Professional Documents
Culture Documents
1342 Am J Clin Nutr 2018;108:1342–1351. Printed in USA. © 2018 American Society for Nutrition. All rights reserved.
VITAMIN D DEFICIENCY AND RENAL HYPERFILTRATION 1343
(15), indicating that vitamin D also has extraskeletal effects. the Division of Chronic Disease Surveillance of the Korea Cen-
Vitamin D also plays pleotropic roles in decreasing the risk ters for Disease Control and Prevention of the Ministry of Health
of several chronic illnesses, including cancers, autoimmune and Welfare to monitor the general health and nutrition status
diseases, infectious diseases, and cardiovascular disease (16–18). of South Koreans (19). The KNHANES is composed of demo-
Notably, vitamin D deficiency is known to be associated with an graphic, anthropometric, nutritional, and personal medical his-
increased risk of adverse clinical outcomes in patients with CKD tory data collected by trained investigators. A sampling frame was
(19, 20). Conversely, decreased renal function is associated with developed based on multistage probability stratified according to
a reduced synthesis of 1,25-dihydroxyvitamin D, the active form geographic location, sex, and age. Each KNHANES consists of
of vitamin D, and consequently results in disease progression independent sets of participants from the South Korean popula-
(21). Moreover, the renoprotective effects of vitamin D have tion. All participants were randomly selected from 600 randomly
been demonstrated in human clinical trials. Oral vitamin D or selected districts of cities and provinces in South Korea.
vitamin D analog supplementation enhanced the antiproteinuric Participants in KNHANES IV–VI were screened for this study.
effects of the renin angiotensin system (RAS) inhibitors and A total of 53,105 subjects were initially screened and participants
reduced albuminuria in patients with diabetic nephropathy and <18 y old or with missing data were excluded. Given the aim of
METHODS
Anthropometric and laboratory data
Study subjects Anthropometric measurements were obtained by trained ex-
This study was based on data from the 4th , 5th , and 6th editions perts following standardized protocols. The BMI was calculated
of the Korean NHANES (KNHANES IV, V, and VI, 2008– as weight (kg) divided by the square of height (m2 ). Activity
2015). KNHANES is a nationwide, population-based, and cross- restriction was determined if the subject answered “yes” when
sectional health examination and survey regularly conducted by questioned whether they had any problems associated with daily
physical activity. Systolic blood pressure (SBP) was measured >95th percentile after adjustment for logarithm-transformed age,
twice on the right arm at 5-min intervals using a standard mercury sex, height, weight, and history of hypertension or diabetes
sphygmomanometer (Baumanometer; Baum) and recorded as was defined as RHF. The eGFR was calculated from the
average values. Hypertension was defined as SBP ≥140 mm Hg, Korean version of the Chronic Kidney Disease Epidemiology
diastolic blood pressure ≥90 mm Hg, use of antihypertensive Collaboration equation (30, 31).
medications, or previous diagnosis by a physician. Diabetes
mellitus was defined as serum fasting blood glucose ≥126 mg/dL,
use of antidiabetic medication, or previous diagnosis by a Statistical analysis
physician. All statistical analyses were performed with IBM SPSS
Blood samples were collected after an ≥8-h fast and were software for Windows version 23.0 (IBM Corporation, Armonk,
transported to a central laboratory within 24 h. Serum 25- NY) and R software 3.3.1 (http://www.R-project.org). Contin-
hydroxyvitamin D [25(OH)D] concentration and ferritin were uous variables are expressed as mean ± SD or median (IQR),
determined by radioimmunoassay (DiaSorin Inc.) using a γ - and categoric variables as absolute numbers with percentages.
counter (1470 Wizard; PerkinElmer). The interassay coefficient Comparisons between groups were made using ANOVA or
Vitamin D concentration
estimated glomerular filtration rate; FPG, fasting plasma glucose; HbA1c, glycated hemoglobin; MI, myocardial infarction; PTH, parathyroid hormone; SBP,
systolic blood pressure; 25(OH)D, 25-hydroxyvitamin D.
2 PTH was measured in 10,136 participants. The values for the groups with <10, 10–19.9, 20–29.9, and ≥30 ng/mL were 508, 5161, 3532, and 935,
respectively.
were lower, whereas PTH concentrations were higher, in the in the other groups [adjusted mean eGFR = 100.2 mL · min–1 ·
25(OH)D <10 ng/mL group than in the ≥30 ng/mL group. PTH 1.73 m–2 (95% CI: 99.3, 101.1 mL · min–1 · 1.73 m–2 ), 98.6 mL ·
was only measured in 10,136 participants and its mean value for min–1 · 1.73 m–2 (95% CI: 98.3, 98.9 mL · min–1 · 1.73 m–2 ), 97.4
all subjects was 66.4 ± 27.1 pg/mL. mL · min–1 · 1.73 m–2 (95% CI: 96.9, 97.8 mL · min–1 · 1.73 m–2 ),
and 98.1 mL · min–1 · 1.73 m–2 (95% CI: 96.8, 99.3 mL · min–1
· 1.73 m–2 ) in the groups with 25(OH)D concentrations of <10,
Association of adjusted mean of eGFR and serum vitamin D 10–19.9, 20–29.9, and ≥30 ng/mL, respectively, P-trend < 0.001;
concentrations Figure 2B].
