Association of Body Mass Index with Estimated Glomerular
Filtration Rate and Incident Proteinuria
Seung Min Lee
1
, Minseon Park
2
and Hyung-Jin Yoon
1,*
1
Department of Biomedical Engineering, Seoul National University College of Medicine, Seoul, Korea
2
Department of Family Medicine, Seoul National University Hospital, Seoul, Korea
*corresponding author
Keywords: Glomerular Filtration Rate, Body Mass Index, Chronic Kidney Disease, Clinical Epidemiology, Incident
Proteinuria.
Abstract: Obesity has been one of the most important risk factors of chronic kidney disease (CKD). But the association
of body mass index (BMI) with estimated glomerular filtration rate (eGFR) and incident proteinuria has not
been studied well. The goal of this study was to elucidate the association of BMI with eGFR and proteinuria
using nationwide health examination data. These associations were investigated with data of Korean adults
who had undergone health screenings at least three times between 2009 and 2014. eGFR was calculated with
Chronic Kidney Disease Epidemiology collaboration equation based on serum creatinine level. The
association between BMI and eGFR was analysed with a generalized addictive model adjusting for possible
confounders. Similarly, the association between BMI and incident proteinuria was analysed with Cox hazard
model adjusting for possible confounders. As a result, a V-shape relationship between BMI and eGFR was
observed. The nadir was around 29 kg/m
2
. With subgroup analyses for the association between BMI and
eGFR, a V-shape association was observed in men and younger age group and an inverse association was
observed in women and older age group. A reverse J-shape association between BMI and the adjusted hazard
ratio of incident proteinuria was observed. The nadir was approximately estimated around 22 kg/m
2
.
1 INTRODUCTION
Obesity has been one of the most important risk
factors of Chronic Kidney Disease (Mahmoodnia,
2017). CKD is an emerging health issue because of its
high prevalence and incidence worldwide and its
association with cardiovascular morbidity and
mortality. The relationship between body mass index
(BMI), the most representative index of obesity, and
estimated glomerular filtration rate (eGFR) has not
been studied well. Several studies with small sample
size have reported the inverse linear association
between BMI and eGFR (Grubbs, 2014). Obesity has
been associated with abnormally increased eGFR and
decrease of eGFR or measured GFR after successful
weight reduction, such as Bariatric surgery is rather
well-known (Li, 2016). These observations are not
consistent to the observed inverse linear association
between BMI and eGFR. Similarly, only a few
prospective cohort studies observed an association
between overweight and obesity and increased risk of
incident proteinuria, and the nature of the relationship
between BMI and incident proteinuria has not been
reported yet. A cross-sectional study has observed the
U-shape association between BMI and prevalent
proteinuria (Sato, 2014). Although a sex-specific
association of incident proteinuria with underweight
and no association with overweight and obesity after
2 years’ follow-up has been reported, the follow-up
period of that study might not be long enough (Jang,
2014).
The goal of this study was to elucidate the
association of BMI with eGFR and incident
proteinuria using nationwide health examination data.
2 METHODS
Regular health screening at designated screening
hospital across the country is obligatory for adult
Koreans. The number of eligible subjects between
2009 and 2014 was between 15,036,607 and
16,456,214. The participation rate was between
66.0% and 74.8% (Cheol, 2016). The data of Korean
Lee, S., Park, M. and Yoon, H-J.
Association of Body Mass Index with Estimated Glomerular Filtration Rate and Incident Proteinuria.
