Circulating Klotho levels can predict long-term macrovascular outcomes in type 2 diabetic patients

Atherosclerosis 276 (2018) 83e90

Contents lists available at ScienceDirect

Atherosclerosis journal homepage: www.elsevier.com/locate/atherosclerosis

Circulating Klotho levels can predict long-term macrovascular outcomes in type 2 diabetic patients Heng-Chih Pan a, Kuei-Mei Chou b, Chin-Chan Lee a, c, Ning-I Yang c, d, Chiao-Yin Sun a, c, * a

Division of Nephrology, Department of Internal Medicine, Keelung Chang Gung Memorial Hospital, Keelung, Taiwan Division of Endocrinology and Metabolism, Department of Internal Medicine, Keelung Chang Gung Memorial Hospital, Keelung, Taiwan c Chang Gung University College of Medicine, Taoyuan, Taiwan d Division of Cardiology, Department of Internal Medicine, Keelung Chang Gung Memorial Hospital, Keelung, Taiwan b

a r t i c l e i n f o

a b s t r a c t

Article history: Received 30 November 2017 Received in revised form 11 June 2018 Accepted 5 July 2018 Available online 17 July 2018

Background and aims: Type 2 diabetes is a global health problem that is associated with a wide variety of vascular complications and associated morbidity and mortality. Klotho is an enzyme and transmembrane protein, and increasing evidence suggests that Klotho may contribute to reduced oxidative stress, improved endothelial function, and vasoprotection. To date, the physiological role of Klotho in vascular complications associated with diabetes is unclear. Methods: We prospectively recruited 252 patients with type 2 diabetes, who visited an outpatient clinic at our hospital between 2009 and 2011. Patients in the top and bottom tertiles of circulating Klotho levels were enrolled for analysis. Results: Of the 168 patients enrolled, 45.8% were male, the mean age was 57.2 years, and the average duration of diabetes was 7.58 years. In multiple regression analysis, a high Klotho level was associated with a reduced risk of developing coronary artery disease and cerebrovascular accidents. Klotho level was also an independent predictor for the development of macroangiopathies within the 7-year study period. Conclusions: Our results suggest that circulating Klotho level is a predictor of long-term macrovascular outcomes in patients with type 2 diabetes. © 2018 The Authors. Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).

Keywords: Klotho Circulating Type 2 diabetes Macroangiopathy Long-term Vascular complication

1. Introduction Patients with type 2 diabetes often have multiple comorbidities, especially those with poorly controlled blood glucose, blood pressure, and serum lipid profile [1,2]. Vascular endothelial cell dysfunction, a complication of diabetes characterized by macroand microangiopathies, affects multiple organs and causes considerable morbidity and mortality in patients with diabetes mellitus [3,4]. Accelerated atherosclerosis of coronary, carotid, and peripheral arteries plays a key role in macroangiopathies and can

Abbreviations: ABI, ankle-brachial index; BMI, body mass index; CAD, coronary artery disease; CI, confidence of interval; CVA, cerebrovascular accident; FGF23, fibroblast growth factor-23; LVH, Left ventricular hypertrophy; OR, odds ratio; PAOD, peripheral artery occlusive disease; RAAS, renin-angiotensin-aldosterone system; SD, standard deviations. * Corresponding author. Division of Nephrology, Department of Internal Medicine, Chang Gung Memorial Hospital, 222 Mai-Chih Road, Keelung, 20401, Taiwan. E-mail address: fi[email protected] (C.-Y. Sun).

lead to coronary artery disease (CAD), cerebrovascular accidents (CVAs), and peripheral artery occlusive disease (PAOD). Retinal, renal, and vasa nervorum microangiopathies play important roles in the pathogenesis of retinopathy, nephropathy, and neuropathy, respectively. Such macro- and microvascular complications can result in a poor prognosis [1]. Early interventions to treat such abnormalities are therefore important, and an early biomarker to allow for such interventions is needed. Klotho is an enzyme and a transmembrane protein encoded by the KL gene in humans. In mice, it has been shown to be involved in the development of a syndrome resembling human aging [5]. The KL gene is mainly expressed in the kidneys and brain [6]. It provides some control over an organism's sensitivity to insulin, and appears to play a role in nitric oxide production and hyperphosphatemia [7,8]. Klotho deficiency in mice has been shown to lead to multiple disorders including atherosclerosis, vascular calcification, stroke, osteoporosis, ectopic calcification, skin atrophy, and chronic vascular disease, in addition to a short life-span and infertility [9,10]. Membrane-bound and secreted Klotho may have distinct

https://doi.org/10.1016/j.atherosclerosis.2018.07.006 0021-9150/© 2018 The Authors. Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).

