the three treatment types. However, HD patients had significantly lower allopurinol doses than CCM (p = 0.03) and PD (p = 0.048) patients. At the same time, CCM patients received furosemide more frequently than HD patients (p = 0.02), and PD patients received metolazone more frequently than CCM and HD patients (p < 0.001) to optimize urine output. Finally, CCM had a mean creatinine clearance of 8.5 (±3.1) mL/min/1.73 m².

Table 2. Characteristics of patients with CCM, HD and PD.

Variable	CCM °	HD ^	PD *	p
Age (years)	81 (79.5–85)	78 (75–78)	78 (75.2–82)	°^ 0.08, °* 0.09, ^* 0.87
Treatment vintage (months)	10 (9.5–18)	14 (10–16)	12 (8–24)	°^ 0.34, °* 0.25, ^* 0.43
BMI (kg/m2)	25.3 (22.6–27.9)	23.1 (21.3–26.9)	24.4 (22.3–25.5)	°^ 0.09, °* 0.31, ^* 0.52
Male	24 (53.3%)	22 (66.7%)	17 (85%)	°^ 0.24, °* 0.015, ^* 0.14
Diabetes	24 (53.3%)	18 (54.5%)	6 (30%)	°^ 0.91, °* 0.006, ^* 0.08
Hypertension	44 (97.8%)	33 (100%)	20 (100%)	°^0.39, °* 0.5, ^* 0.99
Previous Acute myocardial ischemia	19 (42.2%)	13 (39.4%)	10 (50%)	°^ 0.8, °* 0.56, ^* 0.45
History of cancer	23 (51.1%)	7 (21.2%)	3 (15%)	°^ 0.007, *° 0.006, ^* 0.57
Peripheral vascular disease	13 (28.9%)	16 (48.5%)	8 (40%)	°^0.06, °* 0.38, ^* 0.48
Previous stroke	7 (15.6%)	9 (27.3%)	1 (5%)	°^ 0.2, °* 0.23, ^* 0.04
COPD	4 (6.7%)	3 (9.1%)	3 (15%)	°^ 0.69, °* 0.28, ^* 0.51
Potassium Binder	15 (33.3%)	5 (15.2%)	8 (40%)	°^ 0.07, °* 0.6, ^* 0.04
Phosphate Binder	22 (48.9%)	27 (81.8%)	13 (65%)	°^ 0.03, °* 0.23, ^* 0.17
Loop diuretics	43 (95.6%)	26 (78.8%)	18 (90%)	°^ 0.02, °* 0.39, ^* 0.29
Thiazide diuretics	0 (0%)	0 (0%)	6 (30%)	°^ 0.22, °* <0.001, ^* <0.001
Antialdosterone diuretics	4 (8.9%)	5 (15.2%)	3 (15%)	°^ 0.39, °* 0.46, ^* 0.98
Allopurinol	28 (62.2%)	19 (57.6%)	15 (75%)	°^ 0.68, °* 0.19, ^* 0.11
Allopurinol dose (mg/day)	75 (0–100)	50 (0–100)	75 (12.5–137.5)	°^ 0.03, °* 0.77, ^* 0.048
Uric acid (mmol/L)	0.37 (±0.1)	0.33 (±0.1)	0.33 (±0.08)	°^ 0.08, °* 0.09, ^* 0.92
Creatinine (µmol/L)	460 (±147)	598 (±129)	646 (±164)	°^ <0.001, °* <0.001, ^* 0.27
Urea (mmol/L)	20.5 (17.9–25.4)	26 (22.1–32.4)	22.1 (20.5–23.9)	°^ <0.001, °* 0.56, ^* 0.24
Albumin (g/dL)	3.8 (3.4–4.1)	3.6 (3.3–4)	3.2 (2.9–3.5)	°^ 0.26, °* 0.001, ^* 0.01
Sodium (mmol/L)	138 (137–140.5)	137 (134–140)	138 (136–142)	°^ 0.03, °* 0.71, ^* 0.17
Potassium (mmol/L)	4.3 (4–4.8)	4.9 (4.2–5.3)	4.3 (3.8–4.7)	°^ 0.08, °* 0.66, ^* 0.09
Hemoglobin (g/L)	11.1 (10.5–11.8)	10.6 (9.8–11.3)	11.2 (10.2–11.9)	°^ 0.14, °* 0.57, ^* 0.08
Calcium (mmol/L)	2.3 (2.2–2.4)	2.25 (2.1–2.3)	2.3 (2.17–2.4)	°^ 0.07, °* 0.8, ^* 0.09
Phosphate (mmol/L)	1.3 (1.17–1.55)	1.47 (1.27–1.84)	1.68 (1.45–1.97)	°^ 0.02, °* <0.001, ^* 0.07
PTH (ng/L)	218 (129–403)	212 (151–329)	145 (82.5–223)	°^ 0.89, °* 0.026, ^* 0.02
Bicarbonate (mmol/L)	24 (±4.4)	22.8 (±4)	25.1 (±2.7)	°^ 0.56, °* 0.33, ^* 0.15
Urine Output (cc/day)	1600 (1350–1800)	1000 (650–1500)	1000 (462–1500)	°^ <0.001, °* <0.001, ^* 0.32

