Long-term glycemic variability and the risk of mortality in diabetic patients receiving peritoneal dialysis.

BACKGROUND: The large amount of glucose in the dialysate used in peritoneal dialysis (PD) likely affects the glycemic control. The aim of this study was to investigate the association between HbA1c variability, as a measure of long-term glycemic variability, and the risk of all-cause mortality in diabetic patients with PD.
METHODS: 325 patients with diabetes and ESRD were followed (2008-2018) in the Swedish Renal Registry. Patients were separated in seven groups according to level of HbA1c variability. The group with the lowest variability was denoted the reference. The ratio of the standard deviation (SD) to the mean of HbA1c, HbA1c (SD)/HbA1c (mean), i.e. the coefficient of variation (CV), was defined as HbA1c variability. Hazard ratios (HR) and 95% confidence intervals (CI) were examined using Cox regression analyses.
RESULTS: During follow-up, 170 (52%) deaths occurred. The highest mortality was among patients with the second highest HbA1c variability, CV≥2.83 [n = 44 of which 68% patients died]. In the multivariate analyses where lowest HbA1c variability (CV≤0.51) was used as the reference group, HbA1c CV 2.83-4.60 (HR 3.15, 95% CI 1.78-5.55; p<0.001) and CV> 4.6 (HR 2.48, 95% CI 1.21-5.11; p = 0.014) were associated with increased risk of death.
CONCLUSION: The high risk of all-cause mortality in patients with diabetes and PD increased significantly with elevated HbA1c variability, as measure of long-term glycemic control. This indicates that stable glycemia is associated with an improvement of survival; whereas more severe glycemic fluctuations, possibly caused by radical changes in dialysis regimes or peritonitis, are associated with a higher risk of mortality in diabetic patients with PD.

Introduction

Globally, diabetes mellitus (DM) is one of the main causes of end-stage renal disease (ESRD) [1, 2]. Patients suffering from DM and chronic kidney disease (CKD) with renal replacement therapy have higher mortality rates compared to the general population [3, 4]. Peritoneal dialysis (PD) is an established renal replacement therapy (RRT) in ESRD patients [5, 6]. The results of most studies in diabetic patients with ESRD have suggested that mortality is approximately similar in patients treated with PD compared with those receiving hemodialysis (HD) [7, 8].

In DM with normal kidney function, elevated mean glycated hemoglobin (HbA1c) levels as well as high glycemic variability were identified as risk factors for worse clinical outcomes and increased mortality [9, 10]. Also in patients with ESRD and dialysis treatment, HbA1c is the most useful marker of long-term glycemic control [11, 12], despite the shortened erythrocyte life span in ESRD [13]. Moreover, high level of HbA1c variability, as a measure of glycemic variability, has been associated with increased risk of diabetic microvascular and macrovascular complications as well as all-cause mortality [14, 15].

Earlier studies evaluated the association between the mean value of HbA1c and the risk of cardiovascular events or mortality in patients with DM and PD [16-18]. However, the impact of the glucose load from the PD dialysate on glycemic variability in relation to mortality has previously not been investigated in diabetic patients.

The aim of the present study was to investigate the association between HbA1c variability, a measure of glycemic variability, and the risk of all-cause mortality in diabetic patients with maintenance PD.

Material and methods

The Regional Ethics Review Board at the University of Gothenburg approved the study (Dnr 698–17, Exp 2017-09-21) that was performed in accordance with the Declaration of Helsinki. In this study all data were fully anonymized before we accessed them. All patients were informed adequately by hospitals or dialysis centers about the Swedish Renal Register (SRR). The patients were completely anonymous.

In this study, 795 patients (age 66±13 years, 71% men) with DM and ESRD receiving PD treatment were followed 2008–2018. The patients with less than three months on PD were excluded from the study (n = 48). Of the remaining patients, 325 had at least two values of HbA1c and were included in the analyses of HbA1c variability. The mean follow-up period was 3.0±3.2 years. HbA1c was 6.8% ± 2.4% on average (median 6.6%, range 4.2%-19.7%). At baseline, 234 (72%) of the patients received treatment with Continuous ambulatory peritoneal dialysis (CAPD) and 91 (28%) of the patients were on Automated Peritoneal Dialysis (APD).

The data were extracted from the Swedish Renal Register (SRR). SRR is a computerized, web-based quality register in Sweden for patients with chronic renal failure. SRR is based on annual cross-sectional surveys from all nephrology departments and dialysis units in Sweden.

Clinical classification

We defined all patients in the SRR with a diagnosis of DM type 1 or type 2 diabetes or diabetes nephropathy as a patient with DM. Age was defined as the baseline age, whereas the mean values during follow-up were used in the analyses in terms of body mass index (BMI), blood pressure (BP), and laboratory variables.

In this study at baseline, 292 (90%) of the patients had at least one medication for treatment of high blood pressure. We defined hypertension as systolic blood pressure (SBT) > 140 mmHg or diastolic blood pressure (DBP) > 90 mmHg. According to this definition, 146 (47%) of the patients had hypertension at baseline.

All data of cardiovascular disease (CVD) and malignancy were obtained from the SRR.

Statistical methods

Baseline clinical and biochemical characteristics are presented as mean values ± standard deviation (SD) or as frequencies and proportions (n, %). For univariate comparisons with respect to continuous variables, the t-test for independent samples was used and for comparisons with respect to categorical variables, the Chi-2 test was used. HbA1c variability was determined as the coefficient of variation (CV) for HbA1c for each patient (the ratio between the standard deviation (SD) of HbA1c and the mean of HbA1c, i.e. HbA1c (SD)/ HbA1c (mean)). To examine the relationship between HbA1c variability and all-cause mortality in univariate analyses we used Kaplan-Meier survival analysis and for further multivariate analysis we performed Cox regression models and hazard ratios (HR) with 95% confidence intervals (CI) were estimated. Proportional hazards assumption was controlled, and it was not violated. The patients were divided in seven groups according to HbA1c-level. The number of groups was defined based on the purpose to include sufficient number of patients in each class and at the same time yield a wideness in CV for HbA1c. The group of patients with the lowest HbA1c variability CV≤0.51 was used as the reference group for calculations of HR and statistical comparisons. Models were adjusted for the all covariates that turned out as statistically significant or when p-value<0.1 in the univariate analysis. Interactions between HbA1c variability and the covariates were tested and were found to be non-significant for all covariates. All statistical analyses were performed by statistical package IBM SPSS v.25 and p-value<0.05 were considered as statistically significant.