DOI: 10.5220/0006716905870590
In Proceedings of the 11th International Joint Conference on Biomedical Engineering Systems and Technologies (BIOSTEC 2018) - Volume 5: HEALTHINF, pages 587-590
ISBN: 978-989-758-281-3
Copyright © 2018 by SCITEPRESS Science and Technology Publications, Lda. All rights reserved
587
adults (27,448,308 participants) who had undergone
health screenings between 2009 and 2014 were
analysed. Health screening tests performed on the
participants under 20 years old and foreign citizens
were excluded (728,569 participants). Subjects with
missing information on BMI, eGFR and other
confounders were also excluded. For more accurate
analysis, we set the minimum time interval between
the first health screening of each individual to the next
his/her health screening to 6 months. Furthermore,
persons who had undergone health screenings at least
three times between 2009 and 2014 were included in
this study (12,211,062 adults). BMI was calculated
from the measured height and weight. The eGFR
(mL/min/1.73 m
2
) was calculated by the Chronic
Kidney Disease Epidemiology Collaboration
equation based on serum creatinine level (Inker,
2012). Depending on smoking habit, persons were
categorized as non-smoker, ex-smoker, and smoker.
Persons who performed moderate or high intensity
exercise more than three times a week were classified
as ‘yes’ in the category of regular exercise. Alcohol
consumption were categorized as non-drinker, 1~2
times a week, 3~4 times a week and heavy drinker
(almost daily).
To observe any differential association between
BMI and eGFR according to age or sex, the
participants were divided into subgroups by sex and
sex-specific median age. The relationship of BMI with
eGFR was analysed with a generalized additive model
adjusting for possible confounding variables at
baseline such as age, sex, systolic blood pressure,
fasting serum glucose, serum triglycerides, serum
high-density lipoprotein- cholesterol, smoking status,
regular alcohol consumption, regular exercise, and
known history of medication for diabetes and hypertension.
To investigate the association between BMI and
incident proteinuria with national health screening
data, we analysed the
Table 1: Characteristics according to body mass index categories.
Total
(n=12,211,062)
BMI
1
< 18.5 kg/m
2
(n=415,403)
18.5 kg/m
2
BMI < 25 kg/m
2
(n=7,841,447)
BMI ≥ 25 kg/m
2
(n=3,954,212)
P-value
Age (years)
46.9 ± 13.6
39.3 ± 14.9
46.4 ± 13.6
48.56 ± 13.0
<.001
Sex
Male
Female
5,541,961 (45.4%)
6,669,101 (54.6%)
280,406 (67.5%)
134,997 (32.5%)
3,783,941 (48.3%)
4,057,506 (51.7%)
1,477,614 (37.4%)
2,476,598 (62.6%)
<.001
SBP
2
(mmHg)
122.2 ± 14.8
113.0 ± 13.5
120.3 ± 14.3
127.1 ± 14.4
<.001
FSG
3
(mg/dL)
96.9 ± 22.4
90.2 ± 17.9
95.2 ± 20.9
101.0 ± 24.9
<.001
TG
4
(mg/dL)
132.6 ± 91.9
82.1 ± 49.1
119.2 ± 80.4
164.7 ± 106.6
<.001
HDL
5
(mg/dL)
55.0 ± 13.6
63.0 ± 14.3
56.4 ± 13.7
51.3 ± 12.4
<.001
Smoking
Never
Formal
Current
7,366,297 (60.3%)
1,821,408 (14.9%)
3,023,357 (24.8%)
297,972 (71.7%)
27,479 (6.6%)
89,952 (21.7%)
4,892,862 (62.4%)
1,073,589 (13.7%)
1,874,996 (23.9%)
2,175,463 (55%)
720,340 (18.2%)
1,058,409 (26.8%)
<.001
Regular exercise
6
Yes
No
2,926,581 (24%)
9,284,481 (76%)
60,088 (14.5%)
355,315 (85.5%)
1,867,823 (23.8%)
5,973,624 (76.2%)
998,670 (25.3%)
2,955,542 (74.7%)
<.001
Alcohol consumption
7
None
1 ~ 2
3 ~ 4
> 4
6,234,122 (51.1%)
4,383,086 (35.9%)
1,155,360 (9.5%)
438,494 (3.6%)
234,699 (56.5%)
143,920 (34.6%)
25,504 (6.1%)
11,280 (2.7%)
4,078,439 (52%)
2,793,094 (35.6%)
693,639 (8.8%)
276,275 (3.5%)
1,920,984 (48.6%)
1,446,072 (36.6%)
436,217 (11%)
150,939 (3.8%)
<.001
Anti-HT
8
Yes
No
1,679,739 (13.8%)
10,531,323 (86.2%)
13,916 (3.3%)
401,487 (96.7%)
829,837 (10.6%)
7,011,610 (89.4%)
835,986 (21.1%)
3,118,226 (78.9%)
<.001
Anti-Diabetic
9
Yes
No
556,411 (4.6%)