Footnotes: Categorical variable reported as number (perceptual value), normally distributed variable reported as mean (standard deviation), non-normally distributed variable reported as median (Interquartile range), BMI body mass index, COPD chronic obstructive pulmonary disease, PTH parathyroid hormone. °^ CCM versus HD comparison, °* CCM versus PD comparison, ^* HD versus PD comparison.

presents the primary clinical and laboratory characteristics by ESKD treatment type.

3.2. Uric Acid Predictors in ESKD Patients

In ESKD patients, the univariable linear regression analysis identifies only BMI and treatment type as predictors of UA levels. Although potassium levels, allopurinol, Thiazide diuretics and potassium binders show a trend toward statistical significance.

Table 3. Univariable linear regression analysis in the entire population.

Covariate	B	p	95% CI
Type of treatment	−0.27	0.05	−0.54 −0.0001
Diabetes	0.004	0.89	−0.38 0.046
Age	0.004	0.13	−0.001 0.008
BMI	−0.05	0.038	−0.009 −0.0001
Creatinine	−0.00007	0.25	−0.00002 0.00005
Urea	0.001	0.41	−0.002 0.004
Albumin	0.003	0.87	−0.036 0.043
Sodium	0.001	0.72	−0.005 0.007
Potassium	−0.029	0.078	−0.06 0.003
Phosphate	0.01	0.74	−0.05 0.07
Bicarbonate	−0.001	0.72	−0.008 0.005
Urine output	0.023	0.19	−0.012 0.058
Allopurinol dose	−0.0001	0.079	−0.0001 0.000
Loop Diuretics	−0.015	0.65	−0.08 0.052
Thiazide Diuretics	0.006	0.13	−0.08 0.09
Antialdosterone Diuretics	0.018	0.59	−0.047 0.08
Potassium Binders	−0.034	0.16	−0.08 0.015
Phosphate Binders	0.047	0.65	−0.034 0.054

Footnotes: B, regression coefficient; BMI stands for body mass index.

details all the B coefficients in the prediction of UA.

Therefore, in our series, only BMI was an independent predictor of UA. Specifically, BMI was inversely associated with UA. Nevertheless, the type of treatment for kidney failure approached statistical significance (p = 0.054).

Table 4. Multivariable linear regression analysis in the entire population.

Covariate	Beta	p	VIF
Type of treatment	−0.19	0.054	1.003
BMI	−0.22	0.03	1.009
Potassium	−0.16	0.11	1.007
Allopurinol dose	−0.14	0.16	1.006

Footnotes: Beta, standardized regression coefficient; VIF, variance inflation factor; BMI, body mass index.

reports multivariable details.

3.3. Uric Acid Predictors in CCM Patients

In CCM patients, BMI, residual kidney function and 24 h urea excretion were significantly predictive of UA levels. At the same time, phosphate binder treatment showed a trend towards statistical significance.

Table 5. Univariable regression analysis in CCM patients.

Covariate	B	p	95% CI
Prescribed diet	0.27	0.15	−0.64 0.099
Diabetes	−0.12	0.72	−0.81 0.56
Age	0.001	0.72	−0.07 0.09
BMI	−0.006	0.028	−0.011 −0.01
Creatinine Clearance	−0.015	0.029	−0.028 −0.002
Urea	−0.001	0.86	−0.006 0.005
Albumin	0.026	0.43	−0.039 0.091
Sodium	0.001	0.85	−0.009 0.011
Potassium	0.014	0.51	−0.042 0.07
Phosphate	0.054	0.32	−0.054 0.163
Bicarbonate	−0.001	0.79	0.009 0.007
U-Urea/day	0.021	0.003	0.008 0.034
Urine output	0.036	0.32	−0.036 0.11
Allopurinol dose	−0.15	0.31	−0.001 0.0001
Loop Diuretics	0.011	0.89	−0.154 0.176
Antialdosterone Diuretics	0.018	0.76	−0. 0.138
Potassium Binder	−0.039	0.28	−0.11 0.33
Phosphate Binder	0.058	0.08	−0.008 0.12
Keto-analogs	0.014	0.73	−0.68 0.096

Footnotes: B, regression coefficient; BMI stands for body mass index.

details all the B coefficients in the prediction of UA.