Results

Clinical and biochemical characteristics

Table 1 presents clinical and biochemical characteristics of the included patients. Mean age was 66 ±14 years, mean HbA1c was 6.8 ± 2.4%, and the mean values of SBP and DBP were 137 ± 17 mmHg and 76 ± 9 mmHg, respectively. The patients had a dialysis vintage of 4.3±4.3 years. The majority of the patients had slightly low serum albumin (31 ± 5 g/L). The patients mostly followed Chronic kidney disease–mineral and bone disorder (CKD–MBD) KGIDO [19] as the mean serum phosphate was 4.8 ± 1.0 mg/dl and the mean PTH was 268 ± 170 pg/mL. Twenty-five percent of the patients had a previous history of CVD and three percent had a previous history of malignancy.

Table 1

Clinical and biochemical characteristics in the study population (n = 325) of diabetic patients receiving PD treatment.

	All patients n = 325	Dead n = 170	Alive n = 155	p-value1
Age (years)	65.9 ± 13	69.5±10.6	61.9±13.9	<0.001
Men n (%)	230 (71%)	128 (75%)	102 (66%)	0.060
Time since PD-start (years)	4.3±4.3	4.5±3.5	4.0±5.0	0.316
HbA1c %	6.8± 2.4	6.9±2.5	6.8±2.6	0.292
Systolic blood pressure (mmHg)	137 ± 17	137±19	136±15	0.666
Diastolic blood pressure (mmHg)	76 ± 9	73±9	78±9	<0.001
MAP	96±10	95±10	97±9	0.012
Antihypertensive treatment, n(%)	311 (96%)	160 (94%)	151 (97%)	0.143
BMI (kg/m 2 )	27 ± 4.6	27±5	27±4	0.952
Hemoglobin (g/L)	118 ± 10	117±10	119±10	0.327
CRP (mg/mL)	11.5 ± 14.8	13.5±16.6	9.2±12.2	0.008
Serum albumin (g/L)	31 ± 4.7	29.8±4.3	32.3±4.7	<0.001
Serum phosphate (mg/dL)	4.8 ± 1.0	4.8±1.1	4.9±1.1	0.452
PTH (pg/mL)	268±170	267±190	280±159	0.268
Total cholesterol (mmol/L)	4.5 ± 1.3	4.5±1.2	4.6±1.3	0.480
Previous history of CVD n(%)	83 (25%)	55 (32%)	28 (18%)	0.003
Previous history of malignancy n(%)	10 (3%)	4 (2%)	6 (4%)	0.429

Data are presented as means ± SD or frequencies (%). Age was defined as the baseline age, whereas BMI, blood pressure, and biochemical variables for each patient were defined as the average of all measurements during the study period.

1p-value consider comparison between dead and alive patients.

In the analyses of HbA1c CV, there were different number of HbA1c measurements in individual patients (between 2 and 12 measurements), but the number of measurements was associated with the level of HbA1c variability, Table 2.

Table 2

Distribution of number of HbA1c measurements per patient according to classes of HbA1c variability.

	Number of measurements per patient % (n)
Coefficient of variation (CV) for HbA1c	2 measurements	3–4 measurements	5–12 measurements
CV≤ 0.51; n = 65	55.4% (36)	21.5% (14)	23.1% (15)
0.51 < CV ≤ 0.77; n = 52	36.5% (19)	38.5% (20)	25.0% (13)
0.77 < CV ≤ 1.32; n = 53	32.1% (17)	39.6% (21)	28.3% (15)
1.32 < CV ≤ 1.77; n = 35	54.3% (19)	22.9% (8)	22.9% (8)
1.77 < CV ≤ 2.83; n = 53	37.7% (20)	50.9% (27)	11.3% (6)
2.83 < CV ≤ 4.60; n = 44	75.0% (33)	22.7% (10)	2.3% (1)
CV > 4.60; n = 23	65.2% (15)	34.8% (8)	0% (0)
Total; n = 325	48.9% (159)	33.2% (108)	17.8% (58)

All-cause mortality

During the follow-up, 170 (52%) of the patients died. In the subgroup of patients > 75 years old (n = 77), mortality was 70% and in patients ≤ 75 years old (n = 248), mortality was 47%. The patients who died during the study compare to all patients were older 69.5±10.6 years, more men (75%), lower DBP 73±9 mmHg and 32% had previous history of CVD.

Association between HbA1c variability and the risk of all-cause mortality

The association between HbA1c variability and the risk of all-cause mortality is shown in Table 3 and Fig 1. The patients with the lowest HbA1c variability (the reference group) had the lowest rate of mortality (CV≤0.51, n = 25, 38% died). The highest incidence of mortality was observed in the group with CV≥2.83 (n = 44, 68%).

Table 3

Univariate and multivariate analyses of the association between HbA1c variability and the risk of all-cause mortality.

HbA1c Coefficient of variation (CV)	Univariate analyses		Multivariate analyses
	HR (95% CI)	p-value	HR (95% CI)	p-value
CV≤ 0.51; n = 65	reference		reference
0.51 < CV ≤ 0.77; n = 52	1.23 (0.69–2.20)	0.474	1.12 (0.62–2.01)	0.701
0.77 < CV ≤ 1.32; n = 53	1.42 (0.83–2.41)	0.197	1.29 (0.75–2.20)	0.350
1.32 < CV ≤ 1.77; n = 35	1.62 (0.89–2.95)	0.117	1.35 (0.73–2.49)	0.336
1.77 < CV ≤ 2.83; n = 53	1.50 (0.88–2.56)	0.132	1.37(0.80–2.35)	0.250
2.83 < CV ≤ 4.60; n = 44	3.66 (2.11–6.33)	<0.001	3.15 (1.78–5.55)	<0.001
CV > 4.60; n = 23	3.09 (1.56–6.12)	0.001	2.48 (1.21–5.11)	0.014

Hazard ratios (HR) with 95% confidence intervals (CI) were calculated using Cox regression analyses. In the multivariate analyses, adjustments were made for the all variables that turned out as statistically significant in Table 1.

Fig 1

All-cause survival (Kaplan-Meier curves) according to HbA1c variability.

Kaplan–Meier curves showing the survival reduced with increase of HbA1c variability.

In the multivariate analyses, in which adjustments were made for age, sex, MAP, CRP, serum albumin and CVD, the risk of all-cause mortality was significantly increased in the HbA1c variability group with CV 2.83–4.60 (HR 3.15, 95% CI 1.78–5.55, p-value<0.001), and the HbA1c variability group with CV> 4.6 (HR 2.48, 95% CI 1.21–5.11, p-value = 0.014), Table 3.

Median survival time in total population was (HR 3.1, 95% CI 2.7–3.4). The two groups with highest CV analyses had lowest median survival time; CV 2.83–4.60 (HR 1.9, 95% CI 1.6–2.2) and CV> 4.60 (HR 2.0, 95% CI 1.4–2.6). The group of patients with lowest CV≤0.5 had highest median survival time (HR 4.3, 95% CI 2.6–5.9), p-value<0.001 (Table 4).

