Interaction between central obesity and frailty on the clinical outcome of peritoneal dialysis patients.

BACKGROUND: Frailty and obesity contribute to the adverse clinical outcome of peritoneal dialysis (PD) patients, but the interaction between frailty and obesity remains uncertain.
OBJECTIVE: To examine the interaction between frailty and obesity on the clinical outcome of PD patients.
DESIGN: Single centre prospective observational cohort study. PATIENTS: 267 prevalent Chinese PD patients were recruited. MEASUREMENTS: Frailty was identified by a standard score. General and central obesity were determined by body mass index (BMI) and waist-hip ratio (WHR), respectively. Body composition was assessed by bioimpedance spectroscopy. All patients were followed for two years. Outcome measures included all-cause as well as cardiovascular mortality and hospitalization.
RESULTS: Of the 267 patients, 120 (44.9%) were frail. Frail individuals were more likely to have central obesity (p < 0.001) but not general obesity. Although WHR did not predict patient survival, there was a significant interaction between WHR and frailty on patient survival and cardiovascular survival (p = 0.002 and p = 0.038, respectively). For patients without frailty, the two-year cardiovascular survival was 91.3% and 74.4% for those with and without central obesity, respectively (p = 0.002). For patients with frailty, however, the two-year cardiovascular survival was 64.6% and 66.7% for those with and without central obesity, respectively (p = 0.6). For patients without frailty, the number of hospital admission for cardiovascular disease over 2 years were 0.12 ± 0.37 and 0.34 ± 0.72 for those with and without central obesity, respectively (p = 0.03). For frail patients, however, the number of hospital admission was similar between those with and without central obesity.
CONCLUSION: There is a significant interaction between frailty and central obesity on the outcome of PD patients. The protective role of central obesity is only apparent in PD patients without frailty but not the frail ones, and there is a little prognostic value of general (non-central) obesity.

Introduction

Background

Frailty is described by Fried et al. [1] as a clinical syndrome involving at least three of the following features: unintentional weight loss (10 lbs in one year), self-reported exhaustion, weakness (as determined by grip strength), slow walking speed, and low physical activity. It is highly prevalent among patients on dialysis, with a prevalence of up to 67.7% in previous report [2].

Physical frailty has important health consequences. Frail patients are more susceptible to the development of adverse outcomes when faced with a stressor. It also leads to a decline in physical function, gait disability, increase in the risks of fall. As a result, frail peritoneal dialysis (PD) patients have a higher hospitalization risk and their hospital stay tends to be prolonged [3]. Physical frailty is also a well-established independent predictor of adverse outcomes including mortality in pre-dialysis end stage renal disease patients [4], hemodialysis (HD) patients and PD patients [3, 5].

Physical frailty causes alteration in body composition. Frail patients tend to suffer from obesity [6-9]. On the other hand, PD patients are more susceptible to central obesity [10] due to exposure to glucose-containing dialysate fluid. In contrast to general obesity, central obesity confers a worse survival in dialysis patients [11]. Previous study has shown synergistic effect on mortality between physical frailty and central obesity in HD patients. However, the interaction between physical frailty and body composition has not been explored.

Objectives

The goal of this study of was to dissect the internal relationship between obesity and frailty on the outcome of PD patients.

Methods

Study design

This is a single centre prospective cohort study. The study was approved by the Joint Chinese University of Hong Kong—New Territories East Cluster Clinical Research Ethics Committee. All study procedures were in compliance with the Declaration of Helsinki. All patients provided a written informed consent before enrollment in the study.

Study population

We recruited 267 prevalent PD patients from a single dialysis unit from 1 January 2015 to 31 December 2016. Patients with expected survival of less than 3 months, or those who were planned to receive kidney transplantation in 3 months, were excluded.

After written informed consent, assessments of frailty, anthropometry, body composition and pulse wave velocity were measured at baseline. Other clinical and laboratory data were obtained by chart review. Clinical data comprises of patient’s age, gender, body weight, height, primary diagnosis of renal disease, concomitant chronic medical illnesses including diabetes mellitus, ischemic heart disease, cerebrovascular accident, peripheral vascular disease, chronic hepatitis B and C infection, chronic lung disease, malignancy and immunological diseases. Laboratory assessment included serum albumin level, serum cholesterol (including total cholesterol, high density cholesterol, low density cholesterol) level, dialysis adequacy by Kt/V, residual renal function by measuring the residual glomerular filtration rate from urine. The Charlson Comorbidity Index (CCI) was used to assess the comorbidity load. The Comprehensive Malnutrition Inflammation Score (MIS) [12] and Subjective Global Assessment (SGA) [13] were used to assess nutritional state.

Assessment of frailty

We used a validated Chinese questionnaire that consisted of 30 yes/no questions [3]. The questions involve assessment of subjective assessment of personal health, psychological state, physical state in terms of number of hospital or doctor visit and medication needs to be taken, body weight, need of assistance in different aspects of daily living and mobility. A total score was calculated; physical frailty was defined as a score of 6 or above [3]. To ensure patient adequate understanding on the 30 questions, all the recruited subjects were directly interviewed by designated well-trained interviewers at baseline.

Anthropometric and body composition assessment

Anthropometric data were measured at recruitment. Body mass index (BMI) was used to assess general adiposity; general obesity was defined as BMI more than or equals 30 kg/m2 respectively according to the definition from World Health Organisation (WHO). Waist hip ratio (WHR), as defined by the ratio between waist circumference to hip circumference, was used to assess central adiposity. Waist circumference was measured at the approximate midpoint between the lower margin of the last palpable rib and the top of the iliac crest, while the hip circumference was measured around the widest portion of the buttocks. Central obesity was defined as WHR more than or equals sex-specific sample median (females 0.92, males 0.98).

We used the multi-frequency bioimpedance spectroscopy device (Body Composition Monitor [BCM], Fresenius Medical Care, Germany), which was validated in dialysis patients [14, 15]. Briefly, electrodes were attached to one hand and one foot with the patient in a supine position. The following parameters were computed: extracellular water (ECW), lean tissue mass (LTM), adipose tissue mass (ATM), and volume of overhydration (OH).

Anthropometric and body composition assessments were performed with empty abdomen (before infusion of dialysate) by the same trained research nurse with at the time of recruitment.

Pulse wave velocity

Pulse wave velocity (PWV) is a simple, non-invasive technique of assessing arterial stiffness. High levels of PWV analysis are indicative of increased rigidity and low distensibility of vascular walls, along with poor vascular function [16]. PWV was measured by Vicorder device (SMT Medical GmbH&Co.), an automatic computerised recorder. The results were analyzed by the Complior Analyse program (Artech Medical, Pantin, France). Pressure-sensitive transducers will be placed over neck (carotid artery), wrist (radial artery), and groin (femoral artery), with the patient in the supine position. PWV of the carotid-femoral (CF-PWV) and carotid-radial (CR-PWV) territories will be calculated by dividing the distance between the sensors by the time corresponding to the period separating the start of the rising phase of the carotid pulse wave and that of the femoral and also the radial pulse waves. Measurements were performed at the beginning of study, and 1 year after recruitment.