In multivariable regression analysis, only BMI and 24 h urine urea excretion were independent predictors of UA levels, as reported in

Table 6. Multivariable linear regression analysis in CCM patients.

Covariate	Beta	p	VIF
BMI	−0.52	0.008	1.24
Creatinine Clearance	0.05	0.79	1.17
U-Urea/day	0.37	0.038	1.09
Phosphate Binder	0.28	0.13	1.19

Footnotes: Beta, standardized regression coefficient; VIF, variance inflation factor; BMI, body mass index.

.

3.4. Uric Acid Predictors in HD Patients

In HD patients, univariable linear regression analysis revealed a significant association between UA levels and protein intake recommendations, potassium levels and the presence of allopurinol in pharmacological therapy. In contrast, diuretics showed a trend towards a significant relationship with UA levels.

Table 7. Univariable regression analysis in HD patients.

Covariate	B	p	95% CI
Recommended protein intake	0.178	0.067	−0.013 0.37
Diabetes	0.005	0.9	−0.71 0.81
Age	0.003	0.4	−0.005 0.011
BMI	−0.001	0.85	0.013 0.11
Urea	0.005	0.58	0.000 0.009
Albumin	−0.27	0.43	−0.096 0.042
Sodium	0.001	0.89	−0.11 0.13
Potassium	−0.065	0.008	−0.112 0.018
Phosphate	0.027	0.6	0.077 0.13
Bicarbonate	−0.001	0.96	−0.64 0.62
Urine output	0.029	0.48	−0.054 0.112
Allopurinol dose	−0.39	0.02	−0.002 −0.0001
Loop Diuretics	−0.064	0.156	−0153 0.026
Antialdosterone Diuretics	0.079	0.124	−0.023 0.18
Potassium Binder	0.032	0.54	−0.137 0.073
Phosphate Binder	−0.06	0.21	−0.15 0.036

Footnotes: B, regression coefficient; BMI stands for body mass index.

details all the B coefficients in the prediction of UA.

In HD patients, the suggested protein intake, potassium levels and allopurinol therapy were independent predictors of UA levels, as reported in

Table 8. Multivariable linear regression analysis in HD patients.

Covariate	Beta	p	VIF
Recommended protein intake	0.34	0.021	1.016
Potassium	−0.046	0.008	1.011
Allopurinol dose	−0.39	0.009	1.019

Footnotes: Beta, standardized regression coefficient; VIF, variance inflation factor.

.

3.5. Uric Acid Predictors in PD Patients

In PD patients, a statistically significant association was observed only for the suggested protein intake. On the contrary, the urea levels showed a trend towards statistical significance. Given their VIF exceeding 2.5 in the multivariable analysis, no further analysis was conducted.

Table 9. Univariable regression analysis in PD patients.

Covariate	B	p	95% IC
Recommended protein intake	0.297	0.036	0.22 0.57
Diabetes	0.024	0.56	−0.06 0.11
Age	0.0001	0.94	−0.012 0.013
BMI	−0.002	0.73	−0.014 0.01
Urea	0.007	0.09	−0.001 0.015
Albumin	−0.051	0.22	−0.137 0.035
Sodium	−0.002	0.7	−0.014 0.01
Potassium	−0.025	0.5	−0.1 0.051
Phosphate	0.028	0.65	−0.097 0.152
Bicarbonate	0.01	0.93	−0.015 0.016
Urine output	−0.024	0.42	−0.086 0.031
Allopurinol dose	−0.14	0.56	−0.001 0.0001
Loop Diuretics	0.026	0.67	−0.1 0.15
Thiazide Diuretics	0.04	0.31	−0.04 0.122
Antialdosterone Diuretics	−0.057	0.27	−0.16 0.048
Potassium Binder	−0.049	0.18	−0.124 0.26
Phosphate Binder	0.035	0.35	−0.044 0.115

Footnotes: B, regression coefficient; BMI stands for body mass index.

details all the B coefficients in the prediction of UA.