Table 4

Survival data according to HbA1c variability.

Coefficient of variation (CV) for HbA1c	Nr of patients at risk	Mortality n (%)	Median survival time (years) with 95% CI
CV ≤ 0.51	65	25 (38%)	4.3 (2.6–5.9)
0.51 < CV ≤ 0.77	52	22 (42%)	3.8 (2.9–4.6)
0.77 < CV ≤ 1.32	53	31 (58%)	3.9 (3.0–4.8)
1.32 < CV ≤ 1.77	35	19 (54%)	3.5 (2.2–4.8)
1.77 < CV ≤ 2.83	53	30 (57%)	3.3 (2.5–4.2)
2.83 < CV ≤ 4.60	44	30 (68%)	1.9 (1.6–2.2)*
CV > 4.60	23	13 (56%)	2.0 (1.4–2.6)*
Total	325	170 (52%)	3.1 (2.7–3.4)

*Statistically significant compared in pairwise comparisons vs all other classes (CV<2.83).

Free survival rates between the seven subgroups during the followed-up time by Kaplan–Meier survival curves is shown in Fig 1. The mean survival time was reduced in groups with higher CV of HbA1c. The association between HbA1c variability and all-cause mortality is visualized in Fig 2. Thus, in both the unadjusted and adjusted models, the HR of death increased markedly i with higher HbA1c variability.

Fig 2

Hazard ratio (HR) of all-cause mortality according to variability of HbA1c levels.

The association between HbA1c variability and the risk of all-cause mortality. Hazard ratios (HR) and 95% confidence intervals (CI) based on univariate and multivariate model are presented.

Discussion

In this study, 795 diabetic patients with ESRD and PD treatment were followed for a mean of 3.0± 3.2 years. Of these patients, the 325 who had at least two measurements of HbA1c were included in the analyses of HbA1c variability. 49% of the patients had 2 measurements and 51% had between 3 and 12 measurements. We showed that high HbA1c variability, as a measure of long-term glycemic variability, was associated with a markedly increased risk of all-cause mortality in diabetic patients with ESRD and PD treatment.

The mean age of our population was 66 years, which is almost similar to the general population with ESRD and dialysis treatment. The mean HbA1c value was 6.8%, which indicates that most of the diabetic patients had adequate long-term glycemic control. Furthermore, the patients mostly had acceptable BP during the study (SBP 137 ±17 mmHg, DBP 76 ±9mmHg). The mean serum albumin was also under adequate control (31±5 g/L). As a result, our population was mostly relevant in terms of the aims of the present study.

Previous studies have assessed the relationship between the mean HbA1c level and the risk of CVD or mortality in diabetic patients with ESRD and dialysis [20, 21]. The results are diverse, but in general, HbA1c>8% was associated with higher mortality risk [22]. Moreover, in several studies, HbA1c<6% was associated with elevated risk of mortality in chronic dialysis [20, 23]. Based on these findings, it has been suggested that there is a J- or U-shaped relationship between the absolute HbA1c level and the risk of mortality [11, 14]. However, in the present study of diabetic patients with ESRD and PD treatment, only high HbA1c variability was associated with increased risk of all-cause mortality. In patients with DM and preserved renal function, increased glycemic variability, which suggests more severe glycemic fluctuations, has been associated with increased risk of cardiovascular events [12, 15]. Furthermore, previous studies have indicated that fluctuations in glycemic control could cause more damage to endothelial cells than a persistently high level of glucose [24, 25]. Underlying mechanisms could be that glucose fluctuations lead to oxidative stress and impaired endothelial function [26]. Elevated glucose variability also increases the occurrence of hypoglycemia, which is associated with a higher risk of mortality among patients with DM [27-29].

However, little is known of the importance of glycemic variability in patients with CKD.

Most previous studies have included the patients with mild to moderate renal dysfunction. The results of a study by Lee et al. showed that greater HbA1c variability was associated with an increased risk of cardiovascular events in type 2 DM with preserved renal function, but not in moderate to advanced CKD. This study included 1834 patients with eGFR < 60 min/ml/1.73m2 (mean eGFR 39.9 ± 14.8 min/ml/1.73m2; follow-up 6.3 years), but the patients that received dialysis were not examined separately [30]. In stages 3–4 of CKD, greater HbA1c variability was associated with a decreased risk of progression to dialysis, but this association was not found in CKD stage 5. Although patients receiving PD treatment were not specifically analyzed, this could suggest that in diabetic patients, aggressive glycemic control in CKD stages 3–4 is most important in terms of the clinical outcomes [31].

The results of the Dialysis Outcomes and Practice Patterns (DOPPS) study suggested that higher HbA1c values than those currently recommended for a dialysis population were associated with less variability of glucose and a decreased risk of mortality in patients with DM and HD [14].

In diabetic patients receiving PD, a large amount of glucose is absorbed from the dialysate, which obviously could increase glucose variability as compared with a HD population. Additionally, the impact of changes in the dialysis regime by the use of lower or higher glucose concentration in the dialysate as well as the importance of peritonitis and inflammation on glucose fluctuations in PD is unclear. To our knowledge, the present study is the first to evaluate the association between HbA1c variability, as a measure of long-term glycemic fluctuations, and the risk of all-cause mortality in diabetic patients with PD.

In the present study, we used HbA1c as a measure of long-term glycemic control, but the validity of this marker in patients with ESRD and dialysis is debated [13]. The use of HbA1c in populations with ESRD and dialysis has several limitations since the shortened erythrocyte life span and metabolic acidosis can affect the HbA1c level [32]. Moreover, it has been suggested that glycated albumin and fructosamine could be better markers of long-term glycemic control than HbA1c in a dialysis population [33, 34]. However, the level of glycated albumin is highly affected by inflammation and low serum albumin, which is common in PD [32]. Therefore, as in the present study, HbA1c is generally used to evaluate long-term glycemic control in ESRD patients receiving dialysis treatment.

The major strengths of this cohort study are the nationwide scale and the large number of patients and events, which likely resulted in adequate statistical power and high external validity. Furthermore, the majority of our patients were followed according to national guidelines for PD treatment. Thus, the study population was mostly homogeneous, which suggests high internal validity. Furthermore, high BMI and obesity are important risk factors for mortality among PD patients [35, 36]. Our patients tended to be slightly overweight (BMI 27 ± 4.6 kg/m2), but in the multivariate analyses, the association between high HbA1c variability and increased risk of all-cause mortality remained significant also after adjustment for BMI and multiple other covariates. Finally, the high incidence of mortality (52% of the patients died) is consistent with the fact that 24% of our patients was more than 75 years old.

