Serum uric acid level and all-cause and cardiovascular mortality in peritoneal dialysis patients: A systematic review and dose-response meta-analysis of cohort studies.

BACKGROUND: The association between serum uric acid (SUA) and all-cause and cardiovascular disease (CVD) mortality in peritoneal dialysis (PD) patients is controversial. Therefore, we aimed to determine the relationship between SUA and all-cause and CVD mortality in PD patients.
METHOD: Web of Science, EMBASE, PubMed and the Cochrane Library databases were searched from their inception to 7 April 2021. Effect estimates were presented as hazard ratios (HRs) with 95% confidence intervals (95% CIs) and pooled using random effects model. RESULT: Thirteen cohort studies with 22418 patients were included in this systematic review, of which 9 were included in the meta-analysis. Before switching the reference group, pooled result for the highest SUA category was significantly greater than the median for all-cause mortality (HR = 2.41, 95% CI: 1.37-4.26). After switching the reference group, the highest SUA category did not demonstrate an increased all-cause (HR = 1.40, 95% CI: 0.95-2.05) or CVD (HR = 1.30, 95% CI: 0.72-2.34) mortality compared with the lowest SUA category. Dose-response analysis suggested a nonlinear association between SUA and all-cause mortality risk (Pnonlinearity = 0.002).
CONCLUSION: This meta-analysis didn't find the relationship between SUA levels and all-cause and CVD mortality risk in PD patients. More rigorously designed studies are warranted in the future.

Introduction

Chronic kidney disease (CKD) is a worldwide public health problem with high incidence rate [1, 2] and high mortality [3], which have aggravated the burden of medical care. The consequences of CKD include cardiovascular disease, stroke, end-stage renal disease, renal replacement therapy (RRT) and kidney transplantation, all of which are serious and costly medical events [3]. The global prevalence of CKD is likely to rise further with the aging of the population and the increasing prevalence of diabetes, especially in China [4]. In the treatment of chronic kidney disease, dialysis is a conventional treatment method, including peritoneal dialysis (PD) and hemodialysis (HD). Compared with in-center HD, the use of PD is a more economical dialysis modality, which may potentially decrease infection risk, enhance patient satisfaction and preserve residual renal function, while having a comparable survival rate [5]. Due to the limited health-care resources, combined with the advantages of PD treatment and the support of government policies, PD has been widely used worldwide, including in China, Thailand and the United States [3, 6].

There are many factors that affect the survival rate of patients receiving dialysis treatment, including residual renal function, serum uric acid (SUA), and so on [7]. Uric acid (UA) is the final product in the liver from the degradation of dietary and endogenously synthesized purine or nucleotide compounds [8], about two-thirds of which is excreted by glomerular filtration [9]. Studies have found that UA is closely related to many chronic diseases. Epidemiological evidence demonstrated that higher UA concentration was a strong independent predictor of the incidence of type 2 diabetes mellitus (DM) [10, 11]. The possibility of gout arthritis development is correlated with the levels and the duration of SUA elevation [3]. Persistent hyperuricemia is closely related to cardiovascular, urolithiasis, thyroid dysfunction, psoriasis and hypertension [3]. In addition, research has shown that hyperuricemia is associated with all-cause and cardiovascular disease (CVD) mortality in CKD [12], HD [13] and PD patients [14, 15]. However, the evidence is conflicting.

Here we only focus on the relationship between SUA concentration and mortality (all-cause and CVD) in PD patients. Several studies showed that hyperuricemia was an independent risk factor for all-cause mortality in PD population [14-16], high level of SUA was associated with a high risk of CVD mortality in men treated with PD [15]. However, Lai et al. [17] found that there was an inverse association between the elevated SUA level and all-cause and CVD-associated mortality in women treated with continuous ambulatory PD. In addition, another study reported that hyperuricemia was weakly associated with all-cause and CVD mortality in PD patients [18]. Interestingly, different researchers have tried to explain the relationship between SUA levels and mortality from different directions. For example, different forms of SUA (time-averaged uric acid (TA-UA) [19] had been calculated at 3 months [16] or 6 months [20] after initiating PD; the longitudinal change in SUA [21]) were used to explore the association between mortality and SUA. Unfortunately, the impact of SUA on the survival of PD patients remains unclear.

To date, there has been no systematic review and dose-response meta-analysis to investigate the relationship between SUA and all-cause and CVD mortality in patients who had undergone PD. Hence, the objective of this study is to determine the association between SUA and all-cause and CVD mortality with a detailed analysis of eligible literature.

Materials and methods

Search strategy

One author (Kang) conducted systematic literature searches in electronic databases with no restriction on the language from their inception to 7 April 2021, including Web of Science, EMBASE, PubMed and the Cochrane Library databases. An initial literature search in the mentioned databases used such keywords as “Peritoneal dialysis” and “Uric acid”. The following search criteria were applied for PubMed: (("Peritoneal Dialysis"[Mesh]) OR (((Dialyses, Peritoneal) OR (Dialysis, Peritoneal)) OR (Peritoneal Dialyses))) AND (("Uric Acid"[Mesh]) OR (((((((((((((((Acid, Uric) OR (2,6,8-Trihydroxypurine)) OR (Trioxopurine)) OR (Potassium Urate)) OR (Urate, Potassium)) OR (Acid Urate, Ammonium)) OR (Urate, Ammonium Acid)) OR (Sodium Acid Urate)) OR (Urate, Monosodium)) OR (Urate, Sodium)) OR (Acid Urate, Sodium)) OR (Urate, Sodium Acid)) OR (Sodium Urate Monohydrate)) OR (Monohydrate, Sodium Urate)) OR (Urate Monohydrate))). Besides, the list of reference literature relevant was checked to identify additional eligible studies. All non-English studies were translated by software first, and then the researchers checked whether they could be included. Two reviewers (Kang and Hu) independently screened the abstracts to determine if they met the inclusion criteria, and disagreements were resolved through a third investigator (Huang). The searching strategies for the remaining databases are presented in the S1 Table.

Inclusion criteria

The literature was selected if they met all of the following criteria: (1) the interest design were case-control or cohort studies; (2) study objects and interventions: participants treated with PD with no gender, race, or nationality limitations imposed; (3) the outcomes of interest were all-cause mortality and CVD mortality; (4) literature from which hazard ratios (HRs) data or calculate HRs data could be extracted were included in this meta-analysis. (5) If there were duplicate publications, the one with the largest number of participants or the longest follow-up period was included.

Data extraction

The information was tabulated including the first author, year of publication, national/region where the research was conducted, study design, number of center, sample size, number of all-cause and cardiovascular deaths, follow-up duration, mean or median of SUA concentration, multi-factorial adjusted HRs and its 95% confidence intervals (CIs) of all-cause mortality or CVD mortality, adjusted covariates and quality evaluation information. Two authors (Kang and Hu) performed data extraction independently following the table contents. Discrepancies were resolved by discussion or a third investigator (Huang).

Quality assessment of included studies

The quality of literature was evaluated independently by two authors (Kang and Huang) using the Newcastle-Ottawa Scale (NOS) [22], and the disputes were conformed through discussion with Hu. There are 3 quality parameters of NOS, of which the study population selects parameters worth up to 4 points, comparability parameters worth up to 2 points, and exposure or outcome evaluation parameters worth up to 3 points. The full score of NOS is 9, and studies with scores of 0 ~ 3, 4 ~ 6, and 7 ~ 9 points are defined as low, medium, and high quality ones, respectively.

Data estimation

In this meta-analysis, SUA concentration was given in mg/dL. To use morbidity data to analyze possible dose-response relationships, it is necessary to have the following information for each SUA concentration category: assigned average or median SUA levels, deaths, follow-up person-years, adjusted HRs and 95% CIs. When the number of all-cause and CVD deaths in each subgroup was not directly available in published data, appropriate statistical methods were employed to estimate missing data using the total number of deaths, HRs and the total number of patients in each subgroup [23]. For each SUA category in each study, the "years of follow-up" was calculated by multiplying the number of patients in that SUA category by the median or average follow-up months and dividing by 12. When the HRs and 95% CIs reported in the original study were not based on the lowest SUA group as the reference, we recalculated the relevant HRs and 95% CIs using the lowest dose group as the reference by the method developed by Hamling et al. [24].