84

H.-C. Pan et al. / Atherosclerosis 276 (2018) 83e90

functions. Membrane-bound Klotho has been shown to be a necessary coreceptor for fibroblast growth factor-23 (FGF23) and to regulate calcium and phosphate metabolism in the kidneys [11,12]. Secreted Klotho has been shown to be involved in growth factor signaling pathways, including Wnt transforming growth factor b1 and insulin-like growth factor-1, in ion transport, it has been implicated in regulation of the renin-angiotensin system and has antioxidative effects [11,13]. Recent human studies have shown that a higher circulating level of Klotho is associated with lower risks of metabolic syndrome, renal disease, and cardiovascular disease [14e17]. The protective effects of Klotho on renal insults and renal fibrosis in chronic kidney disease have been well described [18]. Many experimental models have shown positive correlations between the expression of Klotho and statin therapy. In addition, previous studies have suggested that the potential role of Klotho in the reduction of oxidized low-density lipoprotein (ox-LDL)-induced oxidative stress may be correlated to the benefits of statin therapy in ameliorating endothelial dysfunction and atherosclerosis [19,20]. However, clinical studies focusing on Klotho and vascular complications in diabetes have reported inconsistent results [14,21e23], and the clinical impact of Klotho on macro- and microangiopathies is still unclear. We hypothesized that circulating Klotho level could be used as an early biomarker of vascular complications in patients with type 2 diabetes. Therefore, in this study we investigated the associations between circulating levels of Klotho and the development/progression of macro- and microangiopathies based on a longitudinal study design. 2. Materials and methods 2.1. Ethics statement This study was conducted in full compliance with the ethical principles of the Declaration of Helsinki, the Good Clinical Practice guidelines, and the applicable local regulatory requirements. Subjects were invited to participate in this study when they attended a health screening program. A trained endocrinologist examined all of the patients during screening and informed them about the consent procedure. Written informed consent was obtained from all subjects prior to their participation. This study was approved by the Institutional Review Board of Chang Gung Memorial Hospital (approval no. 100e1116B). 2.2. Study design and population This study was performed between October 2009 and January 2011 at Chang Gung Memorial Hospital, Keelung, Taiwan. Patients with type 2 diabetes mellitus between the ages of 18 and 70 years who attended an outpatient clinic in the hospital were recruited for the study. The exclusion criteria were patients with a urinary tract infection over the past 3 months, poorly controlled hypertension, pregnancy, liver cirrhosis, polycystic kidney disease, and systemic lupus erythematosus. The patients completed a baseline survey including demographic and clinical laboratory data. The patients were then invited to attend follow-up visits approximately every year thereafter. The primary study outcomes were the development of macro- and microangiopathies. Follow-up examinations were conducted for 7 years, and physical parameters, assessments of macro- and microangiopathies, as well as blood and urine laboratory test results were recorded yearly.