The present study has some limitations. In our observational study, a cause–effect relationship cannot be established. The data on dialysis effectivity (Kt/v), type of peritoneum (“slow transporters”, “fast transporters”) and dialysis regime were not recorded, and especially the influence of the use of icodextrin or non-glucose based peritoneal dialysis solution (Nutrineal) on glycemic variability would have been of interest to delineate [37-41]. Furthermore, data on peritonitis as a complication of PD, which can affect ultrafiltration and glucose absorption [42], were not available in the present study. All the included patients were treated with insulin, but the daily insulin doses were not available. Finally, of the totally 795 patients with DM and ESRD receiving PD treatment, we could include the 325 patients with at least three months on PD and two or more values of HbA1c.

In conclusion, to our knowledge, the present study is the first to evaluate the association between glycemic variability and the risk of all-cause mortality in diabetic patients receiving PD treatment. In this observational study using data from the SRR of diabetic patients with maintenance PD, high HbA1c variability, as a measure of long-term glycemic control, was significantly associated with increased risk of all-cause mortality. Therefore, higher magnitudes of glycemic fluctuations, which might be caused by radical changes in dialysis regimes or peritonitis, are associated with higher risk of mortality in this group of patients. Further studies are needed to evaluate whether reduced glycemic variability by improved clinical care can reduce the high mortality seen in patients with DM and PD.

14 Jul 2021

PONE-D-21-18957

LONG-TERM GLYCEMIC VARIABLITY AND THE RISK OF MORTALITY IN DIABETIC PATIENTS RECEIVING PERITONEAL DIALYSIS

PLOS ONE

Dear Dr. Peters,

Thank you for submitting your manuscript to PLOS ONE. After careful consideration, we feel that it has merit but does not fully meet PLOS ONE’s publication criteria as it currently stands. Therefore, we invite you to submit a revised version of the manuscript that addresses the points raised during the review process.

Please address all the issues raised by the reviewer. If you do not have the clinical data necessary to make the association that you have proposed in the title robust and consistent with clinical reality, you may shorten the article with a statistical description of what you find from the data available. 

Please submit your revised manuscript by Aug 26 2021 11:59PM. If you will need more time than this to complete your revisions, please reply to this message or contact the journal office at plosone@plos.org. When you're ready to submit your revision, log on to https://www.editorialmanager.com/pone/ and select the 'Submissions Needing Revision' folder to locate your manuscript file.

Please include the following items when submitting your revised manuscript:

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:  http://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,

Vivekanand Jha

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

and

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

2. In your ethics statement in the Methods section and in the online submission form, please provide additional information about the data used in your retrospective study. Specifically, please ensure that you have discussed whether all data were fully anonymized before you accessed them and/or whether the IRB or ethics committee waived the requirement for informed consent. If patients provided informed written consent to have data from their medical records used in research, please include this information.

3. We note that you have indicated that data from this study are available upon request. PLOS only allows data to be available upon request if there are legal or ethical restrictions on sharing data publicly. For information on unacceptable data access restrictions, please see http://journals.plos.org/plosone/s/data-availability#loc-unacceptable-data-access-restrictions.

In your revised cover letter, please address the following prompts:

a) If there are ethical or legal restrictions on sharing a de-identified data set, please explain them in detail (e.g., data contain potentially identifying or sensitive patient information) and who has imposed them (e.g., an ethics committee). Please also provide contact information for a data access committee, ethics committee, or other institutional body to which data requests may be sent.

b) If there are no restrictions, please upload the minimal anonymized data set necessary to replicate your study findings as either Supporting Information files or to a stable, public repository and provide us with the relevant URLs, DOIs, or accession numbers. Please see http://www.bmj.com/content/340/bmj.c181.long for guidelines on how to de-identify and prepare clinical data for publication. For a list of acceptable repositories, please see http://journals.plos.org/plosone/s/data-availability#loc-recommended-repositories.

We will update your Data Availability statement on your behalf to reflect the information you provide.

4. We noticed you have some minor occurrence of overlapping text with the following previous publication, which needs to be addressed:

- https://academic.oup.com/ndt/article/34/Supplement_1/gfz101.SaO059/5515881

In your revision ensure you cite all your sources (including your own works), and quote or rephrase any duplicated text outside the methods section. Further consideration is dependent on these concerns being addressed.

[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

Reviewer #2: Partly

**********

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

Reviewer #1: N/A

Reviewer #2: N/A

**********

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: No

Reviewer #2: 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

Reviewer #2: 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: The authors have studied the long term glycemic variability and risk of mortality in diabetics on peritoneal dialysis (PD), which is a novel and good concept. May I comment the following:

1. As the title mentions 'long term variability', it is preferable to have a follow-up of at least one year on PD with four HbA1c readings. In the manuscript, follow-up of 3.0+/3.2 years has been mentioned, which means that some patients had a very short follow-up and some patients had only two HbA1c readings. Also, since data is not normally distributed, better to present it as median and range

2. Kindly mention the PD regimes in the population studied and adequacy of PD (if possible)

3. Separating the patients in standard five quintiles may be preferable

4. Would like to know the mortality in a similar group of non-diabetics on PD in the Registry

Reviewer #2: 1. This seems to be a retrospective observational study retrieving data from a National (Swedish) Registry. The aim of this study was to investigate the association between HbA1c variability, as a measure of long-term glycemic variability, and the risk of all -cause mortality in diabetic patients with PD.

2. In the first paragraph of the Methods section, the authors state that 795 diabetic PD patients were followed up between 2008-2018, but 48 were excluded as they had been on PD for less than three months. Of the remaining patients, 325 had at least two values of HbA1c and were included in the HbA1c variability analyses with a mean follow-up period of 3.0 �  3.2 years. Would you confirm that 322 were excluded for not having, at least, two values of HbA1c ? Therefore, a total of 370 patients (roughly 46 % of all diabetic PD patients on the Registry) were excluded for either being on PD for less than three months or not having two values or more of HbA1c during the follow-up period (3.0 �  3.2 years).

3. Certainly, all 795 diabetic PD patients followed up between 2008-2018 must have had more than two values of HbA1c analysed during their follow-up period, but most likely not captured for the Registry. Can this be considered a bias ? a limitation of the study ? and what about the variability of the results measured in different laboratories around Sweden ? were all PD patients in Sweden evaluated for their HbA1c levels with the same lab method ?

4. Is it possible to know how many of the 322 patients excluded for not having, at least two values of HbA1c, had no (0) values of HbA1c ?