After EXCEL conversion, the number of pseudo-effective cases, effect size and 95% CIs were used to replace the data provided by the original literature for dose-response meta-analysis [25]. If other relevant information was not available, we contacted the corresponding author via email.

Statistical analysis

This meta-analysis was conducted using Stata software12.0 (version 12.0; Stata Corp, College Station, TX). We conducted a comparison between different SUA levels among included studies, including the highest SUA level compared with the lowest category of SUA level, and the highest / lowest category SUA level compared with the median SUA level. Given studies have reported several different possible relationships (U-shaped [19], inverse [17] or no relationship [18]) between SUA levels and all-cause or CVD mortality, we performed a dichotomy and dose-response meta-analysis. Multi-factor adjusted HRs and 95% CIs were extracted from the included studies and the pooled HRs was calculated using the lowest dose group as reference using the Mantel-Haenszel method developed by Hamling et al. [24]. A two-stage fixed-effects dose-response model was employed to explore the dose-response relationship between SUA levels and mortality in patients treated with PD. The potential linear or nonlinear relationship between SUA concentration and mortality (all-cause and CVD) was assessed using a restricted cubic spline regression model with 4 knots at fixed percentiles (5%, 35%, 65%, and 95%) of SUA concentration distribution [26]. The P value for curve linearity or nonlinearity was calculated by testing the null hypothesis that the coefficients of the second and third spline transformations were equal to zero. If P < 0.05, the null hypothesis was rejected and a nonlinear dose-response relationship was considered to exist. Otherwise, a linear regression model was considered.

The Q test and I2 statistics were used for heterogeneity analysis. If the included literature were considered to have no significant heterogeneity (P > 0.1 and I2 < 50%), the fixed-effects model was applied. When heterogeneity was considered acceptable (P < 0.1 or 50% ≤ I2 < 85%), the random effects model was used. When I2 > 85%, we considered that the results could not be pooled. Subgroup analysis was employed to explore the sources of heterogeneity. Sensitivity analysis, in which 1 study was removed at a time, was performed to evaluate the stability of the results. Egger’s test was used to analyze the possibility of publication bias [27]. P < 0.05 was defined as significant publication bias. It must be mentioned that we did not register for this meta-analysis, but we conducted this systematic review and meta-analysis in strict accordance with the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) statement.

Results

Literature search and selection

The PRISMA flowchart of the screening and selection process was summarized in Fig 1. A total of 971 references were identified to evaluate the relationship between SUA levels and mortality during the initial search but 429 articles were removed due to duplication. A further 523 literatures were excluded for not meeting the inclusion criteria after screening the title and abstract. Nineteen articles were conducted full text assessment, and 6 studies were excluded, including 3 studies [28-30] excluded for the reasons of part of participants treated with hemodialysis and 3 studies [31-33] excluded for exploring the cause of mortality in patients treated with PD. Finally, 13 cohort studies with 22418 patients were included in this systematic review, of which 9 [14, 16–20, 34–36] were included in the meta-analysis. A study [21] was excluded in the meta-analysis because they focused on whether the longitudinal change in SUA affected all- cause mortality (SUA decliner vs SUA non-decliner) and another literature [37] explored the relationship between SUA and PD treatment failure. In addition, 2 studies [15, 38] duplicated publications with Xia [34].

Fig 1

Flow diagram of PRISMA presenting the process of search and selection of studies.

PRISMA, Preferred Reporting Items for Systematic Reviews and Meta-Analyses.

Study characteristics

The main characteristics of the included literature were presented in Table 1. A total of 13 studies consisting of 22418 participants were included. All studies were published from 2013 to 2021 and were designed as cohort studies, of which 9 were retrospective cohort studies (RCS) and three [14, 18, 35] were multicenter. Except for two studies [35, 36], all the studies were conducted in China, two [17, 37] of which were carried out in Taiwan. The median or mean time of follow-up duration was 4.0 to 68.7 months. Five included studies explored the relationship between SUA and mortality by comparing the highest with the lowest SUA level, another four by comparing the lowest or highest with the intermediate level of SUA. The NOS score ranged from 7 to 9, and all included studies were considered to be of high quality (see S2 Table).

Table 1

Characteristics and quality of included studies.

