Association of albumin to non-high-density lipoprotein cholesterol ratio with mortality in peritoneal dialysis patients

Abstract

Objective

Malnutrition and inflammation are associated with mortality in peritoneal dialysis (PD) patients. Serum albumin and non-high-density lipoprotein cholesterol (non-HDL-C) are independently associated with mortality in PD patients. Combining albumin and non-HDL-C with mortality may be more plausible in clinical practice.

Methods

This retrospective cohort study included 1954 Chinese PD patients from 1 January 2009 to 31 December 2016. Kaplan–Meier curve was used to determine the relationship between albumin to non-HDL-C ratio and all-cause mortality. Cox regression analysis was applied to assess the independent predictive value while adjusting for confounding factors. Competitive risk analysis was used to examine the effects of other outcomes on all-cause mortality prognosis.

Results

In the 33-month follow-up period, there were 538 all-cause deaths. Kaplan–Meier analysis presented significant differences in all-cause mortality. Multivariate Cox regression showed that the risk of all-cause mortality was lower in the moderate group (9.36–12.79) (HR, 0.731; 95% CI, 0.593–0.902, p = 0.004) and the highest group (>12.79) (HR, 0.705; 95% CI, 0.565–0.879, p = 0.002) compared to the lowest group (≤9.36). Competitive risk analysis revealed significant differences for all-cause mortality (p < 0.001), while there was no statistical significance for other competing events.

Conclusions

Low albumin to non-HDL-C ratio was associated with a high risk of all-cause mortality in PD patients. It may serve as a potential prognostic biomarker in PD patients.

Keywords:

• Albumin
• non-high-density lipoprotein cholesterol
• mortality
• peritoneal dialysis
• inflammation
• nutrition

Introduction

Malnutrition is a common issue in peritoneal dialysis (PD) patients and has been linked to an increased risk of mortality. Early studies in the US found that 40–66% of PD patients were malnourished. Later research defined protein-energy wasting using serum albumin levels, which revealed a high prevalence of malnutrition in PD patients. PD patients with baseline serum albumin level <3.0 g/dL had a more than 3-fold higher adjusted risk of all-cause and cardiovascular mortality and 3.4-fold higher risk of infection-related mortality [1–7]. Malnutrition and inflammation are strongly associated with mortality and cardiovascular events in PD patients. The causes of low serum albumin in PD patients are multifactorial, including protein loss during dialysis, inflammation, reduced protein intake, chronic acidosis, and psychosocial factors. Recent studies have indicated that hypoalbuminemia, which is negatively linked to patient outcomes, may be more related to inflammation than malnutrition [8,9].

In PD patients, higher levels of non-high-density lipoprotein cholesterol (non-HDL-C) are linked to a greater risk of cardiovascular death and all-cause mortality. Non-HDL-C encompasses all cholesterol that contributes to the formation of atherosclerotic plaques along with low-density lipoprotein cholesterol (LDL-C), Lp (a), very low-density lipoprotein cholesterol (VLDL-C), and other remnants [10–14]. Unlike other measurements, non-HDL-C levels are not influenced by fasting in clinical practice. The sum of cholesterol from all lipoproteins carrying proatherogenic properties, including LDL-C, very low-density lipoprotein, intermediate-density lipoprotein, chylomicrons, and lipoprotein (a), is represented by non-HDL-C. This measurement is closely tied to apolipoprotein B and more accurately reflects all apolipoprotein B-containing lipoproteins compared to LDL-C [15–18]. Measuring non-HDL-C levels is convenient as it can be obtained by subtracting HDL-C from non-HDL-C, without being impacted by fasting.

Inflammation is known to affect lipid metabolism. During inflammation, there is an increase in the levels of pro-inflammatory cytokines like interleukin-6 (IL-6), C-reactive protein (CRP), hypersensitive CRP (hs-CRP), and HDL-associated antioxidant proteins. These factors can alter lipid metabolism, leading to decrease levels of HDL-C. At the same time, inflammation can lead to decreased albumin levels due to increased capillary permeability and greater fractional catabolic rate (FCR). Malnutrition can have adverse effects on the lipid profile. In cases of severe malnutrition, the liver may produce fewer lipoproteins, leading to changes in lipid levels. Lower albumin levels might indicate malnutrition, which can indirectly affect lipid metabolism and increase non-HDL-C levels [19–24].