11,654,651 (95.4%)
5,543 (1.3%)
409,860 (98.7%)
291,845 (3.7%)
7,549,602 (96.3%)
259,023 (6.6%)
3,695,189 (93.4%)
<.001
1
Body mass index;
2
Systolic blood pressure;
3
Fasting serum glucose;
4
Serum triglycerides;
5
Serum high-density lipoprotein-cholesterol;
6
Regular
exercise: moderate or high intensity exercise, more than three times per week;
7
Number of alcohol consumption per week;
8
History of anti-
hypertensive medication;
9
History of anti-diabetic medication
Note: Number (percentage %) for categorical variables; Mean ± standard deviation for continuous variables.
P-value was calculated by ANOVA-test for continuous variables and Pearson’s chi-squared test for categorical variables. P-values < 0.05 denote
statistical significance.
HEALTHINF 2018 - 11th International Conference on Health Informatics
588
data of participants who had eGFR 60 mL/min/1.73
m
2
or above and normo-proteinuria at baseline
(11,559,520 adults). Incident proteinuria was defined
as the occurred event when the result of urine dipstick
test showed a protein 1+ or higher during the follow-
up period in each individual who had negative result
in urine dipstick test at the first health screening.
Figure 1: The relationship between body mass index and
estimated glomerular filtration rate was evaluated with a
general additive model with adjustment for age, sex,
smoking status, regular exercise, regular alcohol
consumption, known history of diabetes or hypertension
medication, systolic blood pressure, fasting serum glucose,
serum triglycerides, and serum high-density lipoprotein-
cholesterol at baseline. Estimated glomerular filtration rate
was calculated with Chronic Kidney Disease Epidemiology
Collaboration equation based on serum creatinine. The
shaded area represents 95% confidence interval.
The Cox hazard model was used to analyze the
association between BMI and incident proteinuria
with the adjustment of possible confounding
variables at baseline, such as age, sex, fasting serum
glucose, serum triglycerides, serum high-density
lipoprotein-cholesterol, systolic blood pressure,
known history of diabetes or hypertension medication,
smoking status, regular alcohol consumption, and
regular exercise. To visualize the association of BMI
with eGFR and incident proteinuria, penalized splines
as the smoothing were implemented by the R function
pspline in package survival (degree of freedom set as
default).
3 RESULTS
As shown in figure 1, the V-shape association between
BMI and eGFR was observed in the analysis of total
population. The nadir was around 29 kg/m
2
. With
subgroup analyses, the V-shape association was
observed only in men and younger age group (sex-
specific median age; 44 years old in men, 49 years old
in women). In women and older age group, the
inverse association between BMI and eGFR was
observed.
Figure 2: The association between body mass index and
incident proteinuria in total population. A reverse J-shape
association between body mass index and adjusted hazard
ratio of incident proteinuria was observed with adjustment
for age, sex, smoking status, regular exercise, regular
alcohol consumption, known history of diabetes or
hypertension medication, systolic blood pressure, fasting
serum glucose, serum triglycerides, and serum high-density
lipoprotein-cholesterol at baseline. The shaded area
represents 95% confidence interval.
Association of Body Mass Index with Estimated Glomerular Filtration Rate and Incident Proteinuria
589
The reverse J-shape association between BMI and the
adjusted hazard ratio of incident proteinuria was
observed. The nadir was approximately estimated
around 22 kg/m
2
as shown in figure 2. With subgroup
analyses, there was no difference according to sex and
age (sex-specific median age; 44 years old in men, 49
years old in women).