baseline survey and were invited to attend yearly follow-up visits. Of those invited, 252 attended and seven declined. During the 7year study period, 22 patients developed CAD, 29 had CVAs (including 27 with ischemic stroke and two with hemorrhagic stroke), 12 developed left ventricle hypertrophy (LVH), 8 developed or had progression of PAOD, 12 developed or had progression of retinopathy, and 67 developed or had progression of neuropathy. At the end of the study, 218 patients had successfully completed the 7year annual follow-up, 22 had died, and 12 were lost to follow-up. 2.4. Definition of outcomes The patients were defined as having CAD if they had angina pectoris with a positive exercise test result, myocardial infarction, angiographic evidence of (>75%) coronary artery stenosis after the administration of intra-coronary nitroglycerine 50e200 mg, percutaneous coronary revascularization, coronary artery bypass grafting or a positive myocardial perfusion scan, as previously described [24]. CVAs were diagnosed based on brain computed tomography findings or magnetic resonance imaging [25], and the patients with a CVA were defined as those with an ischemic or hemorrhagic stroke episode during the study period. LVH was diagnosed based on the results from echocardiography. Left ventricular (LV) mass was calculated using the formula recommend by the American Society of Echocardiography, and LV mass index was calculated as LV mass divided by body surface area (in m2). LVH was defined as a LV mass index >115 g/m2 for men and >95 g/m2 for women [26]. PAOD was diagnosed according to the ankle-brachial index (ABI), and was defined as the lowest values of right and left ABIs as measured by Doppler ultrasound. The right and left ABIs were determined as the right and left ankle systolic pressures divided by the highest brachial systolic pressure, respectively. PAOD was defined as an ABI
0.9 or >1.3 [27]. Retinopathy was diagnosed based on the findings of a fundoscopic examination [28]. Neuropathy was diagnosed based on the findings of a nerve test, including pinprick, temperature sensation, vibration sensation, and 10-g monofilament tests [29]. Doppler ultrasound, fundoscopic examinations, and nerve tests were routinely performed at least yearly to assess PAOD, retinopathy, and neuropathy. The associated tests to diagnose CAD, CVA and LVH were performed when clinical evidence was noted. Macrovascular complications included CAD, CVAs and PAOD; microvascular complications included neuropathy and retinopathy. The treatment targets for blood pressure, LDL, HbA1C and body mass index (BMI) were defined according to current guidelines [29]. 2.5. Measurement of circulating Klotho level Blood samples were centrifuged for 15 min at 3000 rpm within 30 min of collection. Plasma was removed, and the samples were stored at 70  C until analysis. Serum Klotho level was measured using an enzyme-linked immunosorbent assay (Cusabio Human Klotho ELISA kit, CSB-E13235h, China) at baseline and at the end of the study. The inter- and intra-assay coefficients of variation were <8% and <10%, respectively. The detection limit of this assay was 39 pg/mL [30]. The reference interval for the Klotho assay was 4700e437600 pg/mL in a healthy population [31]. In our laboratory, each Klotho assay was performed in duplicate according to the manufacturer's instructions, and the mean value was used for further statistical analysis. 2.6. Statistical analysis

2.3. Follow-up During the recruitment period, 259 patients completed the

Values were expressed as means and standard deviations (SDs) for continuous data, and categorical data were expressed as