5. In the Introduction the authors write ……… Earlier studies evaluated the association between the mean value of HbA1c and the risk of cardiovascular events or mortality in patients with DM and PD [16-18]. However, the impact of the glucose load from the PD dialysate on glycemic variability has not been previously investigated in diabetic patients. In order to evaluate the impact of glucose load from the PD fluid on glycemic variability, it is fundamental to have important clinical information which may significantly impact on HbA1c variability , the most important being:

a. Data on which modality (APD or CAPD) and the daily prescription for each patient. Were there switches from APD to CAPD or from CAPD to APD in any of the patients included, during the observation period ?

b. CAPD daily dialysis prescription: number of exchanges, volume per exchange, glucose concentration in each glucose-based PD solution, use of the non-glucose based PD solutions icodextrin (Extraneal) and or aminoacids solution (Nutrineal)

c. APD daily dialysis prescription: number of exchanges, volume per exchange, total volume used, glucose concentration in each glucose-based PD solution, use of the non-glucose based PD solutions icodextrin (Extraneal) and or aminoacids solution (Nutrineal)

d. Daily insulin prescription in case of insulin-dependent diabetics and other drugs for non-insulin dependent diabetics. How many insulin-dependent and how many non-insulin dependent patients were on the study ?

e. How was the peritoneal membrane transport characteristics distribution among the CAPD and APD patients ? High transporters usually absorb more glucose and this can impact on morbidity and mortality.

f. Information on blood pressure (percentage of hypertensive patients) and/or fluid overload status.

g. Use of BRA, ACEi, aspirin, beta-blockers

h. Presence of co-morbidities ( Charlson index ?)

i. Information on peritonitis rates, Kt/V, residual renal function.

6. In the introduction the Authors wrote…. However, the impact of the glucose load from the PD dialysate on glycemic variability has not been previously investigated in diabetic patients. Information above described from a to h should be provided in order to corroborate this assumption.

7. The authors conclude….” In conclusion, to our knowledge, the present study is the first to evaluate the association between glycemic variability and the risk of all-cause mortality in diabetic patients receiving PD treatment. In our diabetic patients with maintenance PD, high HbA1c variability, as a measure of long-term glycemic control, was significantly associated with increased risk of all-cause mortality. Therefore, higher magnitudes of glycemic fluctuations, which might be caused by radical changes in dialysis regimes or peritonitis, are associated with higher risk of mortality in this group of patients. Further studies are needed to evaluate whether improved clinical care can reduce glycemic variability as well as the high mortality seen in patients with DM and PD. “Are they sure that “they are the first to evaluate the association between glycemic variability and the risk of all-cause mortality in diabetic patients receiving PD treatment.” ?

8. When the authors write “In our diabetic patients with maintenance PD…”, isn’t it more appropriate to write “In the SRR diabetic patients group with maintenance PD…” ?

9. I cannot see support to the following sentence “Therefore, higher magnitudes of glycemic fluctuations, which might be caused by radical changes in dialysis regimes or peritonitis, are associated with higher risk of mortality in this group of patients “ as there is no information or data on dialysis prescriptions or peritonitis rates analysed in this study as well as of a definition of the word “radical” in the context of “radical changes in dialysis regimes or peritonitis “. In order to address this association (changes in dialysis prescriptions or peritonitis) with mortality, what is needed to investigate is primarily not the glycemic variability (a consequence of glucose exposure), but the changes in the daily glucose exposure/load and its clinical impact on different sub-groups of diabetic PD patients. In diabetic PD patients, higher HbA1c levels may indicate greater cumulative peritoneal glucose exposure with its attendant damage to the peritoneal membrane

10. The authors wrote as last sentence in the conclusion…. “Further studies are needed to evaluate whether improved clinical care can reduce glycemic variability as well as the high mortality seen in patients with DM and PD. “. I would challenge this conclusion using as basis, two not recent published randomized trials (Paniagua et al, de Moraes et al), already showing improved clinical care by improving metabolic control, decreasing glucose load and exposure as well as optimizing fluid management in PD patients (both diabetics and non-diabetics). I suggest reading the RCT CAPD study by Ramon Paniagua et al. and the RCT APD study by Thyago de Moraes et al.

Paniagua R, Ventura MD, Avila-Diaz M et al. Icodextrin improves metabolic and fluid management in high and high-average transport diabetic patients. Perit Dial Int 2009; 29: 422–432

de Moraes T.P., Andreoli M.C., Canziani M.E. et al. Icodextrin reduces insulin resistance in non-diabetic patients undergoing automated peritoneal dialysis: results of a randomized controlled trial (STARCH). Nephrol Dial Transplant. 2015; 30: 1905-1911

I also suggest reading the RCT Impendia study by Li PK et al

In the IMPENDIA study, the primary endpoint was change in glycated hemoglobin from baseline. Mean glycated hemoglobin at baseline was similar in both groups. During the six months of the study, in the intention-to-treat population, the mean glycated hemoglobin profile improved in the intervention group but remained unchanged in the control group (0.5% difference between groups; 95% confidence interval, 0.1% to 0.8%; P=0.006).

Li PK, Culleton BF, Ariza A et al. Randomized, controlled trial of glucose sparing peritoneal dialysis in diabetic patients. J Am Soc Nephrol 2013; 24: 1889–1900

As well as reading the paper by McIntyre et al on glycemic control in diabetic CAPD patients assessed by CGMS

A practical approach to reduce disturbances of the carbohydrate metabolism in PD patients is the reduction of glucose exposure by also prescribing glucose-sparing solutions. In a study involving eight diabetic CAPD patients, replacement of a glucose-based regimen with a Physioneal-Extraneal-Nutrineal regimen was associated with a reduction in the 24-hour variability of glucose concentrations as measured by a subcutaneous probe in the interstitial fluid of the abdominal wall.

Marshall J, Jennings P, Scott A, Fluck RJ, McIntyre CW: Glycemic control in diabetic CAPD patients assessed by continuous glucose monitoring system (CGMS). Kidney Int 64: 1480–1486, 2003

11. Lastly, it should be taken into account that in type 1 and 2 diabetics (not in dialysis), HbA1c is not a good predictor of cardiovascular disease (CVD), whereas insulin resistance is predictive of CVD and indeed may be the most important single cause of coronary artery disease (see for example: Home P. Diabetes Care 2019; 42:1615-23; Adeva-Andany MM et al. Diabetes Metab Syndr 2019; 13:1449-55; Shahim B. et al. Diabetes Care 2017:40:1233-40; Eddy D. et al. Diabetes Care 2009; 32; 361-66; Orchard TJ et al. 2003; 26:1374-79). In addition, since patients in dialysis are often affected by subclinical/clinical anemia, which reduces red blood cell survival and hence Hb glycosylation, this introduces an additional bias in the interpretation of HbA1c variability. At this regard, it should be kept in mind that the simple evaluation in dialysis patients of total Hb levels would be of little help in figuring out potential differences in the glycosylation rate as total Hb levels. Indeed, the latter (Hb levels) may be easily corrected by Epo treatment, but it doesn’t tell you anything about survival rate of RBCs, one of the major determinant of the extent of HB glycosylation.