Previous studies have found that lower levels of serum albumin are associated with a higher risk of mortality, while higher levels of non-HDL-C (which is calculated as total cholesterol minus HDL-C) have consistently been linked to a higher risk of mortality in dialysis patients [25,26]. Although the albumin-to-cholesterol ratio and non-HDL-C to HDL-C ratio have been used to predict cardiovascular disease risk in the general population, no studies have examined the association between the albumin to non-HDL-C ratio and mortality in PD patients. To more accurately evaluate the prognosis of PD patients, it is necessary to consider both serum albumin and non-HDL-C levels. Therefore, this study aims to investigate the association between the albumin to non-HDL-C ratio and all-cause mortality in PD patients.

Materials and methods

Data source

In this retrospective cohort study, we recruited 2066 incident Chinese PD patients from 6 PD centers between 1 January 2009 and 31 December 2016. Patients younger than 18 years (n = 17), who received PD for less than 3 months (n = 63), and lost data (n = 32) were excluded. Finally, 1954 patients were enrolled in this study. Data were collected 1 week before. This statement indicates that the study was approved by the Institutional Review Board of six PD centers and that informed consent was not needed as the data used was preexisting medical data from the hospital (Figure 1).

Figure 1. Flow chart-including patient enrollment and outcomes.

Data extraction

The study collected demographic, comorbidity, and laboratory data from six facilities for patients starting PD treatment. Data collected included: age, sex, body mass index (BMI), current smoking, current alcohol consumption, SBP, DBP, comorbidities (diabetes, history of cardiovascular events [CVE], and hypertension), medication use (calcium channel blocker [CCB], beta blocker, angiotensin-converting enzyme inhibitor/angiotensin receptor blocker [ACEI/ARB], insulin, aspirin, and diuretics), and laboratory results (WBC, hemoglobin, albumin, cholesterol, triglycerides, LDL, HDL, non-HDL, creatinine, urea nitrogen, uric acid, sodium, calcium, potassium, chlorine, total Kt/V, and renal residual function [RRF]). The primary outcome was all-cause mortality and the follow-up period was from the start of PD to death, transfer to hemodialysis, transplant, transfer to another center, loss of follow-up, or 31 December 2017. Patients lost to follow-up were censored at the last examination.

Laboratory measurements were obtained using standard methods in the clinical laboratory. Total Kt/V was examined by PD request software 2.0 (Baxter, Deerfield, IL). Patients had quarterly evaluations at the centers and monthly telephone interviews with trained nurses to assess their general condition.

Statistical analysis

The study aimed to explore the possible non-linear relationship between albumin to non-HDL-C ratio and all-cause mortality in PD patients. We used restricted-cubic-spline plots to investigate the shape of association between the albumin to non-HDL-C ratio and mortality. A restricted-cubic-spline function with four knots (located at the 25th, 50th, 75th, and 95th percentiles) was fitted to the data. Based on the findings from our restricted-cubic-spline plots for the primary outcome, we categorized the participants into three groups: Group 1 (Lowest): Participants with an albumin to non-HDL-C ratio of 9.36 or lower. Group 2 (Moderate): Participants with an albumin to non-HDL-C ratio greater than 9.36 but less than or equal to 12.79. Group 3 (Highest): Participants with an albumin to non-HDL-C ratio greater than 12.79 (Figure 2). Continuous variables were reported as the median with the 25th–75th percentile and categorical variables were reported as frequency and percentage. The study analyzed categorical variables using the Chi-square test and skewed continuous variables using the Mann–Whitney U test. The results were considered statistically significant if the p value was less than 0.05. The incidence of all-cause mortality was analyzed using Kaplan–Meier curves and the differences between groups were evaluated using the Log-rank test. The multivariate COX regression models were built using relevant covariates that were found to be significantly associated with all-cause mortality in the multivariate analysis or were listed in Table 1 or considered important for clinical concern.

Figure 2. Restricted-cubic-spline plot of the association between albumin to non-HDL-C ratio and all-cause mortality. The median albumin to non-HDL-C ratio was 10.8.

Table 1. Baseline characteristics of the study population.