4. Discussion

To the best of our knowledge, the present study is the first to suggest the impact of ESKD treatment modalities on UA levels.

Notably, patients receiving CCM had slightly higher UA levels than those undergoing RRT. The findings indicate that covariates affect UA control differently across ESKD treatment strategies. In the overall cohort, UA levels were mainly influenced by both BMI and CKD therapy. For CCM patients, adherence to a low-protein diet and BMI were the only independent predictors of UA levels. Among HD patients, potassium levels and allopurinol treatment were negatively associated with UA, whereas recommended dietary protein intake was positively associated with UA. In PD patients, UA levels were significantly influenced only by recommended dietary protein intake and urea levels.

4.1. Analysis of the Whole Population

The study population demonstrated considerable variability across treatment types in age, comorbidities, residual renal function, the development of CKD complications and the need for pharmacological intervention. This heterogeneity was expected across CCM, HD and PD groups. The choice of the treatment is strictly related to the etiology of kidney disease [61,62,63,64,65], the rate of CKD progression [66,67,68], comorbidities [69,70,71,72,73], residual renal function [74,75] and the balance between risk and benefit [76,77,78]. However, the choice among CCM, RRT, HD and PD is also influenced by factors such as age, performance status, the presence of absolute or relative contraindications and patient preferences, as outlined in clinical guidelines [32,79].

On the other hand, the type of treatment significantly influences the development of clinical symptoms and metabolic control in CKD. In this context, most differences observed among our groups are consistent with prior research. CCM patients tend to be older and have more comorbidities compared to RRT patients [80]. Uncontrolled diabetes often acts as a relative contraindication for PD due to its adverse effect on glycemic regulation [81]. Active urine output and diuretic use vary depending on treatment type [82,83,84,85]. Potassium binders are generally used more frequently in CCM patients [86], whereas phosphate binders are more commonly used in dialysis patients [87,88]. Some of these factors are strongly associated with UA levels, such as age [89] and the primary kidney disease [90,91], but they are also closely associated with the therapeutic strategy.

Across the entire population, treatment type and BMI were predictors of UA control, whereas potassium levels and allopurinol dose were nearly statistically significant. No previous report has compared ESKD treatment types; however, various studies have reported the effects of HD [92], PD [93] and a low-protein diet [94,95,96,97,98] on UA levels. The association between BMI and UA is consistent with previous reports, which have shown an increase in hyperuricemia with increasing BMI [99,100]. As reported in mouse models, adipose tissue enhances UA production via xanthine oxidoreductase activity [101], which may explain the association between higher BMI and elevated UA levels in our and other reports. Although our series did not show a significant impact of Allopurinol doses on UA levels, other reports support its value in ESKD [102,103,104,105]. Given the results of our univariable and multivariable analyses, this is likely due to the small sample size of our exploratory report. Considering the explorative nature of our analysis, we need additional studies to confirm our findings.

4.2. Analysis of the CCM Patients

The higher UA in CCM is likely attributable to the dialysis advantage in UA removal. Certainly, HD and PD procedures have a positive impact on UA balance [45,46,47,48], reducing UA levels modestly but significantly through dialytic clearance. However, CCM is associated with other conditions that could worsen UA control. Specifically, CCM patients were older, had a higher prevalence of non-active cancer history and used phosphate binders less frequently compared to RRT patients. Older age and lower use of phosphate binders may adversely affect UA levels, as reported in previous studies showing that UA increases with age [106,107] and that phosphate binders lower UA [108,109]. While no reports about the impact of inactive previous cancer were reported, its higher prevalence could be related to the preference of CCM patients for this treatment strategy.

Furthermore, our analyses demonstrated that both BMI and protein intake (estimated by the Maroni equation) were independent predictors of UA levels in CCM. This indicates that a restricted dietary protein intake has an overall effect nearly as effective as dialysis in ESKD. Our results across the entire population support this idea, showing that higher protein intake was significantly associated with higher UA levels. In RRT patients, dietary recommendations regarding meat consumption influence amino acid absorption, specifically cysteine, phenylalanine, glutamine, glycine and threonine, which increase purine metabolism and are linked to higher UA levels [1,57]. According to these preliminary results, we plan to investigate whether increased protein intake may diminish the beneficial effects of dialysis clearance. A significant association between BMI and UA levels is also observed in CCM. In these patients, protein restriction and the consequent reduction in the diet’s impact on UA may make the detrimental effects of adipose tissue on purine metabolism more evident, thereby increasing UA levels.