Outcome measures

All patients were followed for 2 years. Primary outcome measures were all-cause mortality and cardiovascular mortality. Secondary outcome measures were number of all-cause hospital admission, number of cardiovascular event-related hospitalization, and total length of hospital stay within study period. Cardiovascular mortality is defined by death due to peripheral vascular disease, myocardial infarction, heart failure, stroke and sudden cardiac death. In this analysis, receiving kidney transplantation, transferal to hemodialysis, transferal to another center and death from non-cardiovascular causes were considered as competing events.

Statistical analysis

Statistical analysis was performed by SPSS for Windows software version 24 (SPSS Inc., Chicago). Descriptive data were presented as mean ± SD if normally distributed, or median (inter-quartile range) otherwise. Patients were grouped according to presence of physical frailty as defined above for analysis. Clinical parameters were compared by Student’s t-test or one way analysis of variance (ANOVA) for continuous variables, and Chi-square test for categorical variables. P<0.05 was considered to be statistically significant in this study. All probabilities were two-tailed.

The number and duration of hospitalization were compared between non-frail and frail individuals. Age, dialysis vintage, Charlson Comorbidity Index, serum albumin, total Kt/V, normalised protein nitrogen appearance (NPNA) and residual glomerular filtration rate (GFR), which were potential confounding factors shown in previous studies [3, 17, 18], were added to linear regression model for adjustment. Backward stepwise elimination was used to determine the independent predictor of hospitalization.

Kaplan Meier plots were constructed for patient and cardiovascular survival rates. Log-rank test was used to compare between survival curves. Receiving kidney transplantation, transferal to hemodialysis, and transferal to another center were considered as competing events. Multivariate Cox regression models were constructed to further identify independent predictors of survival after adjustment of potential confounders with P values below 0.1 with univariate analysis. Backward stepwise analysis was used to remove insignificant variables. P<0.05 was considered to be statistically significant in this study. All probabilities were two-tailed.

Results

Baseline demographic characteristics

A total of 267 PD patients were recruited. Their baseline clinical and demographic data are summarized in Table 1 and S1 Table. Among them, 120 patients (44.9%) were physically frail (Fig 1). As compared to those without frailty, frail PD patients were older, more often required helper assistance in performing PD, had higher CCI, and worse nutritional state. Frail PD patients were also more likely to be diabetic, had lower residual renal function and hemoglobin level.

Table 1

Baseline clinical and demographic data.

	Not frail	Frail
No. of patients	147	120
Age (year)	61.1 ± 11.8	64.2 ± 12.2
Sex (M:F)	75: 72	56: 64
Duration of dialysis (months)	44.3 ± 46.9	51.8 ± 57.4
Blood pressure (mmHg)
Systolic	142.8 ± 18.9	142.6 ± 20.9
Diastolic	75.8 ± 12.4	73.4 ± 12.7
Renal Diagnosis, no of cases. (%)
Glomerulonephritis	51 (34.7%)	33 (27.5%)
Diabetic nephropathy	42 (28.6%)	56 (46.7%)
HTN	18 (12.2%)	12 (10%)
Polycystic kidney	5 (3.4%)	3 (2.5%)
Urological disease	9 (6.1%)	2 (1.7%)
Other / unknown	22 (15%)	14 (11.6%)
Comorbid disease, no of cases (%)
DM	55 (37.4%)	66 (55.0%)
IHD	19 (12.9%)	19 (15.8%)
CVA	16 (10.9%)	31 (25.8%)
Charlson Comorbidity Index	5.1 ± 2.2	5.6 ± 2.2
Nutritional Status
MIS	6.57 ± 3.13	9.00 ± 3.46
SGA	5.38 ± 0.83	4.84 ± 0.92
Total weekly Kt/V	1.88 ± 0.49	1.76 ± 0.40
Residual GFR (ml/min/1.73m2)	1.70 ± 2.02	1.04 ± 1.74
NPNA (g/kg/day)	1.07 ± 0.25	1.03 ± 0.30
Biochemical parameters
Hemoglobin (g/dL)	9.72 ± 1.31	9.25 ± 1.26
Albumin (g/L)	34.10 ± 4.15	33.46 ± 4.16
Pulse Wave Velocity (m/sec)
CF-PWV	10.93 ± 2.38	11.19 ± 2.40
CR-PWV	10.34 ± 1.75	10.48 ± 1.80
Peritoneal Transport State- 4-hour dialysate/plasma creatinine	0.64 ± 0.13	0.65 ± 0.12
Helper-assisted PD, no of cases (%)	8 (5.4%)	24 (20.0%)
Type of peritoneal dialysis, no of cases (%)
CAPD	127 (86.4%)	106 (88.3%)
CCPD	4 (2.7%)	2 (1.7%)
NIPD	16 (10.9%)	12 (10.0%)
Icodextrin dialysate use, no of cases (%)	39 (26.5%)	34 (28.3%)
Concomitant medications, no of cases (%)
Aspirin	35 (23.8%)	42 (35.0%)
Beta blocker	99 (67.3%)	72 (60.0%)
RAS blocking agent	79 (53.7%)	77 (64.2%)
Calcium channel blocker	125 (85.0%)	90 (75.0%)
Statin	64 (43.5%)	62 (51.7%)
Calcium containing phosphate binder	115 (78.2%)	84 (70.0%)
Vitamin D supplement	64 (43.5%)	62 (51.7%)
Erythropoiesis-stimulating agents	112 (76.2%)	82 (68.3%)

Data are expressed as mean ± standard deviation.

HTN, hypertensive nephrosclerosis; DM, diabetes mellitus; IHD, ischemic heart disease; CVA, cerebrovascular accident; MIS, Malnutrition Inflammation Score; SGA, Subjective Global Assessment; GFR, glomerular filtration rate; NPNA, normalized protein nitrogen appearance; CF-PWV, carotid-femoral pulse wave velocity; CR-PWV, carotid-radial pulse wave velocity; PD, peritoneal dialysis; CAPD, continuous ambulatory peritoneal dialysis; CCPD, continuous cycler peritoneal dialysis; NIPD, nocturnal intermittent peritoneal dialysis; RAS blocking agent, renin angiotensin system-blocking agent.

Fig 1

Flow diagram of subjects' inclusion and their outcomes.

The body composition of patients with and without frailty are summarized and compared in Table 2. Frail patients had significantly higher WHR, and marginally higher BMI, than those without frailty, although the latter was not statistically significant. Among frail patients, the prevalence of central obesity and general obesity was 60.0% and 13.3% respectively, whereas among non-frail patients, the prevalence was 46.9% and 10.2% respectively. Frail patients had marginally higher adipose tissue mass than non-frail patients by bioimpedance spectroscopy, although the difference was not statistically significant. Frail patients also had significantly a higher overhydration volume than those without frailty, but their lean tissue mass was similar.

Table 2

Body anthropometry and composition by bioimpedance spectroscopy.