Study	Design	Region	Subjects	Number of center	Follow-up duration (months)	Outcomes	Concentration range of the SUA categories	Adjustment for covariates	NOS score
Feng, 2013 [16]	RCS	China	156	Single center	31.3	All-cause mortality	Group 1: ≤ 7.0 mg/dL	Age, HTN, DM, serum albumin, CRP, phosphate, RRF and UA group	8
							Group 2: 7.0–10.0 mg/dL (Reference)
							Group 3: ≥ 10.0 mg/dL
Dong, 2014 [18]	PCS	China	2193	Multi-center	26.5	All-cause, CVD mortality	Men:	Age, RRF, SA, hemoglobin, phosphate, CRP, CVD, BMI, mean arterial pressure, LDL-C and center size	9
							Tertile 1: 2.09–5.79 mg/dL (Reference)
							Tertile 2: 5.80–7.38 mg/dL
							Tertile 3: 7.39–16.7 mg/dL
							Women:
							Tertile 1: 1.74–5.37 mg/dL (Reference)
							Tertile 2: 5.38–6.65 mg/dL
							Tertile 3: 6.66–8.08 mg/dL
Xia, 2014 [15]	PCS	China	985	Single center	25.3	All-cause, CVD mortality	Men:	Age, BMI, Davies comorbidity score, hemoglobin, SA, SC, albumin-corrected calcium, SP, total triglyceride, LDL-C; RRF; log-transformed high-sensitivity CRP, total Kt/V, use of allopurinol, ACE inhibitor, or angiotensin receptor blocker and loop diuretics	8
							Tertile 1: ≤ 6.67 mg/dL (Reference)
							Tertile 2: 6.67–7.56 mg/dL
							Tertile 3: > 7.56 mg/dL
							Women:
							Tertile 1: ≤ 6.19 mg/dL (Reference)
							Tertile 2: 6.19–7.13 mg/dL
							Tertile 3: > 7.13 mg/dL
Xia, 2016 [34]	PCS	China	1278 (diabetes:328	Single center	30.7	All-cause, CVD mortality	Diabetic men:	Non-diabetes: age, BMI, history of hypertension and CVD, hemoglobin, SA, SP, SC, HDL-C, RRF, log-transformed high-sensitive CRP, use of allopurinol and Drugs used of allopurinol, ACE inhibitor, or angiotensin receptor blocker;Diabetes: non-diabetic adjustment content and glycated hemoglobin	8
							Tertile 1: < 6.46 mg/dL (Reference)
							Tertile 2: 6.46–7.38 mg/dL
							Tertile 3: ≥ 7.38 mg/dL
							Non-diabetic men:
							Tertile 1: < 7.00 mg/dL (Reference)
							Tertile 2: 7.70–7.89 mg/dL
							Tertile 3: ≥ 7.89 mg/dL
							Diabetic women:
			non-diabetes:950				Tertile 1: < 5.89 mg/dL (Reference)
							Tertile 2: 5.89–7.09 mg/dL
							Tertile 3: ≥ 7.09 mg/dL
							Non-diabetic women:
							Tertile 1: < 6.46 mg/dL (Reference)
							Tertile 2: 6.46–7.48 mg/dL
							Tertile 3: ≥ 7.48 mg/dL
Hsieh, 2017 [37]	RCS	Taiwan, China	371	Single center	36.7	All-cause technique failure, peritonitis-related failure	Group 1: ≤ 8 mg/dL (Reference)Group 2: > 8 mg/dL	Gender, age, BMI, comorbid conditions, and the use of ACE inhibitor, ARB, β-blocker, CCB, hypouricaemic agents, diuretics, BUN, creatinine, HB, ferritin, HbA1c, SA, Ca×P, GPT, RRF, icodextrin use, Balance dialysate use, assistance for dialysate exchanges, peritoneal Kt/V, weekly total Kt/V urea, nPNA, D/P (creatinine) at 4 hours, ultrafiltration, 24-hour urine output, and exit-site infection, tunnel infection, number of exchanges per day and peritonitis rate	7
Lai, 2018 [17]	RCS	Taiwan, China	492	Single center	36.4	All-cause, CVD mortality	Men:	Age, sex, BMI, the pre-dialysis status, smoking status, medications (ACE, ARB, ESA, furosemide, vitamin D, statin, allopurinol, CCB), comorbidities (DM, hypertension, CVD, Charlson score), PD related parameters (weekly total Kt/V urea, nPNA, D/P creatinine at 4 h, ultrafiltration, 24-h urine output, RRF), laboratory data (BUN, creatinine, albumin, GPT, WBC, alkaline phosphate, HB, ferritin, TSC, triglyceride, PTH, calcium, phosphate)	8
							Tertile 1: ≤ 6.8 mg/dL (Reference)
							Tertile 2: 6.9–8.0 mg/dL
							Tertile 3: ≥ 8.1 mg/dL
							Women:
							Tertile 1: ≤ 6.5 mg/dL (Reference)
							Tertile 2: 6.6–7.6 mg/dL
							Tertile 3: ≥ 7.7 mg/dL
Zhang, 2018 [20]	RCS	China	1063	Single center	33.0	All-cause, CVD mortality	Group 1: < 7 mg/dL (Reference)	Age, Scr, P, Alb, BG, iPTH, history of DM, DBP, Charlson score	8
							Group 2: ≥ 7 mg/dL
Chang, 2019 [19]	RCS	China	300	Single center	22.6	All-cause mortality	Group 1: TA-UA < 6 mg/dL	Age, sex, DM, CVD, RRF, BMI, SBP, Hb, Alb, BUN, Cr, Na, K, CO2, cCa, P, LDL-C, CRP, RASi, diuretic	8
							Group 2: TA-UA 6–8 mg/dL (Reference)
							Group 3: TA-UA ≥ 8 mg/dL
Xiang, 2019 [14]	RCS	China	9045	Multi-center	29.4	All-cause, CVD mortality	Quintile 1: < 6.06 mg/dL	Age, sex, BMI, DM, CVD, RRF, hemoglobin, SA, serum potassium, serum natrium, SP, serum calcium, serum parathyroid hormone, SC, and fasting plasma glucose	7
							Quintile 2: 6.06–6.67 mg/dL
							Quintile 3: 6.68–7.27 mg/dL (Reference)
							Quintile 4: 7.28–8.03 mg/dL
							Quintile 5: ≥ 8.04 mg/dL
Chang, 2019 [21]	RCS	China	309	Single center	≥4.0	All-cause mortality	Group 1: SUA decliner	Gender, age, BMI, SBP, Hb, Na, K, Cl, BUN, Cr, CO2, Ca, P, ALB, TG, FBG, CRP, RRF, PET type, Kt/V, CCB, RASi, β-blocker, diuretic	7
							Group 2: SUA non-decliner
Xiao 2020 [38]	RCS	China	802	Single center	68.7	All-cause mortality	Group 1: > 7 mg/dLGroup 2: ≤ 7 mg/dL (Reference)	Age, gender, Charlson comorbidity score, PD vintage, total Kt/V, using of angiotensin-converting enzyme inhibitor or angiotensin II receptor blocker, using of diuretic, using of uric acid-lowering agent, total cholesterol, high-density lipoprotein cholesterol, neutrophil to lymphocyte ratio, intact parathyroid hormone, ECW/TBW ratio ≥0.4, ASMI groups, and ASMI groups × SUA, serum albumin	8
Sugano 2020 [35]	PCS	Japan	4742	Multi-center	12.0	All-cause mortality	Group 1: < 5.0 mg/dL	Age, sex, dialysis duration, BMI, UV, use of ULT, diabetes, history of acute myocardial infarction, cerebral hemorrhage and cerebral infarction comorbid disease, and laboratory data including BUN, Cr, albumin, CRP, and Hb	8
							Group 2: 5.0 to < 5.5 mg/dL
							Group 3: 5.5 to < 6.0 mg/dL
							Group 4: 6.0 to < 6.5 mg/dL
							Group 5: 6.5 to < 7.0 mg/dL
							Group 6: 7.0 to < 7.5 mg/dL (Reference)
							Group 7: 7.5 to < 8.0 mg/dL
							Group 8: 8.0 to < 8.5 mg/dL
							Group 9: ≥ 8.5 mg/dL
Coelho 2020 [36]	RCS	Portugal	682	Single center	31.4	All-cause mortality	Not reported	Age, diabetes, comorbidity and baseline residual kidney function	7

NOS, Newcastle-Ottawa Scale; RCS, retrospective cohort study; PCS, prospective cohort study; DM, diabetic mellitus; CVD, cardiovascular disease; RRF, residual renal function; BMI, body mass index; HTN, underlying hypertensive nephropathy; UA, uric acid; Alb, albumin; BUN, blood urea nitrogen; LDL-C, low-density lipoprotein cholesterol; HDL-C, high-density lipoprotein cholesterol; CRP, C-reactive protein; SA, serum albumin; SC, serum creatinine; SP, serum phosphorus; ACE, angiotensin-converting enzyme; ARB, inhibitors/angiotensin II receptor blocker; ESA, erythropoiesis stimulating agents; CCB, calcium channel blocker; nPNA, normalized protein nitrogen appearance; GPT, glutamic-pyruvic transaminase; WBC, white blood cell counts; PTH, intact parathyroid hormone; TSC, transferrin saturation, cholesterol, HB, hemoglobin; PD, peritoneal dialysis; ECW/TBW, extracellular water/total body water; ASMI, appendicular skeletal muscle mass index; SUA, serum uric acid; UV, urinary volume; D/P, dialysate-to-plasma; RASi, renin-angiotensinsystem inhibitor; Kt/V, urea clearance index; DBP, diastolic blood pressure; FBG, fasting blood glucose; SBP, systolic blood pressure; TG, triglyceride; PET, peritoneal equilibration test; ULT, urate-lowering treatment.

Different SUA forms with all-cause and CVD mortality

UA values at different time points were used to explore the relationship with all-cause and CVD mortality. Eight studies [14–18, 34, 36, 38] used UA within 3 months of PD as a baseline to explore the relationship with mortality. Zhang et al. [20] compared the effect of high UA (≥ 420μmol/L) with normal UA (≤ 420μmol/L) on mortality after 6 months of PD. In addition, in a recently published study [19], the average time UA was calculated to investigate the association between TA-UA and all-cause mortality in PD patients, taking into account that UA concentration was easily affected by dialysis efficiency, diet, and medication. Furthermore, the effect of SUA decline and non-decline during PD on mortality has been studied by Chang et al. [21], showing that the decline of SUA meant a higher risk of all-cause mortality. In contrast, Feng et al. [16] found that unchanged SUA levels were associated with greater risk of death. A long-term observational cohort study [37] reported that the rate of all-cause technical failure was significantly higher in the hyperuricemia group than in the normal uricemia group. Hyperuricemia was an independent risk factor for a higher risk of all-cause technical failure, which meant higher risk of death.

Impact of UA on mortality in male and female treated with PD

The adverse effect of hyperuricemia on all-cause mortality was more prominent in the men’s group, but there was no significant difference between male and female [14] on CVD mortality. For patients treated with continuous ambulatory PD, elevated UA levels were associated with reduced all-cause and CVD mortality in women, while there were no significant differences in men [17]. Hyperuricemia, however, is an independent predictor of all-cause and CVD mortality in males treated with PD [15].

Impact of UA in PD population with diabetes or non-diabetes

The adverse effect of hyperuricemia on all-cause mortality was more prominent in patients without DM, while the effect was not significant between diabetes and non-diabetes subgroups for CVD mortality [14]. Another study showed that diabetes mellitus combined with peritoneal dialysis was an independent risk factor for death from cardiovascular events [20]. A multi-center cohort study of 2264 PD participants (37.7% of whom had DM) showed that each 1 mg/dL increase in SUA predicted a 10% increase in the CVD mortality rate in DM patients and a 12% increase in non-diabetic patients [18]. Elevated SUA was an independent risk factor for CVD mortality in males treated with PD, as well as predicted a higher risk of all-cause mortality in non-diabetic males, but not for females in predicting the risk of all-cause and CVD mortality [34].

Relationship between SUA and all-cause mortality (before recalculated the HRs and 95% CIs)

High vs low

Five studies [17, 18, 20, 34, 36] reported HRs and 95% CIs of all-cause mortality for the highest SUA category compared with the lowest. As presented in Fig 2a, all-cause mortality (HR = 1.19, 95% CI = 0.82–1.71, I2 = 81.3%) was not significantly elevated compared with the lowest category of patients with PD. No obvious publication bias was found (t = -0.23, P = 0.83).