H.-C. Pan et al. / Atherosclerosis 276 (2018) 83e90

percentages. In the primary analysis, patients with high and low levels of Klotho were compared. Continuous variables were tested for normal distribution using the Kolmogorov-Smirnov test. The Student's t-test was used to compare the mean values of continuous variables and normally distributed data; in other cases, the Mann-Whitney U test was used. Categorical data were tested using the chi-square test. We assessed the risk factors for 7-year outcomes using univariate and multivariate analyses. Five multiple logistic regression models with the enter method were used to evaluate the impact of Klotho level on macrovascular complications. Correlations between paired-groups of variables were assessed using Spearman's rank correlation analysis. All statistical analyses were two-tailed, and a p value of <0.05 was considered to be statistically significant. All data were analyzed using the Statistical Package for Social Sciences software, version 20.0 for Windows (SPSS, Inc., Chicago, IL, USA). 3. Results 3.1. Baseline study characteristics A total of 252 patients were included in this study (mean age 56.6 years; 50.4% male). The patients were equally divided into three groups according to their serum Klotho level. The cut-off value for the top tertile (high Klotho group) of the patients was 71050.8 pg/mL, whilst that for the bottom tertile (low Klotho group) was 32589.4 pg/mL. The baseline characteristics of the high and low Klotho groups are shown in Table 1. The high Klotho group was older, had a longer duration of diabetes, and higher percentages of retinopathy and neuropathy, whilst the low Klotho group had more male patients and higher percentages of smokers and dyslipidemia. There were no significant differences in the prevalence of hypertension and metabolic syndrome between the two groups. The average levels of total cholesterol, LDL, high-density lipoprotein (HDL), blood urea nitrogen, and urine albumin-tocreatinine ratio were significantly higher in the high Klotho group. There were no significant differences in the use of medications including angiotensin-converting enzyme inhibitors, angiotensin II receptor blockers or Ezetrol. However, the use of statins was significantly higher in the high Klotho group, whereas the use of fibrates was significantly higher in the low Klotho group. There were no significant differences in the control of blood pressure, LDL, HbA1C, and BMI between the two groups. 3.2. Long-term outcomes of the patients Table 2 presents the complications of diabetes developed by the patients during the 7-year follow-up period. The 7-year incidence rates of new-onset CAD, CVAs, and LVH were significantly higher in the low Klotho group, while there were no significant differences in the development/progression of PAOD, retinopathy, and neuropathy between the low and high Klotho groups. Fig. 1 illustrates the 7year incidence rates and extent of macrovascular complications. In the low Klotho group, 22 patients (26.2%) developed only one endorgan complication and 5 patients (6.0%) had two end-organ complications; while in the high Klotho group, 6 patients (7.1%) had one end-organ complication and 2 patients (2.4%) had two end-organ complications. None of the study patients had all three end-organ complications. The high Klotho group had significantly lower rates of the development of macrovascular complications during the 7-year follow-up period. Sixty-nine of the 84 patients in the low Klotho group (82.1%) and 76 of the 84 patients in the high Klotho group (90.5%) completed the 7-year annual follow-up visits, and their characteristics at the 7th annual follow-up visit are shown in Supplement Table 1. The

85

low Klotho group had higher percentages of dyslipidemia and metabolic syndrome, while the average levels of LDL and HDL were significantly higher in the high Klotho group. The use of medications and the attainment rates of each target were not significantly different between the two groups. However, when all four targets were taken into account, the high Klotho group had a significantly higher attainment rate than the low Klotho group. 3.3. Risk factors for long-term macrovascular complications Univariate analysis showed that age, male gender, smoking, estimated glomerular filtration rate, and Klotho level were good prognostic indicators for the development of macrovascular complications (Table 3A). To adjust for imbalances in baseline characteristics and multiple associations between Klotho and other variables, we used five multivariate models to evaluate the impact of serum Klotho level on macrovascular complications (Table 3B). The results of the multivariate analysis revealed that Klotho levels were significantly associated with macrovascular complications. Model 1 (uncorrected): odds ratio [OR] ¼ 0.471, 95% confidence interval [CI] ¼ 0.307e0.725, p ¼ 0.001; Model 2 (corrected for age and sex): OR ¼ 0.354; 95% CI ¼ 0.210e0.596, p < 0.001; Model 3 (corrected for age, sex, hypertension, smoking, LDL, HDL and statin usage): OR ¼ 0.402; 95% CI ¼ 0.238e0.679, p ¼ 0.001; Model 4 (corrected for age, sex, hypertension, smoking, LDL, HDL, statin usage and estimated glomerular filtration rate): OR ¼ 0.397; 95% CI ¼ 0.227e0.696, p ¼ 0.001; Model 5 (corrected for all available variables): OR ¼ 0.394; 95% CI ¼ 0.224e0.690, p ¼ 0.001). The impact of serum Klotho on new-onset CAD, CVAs, and LVH was also evaluated. 3.4. Correlation between baseline Klotho level and long-term diabetes outcomes Spearman's rank correlation analysis revealed that circulating Klotho levels were significantly negatively correlated with newonset CAD (r ¼ 0.230; p < 0.001), CVAs (r ¼ 0.221; p < 0.001), and LVH (r ¼ 0.185; p < 0.05). However, circulating levels of Klotho were not correlated with the development/progression of PAOD, retinopathy, or neuropathy within the 7-year follow-up period (Table 4). 4. Discussion In this study, we demonstrated that low baseline circulating Klotho levels were associated with an increased risk of long-term outcomes of macrovascular complications in our patients with type 2 diabetes. The results also showed that high levels of circulating Klotho were strongly negatively correlated with the development of CAD, CVAs, and LVH during the 7-year follow-up period. Multiple logistic regression analysis demonstrated that the level of circulating Klotho was an independent predictor of the development of macrovascular complications. Klotho is a single-pass transmembrane protein that is highly expressed in the kidneys and acts as a co-receptor for FGF23 [22]. Circulating Klotho can be generated directly by alternative RNA splicing or proteolytic cleavage [18]. Secreted Klotho acts as a hormonal factor and plays an important role in anti-oxidation and ion transport modulation [22,32]. The beneficial effects of Klotho have been associated with significant increases in the number of bcells, insulin storage levels in pancreatic islets, and glucosestimulated insulin secretion from pancreatic islets, all of which were shown to increase blood insulin levels in a diabetic mice model [33]. Furthermore, the b-cell-specific expression of Klotho has been shown to increase the expressions of Pdx-1 (an insulin