	Not frail	Frail	P value
No. of patients	147	120
Body weight (kg)	64.2 ± 12.9	66.3 ± 12.5	p = 0.2
Body height (cm)	161.0 ± 7.5	160.0 ± 8.3	p = 0.2
Body mass index (kg/m2)	25.7 ± 4.1	26.0 ± 4.3	p = 0.07
Waist hip ratio (WHR)	0.95 ± 0.07	0.99 ± 0.11	p < 0.001
Overhydration (OH)
Volume (L)	2.90 ± 3.09	3.62 ± 2.99	p = 0.08
>1 L, no. of case (%)	101 (72.7%)	96 (87.3%)	p = 0.005
Adipose tissue mass (kg)	21.1 ± 11.6	23.1 ± 11.0	p = 0.16
Lean tissue mass (kg)	38.0 ± 9.66	36.7 ± 9.67	p = 0.3

There was no significant difference in baseline CF-PWV and CR-PWV between patients with and without frailty. After 1 year of PD, CF-PWV increased from 11.04 ± 2.39 to 12.19 ± 3.17 m/sec (paired Student’s t test, p = 0.003), whereas there was no significant change in CR-PWV (from 10.40 ± 1.77 to 10.73 ± 2.14 m/sec, paired Student’s t test, p = 0.8). The baseline frailty score did not correlate with the change in CF-PWV (r = 0.145, p = 0.3) or CR-PWV (r = 0.081, p = 0.6) over one year. In central obese patients, there was a modest correlation between physical frailty and the change in CF-PWV although it did not reach statistical significance (r = 0.359. p = 0.08). There was no correlation between physical frailty and change in CR-PWV (r = 0.001. p = 0.9).

Patient survival

During the study period, 83 patients (31.1%) died. During this period, 16 patients (6.0%) underwent kidney transplantation, 12 (4.5%) were switched to chronic hemodialysis, and 3 (1.1%) were transferred to other renal center. The 2-year overall survival rate was 58.3% in frail and 77.6% in non-frail patients (log-rank test, p = 0.001). The 2-year cardiovascular survival was 65.8% in frail and 82.3% in non-frail individuals (log-rank test, p < 0.001). After adjusting for potential confounding factors by multivariate Cox regression analysis (Table 3), physical frailty remained a significant predictor for all-cause mortality and cardiovascular mortality. In these models, other significant predictive factors of all-cause mortality included age, CCI and serum albumin level. For cardiovascular mortality, CCI was the only other independent predictor.

Table 3

Univariate and multivariate Cox regression analysis.

(A) Patient survival
	Univariate analysis		Multivariate analysis
Factors	AHR (95% CI)	P value	AHR (95% CI)	P value
Physical frailty	2.126 (1.369–3.330)	p = 0.001	1.793 (1.094–2.939)	p = 0.021
CCI	1.350 (1.246–1.464)	p < 0.001	1.131 (1.015–1.259)	p = 0.025
Age	1.067 (1.043–1.091)	p < 0.001	1.069 (1.034–1.104)	p < 0.001
Albumin	0.881 (0.837–0.927)	p < 0.001	0.915 (0.859–0.975)	p = 0.006
CF-PWV	1.172 (1.070–1.284)	p = 0.001
Overhydration	1.085 (1.019–1.156)	p = 0.011
(B) Cardiovascular survival
	Univariate analysis		Multivariate analysis
Factors	AHR (95% CI)	P value	AHR (95% CI)	P value
Physical frailty	2.959 (1.277–6.860)	p = 0.011	2.652 (1.134–6.200)	p = 0.024
CCI	1.282 (1.111–1.479)	p = 0.001	1.294 (1.109–1.509)	p = 0.001
Age	1.052 (1.011–1.095)	p = 0.012
Albumin	0.921 (0.839–1.011)	p = 0.085

CCI, Charlson Comorbidity Index; CF-PWV, carotid-femoral pulse wave velocity; BMI, body mass index; AHR, adjusted hazard ratio; CI, confidence interval; CV, cardiovascular.

Interaction with WHR on survival

Neither WHR nor BMI predicted overall survival or cardiovascular survival. However, there was a significant interaction between central obesity and frailty on patient survival and cardiovascular survival (p = 0.002 and p = 0.038, respectively, for interaction), while the interaction between general obesity and frailty on survival did not reach statistical significance (overall survival, p = 0.057 and cardiovascular survival, p = 0.156, respectively, for interaction). The 2-year patient survival of patients with central obesity was marginally lower than those without central obesity for the frail group (54.2% vs 64.6%, p = 0.3), but marginally higher than those without central obesity for the non-frail group (81.2% vs 74.4%, p = 0.2), although neither of the difference was statistically significant (Fig 2). Similarly, the 2-year cardiovascular survival of patients with central obesity was similar to those without central obesity for the frail group (64.6% vs 66.7%, p = 0.6), but significantly higher than those without central obesity for the non-frail group (91.3% vs 74.4%, p = 0.002).

Fig 2

Kaplan-Meier plot of patient survival according to the frailty state and central obesity.

Further subgroup analysis on 2-year patient survival was performed for patients on short duration of peritoneal dialysis (less than 1 year use). In short, the survival rate was higher among non-frail individuals, which is in line with the results of our whole cohort. Non-frail, central obese individuals had the best survival rate while frail, central obese individuals had the worst survival rate. The survival rates decreased in a stepwise manner, from 100% for non-frail central obese, 82.4% for non-frail non-central obese, 76.9% for frail non-central obese, to 61.5% for non-frail non-central obese (S1 Fig) though the difference did not reach significance (log-rank test, p = 0.069).

Hospitalization

After 2 years of follow up, there were in total 984 episodes of hospital admission for a total of 10,695 days in our cohort; 45 patients (16.8%) did not require any hospitalization during the study period (Table 4). Among all, 88 episodes of hospitalization admission for a total of 1,190 days had cardiovascular events as the primary diagnosis. In short, frailty score had modest correlations with the number of hospitalization for all-cause (r = 0.158, p = 0.025) or cardiovascular event (r = 0.133, p = 0.049), total duration of hospitalization (r = 0.176, p = 0.068), although the last correlation did not reach statistical significance. In contrast, WHR did not correlate with number of hospitalization for all-cause (p = 0.4) or cardiovascular-event (p = 0.7), or the duration of hospitalization (p = 0.4). By using linear regression model after log transformation to adjust for potential confounders, frailty score remained an independent factor associated with the number of hospital admission for all-cause and total duration of hospitalization (Table 5).

Table 4

Relation between physical frailty and hospitalization.

	Not frail	Frail	P value
Number of patients	147	120
Without hospitalization, no. of patients (%)	24 (16.3%)	21 (17.5%)	p = 0.799a
Number of hospitalization
All-cause	3.13 ± 3.51	4.37 ± 4.27	p = 0.025b
Cardiovascular event	0.24 ± 0.59	0.44 ± 0.92	p = 0.049b
Total duration of hospitalization (days)	31.05 ± 42.48	51.08 ± 68.99	p = 0.068b

Data are compared by

aChi-square test and

bMann Whitney U test.

Table 5

Multivariate linear regression analysis of hospitalization after log transformation.

(A) Number of hospitalization
Factors	Unstandardized B	95% CI	P value
Physical frailty	0.998	0.024–1.971	p = 0.045
CCI	0.372	0.101–0.644	p = 0.007
GFR	-0.348	-0.602 –-0.094	p = 0.007
(B) Number of cardiovascular event-related hospitalization
Factors	Unstandardized B	95% CI	P value
CCI	0.060	0.017–0.104	p = 0.007
Kt/V	-0.191	-0.355 –-0.027	p = 0.022
(C) Total duration of hospitalization
Factors	Unstandardized B	95% CI	P value
Physical frailty	14.295	0.032–28.558	p = 0.049
CCI	4.090	0.954–7.227	p = 0.011
Kt/V	-14.483	-26.696 –-2.270	p = 0.020

CI, confidence interval; CCI, Charlson Comorbidity Index; GFR, glomerular filtration rate.