		Albumin to non-HDL-C ratio
	Total	Lowest (≤9.36)	Moderate (9.36–12.79)	Highest (>12.79)
	(n = 1954)	(n = 652)	(n = 651)	(n = 651)	p Value
Demographic data
Age (Y)	51 (40,62)	55 (44,65)	50 (40,61)	48 (38,60)	<0.001
Male sex (%)	1082 (55.4)	322 (49.4)	364 (55.9)	396 (60.8)	<0.001
BMI (kg·m^−2)	21.8 (20.0,24.0)	22.1 (20.1,24.4)	21.8 (20.1,24.0)	21.5 (19.8,23.7)	0.093
Systolic BP (mmHg)	141 (131,157)	142 (133,159)	140 (132,154)	141 (130,156)	0.028
Diastolic BP (mmHg)	85 (79,94)	84 (79,94)	85 (78,93)	86 (79,94)	0.876
Current smoking (%)	81 (4.1)	27 (4.1)	31 (4.8)	23 (3.5)	0.539
Current alcohol consumption (%)	32 (1.6)	13 (2.0)	9 (1.4)	10 (1.5)	0.664
Comorbid
Hypertension (%)	1388 (71.0)	475 (72.9)	467 (71.7)	446 (68.5)	0.200
Diabetes (%)	396 (20.3)	193 (29.6)	111 (17.1)	92 (14.1)	<0.001
History of CVE (%)	216 (11.1)	109 (16.7)	65 (10.0)	42 (6.5)	<0.001
Hyperlipidemia (%)	95 (4.9)	69 (10.6)	20 (3.1)	6 (0.9)	<0.001
Peritonitis (%)	298 (15.3)	110 (5.6)	115 (5.9)	73 (3.7)	0.002
Treatments
CCB (%)	1446 (74.0)	510 (78.2)	504 (77.4)	432 (66.4)	<0.001
Beta-blocker (%)	706 (36.1)	259 (39.7)	248 (38.1)	199 (30.6)	0.001
Diuretics (%)	136 (7.0)	61 (9.4)	46 (7.1)	29 (4.5)	0.002
ACEI/ARB use (%)	780 (39.9)	284 (43.6)	297 (45.6)	199 (30.6)	<0.001
Aspirin (%)	191 (9.8)	86 (13.2)	57 (8.8)	48 (7.4)	0.001
Insulin (%)	274 (14.0)	137 (21.0)	74 (11.4)	63 (9.7)	<0.001
Statin (%)	221 (11.3)	94 (14.4)	70 (10.8)	57 (8.8)	0.005
Laboratory variables
WBC (*10^9·L^−1)	6.2 (4.9,7.7)	6.7 (5.5,8.2)	6.1 (5.0,7.7)	5.7 (4.5,7.0)	<0.001
Hemoglobin (g·L^−1)	89 (75,107)	94 (79,110)	91 (75,107)	83 (70,98)	0.004
Albumin (g·L^−1)	35.7 (32.0,39.0)	33.3 (29.3,37.0)	36.0 (32.5,39.3)	37.1 (34.2,40.5)	<0.001
Albumin to non-HDL-C ratio	10.8 (8.6,14.1)	7.8 (6.5,8.6)	10.8 (10.1,11.8)	15.8 (14.1,18.8)	<0.001
Cholesterol (mmol·L^−1)	4.5 (3.6,5.3)	5.6 (5.0,6.3)	4.5 (4.0,4.9)	3.4 (3.0,3.9)	<0.001
Triglycerides (mmol·L^−1)	1.3 (0.9,1.9)	1.7 (1.2,2.4)	1.4 (1.0,1.8)	1.0 (0.8,1.4)	<0.001
LDL (mmol·L^−1)	2.6 (2.0,3.2)	3.4 (2.9,4.0)	2.6 (2.2,3.0)	1.9 (1.5,2.2)	<0.001
HDL (mmol·L^−1)	1.1 (0.9,1.4)	1.1 (0.9,1.4)	1.1 (0.9,1.4)	1.1 (0.9,1.4)	0.719
non-HDL (mmol·L^−1)	3.3 (2.5,4.1)	4.4 (3.9,5.1)	3.3 (2.9,3.7)	2.3 (1.9,2.6)	<0.001
Creatinine (μmol·L^−1)	702 (541,940)	651 (504,871)	730 (561,969)	734 (566,967)	<0.001
Urea Nitrogen (mmol·L^−1)	19.4 (14.8,25.4)	17.4 (13.5,21.8)	19.8 (15.1,25.4)	21.5 (16.4,29.0)	<0.001
Uric acid (mmol·L^−1)	412 (346,489)	406 (345,473)	413 (345,491)	418 (348,509)	0.110
Sodium (mmol·L^−1)	140 (138,142)	140 (138,142)	140.0 (138.0,142.2)	140.0 (138.0,142.0)	0.298
Calcium (mmol·L^−1)	2.1 (1.9,2.3)	2.2 (2.0,2.3)	2.1 (2.0,2.3)	2.1 (1.9,2.2)	0.067
Potassium (mmol·L^−1)	4.1 (3.6,4.7)	3.9 (3.4,4.5)	4.1 (3.6,4.7)	4.2 (3.7,4.8)	0.001
Chlorine (mmol·L^−1)	103.0 (99.0,106.9)	102.0 (98.8,105.0)	102.0 (99.0,106.0)	104.0 (100.0,109.0)	0.133
Total Kt/V	2.2 (1.8,2.7)	2.3 (1.9,2.7)	2.2 (1.8,2.7)	2.2 (1.7,2.6)	0.086
RRF (ml·min^−1·1.73m^−2)	3.2 (1.9,5.0)	3.1 (1.9,5.0)	3.1 (1.9,4.9)	3.5 (2.0,5.1)	0.840
hs-CRP (mg/L)	5.7 (2.0,13,7)	5.3 (1.9,13.2)	5.8 (2.0,13)	6.0 (2.0,15.7)	0.626
nPCR (g·kg^−1·d^−1)	17.2 (1.1,38.0)	2.1 (0.9,30.2)	12.9 (1.1,36.7)	32.4 (8.6,42.9)	<0.001
D/P creatinine	0.69 (0.66,0.72)	0.70 (0.66,0.74)	0.69 (0.66,0.73)	0.67 (0.65,0.69)	<0.001