In my opinion the paper needs a major revision. If the Authors do not have the clinical data necessary to make the association in the title robust and as a reflection of the clinical reality, you may shorten the article with a statistical description of what you see from the data available. I am not an expert in Statistics, but as a clinician, I consider the clinical data I described above of utmost importance in order to keep the paper in the present format. So I really hope you have the data I am suggesting to add to the paper.

**********

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

Reviewer #2: Yes: Jose Carolino Divino-Filho

[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.

26 Sep 2021

Responses to Reviewers comments

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

Response to comment 1: Yes, we have checked PLOS ONE´s style requirements.

2. In your ethics statement in the Methods section and in the online submission form, please provide additional information about the data used in your retrospective study. Specifically, please ensure that you have discussed whether all data were fully anonymized before you accessed them and/or whether the IRB or ethics committee waived the requirement for informed consent. If patients provided informed written consent to have data from their medical records used in research, please include this information.

Response to comment 2: In this study all data were fully anonymized before we accessed them. All patients were informed adequately by hospitals or dialysis centers about SRR. The patients were completely anonymous.

3. We note that you have indicated that data from this study are available upon request. PLOS only allows data to be available upon request if there are legal or ethical restrictions on sharing data publicly. For information on unacceptable data access restrictions, please see http://journals.plos.org/plosone/s/data-availability#loc-unacceptable-data-access-restrictions.

In your revised cover letter, please address the following prompts:

a) If there are ethical or legal restrictions on sharing a de-identified data set, please explain them in detail (e.g., data contain potentially identifying or sensitive patient information) and who has imposed them (e.g., an ethics committee). Please also provide contact information for a data access committee, ethics committee, or other institutional body to which data requests may be sent.

Response to comment 3:

The raw data used in the current study are restricted in order to protect participant privacy, as required by data protection acts in Sweden. Data can be made accessible by request for researchers after permission from the Swedish Ethics Review Authority. Data cannot be shared publicly because of data policy regulations at Skaraborg hospital. Data are available from the Skaraborg hospital Institutional Data Access (contact via skas.dso@vgregion.se) for researchers who meet the criteria for access to confidential data.

b) If there are no restrictions, please upload the minimal anonymized data set necessary to replicate your study findings as either Supporting Information files or to a stable, public repository and provide us with the relevant URLs, DOIs, or accession numbers. Please see http://www.bmj.com/content/340/bmj.c181.long for guidelines on how to de-identify and prepare clinical data for publication. For a list of acceptable repositories, please see http://journals.plos.org/plosone/s/data-availability#loc-recommended-repositories.

We will update your Data Availability statement on your behalf to reflect the information you provide.

4. We noticed you have some minor occurrence of overlapping text with the following previous publication, which needs to be addressed:

- https://academic.oup.com/ndt/article/34/Supplement_1/gfz101.SaO059/5515881

In your revision ensure you cite all your sources (including your own works), and quote or rephrase any duplicated text outside the methods section. Further consideration is dependent on these concerns being addressed.

Response to comment 4: The above-mentioned publication was an abstract presented at ERA-EDTA-congress in Budapest/Hungary in 2019.

Reviewer 1

1. As the title mentions 'long term variability', it is preferable to have a follow-up of at least one year on PD with four HbA1c readings. In the manuscript, follow-up of 3.0+/3.2 years has been mentioned, which means that some patients had a very short follow-up and some patients had only two HbA1c readings. Also, since data is not normally distributed, better to present it as median and range?

Response from authors: In this study, we used HbA1c values to estimate glycemic variability, As HbA1c reflect the glycemic control during the last two to three months, and that we in addition had a follow-up of 3.0 ± 3.2 years, we believe that 'long-term glycemic variability' is adequate in the title. In this study, the median HbA1c was 6.6% and the range was between 4.2-19.7% (Material and Methods, first paragraph, fifth sentence).

2. Kindly mention the PD regimes in the population studied and adequacy of PD (if possible):

Response from authors: A. In a large number of patients, PD regimes were changed during the follow-up period. Furthermore, in a significant number of patients, PD regimens were even changed more than one time. It is therefore very difficult to present data on the specific PD regimes, which is the reason why we did not include such data in the manuscript. However, at baseline, 72% of the patients received treatment with CAPD regime and 28% was on APD, which is now stated in the manuscript (Material and Methods, first paragraph, last sentence). B. We do not have data on dialysis adequacy of PD. The study limitations are now described more extensively in the second last paragraph of Discussion.

3. Separating the patients in standard five quintiles may be preferable?.

Response from authors: We have also analyzed HbA1c variability according to five quintiles. This analysis showed that the last quintile (CV>3.2) was associated with higher risk of mortality, HR=3.2 compared to first quantile (CV<0.67) as reference category. However, our aim was to evaluate whether very low and very high degree of HbA1c variability are related to increased risk of mortality. As a result, we believe that presenting Hba1c variability in seven categories is relevant.

4. Would like to know the mortality in a similar group of non-diabetics on PD in the Registry: Response from authors: Unfortunately, we have no data on non -diabetic patients.

Reviewer 2

1. This seems to be a retrospective observational study retrieving data from a National (Swedish) Registry. The aim of this study was to investigate the association between HbA1c variability, as a measure of long-term glycemic variability, and the risk of all -cause mortality in diabetic patients with PD.

Response from authors. We thank the Reviewer for this description of the design and aims of our study.

2. In the first paragraph of the Methods section, the authors state that 795 diabetic PD patients were followed up between 2008-2018, but 48 were excluded as they had been on PD for less than three months. Of the remaining patients, 325 had at least two values of HbA1c and were included in the HbA1c variability analyses with a mean follow-up period of 3.0 ± 3.2 years. Would you confirm that 322 were excluded for not having, at least, two values of HbA1c ? Therefore, a total of 370 patients (roughly 46 % of all diabetic PD patients on the Registry) were excluded for either being on PD for less than three months or not having two values or more of HbA1c during the follow-up period (3.0 ± 3.2 years).

Response from authors: The Reviewer is correct. Thus, the majority of the patients that were excluded had only one measurement of HbA1c and could therefore not be included in the analyses of HbA1c variability. We believe that this is clearly described in the manuscript in the first paragraph of Material and Methods.