Figure 3: The association between body mass index and
incident proteinuria in subgroups; A reverse J-shape
association between body mass index and the adjusted hazard
ratio of incident proteinuria was not differential according to
sex and age with adjustment for age, sex, smoking status,
regular exercise, regular alcohol consumption, known history
of diabetes or hypertension medication, systolic blood
pressure, fasting serum glucose, serum triglycerides, and serum
high-density lipoprotein-cholesterol at baseline. The shaded
area represents 95% confidence interval.
4 CONCLUSIONS
The association of BMI and eGFR was not linear and
differential according to sex and age. Between BMI
and incident proteinuria, the reverse J-shape
association was observed and its nadir between 22
and 23 kg/m
2
. It is necessary to consider non-linear
association of BMI and eGFR and incident
proteinuria when the association of obesity with renal
function or incident proteinuria is evaluated. Clinical
implications of these observations need to be studied
with future studies.
ACKNOWLEDGEMENTS
This research was supported by Next-Generation
Information Computing Development Program
through the National Research Foundation of Korea
(NRF) funded by the Ministry of Science, ICT (NRF-
2017M3C4A7083412). The National Health
Information Database made by the National Health
Insurance Service (NHIS) of Korea was used.
REFERENCES
Grubbs, V., Lin, F., Vittinghoff, E., Shlipak, M., et al. (2014).
Body Mass Index and Early Kidney Function Decline in
Young Adults: A Longitudinal Analysis of the CARDIA
(Coronary Artery Risk Development in Young Adults)
Study. American Journal of Kidney Diseases, 63(4),
pp.590-597.
Hobbs, H., Farmer, C., Irving, J., et al. (2011). Is high body
mass index independently associated with diminished
glomerular filtration rate? An Epidemiological study.
Journal of Renal Care, 37(3), pp.148-154.
Inker LA, Schmid CH, Tighiouart H, et al. (2012). Estimating
glomerular filtration rate from serum creatinine and
cystatin C. N Engl J Med.;367:20-29.
Li, K., Zou, J., Ye, Z., Di, J., et al. (2016). Effects of Bariatric
Surgery on Renal Function in Obese Patients: A Systematic
Review and Meta Analysis. PLOS ONE, 11(10),
p.e0163907.
Mahmoodnia, L. and Tamadon, M. (2017). On the occasion of
world kidney day 2017; obesity and its relationship with
chronic kidney disease. Journal of Nephropathology, 6(3),
pp.105-109.
Kawamoto, R., Kohara, K., Tabara, Y., Et Al. (2008). An
Association between Body Mass Index and Estimated
Glomerular Filtration Rate. Hypertension Research, 31(8),
pp.1559-1564.
Kansui, Y., Ohtsubo, T., Goto, K., et al. (2012). Association of
Body Mass Index with Glomerular Filtration Rate in
Japanese: A Cross-Sectional Study in Work-Site
Population. Clinical and Experimental Hypertension,
34(2), pp.140-144.
Cheol Seong, S., Kim, Y., Khang, Y., et al. (2016). Data
Resource Profile: The National Health Information
Database of the National Health Insurance Service in
South Korea. International Journal of Epidemiology,
p.dyw 253.
Sato, Y., Fujimoto, S., Konta, T., Iseki, K., et al. (2013). U-
shaped association between body mass index and
proteinuria in a large Japanese general population sample.
Clinical and Experimental Nephrology, 18(1), pp.75-86.
Jang, C., Hyun, Y., Lee, K. and Kim, H. (2014). The
association between underweight and the development of
albuminuria is different between sexes in relatively healthy
Korean subjects. Nephrology Dialysis Transplantation,
29(11), pp.2106-2113.
HEALTHINF 2018 - 11th International Conference on Health Informatics
590