Fig 2

Forest plots for relationship between SUA and all-cause mortality in PD patients.

(a) the highest SUA category versus the lowest. (b) the highest SUA category versus median. (c)the lowest SUA category versus median. SUA, serum uric acid; PD, peritoneal dialysis.

Sensitivity analysis

The combined HR changed (HR = 1.39, 95% CI: 1.08–1.79) and the heterogeneity was reduced after excluding Lai et al. [17] from the meta-analysis (from I2 = 81.3% to I2 = 56.9%), which could explain the part source of heterogeneity (see S1 Fig).

High or low vs median

Four researches [14, 16, 19, 35] reported HRs and 95% CIs of all-cause mortality for the highest or lowest SUA category compared with the median. As shown in Fig 2b, the pooled result for the highest SUA category was significantly greater than the median (HR = 2.41, 95% CI: 1.37–4.26, I2 = 72.0%). No significant publication bias was found (t = 3.62, P = 0.07). But the lowest versus median levels of SUA were not associated with the all-cause death risk (HR = 1.45, 95% CI: 0.96–2.18, I2 = 58.9%, Fig 2c). No significant publication bias was found (t = 0.98, P = 0.43).

The results of sensitivity analysis confirmed the stability of our result (see S2 and S3 Figs).

Relationship between SUA and CVD mortality (before recalculated the HRs and 95% CIs)

Four studies [17, 18, 20, 34] reported HRs and 95% CIs of CVD mortality for the highest SUA category compared with the lowest. As shown in Fig 3, we did not find any association of SUA level and CVD mortality (HR = 1.48, 95% CI: 0.80–2.74, I2 = 79.6%). No obvious publication bias was found (t = -0.05, P = 0.96).

Fig 3

Forest plot for relationship between SUA and cardiovascular mortality in PD patients.

The highest SUA category versus the lowest. SUA, serum uric acid; PD, peritoneal dialysis.

The combined HR changed (HR = 1.93, 95% CI: 1.39–2.68) and the heterogeneity was reduced after excluding Lai et al. [17] from the meta-analysis (from I2 = 79.6% to I2 = 13.9%), which could explain the part source of heterogeneity (see S4 Fig).

Only one research [14] reported HRs and 95% CIs of CVD mortality for the highest (HR = 1.14, 95% CI: 0.79–1.67) or lowest (HR = 1.17, 95% CI: 0.82–1.66) SUA category compared with the median.

Relationship between SUA and all-cause mortality (after recalculated the HRs and 95% CIs)

The pooled HR and 95% CI of all-cause mortality comparing the highest versus the lowest category was 1.40 (95% CI: 0.95–2.05), presented in Fig 4. There was no significant difference in all-cause mortality between the highest and lowest subgroup of SUA level, which was consistent with the result before we recalculated. No obvious publication bias was found (t = 0.53, P = 0.62), but there was significant heterogeneity (I2 = 83.1%, P < 0.001) between the studies.

Fig 4

Forest plot about the relationship between SUA and all-cause mortality in PD patients.

SUA, serum uric acid; PD, peritoneal dialysis.

Subgroup analysis

Significant associations were found for subgroups by study design (PCS or RCS), number of center (multi-center or single center), publication years (2013–2016 or 2017–2020), sample size (< 900 or > 900), follow-up duration (< 30 months or > 30 months), male proportion (< 50% or ≥ 50%), or whether the results were adjusted for diabetes status and BMI (see Table 2).

Table 2

Subgroup analysis of the relationship between serum uric acid and all-cause mortality.

	Serum uric acid
	Number of study	HR (95% CI)	Heterogeneity (I2)
Study design
Prospective cohort study	4	1.40(1.04, 1.88)	46.0%
Retrospective cohort study	4	1.44(0.57, 3.61)	91.5%
Number of center
Multi-center	3	1.23(1.04, 1.45)	0.0%
Single center	5	1.63(0.70, 3.83)	90.0%
Publication years
2013–2016	4	2.08(1.16, 3.73)	82.4%
2017–2020	4	0.93(0.55, 1.56)	80.4%
Sample size
< 900	4	1.50(0.50, 4.54)	91.6%
> 900	4	1.33(1.05, 1.69)	46.0%
Follow-up duration(months)
< 30	4	1.23(1.05, 1.45)	0.0%
> 30	4	1.67(0.63, 4.45)	92.5%
Male (%)
≥ 50%	6	1.75(1.17, 2.62)	76.7%
< 50%	2	0.73(0.27, 2.02)	90.6%
Adjust for diabetes
Yes	5	1.33(0.69, 2.53)	88.8%
No	3	1.53(1.08, 2.18)	48.9%
Adjust for BMI
Yes	7	1.18(0.86, 1.62)	73.9%
No	1	6.02(2.93, 12.37)	-

HR, hazard ratio; CI, confidence interval; BMI, body mass index.

The combined HR changed (HR = 1.62, 95% CI: 1.17–2.24) and the heterogeneity was reduced after excluding Lai et al. [17] from the meta-analysis (from I2 = 83.1% to I2 = 73.0%), which could explain the part source of heterogeneity (see S5 Fig).

Dose-response analysis

The dose-response relationship between SUA and all-cause mortality was analyzed by random effect nonlinear model. Fig 5 reveals the nonlinear dose-response relationship between SUA concentration and all-cause mortality in PD population, suggesting higher SUA level was associated with increasing all-cause mortality (Pnonlinearity = 0.002).

Fig 5

Dose-response relation between SUA concentration and all-cause mortality in PD patients.

The solid line and the dash line represent the estimated hazard risk and its 95% confidence interval. SUA, serum uric acid; PD, peritoneal dialysis.

Relationship between SUA and CVD mortality (after recalculated the HRs and 95% CIs)

There was no significant difference in CVD mortality between the highest and lowest subgroups of SUA (HR = 1.30, 95% CI: 0.72–2.34) (Fig 6), which was consistent with the result (the high vs low) before we recalculated. No obvious publication bias was found (t = 0.72, P = 0.52), but there was significant heterogeneity (I2 = 80.8%, P < 0.001) between the studies.

Fig 6

Forest plot about the relationship between SUA and cardiovascular mortality in PD patients.

SUA, serum uric acid; PD, peritoneal dialysis.

As shown in Table 3, significant associations were found for subgroups by study design, publication years, or whether the results were adjusted for diabetes status. Non-significant associations of SUA level with CVD mortality were detected in the subgroup analysis conducted by number of center, sample size, follow-up duration, and male proportion.

Table 3

Subgroup analysis of the relationship between serum uric acid and cardiovascular mortality.

	Serum uric acid
	Number of study	HR (95% CI)	Heterogeneity (I2)
Study design
Prospective cohort study	3	2.06(1.27, 3.34)	38.7%
Retrospective cohort study	2	0.70(0.33, 1.49)	77.0%
Number of center
Multi-center	2	1.07(0.78, 1.45)	0.0%
Single center	3	1.45(0.44, 4.79)	89.5%
Publication years
2013–2016	3	2.06(1.27, 3.34)	38.7%
2017–2020	2	0.70(0.33, 1.49)	77.0%
Sample size
< 900	2	1.00(0.21, 4.88)	91.2%
> 900	3	1.52(0.80, 2.87)	76.1%
Follow-up duration(months)
< 30	2	1.07(0.78, 1.45)	0.0%
> 30	3	1.45(0.44, 4.79)	89.5%
Male (%)
≥ 50%	3	1.81(0.85, 3.86)	81.2%
< 50%	2	0.78(0.27, 2.30)	83.4%
Adjust for diabetes
Yes	2	0.70(0.33, 1.49)	77.0%
No	3	2.06(1.27, 3.34)	38.7%
Adjust for BMI
Yes	5	1.30(0.72, 2.34)	80.8%
No	0	-	-

HR, hazard ratio; CI, confidence interval; BMI, body mass index.

The results showed that the pooled HR value was close before and after removing any study, indicating that the result was relatively stable (see S6 Fig).