All continuous variables were skewed distribution, the values for continuous variables were given as P50 (P25, P75).

Abbreviations: BMI: body mass index; BP; blood pressure; CVE: cardiovascular events; CCB: calcium channel blocker; WBC: white blood cell; ACEI/ARB: angiotensin-converting enzyme inhibitor/angiotensin receptor blocker; HDL: high-density lipoprotein cholesterol; LDL: low-density lipoprotein cholesterol; Total Kt/V, K, dialyzer clearance of urea; t: dialysis time; V: the volume of distribution of urea; RRF: renal residual function; hs-CRP: hypersensitive C-reactive protein; nPCR: normalized protein catabolic rate.

The COX regression models considered the time at risk to be the period starting from study entry until the occurrence of one of the following end events: all-cause death, transfer to hemodialysis therapy, transfer to kidney transplantation, transfer of care from the study center, or the end of the study on 31 December 2017. Competitive risk analysis was used to account for the potential impact of the above follow-up end events on all-cause mortality, which allows for the effects of these events to be excluded.

To further explore the potential effect of imbalanced confounders, we performed sensitivity analyses that were categorized into four groups by the median values of non-HDL-C (3.25 mmol/L) and albumin (35.7 g/L). The statistical analysis for the study was carried out using SPSS version 26.0 (SPSS Inc., Chicago, IL) and R software version R-4.2.2. The tests were performed with a bilateral approach and a significance level of p < 0.05 was used to determine statistical significance.

Results

Participants

In this study, 1954 patients were included in the analysis. The median follow-up period was 33 months. During this period, 265 (13.5%) patients had new-onset cardiovascular events, 221 (11.3%) had cardiovascular disease deaths, 538 (27.5%) had all-cause mortality, and other endpoints included transfer to hemodialysis (n = 234), kidney transplantation (n = 90), transfer to other PD centers (n = 25), and loss of follow-up (n = 39). Finally, 1028 patients were followed up to 31 December 2017.