Pearson’s correlation analysis indicated that vitamin D con-
centrations were negatively associated with eGFR (r = −0.19,
P < 0.001). In a multivariable linear regression model, vitamin Prevalence of RHF according to vitamin D status
D concentrations were related to clinical factors: they were Because the adjusted mean eGFR was higher in the vitamin
negatively associated with eGFR (β = −0.02; 95% CI: −0.02, D–deficient groups, the prevalence of RHF according to vitamin
−0.01; P < 0.001), as well as BMI, SBP, and fasting plasma D concentration was assessed. RHF was defined as a logarithm-
glucose, whereas they were positively associated with age and transformed eGFR with an adjusted residual >95th percentile.
hemoglobin (Supplemental Table 1). Next, we examined the The results showed that the prevalence of RHF was significantly
association between adjusted mean eGFR and serum vitamin higher in the 25(OH)D <10 ng/mL group than in the others
D concentrations treated as linear covariate. The adjusted mean (5.8%, 4.6%, 4.4%, and 5.0% in the groups with 25(OH)D
eGFR significantly increased with decreasing amounts of vitamin <10, 10–19.9, 20–29.9, and ≥30 ng/mL, respectively; Figure 3).
D after adjustment for age, sex, and history of hypertension Furthermore, we evaluated baseline differences among the
or diabetes (Figure 2A). A general linear regression model 25(OH)D <10 ng/mL group with or without RFH (n = 2693).
was used to further assess the relationship between eGFR and The subjects with RHF showed significantly higher proportions
vitamin D status groups, adjusting for age, sex, BMI, history of men and history of hypertension, diabetes, and stroke,
of hypertension or diabetes, hemoglobin, and LDL cholesterol. and higher BMIs and SBP than did those without RHF. The
When the subjects were categorized into 4 groups according subjects with RHF also showed significantly higher amounts of
to vitamin D concentration, the adjusted mean eGFR was eGFR, fasting plasma glucose, glycated hemoglobin, and alkaline
significantly higher in the group with 25(OH)D <10 ng/mL than phosphates than did those without RHF (Supplemental Table 2).
1346 JHEE ET AL.
and hypertension, and obesity. We defined obesity as BMI 20–29.9, and ≥30 ng/mL groups, respectively; Supplemental
(in kg/m2 ) ≥25 according to WHO recommendations for Figure 2]. Next, the ORs of RHF were evaluated across
Asian populations (32). Increased ORs of RHF associated with each group. The highest risk of RHF in the 25(OH)D <10
25(OH)D concentrations <10 ng/mL were consistent in sub- ng/mL group remained a consistent event after adjustment
groups with young ages, no diabetes, and nonobese, but showed for vitamin supplementation (OR: 2.28; 95% CI: 1.63, 3.18;
no association with gender and hypertension; no significant P < 0.001; Supplemental Figure 3). Furthermore, we performed
interactions were found across all subgroups (Figure 6). propensity score matching (PSM) analysis. In PSM, subjects
were classified into 2 groups with or without vitamin D
supplements. Subsequently, we determined the propensity score
Sensitivity analysis using binary logistic regression with greedy nearest-neighbor
matching technique without replacement. We used a caliper of
The impact of vitamin D deficiency on RHF was further
0.2 times the SD. Covariates for PSM were age, sex, BMI,
evaluated in the subgroup that included subjects taking vitamin
SBP, histories of hypertension or diabetes, LDL cholesterol, and
supplements (n = 28,957). First, the prevalence of RHF among
hemoglobin. PSM yielded 10,424 matched pairs (1:1 ratio) of
groups stratified according to vitamin D serum concentrations
was assessed and the results consistently showed that the
25(OH)D <10 ng/mL group had the highest prevalence of RHF
[6.0%, 5.7%, 5.2%, and 4.5% in the 25(OH)D <10, 10–19.9,
FIGURE 4 Association of serum vitamin D concentrations with ORs FIGURE 5 Adjusted ORs of RHF according to groups with different
for RHF in fully adjusted logistic regression model using restricted cubic vitamin D concentrations. Black squares indicate ORs, and the error bars
splines. Black line shows ORs, and dotted lines represent 95% CIs. RHF, represent 95% CIs. ∗ P < 0.001 compared with reference group. RHF, renal
renal hyperfiltration; 25(OH)D, 25-hydroxyvitamin D. hyperfiltration; 25(OH)D, 25-hydroxyvitamin D.