3. Certainly, all 795 diabetic PD patients followed up between 2008-2018 must have had more than two values of HbA1c analyzed during their follow-up period, but most likely not captured for the Registry. Can this be considered a bias ? a limitation of the study ? and what about the variability of the results measured in different laboratories around Sweden ? were all PD patients in Sweden evaluated for their HbA1c levels with the same lab method ? It is

Response from authors: A few dialysis centers have not regularly reported their HbA1c values to the Swedish Renal Register (SRR). In this study, we only included the patients with at least two HbA1c values. It is now acknowledged in Discussion (second last paragraph, last sentence) that it is a study limitation that only 325 of the totally 795 patients were included in the present study. Finally, all laboratories in Sweden use the same method to analyze HbA1c. Thus, assay variability was likely of relatively small importance in terms of the accuracy of the data presented in the manuscript. Finally, in the new Table 2, the distribution of number of HbA1c measurements per patient according to classes of HbA1c variability is presented. As seen in the new Table 2, the number of HbA1c measurements did not appear to be associated with the level of HbA1c variability.

4. It possible to know how many of the 322 patients excluded for not having, at least two values of HbA1c, had no (0) values of HbA1c ?

Response from authors: All the totally 795 patients had at least one HbA1c value. However, to evaluate the association between HbA1c variability and mortality, we only included the patients with at least three months on PD treatment and two or more values of HbA1c (n=325).

5. In the Introduction the authors write ……… Earlier studies evaluated the association between the mean value of HbA1c and the risk of cardiovascular events or mortality in patients with DM and PD [16-18]. However, the impact of the glucose load from the PD dialysate on glycemic variability has not been previously investigated in diabetic patients. In order to evaluate the impact of glucose load from the PD fluid on glycemic variability, it is fundamental to have important clinical information which may significantly impact on HbA1c variability , the most important being:

a. Data on which modality (APD or CAPD) and the daily prescription for each patient. Were there switches from APD to CAPD or from CAPD to APD in any of the patients included, during the observation period ?

Response from authors: At baseline, among the 325 patients included in the present study, 235 (72%) were on CAPD and 90 (28%) were on APD. Switches were common during the follow-up period and were mostly temporary. At the end of study, 194 (60%) of the patients had CAPD and 131 (40%) of the patients were on APD .

b. CAPD daily dialysis prescription: number of exchanges, volume per exchange, glucose concentration in each glucose-based PD solution, use of the non-glucose based PD solutions icodextrin (Extraneal) and or aminoacids solution (Nutrineal)?

Response from authors: We do not have data on dialysate solutions, which is now acknowledged as a study limitation (Discussion, second last paragraph, third sentence).

c. APD daily dialysis prescription: number of exchanges, volume per exchange, total volume used, glucose concentration in each glucose-based PD solution, use of the non-glucose based PD solutions icodextrin (Extraneal) and or aminoacids solution (Nutrineal).

Response from authors: These data were not available, which is now acknowledged in Discussion, second last paragraph, third sentence.

c. Daily insulin prescription in case of insulin-dependent diabetics and other drugs for non-insulin dependent diabetics. How many insulin-dependent and how many non-insulin dependent patients were on the study ?

Response from authors: All patients were treated with insulin. Data on the daily insulin doses were not available (Discussion, second last paragraph, second last sentence).

d. How was the peritoneal membrane transport characteristics distribution among the CAPD and APD patients ? High transporters usually absorb more glucose and this can impact on morbidity and mortality

Response from authors: We used data from Swedish Renal register (SRR), in which data on type of peritoneal membrane transport are not included. This has been included in the manuscript as a study limitation (Discussion, second last paragraph, third sentence).

Information on blood pressure (percentage of hypertensive patients) and/or fluid overload status.

Response from authors: In this study, 292 (90%) of the patients had at least one medication for treatment of high blood pressure. We defined hypertension as systolic blood pressure (SBT) > 140 mmHg or diastolic blood pressure (DBP) > 90 mmHg. Using this definition, 146 (47%) of the patients had hypertension (Clinical classification, second paragraph).

e. Use of BRA, ACEi, aspirin, beta-blockers

Response from authors: Unfortunately, we do not have data about these medications.

f. Presence of co-morbidities ( Charlson index ?).

Response from authors: In this study, we only have data on cardiovascular disease and previous malignancy. Cardiovascular disease included as coronary heart disease (CHD), stroke or peripheral arterial disease (PAD).

g. Information on peritonitis rates, Kt/V, residual renal function.

Response from authors: We have no data on dialysis effectivity and Kt/V, which is stated in Discussion, second last paragraph, third sentence.

6. In the introduction the Authors wrote…. However, the impact of the glucose load from the PD dialysate on glycemic variability has not been previously investigated in diabetic patients. Information above described from a to h should be provided in order to corroborate this assumption. ??

Response from authors: In response to this comment, we have slightly modified this statement in Introduction to: “However, the impact of the glucose load from the PD dialysate on glycemic variability in relation to mortality has previously not been investigated in diabetic patients.” To our knowledge, this has previously not been investigated in diabetic patients.

7. The authors conclude….” In conclusion, to our knowledge, the present study is the first to evaluate the association between glycemic variability and the risk of all-cause mortality in diabetic patients receiving PD treatment. In our diabetic patients with maintenance PD, high HbA1c variability, as a measure of long-term glycemic control, was significantly associated with increased risk of all-cause mortality. Therefore, higher magnitudes of glycemic fluctuations, which might be caused by radical changes in dialysis regimes or peritonitis, are associated with higher risk of mortality in this group of patients. Further studies are needed to evaluate whether improved clinical care can reduce glycemic variability as well as the high mortality seen in patients with DM and PD. “Are they sure that “they are the first to evaluate the association between glycemic variability and the risk of all-cause mortality in diabetic patients receiving PD treatment.” ?

Response from authors: We have evaluated previous studies and we did not find any study that assessed the aim of our study ``association between HbA1c variability and the risk of mortality in patients with diabetes mellitus on PD´´. Most of the earlier studies in PD populations examined the association between the mean value of HbA1c and the risk of mortality or cardiovascular events. However, we have slightly modified the last paragraph of Discussion in response to this comment.

8. When the authors write “In our diabetic patients with maintenance PD…”, isn’t it more appropriate to write “In the SRR diabetic patients group with maintenance PD…” ?

Response from authors: In response to this comment from the Reviewer, we have changed this sentence to ´´In this observational study using data from the SRR of diabetic patients with maintenance PD, high HbA1c variability, as a measure of long-term glycemic control, was significantly associated with increased risk of all-cause mortality´´.