The P values of the overall null hypothesis tests of model parameters were all > 0.05, and the P values of the goodness-of-fit tests of model were all < 0.05, indicating that neither liner (P = 0.360, random effect linear model) nor nonlinear (P = 0.476, random effect nonlinear model) relationship between SUA and CVD mortality was observed.

Discussion

To the best of our knowledge, this is the first study to determine the association between SUA and all-cause and CVD mortality in PD patients based on dose-response meta-analysis of cohort studies. Before and after recalculating the relevant HRs and 95% CIs using the lowest dose group as the reference, the pooled results were consistent. We found that higher SUA level was significantly associated with higher all-cause mortality compared with the median level. In addition, the results from this meta-analysis did not detect any relationship between different SUA levels and the risk of all-cause and CVD mortality. The dose-response analysis suggested a J-shaped nonlinear relationship between SUA concentrations and all-cause mortality. When Lai et al.’s study was removed, we detected that compared with lowest SUA levels, highest SUA levels were associated with an increased risk of all-cause mortality in PD patients.

SUA was a clinically powerful nutritional marker and as well as an independent predictor of all-cause and CVD death risk [39]. A systematic review and meta-analysis of hemodialysis death risk factors published in 2017 showed that all-cause and cardiovascular death are affected by multiple factors (age, gender, diabetes, CRP, CV, HbA1c, etc.), but did not explore the relationship between SUA and mortality [40]. In addition, Anderson [41] reported a systematic review and meta-analysis of death risk prediction for patients starting dialysis, unfortunately, the relationship between SUA and death was not explored. A recent meta-analysis by Xue et al. [42] explored the relationship between SUA and all-cause and CVD mortality in PD patients and found that the results of prospective and retrospective cohort studies were inconsistent. However, they did not research the impact of intermediate levels of SUA on mortality compared with the highest and lowest levels. Another meta-analysis by Liu et al. [43] found that high SUA levels were associated with an increased risk of all-cause mortality in PD patients compared with middle SUA levels, but SUA levels may not be associated with CVD mortality, which is consistent with our results before recalculating HRs and 95% CIs. What is more, it may be a pity that both them did not recalculate the original data using the lowest dose group as the reference and perform the dose-response analysis. Liu et al. [7] explored the relationship between patient characteristics and risk factors of early and late mortality in PD patients, which has demonstrated that higher UA level was associated with early death, therefore specific intervention according to risk factors at the initiation of PD should be established to improve the survival of PD patients.

Although the underlying mechanism between SUA and mortality in PD patients is still unclear, some research advances have provided us with clues. Some studies support the association between high UA levels and high mortality. Animal experiments and clinical studies have confirmed that UA is an endothelial toxin and causes endothelial dysfunction. Hyperuricemia suppresses the production of nitric oxide [44], leading to activation of the renin-angiotensin system, which ultimately leads to endothelial damage [45, 46]. UA-lowing drug (allopurinol and xanthine oxidase inhibitor) treatment resulted in a decrease in SUA. Studies have demonstrated that allopurinol can significantly improve endothelial function in patients with CKD or chronic heart failure [47] and xanthine oxidase inhibitor reduce the incidence of adverse CV events [48]. C-reactive protein (CRP) is the most commonly used inflammatory parameter primarily produced by hepatocytes [49], and elevated CRP levels are independent risk factor for CKD [50]. Evidence suggests that there is a positive correlation between SUA and serum CRP levels in healthy populations, patients with acute coronary syndrome [51], and CKD patients undergoing peritoneal dialysis [52]. CRP is becoming a clinical marker for many noncommunicable diseases (atherosclerosis, CVD, ischemic stroke, hypertension, insulin resistance, and metabolic syndrome) and can independently predict adverse cardiovascular events in individuals, including ischemic stroke, myocardial infarction and sudden cardiac death [53]. In addition, CRP can independently predict all-cause mortality in China’s middle-aged and elderly population [54]. The effect of SUA on RRF is an important factor affecting mortality. Elevated SUA is common in patients with PD, and it is inversely related to the decrease in RRF [55]. A study from Taiwan revealed that the UA level has a U-shaped relationship with the decline rate of RRF in continuous ambulatory peritoneal dialysis patients, with a faster decline rate in those of higher and lower UA groups [56]. The rate of decline of RRF is a powerful predictive factor associated with lower survival and technical failure in patients treated with PD [57, 58], which may explain one of the reasons for the effect of both higher and lower SUA concentrations on the death of patients receiving PD treatment.

In addition, research has also reported the role of low UA levels in mortality. Malnutrition is an independent risk factor for the prognosis of patients with PD [23, 59]. Patients with malnutrition have low immunity, are prone to various infections and are difficult to control. In addition, malnutrition is closely related to cardiovascular events [60]. SUA levels are associated with nutritional risk and independently predict all-cause and CVD death risk [39]. Bae et al. [29] have found that SUA < 5.5mg/dL is associated with all-cause mortality. SUA is not only a simple biomarker which indicates the nutritional status of patients treated with chronic dialysis, but also one of the most important antioxidants in human biological fluids, removing excess oxygen free radicals from the body [29]. SUA levels were correlated with the total antioxidant capacity in population treated with dialysis, hence, hypouricemia may lead to a decrease in total antioxidant capacity in dialysis patients [61]. Lai et al. [17] reported that high UA levels were associated with low all-cause and cardiovascular mortality in female populations undergoing continuous PD, which may be explained by the antioxidant capacity of UA. UA plays a major antioxidant role in the plasma, but a major pro-oxidant role when it enters cells and a pro-oxidant role in the development of cardiovascular disease [62].

In summary, research evidence shows that SUA has a dual biological effect on the human body, which may partly explain why the results of the included studies are contradictory and the pooled results are not clear. Furthermore, we need to clarify which effect is greater in PD population. If the harmful effect is greater, we can reduce the concentration of SUA through drugs, diet or other measures. Meanwhile, the protective role of SUA, such as antioxidant capacity, can be played by dietary supplements [62]. In order to clarify the association between SUA and mortality (all-cause and CVD) and to guide clinical treatment, strict design, large sample size and multi-center cohort studies are required to collect as much information as possible, such as SUA at different time points, changes in SUA level during PD treatment and populations with different characteristics (gender, diabetes and non-diabetes, etc).

Several limitations should be noted. First, we found significant heterogeneity in our study. The “leave-one-out” sensitive analysis indicated that the study conducted by Lai et al. [17] had a great influence on the combined HRs and maybe was the pivotal contributor to heterogeneity, which may be attributed to its opposite results with other studies (a higher SUA level was associated with a lower risk of all-cause and CVD mortality). The reason may be the longer median follow-up period (> 3 years), the relationship between low SUA levels and malnutrition, and increased oxidative stress. Furthermore, results of subgroup analyses suggested that study design (prospective or retrospective, single center or multi-center cohort study), year of publication, sample size, duration of follow-up, male proportion and whether adjusted for diabetes status and BMI may be sources of heterogeneity. It is interesting that there is no significant difference in the subgroup analysis when adjusted with diabetes and BMI, suggesting there is a difference in the impact of UA in these subgroups. Dong et al. [18] found that the associations of UA and CVD/all-cause mortality disappeared with additional adjustment for traditional CV factors such as CVD history, diabetes, BMI, and low-density lipoprotein cholesterol. This may indicate that the association between UA and CVD/all-cause mortality is not independent, but related to traditional CV risk factors in the PD population. At the same time, we found that the pooled HR of SUA and all-cause mortality was significant when the proportion of men was ≥ 50%, which was consistent with some previous studies [14, 15, 19]. In addition, UA measured at different time points after PD initiation (including three months [14], sixth month [20] or time-average SUA [19]) may also affect the results. Second, covariate adjustment may affect the correlation between SUA and all-cause and CVD mortality, although we extracted the HRs that adjusted the greatest degree of potential confounders. Furthermore, although most of the original studies adjusted for many important confounding factors, the effects of residual and unknown confounding factors on the results cannot be completely excluded. Finally, all the studies included in the meta-analysis were conducted in Asian countries except one study, which greatly limits the applicability of the results to the global population.