Table 1 summarizes the baseline characteristics of different albumin to non-HDL-C ratio levels in the study population. The median age of the patients was 51 years (range 40–62), with 1082 male and 872 female patients. A total of 1388 (71.0%) of the patients had a history of hypertension, 396 (20.3%) had diabetes, 298 (15.3%) had peritonitis, and 216 (11.1%) had a history of cardiovascular disease. The median albumin to non-HDL-C ratio value is 10.8 (range 8.6–14.1). The lowest group has low levels of serum albumin and high levels of non-HDL-C, while the highest group has high levels of serum albumin and low levels of non-HDL-C. Compared to the moderate group, the highest albumin to non-HDL-C ratio group had younger age and lower prevalence of diabetes, history of cardiovascular disease, hyperlipidemia, and use of medication such as calcium antagonists, beta-blockers, ACE inhibitors or ARBs, diuretics, aspirin, insulin, statins. The highest group also had lower systolic blood pressure, white blood cell count, hemoglobin, estimated glomerular filtration rate, cholesterol, triglycerides, and low-density lipoprotein cholesterol, but higher creatinine, blood urea nitrogen, and potassium levels. On the other hand, the low albumin to non-HDL-C ratio group had a higher prevalence of diabetes, a history of cardiovascular disease, and hyperlipidemia (Figure 1).

Albumin to non-HDL-C ratio predicts the prognosis of all-cause mortality in PD patients

The restricted cubic spline plot had a knot at 10.8. When the value was below 10.8, the hazard ratio (HR) was greater than 1 and decreased steeply as the albumin to non-HDL-C ratio increased. When the value exceeded 10.8, the HR was less than 1, and showed a relatively flat trend (Figure 1). The results from the Kaplan–Meier analysis demonstrated significant differences in all-cause mortality rates between the lowest and moderate groups (log-rank = 24.35, p < 0.001) as well as between the lowest and highest groups (log-rank = 36.03, p < 0.001). However, there was no significant difference observed between the moderate and highest groups (log-rank = 1.128, p = 0.288) (Figure 3). Table 2 shows the associations of albumin to non-HDL-C ratio with all-cause mortality. Multivariate Cox regression suggested that decreased albumin to non-HDL-C ratio is an independent risk factor for all-cause mortality in patients undergoing PD, after adjusting for other factors such as complications, medication, age, sex, BMI, and biochemical examination. Compared to the lowest group (≤9.36), the risk of all-cause mortality was significantly lower in both the moderate group (9.36–12.79) (HR, 0.731; 95% CI, 0.593–0.902, p = 0.004) and the highest group (>12.79) (HR, 0.705; 95% CI, 0.565–0.879, p = 0.002) (Table 2).

Figure 3. The Kaplan–Meier curves with all-cause mortality by category of the level of albumin to non-HDL-C ratio. The curves were constructed using the Kaplan–Meier method and compared using the Mantel-Cox log-rank test. Patients in the highest albumin to non-HDL-C ratio group (albumin to non-HDL-C ratio > 12.79) had a lower risk of all-cause mortality. Log-rank test of lowest vs. moderate, lowest vs. highest, and moderate vs. highest, was 24.35 (p < 0.001), 36.03 (p < 0.001), and 1.124 (p = 0.288), separately.

Table 2. Albumin to non-HDL-C ratio predicts the prognosis of all-cause mortality in PD patients.

Reference group: lowest Albumin to non-HDL-C ratio group
	Moderate group		Highest group
	HR (95%CI)	p Value	HR (95%CI)	p Value
Unadjusted	0.605 (0.494,0.742)	<0.001	0.538 (0.438,0.662)	<0.001
Model 1	0.725 (0.590,0.891)	0.002	0.675 (0.547,0.833)	<0.001
Model 2	0.757 (0.616,0.932)	0.009	0.744 (0.602,0.920)	0.006
Model 3	0.737 (0.598,0.908)	0.004	0.720 (0.580,0.893)	0.003
Model 4	0.731 (0.593,0.902)	0.004	0.705 (0.565,0.879)	0.002

Model 1: sex, age, BMI; Model 2: Model 1 plus hypertension, history of CVE, diabetes; Model 3: Model 2 plus CCB, beta blocker, ACEI/ARB, insulin, aspirin, diuretics; Model 4: Model 3 plus hemoglobin, creatinine, urea nitrogen, total Kt/V, and RRF.

Comparison with other inflammation indicators

The areas under the ROC curve (AUC) of albumin/non-HDL-C, non-HDL-C, albumin, hs-CRP, normalized protein catabolic rate (nPCR), and BMI were 0.604 (95%CI: 0.368–0.425, p < 0.001), 0.583 (95%CI: 0.554–0.612, p < 0.001), 0.552 (95%CI: 0.523–0.581, p < 0.001), 0.609 (95%CI: 0.504–0.710, p < 0.001), 0.523 (95%CI: 0.493–0.552, p = 0.123), and 0.502 (95%CI: 0.473–0.530, p = 0.907), respectively (Figure 4).