We further performed subgroup analysis on hospitalization data according to physical frailty. In short, general obese patients were hospitalized more frequently with a longer hospitalization period, while central obese patients were hospitalized less frequently with a shorter period (Table 6). In non-frail patients, those with general obesity had more than a doubled number and duration of cardiovascular-event related hospitalization as compared to those without general obesity, while those with central obesity had halved number and duration of cardiovascular-event related hospitalization as compared to those without central obesity. A similar trend was also observed in frail patients, but the difference was not significant (S2 Table).

Table 6

Relation between hospitalization and general versus central obesity in non-frail patients.

	No central obesity	Central obesity	P value*	No general obesity	General obesity	P value*
Number of patients	78	69		132	15
Number of hospitalization
All-cause	3.37 ± 3.94	2.86 ± 2.95	p = 0.5	3.08 ± 3.59	3.60 ± 2.75	p = 0.189
Cardiovascular event	0.34 ± 0.72	0.12 ± 0.37	p = 0.030	0.20 ± 0.56	0.60 ± 0.74	p = 0.002
Duration of hospitalization (days)
All-cause	35.22 ± 51.53	26.34 ± 28.70	p = 0.9	29.67 ± 41.73	43.27 ± 48.39	p = 0.076
Cardiovascular event	3.54 ± 9.16	1.45 ± 5.38	p = 0.040	1.92 ± 7.04	8.13 ± 10.62	p = 0.001

*Data are compared by Mann Whitney U test.

Discussion

The present study identified frailty is prevalent among PD patients, and it is an independent predictor of adverse outcomes including a higher hospitalization rate, a longer duration of hospitalization, as well as a worse survival with higher all-cause and cardiovascular mortality rates. Our results also showed presence of significant interactions between central obesity and frailty in terms of hospitalization and survival in PD patients.

Physical frailty is traditionally regarded as a geriatric disease as it primarily affects elderly patients. As predicted, frail individuals are generally older in our cohort as shown in Table 1. When compared to HD, performing PD is relatively simpler and it does not require as much technical support as in HD. PD also does not require vascular access creation, and it offers a better cardiovascular stability [19-21]. PD is therefore commonly regarded as a better dialysis option over HD in elderlies. More than 30% to 75% of elderlies preferred PD over HD in different countries [22, 23]. Assisted PD is particularly more preferred to hemodialysis as assisted PD provides a better quality of life to frail elderly on dialysis [24-26]. Therefore, it is important to recognize and quantify the impact of physical frailty in PD patients. In fact, routine frailty screening has been suggested to be part of assessment for and maintenance care on PD [27]. However, most of the currently available literature in the field of physical frailty in dialysis population involved either HD [28] or a mixture of HD and PD cases [29]. This evidence are unable to reflect the burden of physical frailty in PD population, and cannot be applied to certain countries like Hong Kong, Taiwan, Thailand, Vietnam, as well as New Zealand and Australia, which have a relatively higher PD to HD case ratio compared to other western countries like the USA and Canada [30].

Patients with chronic kidney disease have a different body composition compared with the general population [31, 32]. The causes are multifold. Firstly, uremia, defects in homeostasis and metabolic derangements in chronic kidney disease cause anorexia, which results in malnutrition and protein energy wasting [33]. Secondly, dialysis treatment also induces protein loss and muscle catabolism [34, 35]. In particular, PD causes central obesity [10, 36] due to the exposure to glucose-containing dialysate fluid. On the other hand, physical frailty itself is also associated with similar change in body composition i.e. development of obesity with a reduction in muscle mass [7-9]. The overall combined effect on body composition by renal failure, dialysis treatment and physical frailty in PD patients were uncertain. Our study demonstrated a comparable total adipose tissue mass between frail and non-frail PD patients. It is worth to note that despite the total adipose tissue was similar between two groups, when comparing between general and central obesity, frail individuals were more central obese but not general obese. This observation highlights frail individual had a different fat composition, with a relatively higher visceral fat to subcutaneous fat ratio when compared to non-frail PD patients. This may be the potential cause for an inferior survival in frail PD patients, which will be further elaborated.

In contrast to general obesity, central obesity correlates with a worse survival in ESRD and HD patients [11, 37]. There were only a few published data in PD patients [11, 38, 39], with limited power due to small sample size. One Korean study [40] advocated the use of Sagittal Abdominal Diameter (SAD) measured by X-ray, and found correlation of central obesity with all-cause and cardiovascular mortality. However, its use is limited by the need of a dedicated X-ray machine and it involves additional radiation exposure, which may not be feasible in a clinic setting. Two small scaled studies by Canadian group [38] with 22 subjects, and Korean group [11] with 84 subjects showed controversial mortality risk from central obesity in PD patients. Another study with a larger sample size from Brazil [39] showed a worse short-term 1-year survival in central obese PD patients. However, the long-term survival could not be assessed due to limited study period. In our study, we recruited 267 subjects, which represented the largest scaled study compared to other studies in same field. Our results reflected neither general nor central obesity predicted 2-year all-cause and cardiovascular disease-related mortality in PD patients.

Central obesity has been reported to have synergistic effect on mortality with physical frailty in HD patients [41]. However, such effect is not yet explored in PD patients. Our study is the first study to prove an additive effect on mortality by physical frailty and central obesity in PD patients. The underlying pathophysiology is probably related to the potentiation of pro-inflammatory state by certain cytokines. Visceral fat in central obese patients secretes numerous pro-inflammatory cytokines including C-reactive protein (CRP), tumor necrosis factor (TNF)-alpha and interleukin (IL)-6 [42, 43]. These cytokines modulate lipid and carbohydrate metabolism [44], and orchestrate the inflammatory pathway by stimulation of T cell proliferation, B cell-related immunoglobulin production and hepatic synthesis of acute phase protein [42]. At the same time, physical frailty is associated with an inflammatory state with high levels of CRP and IL-6. Study has shown elevated CRP and IL-6 levels predict mortality, cardiovascular events and composite outcomes in dialysis patients [42]. Although vascular calcification and arterial stiffening, which could be secondary to chronic inflammation, play a prognostic role on survival of PD patients [45], the cross linkage between frailty, central obesity and mortality appears to be independent from arterial stiffening as reflected by a similar CF-PWV and CR-PWV between frail and non-frail individuals (Table 1, S1 Table), and a lack of correlation between progression of the two parameters with time between frail and non-frail individuals, both in central and non-central obese subgroups. Activation of the inflammatory cascade can also lead to impairment of immune response [46] and subsequent activation of coagulation system causing a higher thrombotic tendency and cardiovascular risk [47]. This may explain why physical frailty had a much higher hazard ratio on all-cause and cardiovascular disease-related mortality in central obese PD patients when compared to all PD patients. It is worthwhile to note that from our result, physical frailty did not cause a worse survival in non-central obese PD patients. The underlying cause remains to be elucidated and further studies should be undertaken in the future to clarify the mechanism.