Our analysis has several strengths. First, we comprehensively considered the relationship between different SUA levels (the highest versus the lowest; the highest or lowest versus the median) and mortality. Second, considering the characteristics of SUA, we hypothesized that very low and very high SUA levels may increase the risk of death. Therefore, we recalculated the relevant HRs and 95% CIs using the lowest dose group as the reference, and conducted a dose-response curve instead of only comparing the effects of the highest versus the lowest levels of SUA on mortality to explore the range of SUA concentration associated with the lowest mortality in patients with PD. The dose-response analysis suggested a J-shaped nonlinear relationship between SUA concentrations and all-cause mortality although there was no significant difference in all-cause mortality between the highest and lowest subgroup of SUA level in the main analysis results. This may suggest that there is a group of people with the lowest risk in the distribution of death risk of PD population caused by exposure, so finding them has very important public health significance.

Conclusion

This meta-analysis did not find there is any relationship between SUA levels and the risk of all-cause and CVD death in PD patients. More rigorously designed studies in the future will be needed to determine the relationship between SUA and cardiovascular and all-cause mortality.

PRISMA checklist.

(DOCX)

Click here for additional data file.

Sensitivity analysis.

For relationship between SUA by categories (the highest SUA category vs the lowest) and all-cause mortality in PD patients before recalculated the HRs and 95% CIs. HR, hazard ratio; CI, confidence interval.

(TIF)

Click here for additional data file.

For relationship between SUA by categories (the highest SUA category vs the median) and all-cause mortality in PD patients before recalculated the HRs and 95% CIs. HR, hazard ratio; CI, confidence interval.

(TIF)

Click here for additional data file.

For relationship between SUA by categories (the lowest SUA category vs the median) and all-cause mortality in PD patients before recalculated the HRs and 95% CIs. HR, hazard ratio; CI, confidence interval.

(TIF)

Click here for additional data file.

For relationship between SUA by categories (the highest SUA category vs the lowest) and cardiovascular mortality in PD patients before recalculated the HRs and 95% CIs. HR, hazard ratio; CI, confidence interval.

(TIF)

Click here for additional data file.

For relationship between SUA by categories (the highest SUA category vs the lowest) and all-cause mortality in PD patients after recalculated the HRs and 95% CIs. HR, hazard ratio; CI, confidence interval.

(TIF)

Click here for additional data file.

For relationship between SUA by categories (the highest SUA category vs the lowest) and cardiovascular mortality in PD patients after recalculated the HRs and 95% CIs. HR, hazard ratio; CI, confidence interval.

(TIF)

Click here for additional data file.

Search strategies for electronic databases.

(DOCX)

Click here for additional data file.

Quality assessment of the included studies utilizing the Newcastle-Ottawa Scale (NOS).

RCS, retrospective cohort study; PCS, prospective cohort study.

(DOCX)

Click here for additional data file.

5 Nov 2021

A rebuttal letter that responds to each point raised by the academic editor and reviewer(s). You should upload this letter as a separate file labeled 'Response to Reviewers'.

A marked-up copy of your manuscript that highlights changes made to the original version. You should upload this as a separate file labeled 'Revised Manuscript with Track Changes'.

An unmarked version of your revised paper without tracked changes. You should upload this as a separate file labeled 'Manuscript'.

If you would like to make changes to your financial disclosure, please include your updated statement in your cover letter. Guidelines for resubmitting your figure files are available below the reviewer comments at the end of this letter.

If applicable, we recommend that you deposit your laboratory protocols in protocols.io to enhance the reproducibility of your results. Protocols.io assigns your protocol its own identifier (DOI) so that it can be cited independently in the future. For instructions see: https://journals.plos.org/plosone/s/submission-guidelines#loc-laboratory-protocols. Additionally, PLOS ONE offers an option for publishing peer-reviewed Lab Protocol articles, which describe protocols hosted on protocols.io. Read more information on sharing protocols at https://plos.org/protocols?utm_medium=editorial-email&utm_source=authorletters&utm_campaign=protocols.

We look forward to receiving your revised manuscript.

Kind regards,

Girish Chandra Bhatt, MD, FASN

Academic Editor

PLOS ONE

Journal Requirements:

When submitting your revision, we need you to address these additional requirements.

1. Please ensure that your manuscript meets PLOS ONE's style requirements, including those for file naming. The PLOS ONE style templates can be found at

https://journals.plos.org/plosone/s/file?id=wjVg/PLOSOne_formatting_sample_main_body.pdf and

https://journals.plos.org/plosone/s/file?id=ba62/PLOSOne_formatting_sample_title_authors_affiliations.pdf

2. We note that you have stated that you will provide repository information for your data at acceptance. Should your manuscript be accepted for publication, we will hold it until you provide the relevant accession numbers or DOIs necessary to access your data. If you wish to make changes to your Data Availability statement, please describe these changes in your cover letter and we will update your Data Availability statement to reflect the information you provide.

Additional Editor Comments:

Details of how NOS quality assessment of each study should be provided in the table.

Wherever unexplained high heterogeneity is present i.e >50% , the authors should use methods such as Bejout;s curve, influential analysis and leave one out analysis.

Visual inspection of the funnel plot for publication bias is used when the number of the studies is more than or equal to 10 for a particular outcome. Moreover, other test such as Egger;s test should be usede and provided as being significant or non -significant.

[Note: HTML markup is below. Please do not edit.]

Reviewers' comments:

Reviewer's Responses to Questions

Comments to the Author

1. Is the manuscript technically sound, and do the data support the conclusions?

The manuscript must describe a technically sound piece of scientific research with data that supports the conclusions. Experiments must have been conducted rigorously, with appropriate controls, replication, and sample sizes. The conclusions must be drawn appropriately based on the data presented.

Reviewer #1: Yes

**********

2. Has the statistical analysis been performed appropriately and rigorously?

Reviewer #1: Yes

**********

3. Have the authors made all data underlying the findings in their manuscript fully available?

The PLOS Data policy requires authors to make all data underlying the findings described in their manuscript fully available without restriction, with rare exception (please refer to the Data Availability Statement in the manuscript PDF file). The data should be provided as part of the manuscript or its supporting information, or deposited to a public repository. For example, in addition to summary statistics, the data points behind means, medians and variance measures should be available. If there are restrictions on publicly sharing data—e.g. participant privacy or use of data from a third party—those must be specified.

Reviewer #1: Yes

**********

4. Is the manuscript presented in an intelligible fashion and written in standard English?

PLOS ONE does not copyedit accepted manuscripts, so the language in submitted articles must be clear, correct, and unambiguous. Any typographical or grammatical errors should be corrected at revision, so please note any specific errors here.

Reviewer #1: Yes

**********

5. Review Comments to the Author

Please use the space provided to explain your answers to the questions above. You may also include additional comments for the author, including concerns about dual publication, research ethics, or publication ethics. (Please upload your review as an attachment if it exceeds 20,000 characters)

Reviewer #1: Serum uric acid level and all cause mortality and cardiovascular mortality in peritoneal dialysis patients: A systematic review and meta-analysis of cohort studies.

Introduction:

The introduction highlights the facts that serum uric acid is a know cause on increased mortality in patients on dialysis. There is conflicting data on the association of uric acid with mortality in patients on PD and the method of estimation of uric acid. The authors may re-frame the sentences to allow easy comprehension and improve clarity.

The objective may also be re-framed “to determine the association…..”

Methods:

The methods mention that studies in all languages were included, but the interpretation of studies in a foreign language has not been described. Were any attempts made to search for unpublished data (conference proceedings etc?). Cochrane library was not included in the literature search. The systematic review /meta-analysis has not been registered in any database.

The systematic review has assessed the bias of risk, assessed quality of the data and accounted for publication bias.

Results:

The results have been discussed well. The Forest plots however are not very clear.

The PRISMA reporting guidelines have been followed.

Discussion: The fact that most of the studies were from Asian countries limits its generalizability to the global population

Conclusions: The conclusions are aligned to the objectives

**********

6. PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy.

Reviewer #1: No

[NOTE: If reviewer comments were submitted as an attachment file, they will be attached to this email and accessible via the submission site. Please log into your account, locate the manuscript record, and check for the action link "View Attachments". If this link does not appear, there are no attachment files.]

While revising your submission, please upload your figure files to the Preflight Analysis and Conversion Engine (PACE) digital diagnostic tool, https://pacev2.apexcovantage.com/. PACE helps ensure that figures meet PLOS requirements. To use PACE, you must first register as a user. Registration is free. Then, login and navigate to the UPLOAD tab, where you will find detailed instructions on how to use the tool. If you encounter any issues or have any questions when using PACE, please email PLOS at figures@plos.org. Please note that Supporting Information files do not need this step.