86

H.-C. Pan et al. / Atherosclerosis 276 (2018) 83e90

Table 1 Characteristics of the patients at the beginning of the study.

Demographics Age, years Male gender, n (%) DM duration, months [median] Hypertension Smoking Metabolic syndrome Dyslipidemia CAD, n (%) CVA, n (%) LVH, n (%) PAOD, n (%) Retinopathy, n (%) Neuropathy, n (%) Laboratory results MAP, mmHg BMI, kg/m2 Total cholesterol, mg/dL [median] LDL cholesterol, mg/dL [median] HDL cholesterol, mg/dL [median] Triglyceride, mg/dL [median] BUN, mg/dL Creatinine, mg/dL eGFR, ml/min/1.73m2 UACR, mg/g Fasting glucose, mg/dL HbA1C, % Insulin, mIU/ml [median] C-peptide, ng/ml HOMA-IR [median] HS-CRP, mg/L [median] Medication ACEI, n (%) ARB, n (%) Statin, n (%) Ezetrol, n (%) Fibrate, n (%) Achievement of treatment goal BP < 130/80 mmHg LDL < 100 mg/dL HbA1C < 7.0% 18.5 & BMI < 24 All above four factors

Total (n ¼ 168)

Low Klotho (n ¼ 84)

High Klotho (n ¼ 84)

p value

57.2 ± 10.3 77 (45.8%) 90.9 [67.0] 62 (36.9%) 47 (28.0%) 143 (85.1%) 161 (95.8%) 15 (9.0%) 25 (15.0%) 10 (6.0%) 17 (10.2%) 35 (21.0%) 26 (15.6%)

53.4 ± 10.4 50 (59.5%) 73.3 [55.0] 30 (35.7%) 36 (43.2%) 74 (88.1%) 84 (100.0%) 11 (13.1%) 11 (13.1%) 6 (7.1%) 8 (9.5%) 11 (13.1%) 7 (8.3%)

61.0 ± 8.7 27 (32.1%) 108.6 [86.5] 32 (38.1%) 11 (13.1%) 69 (82.1%) 77 (91.7%) 4 (4.8%) 14 (16.9%) 4 (4.8%) 9 (10.8%) 24 (28.9%) 19 (22.9%)

< 0.001 < 0.001 0.003 NS (0.749) <0.001 NS (0.278) 0.014 NS (0.102) NS (0.523) NS (0.746) NS (0.804) 0.014 0.011

94.8 ± 8.8 27.1 ± 4.4 184.0 [177.0] 104.6 [101.0] 42.5 [40.0] 182.5 [152.5] 16.5 ± 8.5 0.9 ± 0.4 88.9 ± 31.7 287.3 [22.9] 152.6 ± 54.8 7.7 ± 1.6 24.4 [9.1] 3.2 ± 2.0 9.0 [3.2] 2.3 [1.3]