On top of the proven 1-year mortality risk from frailty in PD patients [5], this study illustrated a longer term mortality risk from frailty in PD patients. However, physical frailty not only confers mortality risk to PD patients, but also imposes a huge burden to the economy and healthcare system as frail individuals are often high users of healthcare resources. In our cohort, frail PD patients were more often required helper assistance in performing dialysis. They also tend to hospitalize more frequent and longer than non-frail individuals, which is in line with other published studies [28, 29, 48]. Notably, frail central obese PD patients required nearly twice as many hospital admissions than non-frail central obese PD patients, and their total hospital length of stay was also nearly twice in our study. Aside from physical frailty, obesity may also influence clinical outcome in terms of hospitalization in dialysis patients. Patients with higher BMI were associated with a higher incidence of peritonitis-related hospitalization but not in non-peritonitis-related hospitalization [49]. Similarly, central obesity was also identified to be a predictor for cardiovascular events, all-cause hospitalization and mortality [37, 39]. Our data is the first published result both evaluating and comparing the impact of both general and central obesity, as well as identifying the potential interaction with physical frailty on hospitalization parameters in PD patients. In contrast to previous study [37, 39], we identified central obesity as a protective factor associated with a fewer hospital admission and higher survival rate in non-frail PD patients. Similar findings have been reported in patients with stable coronary artery disease patients [50], and heart failure patients especially those with reduced ejection fraction [51, 52]. The proposed mechanisms are related to the extra energy reserve provided by visceral fat [51], and the anti-inflammatory and endotoxin-neutralizing effects mediated by a raised lipoprotein level [52]. In addition, central obese individuals also seek medical attention at the earlier stage of acute illnesses as they tend to be more symptomatic, and their renin-angiotensin-aldosterone system is more attenuated, leading to a higher blood pressure and a better tolerance to cardioprotective medications [52]. On the other hand, general obesity in non-frail PD patients were hospitalized more frequently with a longer duration, regardless of the cause of hospitalization. While most of the available literature focused on recognition of frailty and the associated adverse outcomes in PD patients, our result addressed the impact of body composition in non-frail group and this should not be overlooked. Identification of mechanism underlying such difference between general and central obesity is beyond our scope, but this should be evaluated in future study.

Our study has several limitations. Our data collection is retrospective in nature from a single center with a limited number of study subjects, which may lead to the concern on generalisability. We believe our cohort is a good representation of a Chinese PD cohort as our centre is one of the largest tertiary-care and university-affiliated medical centres. It is also part of the public healthcare system which takes care of 94% of dialysis patients in Hong Kong [53]. We also included both incident and prevalent PD patients, which can lead to a few concerns. Firstly, body composition, fat mass and waist circumference can be altered by the long duration of peritoneal dialysis due to exposure of glucose-containing dialysate [10] and that may have prognostic implications too [39]. However, the marker of central obesity we used i.e. WHR takes both waist and hip circumference into account. Whether there is any change with time depends on the proportion of difference in both parameters. Moreover, univariate analysis reflected PD duration did not predict hospitalization (p = 0.364) and mortality (p = 0.210). All other results reported in this study were adjusted with the duration on PD in their corresponding statistical analyses. On the other hand, we also performed subgroup analysis on patients who were on short PD duration of less than 1 year with a shorter exposure time to glucose-containing dialysate. The survival rate trend is the same as the trend in our whole cohort (S1 Fig) though the difference in survival rates of the 4 groups in subgroup analysis did not reach statistical significance. Together with the other insignificant or barely significant results we presented, this can be attributed to the relatively small sample size in our cohort, which may raise the issue regarding the weight that this study can give. Despite so, our results, as the first published data in this aspect, still give valuable information regarding potential interaction between central obesity, frailty and their association with adverse outcomes in PD patients. Base on this, we suggest future research with a larger sample size to identify any significant interaction and association with the outcomes that we investigated. Secondly, there will also be potential lead-time bias and selection bias favoring surviving patients. However, this can be strength of our study as it can be applied more universally. Moreover, the duration of follow-up may not be long enough to detect the effect of physical frailty and body composition in survival. Furthermore, our measurement of frailty and body composition is a one-off measurement performed at the recruitment period. These results may also be transiently affected by acute illnesses. Serial measurements of these parameters may provide a better assessment in corresponding aspects. The cross-sectional nature of our data can only allow us to establish association rather than causal relationship. Although we have sufficient power to identify the independent association between two factors, these results do not have sufficient data to analyses the aspect of effect modification.

In summary, we found that physical frailty is prevalent in Chinese PD patients. Frail PD patients were usually older and they often had higher comorbidity loads and poorer nutritional state. They are also more central obese. Physical frailty was an independent predictor for adverse outcomes in PD patients, in terms of a higher hospitalization rate and longer duration of hospitalization, as well as a worse survival with higher all-cause and cardiovascular mortality rates. In subgroup analysis, central and general obesity caused a differential effect on hospitalization parameters in non-frail PD patients. Furthermore, central obesity conferred a numerically but statistically non-significant inferior patient’s survival and cardiovascular survival in frail PD patients. On the contrary, central obesity conferred a significantly superior cardiovascular survival in non-frail PD patients. We suggest routine screening of physical frailty and central obesity in PD patients, so as to achieve risk stratification by identification of the high-risk group. Further prospective study is urgently needed to explore on potential interventional measures that can be done to these high-risk individuals to improve their clinical outcomes and survival.

(DOCX)

Click here for additional data file.

Kaplan-Meier plot of patient survival among patients on peritoneal dialysis for less than 1 year according to the frailty state and central obesity.

(TIF)

Click here for additional data file.

Baseline clinical and demographic data, statistical tests results.

(DOCX)

Click here for additional data file.

Relation between hospitalization and general versus central obesity in frail patients.

(DOCX)

Click here for additional data file.

27 Jul 2020

PONE-D-20-19119

Interaction Between Central Obesity and Frailty on the Clinical Outcome of Peritoneal Dialysis Patients

PLOS ONE

Dear Dr. Szeto,

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.

This single-centre retrospective study examined the interaction between central obesity and frailty on outcome in a population of peritoneal dialysis patients. The topic is interesting, but a lot of work will need to be done in order to make this paper eligble for publication in PLOS ONE, see notes by the expert reviewers as well. The methodology is somewhat weak, and authors should consider to only include incident patients rather than incident and prevalent patients. The WHO classification for Obesity should be followed in the analyses, much more detail provided regarding the dialysis treatment and type of PD and medications. The paper should be rewritten according to STROBE guidelines. To improve readability, please remove irrelevant details and p values from the text (consider a table if you deem this is important), and make sure to deliver a focused and strong message. References should be updated as much as possible to include more recent relevant literature. I am willing to give the paper the benefit of the doubt and ask you to make revisions as per above and per reviewers' comments. There is no guarantee that a revised version of the paper would be accepted.

Please submit your revised manuscript by Sep 10 2020 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

We look forward to receiving your revised manuscript.

Kind regards,

Frank JMF Dor, M.D., Ph.D., FEBS, FRCS

Academic Editor

PLOS ONE

Journal Requirements:

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

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

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

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

2. Please refer to any post-hoc corrections to correct for multiple comparisons during your statistical analyses. If these were not performed please justify the reasons. Please refer to our statistical reporting guidelines for assistance (https://journals.plos.org/plosone/s/submission-guidelines.#loc-statistical-reporting).

[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

Reviewer #3: No

**********

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

Reviewer #1: Yes

Reviewer #2: No

Reviewer #3: Yes

**********

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

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

Reviewer #1: Yes

Reviewer #2: Yes

Reviewer #3: No

**********

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: No

Reviewer #3: 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: This paper makes an important observation of interaction between central obesity and frailty on outcome in a population of peritoneal dialysis patients.

comments

The paper should be presented according to STROBE guidelines for cohort studies - and in particular there should be a detailed STROBE diagram provided.