9 Dec 2021

Response to Academic Editor:

To explore the source of heterogeneity among studies, subgroup analyses were conducted according to study design (prospective or retrospective cohort study), number of center (multi-center or single center), publication years (2013-2016 or 2017-2020), sample size (< 900 or > 900), follow-up duration (< 30 months or > 30 months), adjustment for gender and adjustment for diabetes status. And we also carefully revised the content and grammar of our article, hoping to improve the article.

Response to additional Editor:

To question 1: We provided details of NOS quality assessment for each study in the S2 Table.

To question 2: Sensitivity analysis, in which 1 study was removed at a time, was performed to evaluate the stability of the results. We added several sensitivity analyses, and the results are shown in S1-S6 Figs.

To question 3: It is really true as editor suggested that visual inspection of the funnel plot for publication bias is used when the number of the studies is more than or equal to 10 for a particular outcome. Therefore, we deleted the funnel plots and recorded the P values of egger’s test in the manuscript.

Response to Reviewer:

To question 1: We re-framed the objective of this study as follows: hence, the objective of this study is to determine the association between SUA and all-cause and CVD mortality with a detailed analysis of eligible literature.

To question 2:

(1) We have re-written this part according to the Reviewer’s suggestion: all non-English studies were translated by software first, and then the researchers checked whether they could be included.

(2) In addition to a comprehensive search, we considered the reference lists of all eligible articles in order to find additional relevant articles. I regret that unpublished data was not considered.

(3) We are very sorry that we did not search the Cochrane Library database before. Therefore, we limited the date filter to April 7, 2021, and finally retrieved 27 literatures (See S1 Table for searching strategies).

(4) We are very sorry that we did not register for this meta-analysis, but we conducted this systematic review and meta-analysis in strict accordance with the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) statement.

To question 3: We have adjusted the resolution and size of the figures, hoping to meet the requirements of the journal.

To question 4: After consideration, we have included a conference abstract by Coelho et al [1] in Portugal, but the pooled results did not change. Nevertheless, almost all the studies included in the meta-analysis were conducted in Asian countries, which greatly limits the applicability of the results to the global population.

Submitted filename: Response to Reviewers.docx

Click here for additional data file.

21 Jan 2022

A rebuttal letter that responds to each point raised by the academic editor and reviewer(s). You should upload this letter as a separate file labeled 'Response to Reviewers'.

A marked-up copy of your manuscript that highlights changes made to the original version. You should upload this as a separate file labeled 'Revised Manuscript with Track Changes'.

An unmarked version of your revised paper without tracked changes. You should upload this as a separate file labeled 'Manuscript'.

If you would like to make changes to your financial disclosure, please include your updated statement in your cover letter. Guidelines for resubmitting your figure files are available below the reviewer comments at the end of this letter.

If applicable, we recommend that you deposit your laboratory protocols in protocols.io to enhance the reproducibility of your results. Protocols.io assigns your protocol its own identifier (DOI) so that it can be cited independently in the future. For instructions see: https://journals.plos.org/plosone/s/submission-guidelines#loc-laboratory-protocols. Additionally, PLOS ONE offers an option for publishing peer-reviewed Lab Protocol articles, which describe protocols hosted on protocols.io. Read more information on sharing protocols at https://plos.org/protocols?utm_medium=editorial-email&utm_source=authorletters&utm_campaign=protocols.

We look forward to receiving your revised manuscript.

Kind regards,

Masaki Mogi

Academic Editor

PLOS ONE

[Note: HTML markup is below. Please do not edit.]

Reviewers' comments:

Reviewer's Responses to Questions

Comments to the Author

1. If the authors have adequately addressed your comments raised in a previous round of review and you feel that this manuscript is now acceptable for publication, you may indicate that here to bypass the “Comments to the Author” section, enter your conflict of interest statement in the “Confidential to Editor” section, and submit your "Accept" recommendation.

Reviewer #1: All comments have been addressed

Reviewer #2: (No Response)

**********

2. Is the manuscript technically sound, and do the data support the conclusions?

The manuscript must describe a technically sound piece of scientific research with data that supports the conclusions. Experiments must have been conducted rigorously, with appropriate controls, replication, and sample sizes. The conclusions must be drawn appropriately based on the data presented.

Reviewer #1: Yes

Reviewer #2: Yes

**********

3. Has the statistical analysis been performed appropriately and rigorously?

Reviewer #1: Yes

Reviewer #2: Yes

**********

4. Have the authors made all data underlying the findings in their manuscript fully available?

The PLOS Data policy requires authors to make all data underlying the findings described in their manuscript fully available without restriction, with rare exception (please refer to the Data Availability Statement in the manuscript PDF file). The data should be provided as part of the manuscript or its supporting information, or deposited to a public repository. For example, in addition to summary statistics, the data points behind means, medians and variance measures should be available. If there are restrictions on publicly sharing data—e.g. participant privacy or use of data from a third party—those must be specified.

Reviewer #1: Yes

Reviewer #2: Yes

**********

5. Is the manuscript presented in an intelligible fashion and written in standard English?

PLOS ONE does not copyedit accepted manuscripts, so the language in submitted articles must be clear, correct, and unambiguous. Any typographical or grammatical errors should be corrected at revision, so please note any specific errors here.

Reviewer #1: Yes

Reviewer #2: Yes

**********

6. Review Comments to the Author

Please use the space provided to explain your answers to the questions above. You may also include additional comments for the author, including concerns about dual publication, research ethics, or publication ethics. (Please upload your review as an attachment if it exceeds 20,000 characters)

Reviewer #1: The authors have responded to all the queries satisfactorily and made the necessary changes in the revised manuscript

Reviewer #2: The authors conducted a very careful systematic review and presented valuable data on association between serum uric acid levels and all-cause mortality or CVD in PD patients. It is very interesting data, and since this study contains enough subjects, subgroup analysis became possible. So, several concerns should be addressed.

Major comments

#1 It is interesting that there is no significant difference in the subgroup analysis (Table 1 and Table2) when adjusted with gender and diabetes, suggesting there is a difference in the impact of uric acid in these subgroups. Are such analyses possible? Or, authors should discuss on these points and data of subgroup analyses (Table 1, Table2).

#2 The patients with hyperuricemia consists of etiologically heterogeneous population. Serum uric acid levels are affected by genetic predisposition for uric acid transporters, and also by environmental factors such as alcohol drinking and obesity/visceral fat accumulation due to overnutrition. The former shows a very high serum uric acid level, and gout and urinary tract stones appear in the foreground, while the latter is prone to atherosclerosis, and may reflect the high CRP that the authors pointed out in Discussion section. Since this study have enough subjects, subgroup analysis is possible, which is very valuable. This time, authors showed the adjusted data by gender and diabetes. Since the authors focused on all-cause mortality and CVD, it may be useful to conduct a subgroup analysis with or without obesity/visceral fat accumulation, and with or without liver dysfunction which reflects NAFLD/NASH. Is it possible that the impact of uric acid will be sharpened by doing so?

Minor comments

#3 It will become more understandable for readers to show data on the concentration range of the uric acid category in each paper, because the meanings of UA<2 and UA <4 are different in terms of dual biological effects of uric acid. Is there a difference in the width of the lowest between each paper?

#4 In that sense, the dose-response analysis data is very significant. Is it possible to show the data related to CVD as a Figure (even if it is negative data)? Author had better to discuss more this dose-response impact on all-cause mortality and CVD in discussion section.

#5 Figure 2 is hard to see. So, authors should increase the resolution.

**********

7. PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy.

Reviewer #1: No

Reviewer #2: No

[NOTE: If reviewer comments were submitted as an attachment file, they will be attached to this email and accessible via the submission site. Please log into your account, locate the manuscript record, and check for the action link "View Attachments". If this link does not appear, there are no attachment files.]

While revising your submission, please upload your figure files to the Preflight Analysis and Conversion Engine (PACE) digital diagnostic tool, https://pacev2.apexcovantage.com/. PACE helps ensure that figures meet PLOS requirements. To use PACE, you must first register as a user. Registration is free. Then, login and navigate to the UPLOAD tab, where you will find detailed instructions on how to use the tool. If you encounter any issues or have any questions when using PACE, please email PLOS at figures@plos.org. Please note that Supporting Information files do not need this step.