Surprisingly, in our series, creatinine clearance was not an independent predictor of UA control. The unexpected lack of a significant impact of kidney function on UA may be related to the small sample size.

4.3. Analysis of the HD Patients

HD patients exhibited a distinct profile compared to CCM and PD patients. Specifically, they had lower use of potassium binders, lower calcium levels and markedly lower allopurinol doses. The different use of potassium binders is strictly related to the higher potassium levels in this patient group. However, given mild hyperkalemia and the three weekly evaluations, the reduced use of potassium binders appears reasonable as a clinical approach. The statistically significant difference in calcium levels is not clinically relevant. Interestingly, HD patients received lower doses of allopurinol. This finding can be attributed to HD’s superior performance in UA removal, as evidenced not only compared to CCM but also relative to PD, as reported by Diez-Lopez et al. [110]. In our multivariable analysis of patients with HD, allopurinol dose, potassium levels and dietary recommendations were independent predictors of UA levels.

In HD patients, who are rarely treated with potassium binders, a higher intake of fruits and vegetables results in higher potassium [1,111,112] and vitamin [1] levels, as well as a higher fiber content [1,113,114,115], which has a positive effect on UA in HD patients [116]. This observation could explain the inverse relationship between potassium and UA levels. Furthermore, as expected, a higher dose of allopurinol lowers the UA level more. Finally, the suggestion to increase protein intake in HD aligns with the observed increase in UA, as reported in further observational studies [117,118].

4.4. Analysis of PD Patients

PD patients exhibited a distinct profile compared with those with CCM and HD. Specifically, this class of patients was more often diabetic, were more often treated with thiazide diuretic, had lower albumin level and lower levels of PTH with worse phosphate control. The biochemical profile appears to be directly related to the dialysis process and corroborates previous reports. Specifically, several studies have reported lower albumin levels than in HD patients, likely due to substantial losses during exchange [119,120]. PD showed suboptimal phosphate removal; in particular, automated peritoneal dialysis (APD) was associated with poorer outcomes [121,122]. At the same time, the lower PTH level may be related to the effects of calcium in the dialysis solution [123,124] or to pharmacological treatment, which were not explored in our study. The higher prevalence of thiazide diuretic use in this patient group may reflect efforts to optimize treatment and preserve urine output, which plays a critical role in PD management. In univariable analysis, we observed a direct association between the suggested protein intake and UA levels. This finding aligns with a previous observation in HD patients, who also exhibit higher protein intake, which negatively affects the control of uremic toxins [125]. By contrast, the lack of effect of allopurinol on UA balance was unexpected [103,104,105]. This missing relationship is likely attributable to the small size of the PD group and its high prevalence of use, which reaches 75% among PD patients. In other words, the widespread use of allopurinol in a small group could mask its impact on UA.

The present study is not devoid of limitations. Firstly, the small sample size may reduce the statistical power of our univariable and multivariable analyses. However, the lack of prior reports on this topic prevented us from estimating an appropriate sample size. Consequently, we conducted this pilot study to assess the key factors influencing UA balance in ESKD. Secondly, adherence to dietary protein intake among dialysis patients was not directly measured, potentially creating a gap between recommended and actual intake. A three-day food diary in dialysis patients may help assess their actual dietary protein intake. Thirdly, we did not evaluate the dosage of all drugs, but the broad spectrum of drugs makes it hard to achieve this goal. We only considered the allopurinol dose, which is the primary determinant of UA in pharmacological treatment. In our report, we did not assess the prevalence of allopurinol allergy or adverse reactions to allopurinol, which hinders its use in patients with uncontrolled UA [126,127]. Given the exploratory nature of the current study, we planned a future study with a sample size of 99 patients per arm. The planned research will involve the introduction of a three-day food diary, assessment of allopurinol allergy and evaluation of the primary pharmacological treatment.

5. Conclusions

This exploratory study on UA control underscores the significance of treatment type in patients with ESKD, particularly emphasizing the influence of diet in both CCM and RRT patients. This report constitutes a preliminary investigation into the relationship between UA control and ESKD treatment modalities, which may modulate the effects of individual risk factors on UA metabolism. Recognizing the varied patterns of hyperuricemia risk among ESKD patients is the first step for personalized medicine.