Time-varying effects of body mass index on mortality among hemodialysis patients

Results from a nationwide korean registry

Sejoong Kim, Jong Cheol Jeong, Shin Young Ahn, Kibbeum Doh, Dong Chan Jin, Ki Young Na

Research output: Contribution to journalArticle

Abstract

Background: Unlike patterns observed in the general population, obesity is associated with better survival among hemodialysis patients, which could be explained by reverse causation or illness-related weight loss. However, the time-varying effect of body mass index (BMI) on hemodialysis survival has not been investigated. Therefore, this study investigated the time-varying effect of BMI on mortality after starting hemodialysis. Methods: In the present study, we examined Korean Society of Nephrology data from 16,069 adult patients who started hemodialysis during or after the year 2000. Complete survival data were obtained from Statistics Korea. Survival analysis was performed using Cox regression and a non-proportional hazard fractional polynomial model. Results: During the median follow-up of 8.6 years, 9,272 patients (57.7%) died. Compared to individuals with normal BMI (18.5-24.9 kg/m2), the underweight group (< 18.5 kg/m2) had a higer mortality hazard ratio (HR, 1.292; 95% confidence interval [CI], 1.203-1.387; P < 0.001) and the overweight group (25.0-29.9 kg/m2) had a lower mortality HR (0.904; 95% CI, 0.829-0.985; P = 0.022). The underweight group had increasing HRs during the first 3 to 7 years after starting hemodialysis, which varied according to age group. The young obese group (< 40 years old) had a U-shaped temporal trend in their mortality HRs, which reflected increased mortality after 7 years. Conclusion: The obese hemodialysis group had better survival during the early post-dialysis period, although the beneficial effect of obesity disappeared 7 years after starting hemodialysis. The young obese group also had an increased mortality HR after 7 years.

Original languageEnglish
Pages (from-to)90-99
Number of pages10
JournalKidney Research and Clinical Practice
Volume38
Issue number1
DOIs
Publication statusPublished - 2019 Mar 1

Fingerprint

Registries
Renal Dialysis
Body Mass Index
Mortality
Survival
Thinness
Obesity
Confidence Intervals
Nephrology
Statistical Models
Survival Analysis
Korea
Causality
Weight Loss
Dialysis
Age Groups
Population

Keywords

  • Body mass index
  • Obesity
  • Renal dialysis
  • Reverse causation
  • Time-varying hazard

ASJC Scopus subject areas

  • Nephrology
  • Urology

Cite this

Time-varying effects of body mass index on mortality among hemodialysis patients : Results from a nationwide korean registry. / Kim, Sejoong; Jeong, Jong Cheol; Ahn, Shin Young; Doh, Kibbeum; Jin, Dong Chan; Na, Ki Young.

In: Kidney Research and Clinical Practice, Vol. 38, No. 1, 01.03.2019, p. 90-99.

Research output: Contribution to journalArticle

Kim, Sejoong ; Jeong, Jong Cheol ; Ahn, Shin Young ; Doh, Kibbeum ; Jin, Dong Chan ; Na, Ki Young. / Time-varying effects of body mass index on mortality among hemodialysis patients : Results from a nationwide korean registry. In: Kidney Research and Clinical Practice. 2019 ; Vol. 38, No. 1. pp. 90-99.
@article{ec58445b965c42cfb62160ee9109ab54,
title = "Time-varying effects of body mass index on mortality among hemodialysis patients: Results from a nationwide korean registry",
abstract = "Background: Unlike patterns observed in the general population, obesity is associated with better survival among hemodialysis patients, which could be explained by reverse causation or illness-related weight loss. However, the time-varying effect of body mass index (BMI) on hemodialysis survival has not been investigated. Therefore, this study investigated the time-varying effect of BMI on mortality after starting hemodialysis. Methods: In the present study, we examined Korean Society of Nephrology data from 16,069 adult patients who started hemodialysis during or after the year 2000. Complete survival data were obtained from Statistics Korea. Survival analysis was performed using Cox regression and a non-proportional hazard fractional polynomial model. Results: During the median follow-up of 8.6 years, 9,272 patients (57.7{\%}) died. Compared to individuals with normal BMI (18.5-24.9 kg/m2), the underweight group (< 18.5 kg/m2) had a higer mortality hazard ratio (HR, 1.292; 95{\%} confidence interval [CI], 1.203-1.387; P < 0.001) and the overweight group (25.0-29.9 kg/m2) had a lower mortality HR (0.904; 95{\%} CI, 0.829-0.985; P = 0.022). The underweight group had increasing HRs during the first 3 to 7 years after starting hemodialysis, which varied according to age group. The young obese group (< 40 years old) had a U-shaped temporal trend in their mortality HRs, which reflected increased mortality after 7 years. Conclusion: The obese hemodialysis group had better survival during the early post-dialysis period, although the beneficial effect of obesity disappeared 7 years after starting hemodialysis. The young obese group also had an increased mortality HR after 7 years.",
keywords = "Body mass index, Obesity, Renal dialysis, Reverse causation, Time-varying hazard",
author = "Sejoong Kim and Jeong, {Jong Cheol} and Ahn, {Shin Young} and Kibbeum Doh and Jin, {Dong Chan} and Na, {Ki Young}",
year = "2019",
month = "3",
day = "1",
doi = "10.23876/j.krcp.18.0094",
language = "English",
volume = "38",
pages = "90--99",
journal = "Kidney Research and Clinical Practice",
issn = "2211-9132",
publisher = "Elsevier Korea",
number = "1",