9. I cannot see support to the following sentence “Therefore, higher magnitudes of glycemic fluctuations, which might be caused by radical changes in dialysis regimes or peritonitis, are associated with higher risk of mortality in this group of patients “ as there is no information or data on dialysis prescriptions or peritonitis rates analysed in this study as well as of a definition of the word “radical” in the context of “radical changes in dialysis regimes or peritonitis “. In order to address this association (changes in dialysis prescriptions or peritonitis) with mortality, what is needed to investigate is primarily not the glycemic variability (a consequence of glucose exposure), but the changes in the daily glucose exposure/load and its clinical impact on different sub-groups of diabetic PD patients. In diabetic PD patients, higher HbA1c levels may indicate greater cumulative peritoneal glucose exposure with its attendant damage to the peritoneal membrane.

Response from authors: The main aim of this study was to evaluate the association between HbA1c variability, as a measure of long-time glycemic control, and the risk of mortality. However, as mentioned in manuscript, there is a need to examine the impact of the changes in the daily glucose exposure/load by dialysis regimes, type peritoneal membrane and even peritonitis, to determine their importance for the association between glycemic variability and mortality in PD populations in future studies.

10. The authors wrote as last sentence in the conclusion…. “Further studies are needed to evaluate whether improved clinical care can reduce glycemic variability as well as the high mortality seen in patients with DM and PD. “. I would challenge this conclusion using as basis, two not recent published randomized trials (Paniagua et al, de Moraes et al), already showing improved clinical care by improving metabolic control, decreasing glucose load and exposure as well as optimizing fluid management in PD patients (both diabetics and non-diabetics). I suggest reading the RCT CAPD study by Ramon Paniagua et al. and the RCT APD study by Thyago de Moraes et al.

Paniagua R, Ventura MD, Avila-Diaz M et al. Icodextrin improves metabolic and fluid management in high and high-average transport diabetic patients. Perit Dial Int 2009; 29: 422–432 de Moraes T.P., Andreoli M.C., Canziani M.E. et al. Icodextrin reduces insulin resistance in non-diabetic patients undergoing automated peritoneal dialysis: results of a randomized controlled trial (STARCH). Nephrol Dial Transplant. 2015; 30: 1905-1911. also suggest reading the RCT Impendia study by Li PK et al

In the IMPENDIA study, the primary endpoint was change in glycated hemoglobin from baseline. Mean glycated hemoglobin at baseline was similar in both groups. During the six months of the study, in the intention-to-treat population, the mean glycated hemoglobin profile improved in the intervention group but remained unchanged in the control group (0.5% difference between groups; 95% confidence interval, 0.1% to 0.8%; P=0.006).

Li PK, Culleton BF, Ariza A et al. Randomized, controlled trial of glucose sparing peritoneal dialysis in diabetic patients. J Am Soc Nephrol 2013; 24: 1889–1900

As well as reading the paper by McIntyre et al on glycemic control in diabetic CAPD patients assessed by CGMS

A practical approach to reduce disturbances of the carbohydrate metabolism in PD patients is the reduction of glucose exposure by also prescribing glucose-sparing solutions. In a study involving eight diabetic CAPD patients, replacement of a glucose-based regimen with a Physioneal-Extraneal-Nutrineal regimen was associated with a reduction in the 24-hour variability of glucose concentrations as measured by a subcutaneous probe in the interstitial fluid of the abdominal wall.

Marshall J, Jennings P, Scott A, Fluck RJ, McIntyre CW: Glycemic control in diabetic CAPD patients assessed by continuous glucose monitoring system (CGMS). Kidney Int 64: 1480–1486, 2003

Response from authors: Many thanks for very interesting articles, which we have read carefully and added as references in our manuscript. The results of these studies show that non-glucose-containing dialysis fluid was associated with better metabolic control. The result of our study indicated that higher HbA1c variability, as measure of long-term glycemic control, was associated with increased risk of mortality. In the manuscript, we suggest that the use of non-glucose-containing dialysis fluid like Icodextrin could possibly be related to lower variability of blood glucose as well as a decrease in the risk of mortality among PD patients with diabetes. This needs to be assessed in further observational and RCT studies.

11. Lastly, it should be taken into account that in type 1 and 2 diabetics (not in dialysis), HbA1c is not a good predictor of cardiovascular disease (CVD), whereas insulin resistance is predictive of CVD and indeed may be the most important single cause of coronary artery disease (see for example: Home P. Diabetes Care 2019; 42:1615-23; Adeva-Andany MM et al. Diabetes Metab Syndr 2019; 13:1449-55; Shahim B. et al. Diabetes Care 2017:40:1233-40; Eddy D. et al. Diabetes Care 2009; 32; 361-66; Orchard TJ et al. 2003; 26:1374-79). In addition, since patients in dialysis are often affected by subclinical/clinical anemia, which reduces red blood cell survival and hence Hb glycosylation, this introduces an additional bias in the interpretation of HbA1c variability. At this regard, it should be kept in mind that the simple evaluation in dialysis patients of total Hb levels would be of little help in figuring out potential differences in the glycosylation rate as total Hb levels. Indeed, the latter (Hb levels) may be easily corrected by Epo treatment, but it doesn’t tell you anything about survival rate of RBCs, one of the major determinant of the extent of HB glycosylation.

Response from authors: As discussed in the manuscript, the use HbA1c as a measure of long-term glycemic control is debated in patients with end stage renal disease (ESRD) and dialysis treatment. However, markers other than HbA1c, like glycated albumin and fructosamine have some major disadvantage in dialysis populations, why we believe that HbA1c still is the best marker of long-term glycemic control.

In my opinion the paper needs a major revision. If the Authors do not have the clinical data necessary to make the association in the title robust and as a reflection of the clinical reality, you may shorten the article with a statistical description of what you see from the data available. I am not an expert in Statistics, but as a clinician, I consider the clinical data I described above of utmost importance in order to keep the paper in the present format. So I really hope you have the data I am suggesting to add to the paper.

Submitted filename: Response to Reviewers PLOS one 210915-R1.docx

Click here for additional data file.

10 Jan 2022

LONG-TERM GLYCEMIC VARIABILITY AND THE RISK OF MORTALITY IN DIABETIC PATIENTS RECEIVING PERITONEAL DIALYSIS

PONE-D-21-18957R1

Dear Dr. Peters,

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,

Vivekanand Jha

Academic Editor

PLOS ONE

Additional Editor Comments (optional):

Reviewers' comments:

17 Jan 2022

PONE-D-21-18957R1

LONG-TERM GLYCEMIC VARIABILITY AND THE RISK OF MORTALITY IN DIABETIC PATIENTS RECEIVING PERITONEAL DIALYSIS

Dear Dr. Peters:

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

Prof Vivekanand Jha

Academic Editor

PLOS ONE