There is limited detail on the timing of the interventions with respect to peritoneal dialysis cycles - ie was abdominal girth measured drained out of PD fluid? When were and how were the interventions conducted - what was the accuracy, repeatability or the tests - who conducted them etc? Were all tests conducted at baseline?

Dialysis treatment data is not included in the analysis - for example prescription data (requirement for hypertonic exchanges) or peritoneal transport status. So essentially the dialysis treatment patients received is being ignored in the analysis. The same goes for hypertensive medications.

The study would be much more powerful if it had been conducted on incident rather than prevalent patients.

Reviewer #2: This was a difficult paper to read and follow. There is a huge amount of data with far too many p values for the number of patients and it was not clear what the 'message' of the paper was meant to be. It is already well recognised that frailty is found in about 50% of most dialysis populations, that it is found in all age groups and that frailty is related to poorer outcomes. The text is difficult as it is full of nonsignificant small differences. Furthermore it is not clear why some things were measured like pulse velocity. Another difficulty was the use of abbreviations - far too many - so one had to keep going back to earlier parts of the paper to find out what they stood for.

There are also problems with generasibility of findings as this is a single centre retrospective study with incident and prevalent patients in Chinese patients. As an example, BMI of >25 would not be defined as obesity in many parts of the world,

Reviewer #3: In the paper entitled “Interaction Between Central Obesity and Frailty on the Clinical Outcome of Peritoneal Dialysis Patients” Szeto et. al. has examined the relationship between overall survival (overall and cardiovascular mortality) and central obesity and frailty in patients on peritoneal dialysis (PD). The parameters used to collect the data are relevant. However, the following are the questions, which I would like the authors to address to fill the gap present in the paper.

1. How long were the patients on PD prior to recruiting them on the study? It has been shown in the previous study that patients on PD do gain wait with time and this is associated with changes in the metabolic profile, increased mortality rate and a higher PD failure rate independent of baseline obesity and fluid status. How do you address this issue in your study?

Ref: Kim JK, Park HC, Song YR, Kim HJ, Moon SJ, Kim SG. Effects of Excessive Body Fat Accumulation on Long-Term Outcomes During Peritoneal Dialysis. Perit Dial Int. 2019 20 May-Jun;39(3):268-275. doi: 10.3747/pdi.2018.00164. Epub 2019 Mar 6.

2. In this study, the Waist/Hip (W/H) ratio was measured once at the time of recruitment. Would the reported outcome have changed if serial measurements were carried out at 6-monthly intervals as the W/H ratio does change with the duration of patients on PD? This needs to be included in the discussion section.

Ref: Waist circumference as a predictor of mortality in peritoneal dialysis patients: a follow-up study of 48 months Br J Nutr. 2017 May;117(9):1299-1303.

3. How many patients in the study group were on CAPD, APD and Assisted PD?

4. In the study, obesity has been defined as a BMI over 25. Would further stratification of BMI (<25; 25- 35; and >35) and subgroup analysis have shown difference in the outcomes?

5. What was the incidence of peritonitis and mortality from sepsis in all 4 groups?

Ref: a. The Relationship Between Body Mass Index and Organism-Specific Peritonitis. Perit Dial Int. May-Jun 2018;38(3):206-214.

b. Impact of Obesity on Modality Longevity, Residual Kidney Function, Peritonitis, and Survival Among Incident Peritoneal Dialysis Patients. Am J Kidney Dis. 2018 Jun;71(6):802-813.

6. Relationship between depression, PD and long-term outcomes previously reported by the same authors have not been discussed in the paper?

7. The role of assisted PD in the management of frail patients is being increasingly recognised and there are several publications currently available, which should be included in the paper.

8. References: Majority of citations are more than 10 years old. There are several relevant publications over last 2 years which have not been included and needs attention.

**********

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: Yes: Martin Wilkie

Reviewer #2: No

Reviewer #3: Yes: Badri Shrestha MD FRCS

[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.

12 Aug 2020

3 August 2020

Editor in Chief

PLoS One

Dear Editor in Chief,

Re: ‘Interaction Between Central Obesity and Frailty on the Clinical Outcome of Peritoneal Dialysis Patients’ (PONE-D-20-19119)

Thanks for your letter on 27 July 2020. The following revisions have been made according to your recommendations:

Reply to Reviewer 1

1. The paper should be presented according to STROBE guidelines for cohort studies - and in particular there should be a detailed STROBE diagram provided.

The STROBE checklist is completed. A flow diagram on subjects’ inclusion and outcome is added to the manuscript as Fig 1.

2. There is limited detail on the timing of the interventions with respect to peritoneal dialysis cycles - ie was abdominal girth measured drained out of PD fluid? When were and how were the interventions conducted - what was the accuracy, repeatability or the tests - who conducted them etc? Were all tests conducted at baseline?

Page 5, paragraph 4, line 1-2: The anthropometric measurements were performed before infusion of dialysate fluid.

Page 5, paragraph 2, line 4-7: The waist and hip circumference were measured at the approximate midpoint between the lower margin of the last palpable rib and the top of the iliac crest, and around the widest portion of the buttocks respectively, according to the WHO 2008 report.

Page 5, paragraph 4, line 1-2: All measurements were performed by the same trained research nurse at baseline. These details are now added to the “Anthropometric and body composition assessment” part of the “Patients and Method” session.

3. Dialysis treatment data is not included in the analysis - for example prescription data (requirement for hypertonic exchanges) or peritoneal transport status. So essentially the dialysis treatment patients received is being ignored in the analysis. The same goes for hypertensive medications.

Page 8, Fig 1; page 33-34, S1 Table: The details for dialysis treatment (in terms of peritoneal dialysis types, and use of icodextrin dialysate), peritoneal transport status (reflected by 4-hour dialysate/plasma creatinine level) and details of medications are added to Table 1 and S1 Table. None of our subject used hypertonic dialysate fluid (i.e. 4.25% dextrose solution) regularly in their PD regimen.

4. The study would be much more powerful if it had been conducted on incident rather than prevalent patients.

Please refer to the reply to Point 1 of Editor comment.

Reply to Reviewer 2

1. This was a difficult paper to read and follow. There is a huge amount of data with far too many p values for the number of patients and it was not clear what the 'message' of the paper was meant to be.

We have modified our manuscript according to the STROBE guideline with appropriate headings and sessions. We have also added new flow diagram, rearranged and simplified our tables to allow readers understand the flow and data of our study.

We have also removed the p values from Table 1, and readers can refer to S1 Table for the corresponding p-value of the data in Table 1. We hope these measures improve readability of our manuscript.

2. It is already well recognised that frailty is found in about 50% of most dialysis populations, that it is found in all age groups and that frailty is related to poorer outcomes. The text is difficult as it is full of nonsignificant small differences.

There are a few published articles on prevalence and outcome of frail dialysis patients. However, only small number of them covered peritoneal dialysis patients. On the other hand, PD is often regarded as a better dialysis option over HD in elderlies, whom frailty is prevalent in this age group. We hope to provide more information, especially the adverse outcomes and interaction with body composition in the field of frailty in PD.