Figure 4. A: ROC curves of 3 indicators: albumin/non-HDL-C, non-HDL-C, albumin. The areas under the ROC curve (AUC) of albumin/non-HDL-C, non-HDL-C, albumin were 0.604 (p < 0.001), 0.583 (p < 0.001), and 0.552 (p < 0.001), respectively. B: ROC curves of 4 indicators: albumin/non-HDL-C, hs-CRP, BMI, and nPCR. The areas under the ROC curve (AUC) of albumin/non-HDL-C, hs-CRP, BMI, and nPCR were 0.523 (p = 0.123), 0.502 (p = 0.907), and 0.609 (p < 0.001), respectively.

Competitive risk analysis

In this case, the event of interest is all-cause mortality and the competing events are being transferred to hemodialysis therapy, being transferred to kidney transplantation, and being transferred to other centers. The cumulative incidence curves for different albumin to non-HDL-C ratio groups show significant variation for all-cause mortality (p < 0.001), but no significant variation for the competing events (Figure 5).

Figure 5. Competitive risk models. Estimated cumulative incidence curves between the all-cause mortality and other competing events for each albumin to non-HDL-C ratio group. The cumulative incidence curves for different albumin to non-HDL-C ratio groups are highly significant for the all-cause mortality (p < 0.001), while there is no statistical significance for the occurrence of other competing events (being transferred to hemodialysis therapy, being transferred to kidney transplantation, being transferred to other centers).

Sensitivity analyses

The whole cohort was categorized into four groups by the median values of non-HDL-C (3.25 mmol/L) and albumin (35.7 g/L). The repeated Kaplan–Meier curves show the statistical significance of the association between the redefined groups and all-cause mortality (log-rank p < 0.001). During the 100-month follow-up period, patients who had non-HDL-C ≤ 3.25 mmol/L and albumin levels > 35.7g/L had the lowest risk of all-cause mortality (Figure 6).

Figure 6. Sensitivity analyses. The whole cohort was categorized into four groups by the median values of non-HDL-C (3.25 mmol/L) and albumin (35.7 g/L). The repeated Kaplan-Meier curves showed the statistical significance of the association between the redefined groups and all-cause mortality (p < 0.001).

Discussion

In this retrospective cohort study, patients with albumin to non-HDL-C ratio greater than 12.79 had the least likelihood of mortality. The study found a positively correlated curve between decreased albumin to non-HDL-C ratio and increased mortality risk. Remarkably, despite younger age and lower prevalence of comorbidities in patients with higher baseline albumin to non-HDL-C ratio compared to those with a moderate or lower ratio, they still had lower mortality risk. This suggests that higher levels of albumin to non-HDL-C ratio may be a stronger indicator of poor prognosis in PD patients.

For patients with end-stage renal disease (ESRD) undergoing chronic dialysis, low serum albumin levels are associated with a higher risk of mortality. According to a prospective cohort study conducted across 14 centers in Canada and the United States, which involved 680 consecutive CAPD patients, an increase in plasma albumin concentration by 1 g/dL was associated with a 6% decrease in the relative risk of mortality [27]. In another prior study involving over 8000 hemodialysis patients with low albumin, low serum albumin levels below 4.0 g/dL were found to be the laboratory variable most strongly associated with the probability of mortality [28]. This risk was partly attributed to inflammation instead of malnutrition, as measured by hs-CRP, 7-point subjective global assessment (SGA), and normalized protein equivalent of nitrogen appearance (nPNA)[8]. The Malnutrition-Inflammation Score (MIS) serves as a marker for the malnutrition-inflammation syndrome, which can accurately predict fatal and non-fatal cardiovascular and infectious events in chronic stable PD patients [29]. Furthermore, while serum albumin was often used as a marker of nutritional status, it may not have been the most accurate measure of nutritional status in HD and PD patients. Therefore, albumin to non-HDL-C ratio is proposed as a potential indicator of both inflammation and nutritional status.

High transporters displayed a trend of having lower serum albumin levels, whereas low-average transporters showed a trend of have higher serum albumin levels. These high transporter peritoneal membrane characteristics were identified as a risk factor for inflammation in individuals with ESRD [30,31]. Our study validated the conclusions, as the lowest group exhibited the highest D/P creatinine levels and the lowest serum albumin levels, while the highest group showed the lowest D/P creatinine levels and the highest serum albumin levels.