6 Feb 2022

Response to Reviewer 2:

Question 1: It is interesting that there is no significant difference in the subgroup analysis (Table 1 and Table2) when adjusted with gender and diabetes, suggesting there is a difference in the impact of uric acid in these subgroups. Are such analyses possible? Or, authors should discuss on these points and data of subgroup analyses (Table 1, Table2)

Answer: First, we carefully checked the variable adjustment of each included article, and ran the heterogeneity test again with Stata software, and the results were consistent with those in the manuscript (Table 2 and Table 3). Dong et al. [1] found that the associations of UA and CVD/all-cause mortality disappeared with additional adjustment for traditional CV factors such as CVD history, diabetes, BMI, and low-density lipoprotein cholesterol. This may indicate that the association between UA and CVD / all-cause mortality is not independent, but related to traditional CV risk factors in the PD population. Second, after careful consideration, we changed the subgroup analysis of whether to adjust gender into the subgroup analysis of male proportion (50% as the cut-off value). As written in paragraph 4 of the “result” section, there may be gender differences in the impact of SUA levels on mortality. Therefore, subgroup analysis based on the proportion of men may be more reasonable. We discussed the results of the subgroup analysis in lines 460-469.

Question 2: Since the authors focused on all-cause mortality and CVD, it may be useful to conduct a subgroup analysis with or without obesity/visceral fat accumulation, and with or without liver dysfunction which reflects NAFLD/NASH. Is it possible that the impact of uric acid will be sharpened by doing so?

Answer: Thank you for your suggestion. We conducted a subgroup analysis according to adjustment for BMI, but the results were similar to those adjusted for diabetes. After careful review of the included literature, NAFLD / NASH was not found in the comorbid conditions of the collected information. For defining NAFLD [2], there must be (1) evidence of hepatic steatosis, either by imaging or histology, and (2) lack of secondary causes of hepatic fat accumulation such as significant alcohol consumption, long-term use of a steatogenic medication, or monogenic hereditary disorders. Unfortunately, there is no data of liver imaging and histological examination in the original article. A recent meta-analysis [3] indicated that controlling the most important confounding factors, including indicators that reflect the current residual renal function of patients, and other confounding factors (including gender, age, diabetes history, CVD history, Kt/V, whether to use UA drugs and serum albumin) are very important for comparability among groups. We noticed that albumin is an index that can reflect the synthesis and reserve function of the liver in a certain period of time. Recent studies have shown NAFLD to be associated with impairments in albumin function, which are associated with impairments in liver function and disease prognosis [4]. However, we found that all studies included in this meta-analysis had adjusted for albumin, so subgroup analysis was not performed.

Question 3: It will become more understandable for readers to show data on the concentration range of the uric acid category in each paper.

Answer: Thank you very much for your advice. We have added concentration range of the SUA categories in Table 1.

Question 4: (1) Is it possible to show the data related to CVD as a Figure (even if it is negative data)? (2) Author had better to discuss more this dose-response impact on all-cause mortality and CVD in discussion section.

Answer: (1) We examined the dose-response relationship between SUA and CVD mortality and confirmed that the dose-response diagram could not be made. The P values of model goodness-of-fit tests were all < 0.05, suggesting that heterogeneity had to be considered. But neither liner (�  2 = 0.84, P = 0.360, random effect linear model) nor nonlinear (�  2 = 1.66, P = 0.476, random effect nonlinear model) relationship between SUA and CVD mortality was observed. The reason why there is no dose-response figure of SUA and CVD is that the random effect model is not significant, rather than the curve. Therefore, we deleted the sentence " We analyzed the linear or non-linear relationship between mortality and SUA levels, and listed only significant curves." in the “statistical analysis” part because it is very misleading.

(2) Thank you very much for your advice. We added the discussion about the dose-response relationship between SUA levels and all-cause mortality. Please see lines 483-492 for details.

Question 5: Figure 2 is hard to see. So, authors should increase the resolution.

Answer: We have adjusted the resolution of figure2, hoping to meet the requirements of the journal.

Once again, thank you very much for your comments and suggestions. Looking forward to hearing from you. Best wishes.

9 Feb 2022

Serum uric acid level and all-cause and cardiovascular mortality in peritoneal dialysis patients: a systematic review and dose-response meta-analysis of cohort studies

PONE-D-21-22383R2

Dear Dr. Lyu,

We’re pleased to inform you that your manuscript has been judged scientifically suitable for publication and will be formally accepted for publication once it meets all outstanding technical requirements.

Within one week, you’ll receive an e-mail detailing the required amendments. When these have been addressed, you’ll receive a formal acceptance letter and your manuscript will be scheduled for publication.

An invoice for payment will follow shortly after the formal acceptance. To ensure an efficient process, please log into Editorial Manager at http://www.editorialmanager.com/pone/, click the 'Update My Information' link at the top of the page, and double check that your user information is up-to-date. If you have any billing related questions, please contact our Author Billing department directly at authorbilling@plos.org.

If your institution or institutions have a press office, please notify them about your upcoming paper to help maximize its impact. If they’ll be preparing press materials, please inform our press team as soon as possible -- no later than 48 hours after receiving the formal acceptance. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information, please contact onepress@plos.org.

Kind regards,

Masaki Mogi

Academic Editor

PLOS ONE

Additional Editor Comments (optional):

No further comment.

Reviewers' comments:

Reviewer's Responses to Questions

Comments to the Author

1. If the authors have adequately addressed your comments raised in a previous round of review and you feel that this manuscript is now acceptable for publication, you may indicate that here to bypass the “Comments to the Author” section, enter your conflict of interest statement in the “Confidential to Editor” section, and submit your "Accept" recommendation.

Reviewer #2: All comments have been addressed

**********

2. Is the manuscript technically sound, and do the data support the conclusions?

The manuscript must describe a technically sound piece of scientific research with data that supports the conclusions. Experiments must have been conducted rigorously, with appropriate controls, replication, and sample sizes. The conclusions must be drawn appropriately based on the data presented.

Reviewer #2: Yes

**********

3. Has the statistical analysis been performed appropriately and rigorously?

Reviewer #2: Yes

**********

4. Have the authors made all data underlying the findings in their manuscript fully available?

The PLOS Data policy requires authors to make all data underlying the findings described in their manuscript fully available without restriction, with rare exception (please refer to the Data Availability Statement in the manuscript PDF file). The data should be provided as part of the manuscript or its supporting information, or deposited to a public repository. For example, in addition to summary statistics, the data points behind means, medians and variance measures should be available. If there are restrictions on publicly sharing data—e.g. participant privacy or use of data from a third party—those must be specified.

Reviewer #2: Yes

**********

5. Is the manuscript presented in an intelligible fashion and written in standard English?

PLOS ONE does not copyedit accepted manuscripts, so the language in submitted articles must be clear, correct, and unambiguous. Any typographical or grammatical errors should be corrected at revision, so please note any specific errors here.

Reviewer #2: (No Response)

**********

6. Review Comments to the Author

Please use the space provided to explain your answers to the questions above. You may also include additional comments for the author, including concerns about dual publication, research ethics, or publication ethics. (Please upload your review as an attachment if it exceeds 20,000 characters)

Reviewer #2: All comments, which the reviewer had raised, have adequately addressed.

The reviewer had no more question and comments.

**********

7. PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy.

Reviewer #2: No

11 Feb 2022

PONE-D-21-22383R2

Serum uric acid level and all-cause and cardiovascular mortality in peritoneal dialysis patients: a systematic review and dose-response meta-analysis of cohort studies

Dear Dr. Lyu:

I'm pleased to inform you that your manuscript has been deemed suitable for publication in PLOS ONE. Congratulations! Your manuscript is now with our production department.

If your institution or institutions have a press office, please let them know about your upcoming paper now to help maximize its impact. If they'll be preparing press materials, please inform our press team within the next 48 hours. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information please contact onepress@plos.org.

If we can help with anything else, please email us at plosone@plos.org.

Thank you for submitting your work to PLOS ONE and supporting open access.

Kind regards,

PLOS ONE Editorial Office Staff

on behalf of

Dr. Masaki Mogi

Academic Editor

PLOS ONE