}

TY - JOUR

T1 - Time-varying effects of body mass index on mortality among hemodialysis patients

T2 - Results from a nationwide korean registry

AU - Kim, Sejoong

AU - Jeong, Jong Cheol

AU - Ahn, Shin Young

AU - Doh, Kibbeum

AU - Jin, Dong Chan

AU - Na, Ki Young

PY - 2019/3/1

Y1 - 2019/3/1

N2 - Background: Unlike patterns observed in the general population, obesity is associated with better survival among hemodialysis patients, which could be explained by reverse causation or illness-related weight loss. However, the time-varying effect of body mass index (BMI) on hemodialysis survival has not been investigated. Therefore, this study investigated the time-varying effect of BMI on mortality after starting hemodialysis. Methods: In the present study, we examined Korean Society of Nephrology data from 16,069 adult patients who started hemodialysis during or after the year 2000. Complete survival data were obtained from Statistics Korea. Survival analysis was performed using Cox regression and a non-proportional hazard fractional polynomial model. Results: During the median follow-up of 8.6 years, 9,272 patients (57.7%) died. Compared to individuals with normal BMI (18.5-24.9 kg/m2), the underweight group (< 18.5 kg/m2) had a higer mortality hazard ratio (HR, 1.292; 95% confidence interval [CI], 1.203-1.387; P < 0.001) and the overweight group (25.0-29.9 kg/m2) had a lower mortality HR (0.904; 95% CI, 0.829-0.985; P = 0.022). The underweight group had increasing HRs during the first 3 to 7 years after starting hemodialysis, which varied according to age group. The young obese group (< 40 years old) had a U-shaped temporal trend in their mortality HRs, which reflected increased mortality after 7 years. Conclusion: The obese hemodialysis group had better survival during the early post-dialysis period, although the beneficial effect of obesity disappeared 7 years after starting hemodialysis. The young obese group also had an increased mortality HR after 7 years.

AB - Background: Unlike patterns observed in the general population, obesity is associated with better survival among hemodialysis patients, which could be explained by reverse causation or illness-related weight loss. However, the time-varying effect of body mass index (BMI) on hemodialysis survival has not been investigated. Therefore, this study investigated the time-varying effect of BMI on mortality after starting hemodialysis. Methods: In the present study, we examined Korean Society of Nephrology data from 16,069 adult patients who started hemodialysis during or after the year 2000. Complete survival data were obtained from Statistics Korea. Survival analysis was performed using Cox regression and a non-proportional hazard fractional polynomial model. Results: During the median follow-up of 8.6 years, 9,272 patients (57.7%) died. Compared to individuals with normal BMI (18.5-24.9 kg/m2), the underweight group (< 18.5 kg/m2) had a higer mortality hazard ratio (HR, 1.292; 95% confidence interval [CI], 1.203-1.387; P < 0.001) and the overweight group (25.0-29.9 kg/m2) had a lower mortality HR (0.904; 95% CI, 0.829-0.985; P = 0.022). The underweight group had increasing HRs during the first 3 to 7 years after starting hemodialysis, which varied according to age group. The young obese group (< 40 years old) had a U-shaped temporal trend in their mortality HRs, which reflected increased mortality after 7 years. Conclusion: The obese hemodialysis group had better survival during the early post-dialysis period, although the beneficial effect of obesity disappeared 7 years after starting hemodialysis. The young obese group also had an increased mortality HR after 7 years.

KW - Body mass index

KW - Obesity

KW - Renal dialysis

KW - Reverse causation

KW - Time-varying hazard

UR - http://www.scopus.com/inward/record.url?scp=85067257592&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85067257592&partnerID=8YFLogxK

U2 - 10.23876/j.krcp.18.0094

DO - 10.23876/j.krcp.18.0094

M3 - Article

VL - 38

SP - 90

EP - 99

JO - Kidney Research and Clinical Practice

JF - Kidney Research and Clinical Practice

SN - 2211-9132

IS - 1

ER -