Non-HDL-C is a type of cholesterol that includes all cholesterol carried by lipoproteins other than HDL. A study conducted on a prospective cohort of 1616 PD patients found that elevated non-HDL-C (>173.75 mmol/L) was a significant independent predictor of mortality in PD patients (95% CI, 1.12–1.39; p < 0.001) [32]. Similar indicators like non-HDL-C/HDL-C ratio and TG/HDL-C ratio can identify high-risk carotid plaque at an early stage, with optimal cutoff values of 1.94 and 2.86, respectively [33]. A study found a U-shaped association between the albumin-to-total cholesterol ratio and mortality in PD patients. This suggests that both low and high ratios may be associated with increased mortality risk, while a moderate ratio may be protective [34]. Another study found hypercholesterolemia was a risk factor for all-cause and cardiovascular disease mortality in patients without malnutrition [35]. This inverse association between total cholesterol levels and mortality may be due to the cholesterol-lowering effects of systemic inflammation and malnutrition, rather than a protective effect of high cholesterol concentrations. Our study has shown a strong, graded, positive association between albumin to non-HDL-C ratio and all-cause mortality in patients. Interestingly, our findings suggest that the masked effect of malnutrition on the association between hypercholesterolemia and mortality may be due to the protective effect of high-density lipoprotein cholesterol.

Based on ROC curves, the albumin-to-non-HDL-C ratio stood out with a relatively high AUC value of 0.604 and a significant p value (<0.001), suggesting its potential as a valuable biomarker for predicting mortality in PD patients. Non-HDL-C and Albumin also performed well but with slightly lower AUC values of 0.583 and 0.552 (p < 0.001, p < 0.001), respectively. The common inflammation marker, hs-CRP, had a poorer performance with an AUC of 0.523 (p = 0.123). The common nutrition marker, BMI, also had a poorer performance with an AUC of 0.502 (p = 0.907). Another common nutrition marker for individuals undergoing dialysis, nPCR, had a slightly better performance with an AUC of 0.609 (p < 0.001). This suggested that the albumin-to-non-HDL-C ratio emerged as a promising new indicator for predicting mortality in PD patients.

In terms of nutritional status, BMI is a commonly used indicator in the general population, but it may not always be an ideal indicator in certain populations, such as patients with chronic kidney disease (CKD) or other conditions that may affect fluid retention and muscle mass. Fluid retention and loss of skeletal muscle mass are common issues in CKD patients, and these factors can influence BMI levels, potentially making it less accurate for assessing nutritional status in such populations. nPCR is a surrogate for daily dietary protein intake and nutritional status in individuals undergoing dialysis. Notably, predictive value of albumin-to-non-HDL-C ratio was found to be comparable to that of nPCR, providing a valuable and complementary insight into mortality prediction, aligning closely with the predictive capabilities demonstrated by nPCR.

In the previous discussion, it was noted that inflammatory marker hs-CRP can impact lipid metabolism, lowering HDL-C levels. High peritoneal transport status is a risk factor for inflammation [30,31]. However, baseline data showed no significant hs-CRP and HDL-C differences among groups. Conversely, the robust nutritional indicator, nPCR, exhibited variations across these groups. D/P creatinine, indirectly indicating inflammation, also showed differences. The highest group had the highest nPCR and the lowest D/P creatinine, and vice versa. Traditionally, malnutrition–inflammation–atherosclerosis (MIA) syndrome is thought to increase the risk of cardiovascular events [36]. Interestingly, the albumin to non-HDL-C ratio was not effective in predicting cardiovascular mortality. Therefore, it can be inferred that this ratio may reflect both malnutrition and inflammatory states, with malnutrition being the more relevant factor.