3. Furthermore it is not clear why some things were measured like pulse velocity.

Page 18, paragraph 1, line 11-17. Measurement of pulse wave velocity is important for us to understand the possible mechanism between frailty, central obesity and mortality in PD patients. This point is elaborated in the Discussion part.

4. Another difficulty was the use of abbreviations - far too many - so one had to keep going back to earlier parts of the paper to find out what they stood for.

We have minimized the use of abbreviations. To improve readability, the full terminology of abbreviation is labelled below tables and figures.

5. There are also problems with generasibility of findings as this is a single centre retrospective study with incident and prevalent patients in Chinese patients. As an example, BMI of >25 would not be defined as obesity in many parts of the world.

Page 19, paragraph 2, line 2-3: We believe our sample is a good representative of PD patients. Our rationale is elaborated in the Discussion part.

Page 5, paragraph 2, line 2-3: As for the issue of BMI, we have modified our definition of general obesity as BMI more than or equals 30 kg/m2 according to the WHO definitions. The results were modified according to the new definition.

Reply to Reviewer 3

1. How long were the patients on PD prior to recruiting them on the study? It has been shown in the previous study that patients on PD do gain wait with time and this is associated with changes in the metabolic profile, increased mortality rate and a higher PD failure rate independent of baseline obesity and fluid status. How do you address this issue in your study?

The mean duration on peritoneal dialysis in our cohort was 47.7 ± 51.9 months.

For the second question, please refer to the reply to Point 1 of Editor comment.

2. In this study, the Waist/Hip (W/H) ratio was measured once at the time of recruitment. Would the reported outcome have changed if serial measurements were carried out at 6-monthly intervals as the W/H ratio does change with the duration of patients on PD? This needs to be included in the discussion section.

Page 19, paragraph 2, line 6-9: It has been reported in the literature that BMI, body fat mass and waist circumference increases with duration of peritoneal dialysis, and such changes have important prognostic implications. However, the change in WHR depends on the degree and proportion of change in waist and hip circumference. Whether WHR changes with time on peritoneal dialysis, together with its prognostic implication are uncertain and should be explored in future study.

3. How many patients in the study group were on CAPD, APD and Assisted PD?

Page 8, Table 1: In our cohort, 233 (87.3%) patients were on continuous ambulatory peritoneal dialysis (CAPD). In the remaining subjects, 6 (2.2% of total) patients were on continuous cyclic peritoneal dialysis (CCPD) and 28 (10.5% of total) patients were on nocturnal intermittent peritoneal dialysis (NIPD).

On the other hand, 32 patients (80%) were on helper-assisted PD in the whole cohort. These details are now added in Table 1 and S1 Table.

4. In the study, obesity has been defined as a BMI over 25. Would further stratification of BMI (<25; 25- 35; and >35) and subgroup analysis have shown difference in the outcomes?

In short, we found no difference in after stratification of BMI and subgroup analysis.

There was no difference between patient survival and cardiovascular survival rates among individuals with BMI <25, 25-35 and >35 (survival rate: 73.5%, 64.8%, 85.7% respectively, log-rank test, p = 0.235) (cardiovascular survival rate: 78.0%, 71.4%, 57.1% respectively, log-rank test, p = 0.269).

In univariate Cox regression analysis, BMI 25-35 and BMI >35 did not predict patient survival (BMI 25-35: hazard ratio [HR] 1.324, 95% confidence interval [CI] 0.791 - 2.215, p = 0.286; BMI >35: HR 1.955, 95% CI 0.595 - 6.420, p = 0.269) and cardiovascular survival (BMI 25-35: HR 1.266, 95% CI 0.760 - 2.109, p = 0.366; BMI >35: HR 2.295, 95% CI 0.699 - 7.534, p = 0.171) with reference to BMI <25.

Among frail individuals, there were again no difference among individuals with BMI <25, 25-35 and >35 (survival rate: 60.4%, 56.3%, 100% respectively, log-rank test, p = 0.184) (cardiovascular survival rate: 66.0%, 66.7%, 66.7% respectively, log-rank test, p = 0.905).

In univariate Cox regression analysis, BMI 25-35 did not predict patient survival (HR 1.130, 95% CI 0.617 - 2.070, p = 0.692), while BMI 25-35 and BMI >35 did not predict cardiovascular survival (BMI 25-35: HR 1.017, 95% CI 0.518 - 1.995, p = 0.962; BMI >35: HR 1.357, 95% CI 0.315 - 5.856, p = 0.682) with reference to BMI <25.

Among non-frail individuals, there were only 1 subject with BMI >35. There were again no difference among individuals with BMI <25, 25-35 and >35 (survival rate: 82.3%, 71.9%, 0% respectively, log-rank test, p = 0.070) (cardiovascular survival rate: 86.1%, 75.4%, 0% respectively, log-rank test, p = 0.191).

In univariate Cox regression analysis, both BMI 25-35 and BMI >35 did not predict patient survival (BMI 25-35: HR 1.663, 95% CI 0.811 - 3.407, p = 0.165; BMI >35: HR 6.834, 95% CI 0.888 - 52.616, p = 0.065) and cardiovascular survival (BMI 25-35: HR 1.635, 95% CI 0.742 - 3.603, p = 0.222; BMI >35: HR 4.265, 95% CI 0.550 - 33.079, p = 0.165) when compared to BMI <25.

5. What was the incidence of peritonitis and mortality from sepsis in all 4 groups?

There were 77 peritonitis episodes (42 in non-frail group, 35 in frail group). 229 subjects (85.8%) were peritonitis-free. The number of peritonitis-free in non-frail non-central obese, frail non-central obese, non-frail central obese and frail central obese individuals were 68 (87.2%), 45 (93.8%), 56 (81.2%) and 60 (84.7%) respectively.

After 2 years of observation, 9 subjects (3.4%) died primarily because of infection. The infection-associated mortality rates in non-frail non-central obese, frail non-central obese, non-frail central obese and frail central obese individuals were 3.8%, 2.1%, 4.3% and 2.8% respectively (Log-rank test, p = 0.419).

6. Relationship between depression, PD and long-term outcomes previously reported by the same authors have not been discussed in the paper?

Page 11, paragraph 1, line 1-2: We added the reference of our previously reported depression and frailty in PD patients. However, the theme of this manuscript is mainly about the interaction between frailty, central obesity and mortality. Therefore, we do not discuss about depression in order to make the paper more focused.

7. The role of assisted PD in the management of frail patients is being increasingly recognised and there are several publications currently available, which should be included in the paper.

Page 16, paragraph 2, line 6-8: Recent publications on the role of assisted PD in frail patients were added to the Discussion Part.

8. References: Majority of citations are more than 10 years old. There are several relevant publications over last 2 years which have not been included and needs attention.

Please refer to the reply to Point 4 of Editor comment.

Reply to Editor Comments

Authors should consider to only include incident patients rather than incident and prevalent patients.

We agree with your suggestion that inclusion of incident patients only will reduce potential lead-time and survivor bias, and increase the power of our study. In view of that, we performed adjustments and subgroup analyses to address this issue.