1348 JHEE ET AL.
the subjects. The 2 groups were well matched for all baseline healthy adults from a large Korean community-based cohort.
characteristics. In the matched cohort, the association between We demonstrated that the adjusted mean eGFR was negatively
vitamin D status and RHF was analyzed with logistic regression associated with serum vitamin D concentrations, and the
analysis. The results were equal to those before PSM, as prevalence of RHF was higher in subjects with 25(OH)D <10
25(OH)D <10 ng/mL still significantly increased the ORs for ng/mL. Finally, the relative risk of RHF was significantly higher
RHF (Supplemental Table 4). in subjects with 25(OH)D <10 ng/mL than in those with ≥30
ng/mL. These findings suggest that 25(OH)D <10 ng/mL is a
robust risk factor for RHF in a relatively healthy population.
DISCUSSION RHF is likely to play a crucial role in the development and
This study observed a significant association between rates progression of CKD (1). In the early 1980s Brenner et al.
of RHF and severe vitamin D deficiency adjusted for age, (33) proposed the “hyperfiltration theory,” in which progressive
sex, height, weight, and history of hypertension or diabetes in deterioration of kidney function was a result of compensatory
VITAMIN D DEFICIENCY AND RENAL HYPERFILTRATION 1349
glomerular hemodynamic changes occurring in response to in the NHANES study (52). Although little is known about
nephron loss. Currently, various conditions have been suggested the exact contributors to the high prevalence of vitamin D
to be associated with RHF, including pregnancy, high-protein deficiency among Korean adults, several plausible explanations
diets, early phases of diabetes, autosomal dominant polycystic have been suggested. Adults of Korean descent are more
kidney disease, and obesity (34–37). The possible mechanisms likely to be dark-skinned than those of Caucasian descent, and
for glomerular hyperfiltration include an increased filtration basal darker-colored skin can reduce the ultraviolet radiation–
fraction, which in turn decreases the effective renal plasma flow mediated synthesis of vitamin D upon the same dose of sun
more than the GFR; vasodilatation of the afferent arteriole; and exposure. Second, vitamin D–fortified foods as well as vitamin D
activation of the renin-angiotensin-aldosterone system (RAAS) supplements are less readily available in Korea than in Western
(38, 39). Glomerular hyperfiltration and increased glomerular countries. Third, cultural habits, such as avoiding direct sun
pressure causing glomerular hypertension lead to progressive exposure and wearing more clothing, result in absorption of
kidney disease as occurs in patients with diabetes, autosomal lower amounts of ultraviolet radiation. All together, it cannot
dominant polycystic kidney disease, and obesity (1). Previous be emphasized enough that early screening and management of
epidemiologic studies have also demonstrated several factors vitamin D deficiency is beneficial to maintaining normal renal
The authors’ contributions were as follows—JHJ and THY: conceived 17. Cantorna MT, Zhu Y, Froicu M, Wittke A. Vitamin D status, 1,25-
of the research idea and designed the study; JHJ: acquired the data; KHN, dihydroxyvitamin D3, and the immune system. Am J Clin Nutr
SYA, MUC, ML, SP, HK, and HRY: analyzed and interpreted the data; 2004;80:1717s–20s.
JHJ and YKK: performed the statistical analysis; JTP, SHH, SWK, and 18. Zittermann A, Schleithoff SS, Tenderich G, Berthold HK, Korfer
R, Stehle P. Low vitamin D status: a contributing factor in the
THY: supervised or mentored; THY: held responsibility to ensure that this
pathogenesis of congestive heart failure? J Am Coll Cardiol 2003;41:
study was reported honestly, accurately, and transparently, that no important 105–12
aspects of the study have been omitted, and that any discrepancies from 19. Pilz S, Iodice S, Zittermann A, Grant WB, Gandini S. Vitamin D status
the study as planned have been explained; and all authors: contributed and mortality risk in CKD: a meta-analysis of prospective studies. Am
important intellectual content during manuscript drafting or revision and J Kidney Dis 2011;58:374–82.
are accountable for the overall work by ensuring that questions pertaining 20. Urena-Torres P, Metzger M, Haymann JP, Karras A, Boffa JJ, Flamant
to the accuracy or integrity of any portion of the work are appropriately M, Vrtovsnik F, Gauci C, Froissart M, Houillier P, et al. Association
investigated and resolved, and read and approved the final manuscript. The of kidney function, vitamin D deficiency, and circulating markers
authors reported no funding received for this study. All authors declare no of mineral and bone disorders in CKD. Am J Kidney Dis 2011;58:
conflicts of interests. 544–53.
21. Uhlig K, Berns JS, Kestenbaum B, Kumar R, Leonard MB, Martin
KJ, Sprague SM, Goldfarb S. KDOQI US commentary on the 2009