95.8 ± 8.8 27.7 ± 4.5 174.5 [171.0] 98.2 [96.5] 37.3 [35.0] 199.1 [165.0] 15.1 ± 6.8 0.9 ± 0.4 93.3 ± 32.3 122.0 [19.1] 150.4 ± 43.0 7.7 ± 1.6 28.4 [10.4] 3.3 ± 2.1 10.8 [4.0] 2.4 [1.5]

93.8 ± 8.8 26.5 ± 4.1 193.5 [188.0] 111.1 [109.0] 47.7 [45.0] 165.8 [138.0] 18.5 ± 10.4 0.9 ± 0.4 84.5 ± 30.6 452.6 [24.7] 154.7 ± 64.7 7.7 ± 1.7 20.2 [8.1] 3.0 ± 1.9 7.1 [2.7] 2.2 [1.0]

NS (0.133) NS (0.075) 0.004 0.020 < 0.001 NS (0.107) 0.015 NS (0.811) NS (0.072) 0.036 NS (0.612) NS (0.989) NS (0.353) NS (0.597) NS (0.242) NS (0.728)

12 85 94 11 22

(7.1%) (50.6%) (56.0%) (6.5%) (13.1%)

6 (7.1%) 42 (50.0%) 40 (47.6%) 5 (6.0%) 17 (20.2%)

6 (7.1%) 43 (51.2%) 54 (64.3%) 6 (7.1%) 5 (6.0%)

NS (1.000) NS (1.000) 0.043 NS (1.000) 0.011

44 (26.2%) 76 (48.1%) 49 (30.2%) 41 (24.4%) 3 (1.8%)

17 (20.1%) 44 (55.0%) 24 (28.9%) 18 (21.4%) 0 (0.0%)

27 (32.1%) 32 (41.0%) 25 (31.6%) 23 (27.4%) 3 (3.6%)

NS NS NS NS NS

(0.079) (0.079) (0.735) (0.369) (0.081)

DM, diabetes mellitus; CAD, coronary artery disease; CVA, cerebrovascular accident; LVH, left ventricle hypertrophy; PAOD, peripheral artery occlusive disease; MAP, mean artery pressure; BMI, body mass index; LDL, low-density lipoprotein; HDL, high-density lipoprotein; BUN, blood urea nitrogen; eGFR, estimated glomerular filtration rate; UACR, urine albumin-to-creatinine ratio; HOMA-IR, homeostatic model assessment-insulin resistance; HS-CRP, high-sensitivity C-reactive protein; ACEI, angiotensinconverting-enzyme inhibitor; ARB, angiotensin receptor blockers; BP, blood pressure. Values in bold are statistically significant (p value < 0.05). There were significant differences in age, DM duration, smoking, dyslipidemia, male gender, retinopathy, neuropathy, total cholesterol, LDL, HDL, BUN, UACR, use of statins and fibrates. Conversion factors: Cholesterol: 1 mg/dL ¼ 0.0259 mmol/L; Triglyceride: 1 mg/dL ¼ 0.0113 mmol/L; BUN: 1 mg/dL ¼ 0.357 mmol/L; Creatinine: 1 mg/dL ¼ 88.4 mmol/L; eGFR: 1 mL/min//1.73 m2 ¼ 0.0167 mL/s/1.73 m2; UACR: 1 mg/g ¼ 0.113 mg/mmol; Glucose: 1 mg/dL ¼ 0.0555 mmol/L; insulin: 1 mIU/mL ¼ 6.945 pmol/L; C-peptide: 1 ng/ mL ¼ 0.333 nmol/L.

transcription factor), proliferating cell nuclear antigen (a marker of cell proliferation), and microtubule-associated protein 1A/1B-light chain 3 (a marker of autophagy) in the pancreatic islets of db/db mice [33]. These results suggest that the b-cell-specific expression of Klotho can improve b-cell function and attenuate the development of type 2 diabetes. Several studies have reported the protective effects of Klotho in cardiovascular disease [16,34]. Klotho has been shown to modulate vascular tone via interactions between the renin-angiotensinaldosterone system (RAAS), nitric oxide production, and calcium and phosphate homeostasis [35,36]. A low Klotho level has been associated with an increased risk of CAD, LVH, poor blood pressure control, vascular calcification, and reduced life span [34,37e40]. However, previous studies aimed at clarifying the role of Klotho in type 2 diabetes have reported contradicting results [14,21,22]. Most