Page 12, paragraph 2, line 1-7: Firstly, we performed a subgroup survival analysis on patients who were on PD therapy for less than 1 year. In short, the survival rates in non-frail central obese, non-frail non-central obese, frail non-central obese, and non-frail non-central obese individuals were 100%, 82.4%, 76.9% and 61.5% (log-rank test, p = 0.069), and the survival rates were grossly decreasing trend. Although the difference did not reach significance, which is likely contributed to a much smaller sample size after excluding patients on long duration of PD, it still gives us the impression that frailty and central obesity may have potential interaction in mortality in this group of patients. This observation should be confirmed with a study with larger sample size.

Page 20, paragraph 1, line 7-9: Secondly, our analysis reflected PD duration did not predict hospitalisation (p = 0.364) and mortality (p = 0.210). All the other reported results were adjusted with patients’ duration on PD.

Page 20, paragraph 1, line 17: Moreover, inclusion of prevalent PD cases could be a strength of our study as it can be applied universally to all PD cases that we encounter, regardless of their time on PD.

These details are now added to the Discussion Part.

The WHO classification for Obesity should be followed in the analyses, much more detail provided regarding the dialysis treatment and type of PD and medications.

The definition of obesity from the WHO classification is adopted in the analyses. The details of analyses with general obesity are modified with the adjusted definition. The details of dialysis treatment and medications are now included in Table 1.

The paper should be rewritten according to STROBE guidelines.

The paper is modified according to the STROBE guideline with a figure showing the flow of our study (Fig 1). The STROBE checklist is also completed.

To improve readability, please remove irrelevant details and p values from the text (consider a table if you deem this is important), and make sure to deliver a focused and strong message. References should be updated as much as possible to include more recent relevant literature.

We have modified our manuscript according to the STROBE guideline with appropriate headings and sessions. New flow diagram is also added in order to allow readers understand the flow of our study. We have also rearranged and simplified our tables. We hope these measures improve readability of our manuscript. The references are also updated with more recent relevant literature replacing the citations that are published more than 10 years ago. There are now only 4 citations that are more than 10 years old left – 2 of them are the landmark studies in frailty and frailty in dialysis (Reference 1 and 2).

We would like to resubmit the manuscript, with the changes highlighted in yellow, for consideration of publication in the PLoS One. Thank you for reviewing our article and we look forward to hearing your favorable reply.

Yours sincerely,

Dr. Gordon CK Chan

for Drs. JKC Ng, KM Chow, BCH Kwan, VWK Kwong, WF Pang, PMS Cheng, MC Law, CB Leung, PKT Li and CC Szeto

Submitted filename: Response to Reviewers.docx

Click here for additional data file.

21 Sep 2020

PONE-D-20-19119R1

Interaction Between Central Obesity and Frailty on the Clinical Outcome of Peritoneal Dialysis Patients

PLOS ONE

Dear Dr. Szeto,

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.

==============================

ACADEMIC EDITOR:

Most of the concerns raised by the reviewers have been addressed by the authors, but i would be keen that the authors address the comments by reviewer 2 in a bit more thorough way and add it to the discussion section:

-can you come with a more crisp and clear clinical message and comment on how much weight can be given to a study in small numbers of patients when outcome differences are barely or not significant.

-There is no explanation why higher Waist:Hip ratio predicts CV events in general population but is supposedly protective in non-frail PD population.

==============================

Please submit your revised manuscript by Nov 05 2020 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

We look forward to receiving your revised manuscript.

Kind regards,

Frank JMF Dor, M.D., Ph.D., FEBS, FRCS

Academic Editor

PLOS ONE

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

Reviewers' comments:

Reviewer's Responses to Questions

Comments to the Author

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

Reviewer #2: All comments have been addressed

Reviewer #3: All comments have been addressed

**********

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

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

Reviewer #2: Partly

Reviewer #3: Yes

**********

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

Reviewer #2: Yes

Reviewer #3: Yes

**********

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

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

Reviewer #2: Yes

Reviewer #3: Yes

**********

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

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

Reviewer #2: Yes

Reviewer #3: Yes

**********

6. Review Comments to the Author

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

Reviewer #2: Most of my concerns have been addressed - but I remain mystified by the clinical message and how much weight can be given to a study in small numbers of patients when outcome differences are barely or not significant. There is no explanation why higher Waist:Hip ratio predicts CV events in general population but is supposedly protective in non-frail PD population

Reviewer #3: The authors have addressed all queries raised by the reviewers and I am happy for the manuscript to be accepted for publication.

**********

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

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

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

Reviewer #2: No

Reviewer #3: No

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

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

8 Oct 2020

2 October 2020

Editor in Chief

PLoS One

Dear Editor in Chief,

Re: ‘Interaction Between Central Obesity and Frailty on the Clinical Outcome of Peritoneal Dialysis Patients’ (PONE-D-20-19119)

Thanks for your letter on 21 September 2020. The following revisions have been made according to your recommendations:

Reply to Editor and Reviewer 2

1. Can you come with a more crisp and clear clinical message and comment on how much weight can be given to a study in small numbers of patients when outcome differences are barely or not significant.

Page 20, paragraph 2, line 16-22: We acknowledged the small number of patients that we recruited in our study is one of our limitations. It is also one of the reasons that may lead to barely significant or insignificant outcome differences. Despite so, our results, which represent the first published data in this aspect, demonstrated the trend of outcome and provided valuable information by suggesting a potential association and interaction with the adverse outcomes. We suggest future research with a larger sample size to look into any association with the outcomes that we investigated. This limitation and recommendation are now included in the Discussion Part.

2. There is no explanation why higher Waist:Hip ratio predicts CV events in general population but is supposedly protective in non-frail PD population.

Page 19, paragraph 1, line 16-23: Abdominal obesity is also reported to be a protective factor against mortality in patients with stable coronary artery disease patients, and heart failure patients especially those with reduced ejection fraction. The proposed mechanisms are related to the extra energy reserve provided by visceral fat, and the anti-inflammatory and endotoxin-neutralizing effects mediated by a raised lipoprotein level. In addition, central obese individuals also seek medical attention at the earlier stage of acute illnesses as they tend to be more symptomatic, and their renin-angiotensin-aldosterone system is more attenuated, causing a higher blood pressure and better tolerance to cardioprotective medications. These findings and proposed mechanism are now included in the Discussion Part.

We would like to resubmit the manuscript, with the changes highlighted in yellow, for consideration of publication in the PLoS One. Thank you for reviewing our article and we look forward to hearing your favorable reply.

Yours sincerely,

Dr. Gordon CK Chan

for Drs. JKC Ng, KM Chow, BCH Kwan, VWK Kwong, WF Pang, PMS Cheng, MC Law, CB Leung, PKT Li and CC Szeto

Submitted filename: Response to Reviewers.docx

Click here for additional data file.

12 Oct 2020

Interaction Between Central Obesity and Frailty on the Clinical Outcome of Peritoneal Dialysis Patients

PONE-D-20-19119R2

Dear Dr. Szeto,

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,

Frank JMF Dor, M.D., Ph.D., FEBS, FRCS

Academic Editor

PLOS ONE

Additional Editor Comments (optional):

Many thanks for making these final changes.

Reviewers' comments:

14 Oct 2020

PONE-D-20-19119R2

Interaction Between Central Obesity and Frailty on the Clinical Outcome of Peritoneal Dialysis Patients

Dear Dr. SZETO:

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

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

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

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

Kind regards,

PLOS ONE Editorial Office Staff

on behalf of

Dr. Frank JMF Dor

Academic Editor

PLOS ONE