Overall, CCBs were the most widely used medication. In the lowest group, the usage rates for CCBs, beta-blockers, diuretics, aspirin, insulin, and statins were the highest, whereas in the highest group, the usage rates for these medications were the lowest. The utilization rate of ACEI/ARB was highest in the moderate group and lowest in the highest group, with the lowest group having a slightly lower usage rate than the moderate group. This may be attributed to the higher proportion of CKD patients with comorbidities such as diabetes, hyperlipidemia, and a history of cardiovascular events in the lowest group, while the occurrence of these comorbidities was lower in the highest group. Regrettably, we did not explore the interrelationships among these medications in our study. Statins effectively lower serum cholesterol, triglycerides, LDL cholesterol, apoB, and oxLDL levels, all of which are adverse lipid factors [37–40]. They benefit patients with stages 1–3 of CKD but do not provide significant advantages for long-term cardiovascular or survival outcomes in advanced CKD and dialysis patients [41]. Long-term use (≥12 months) of ACEI/ARB preserves residual kidney function, especially in continuous ambulatory PD (CAPD) patients [42]. Combining a statin with an ARB is more effective in improving vascular function in nondiabetic PD patients [43]. The benefits of CCBs for pre-dialysis blood pressure and intradialytic hypotension in CKD patients requiring hemodialysis remain uncertain [44].

The study has several strengths, including large sample size, high data integrity, and detailed covariates that allow for the adjustment of potential confounding variables. However, several limitations need to be acknowledged, and the findings should be interpreted with caution. First, this is a retrospective study, and the results can only demonstrate the association between albumin to non-HDL-C ratio and adverse prognosis in patients undergoing PD. The study cannot confirm a causal relationship between the two variables. Second, albumin to non-HDL-C ratio values are baseline data, and there is a lack of follow-up data. This limits our ability to evaluate the longitudinal relationship between the ratio and the prognosis of PD patients. Thirdly, the study lacks data on dietary intake, appetite, and other measures of nutritional status, which could provide a more comprehensive assessment of nutritional status. Fourthly, this study lacks data on peritoneal protein clearance rate, which is another important factor influencing overall risk and cardiovascular mortality risk. Further research is needed in this regard. This limits our ability to evaluate the independent contribution of the ratio to the prognosis of PD patients relative to other factors.

Practical application

Albumin to non-HDL-C ratio is an independent risk factor for all-cause mortality in PD patients, even after adjusting for confounding factors. Clinicians should closely monitor PD patients with inflammation and malnutrition, indicated by high non-HDL-C and low albumin levels, especially when both conditions are present. These patients may require increased medical care and monitoring during follow-up. Further research is needed to confirm the clinical utility of this ratio in guiding PD patient management.

Authors Contribution

YJX: Project development, Manuscript writing; XRF: Data analysis and data collection; YQG: Data analysis; XJZ: Data collection; FFP: Data collection; QZ: Provide statistical support; XFW: Data collection; XYW: Assist in data analysis; NT: Data collection; QDX: Data collection; JBL: Data management; JL: Project development, Manuscript editing, Funding. YQW: Data management, Manuscript editing, Funding. All authors have read and approved the manuscript.

Geolocation information

The research conducted in this study focuses on a specific geographic area located in Guangzhou, Guangdong Province, China.

DeclarationsEthics and consent to participate

All procedures performed in this study involving human participants were following the ethical standards of the institution (IRB approval number 2019-hg-ks-04) and with the 1964 Helsinki Declaration as well as its later amendments or comparable ethical standards. This study was warranted by the Institutional Review Board of two PD.

Consent to participate and publication

Owing to we had collected the existing medical records, written informed consent was not required.

Consent for publication

The data released by our research would not compromise anonymity or confidentiality, nor violate local data protection laws. The researchers disclosed no names or other identifying information related to respondents and their parents or guardian in this manuscript.

Competing Interests

The authors declared that they had no financial conflicts of interest.

Acknowledgments

The authors are grateful to all the participants in this study.

Data availability statement

All data generated or analyzed during this study are included in this published article.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Additional information

Funding

This work was supported by [the Science and Technology Program of Guangzhou, China #1] under Grant [202002030336]; [the Scientific and Technological Project of Combining Traditional Chinese Medicine with Traditional Chinese and Western Medicine of Guangzhou, China #2] under Grant [20182A011017]; [Guangzhou Key Discipline of Urology; Second Affiliated Hospital of Guangzhou Medical University Fund Project #3] under Grant [2021-LCYJ-04]; [Clinical Research Project of the Second Affiliated Hospital of Guangzhou Medical University #4] under Grant [2021-LCYJ-DZX-03, 2022-LCYJ-YYDZX-03]; [The open research funds from the Sixth Affiliated Hospital of Guangzhou Medical University, Qingyuan People’s Hospital #4] under Grant [202201-304] and [The Natural Science Foundation of Guangdong Province, China #5] under Grant [2023A1515011721].