of these studies were cross-sectional investigations that could only present a causal relationship and lacked definitive inferences on the associated variables. In contrast to a study by van Ark et al. [21], our study results showed that a low circulating Klotho level was an independent risk factor for the development of cardiovascular disease in patients with type 2 diabetes. The central pathogenesis of cardiovascular complications is the process of atherosclerosis, which involves increased free radical formation and reactive oxygen species production, and atherosclerosis is stimulated by prolonged hyperglycemia in diabetes [1]. The associations of risk factors related to cardiovascular complications can be difficult to demonstrate at a specific time point or during a short follow-up period, and a longitudinal study design, as used in this study, may be a better method of assessing the development of vascular complications in patients with diabetes, and to more precisely

H.-C. Pan et al. / Atherosclerosis 276 (2018) 83e90

87

Table 2 Seven-year diabetes outcomes of the patients.

CAD event, n (%) CVA event, n (%) LVH development, n (%) PAOD development/progression, n (%) Retinopathy development/progression, n (%) Neuropathy development/progression, n (%) Macrovascular complications Total, n (%) 2 organs, n (%) 1 organ, n (%) Microvascular complications Total, n (%) 2 organs, n (%) 1 organ, n (%)

Total (n ¼ 168)

Low Klotho (n ¼ 84)

High Klotho (n ¼ 84)

p value

15 (8.9%) 20 (11.9%) 9 (5.4%) 7 (4.2%) 8 (4.8%) 45 (26.8%)

13 (15.5%) 16 (19.0%) 8 (9.5%) 3 (3.6%) 3 (3.6%) 21 (25.0%)

2 (2.4%) 4 (4.8%) 1 (1.2%) 4 (4.8%) 5 (6.0%) 24 (28.6%)

0.003 0.004 0.034 NS (1.000) NS (0.720) NS (0.728)

35 (20.8%) 7 (4.2%) 28 (16.7%)

27 (32.1%) 5 (6.0%) 22 (26.2%)

8 (9.5%) 2 (2.4%) 6 (7.1%)

< 0.001 NS (0.443) 0.001

52 (31.0%) 1 (0.6%) 51 (30.4%)

24 (28.6%) 0 (0.0%) 24 (28.6%)

28 (33.3%) 1 (1.2%) 27 (32.1%)

NS (0.617) NS (1.000) NS (0.737)

Macrovascular complications included the development or progression of CAD, CVAs, and PAOD; microvascular complications included the development or progression of neuropathy and retinopathy. Abbreviations: CAD, coronary artery disease; LVH, left ventricle hypertrophy; CVA, cerebrovascular accident; PAOD, peripheral artery occlusive disease. Values in bold are statistically significant (p value < 0.05). There were significant differences in the development of CAD, CVAs, and LVH. Macrovascular complications: the numbers of total patients and patients with single organ involvement were significantly lower in the high Klotho group. Microvascular complications: there was no significant difference between the low and high Klotho groups.

Fig. 1. The incidence and extent of the macrovascular complications developed within the 7-year study period. Patients with high baseline circulating Klotho levels had significantly lower incidence of macrovascular complications than the patients with low baseline circulating Klotho levels within the 7-year study period. * Macrovascular complications included CAD, CVAs, and PAOD development/progression.

explore the clinical associations between levels of circulating Klotho and vascular complications in these patients. In the current study, the patients with low Klotho levels also had a higher incidence of CVAs during the 7-year follow-up period. The KL gene is mainly expressed in the kidneys and brain [41]. Klothodeficient mice have been shown to exhibit increased oxidative stress and disturbed calcium homeostasis in neurons [42]. Emerging evidence also suggests the possible involvement of Klotho function in the neuronal degeneration process [43e45], which may proceed either rapidly or slowly and then present as a CVA or cognitive impairment [46]. Kim et al. also reported an association between polymorphisms in Klotho with ischemic stroke caused by cardioembolism [47]. These findings could explain the highly significant correlations between low Klotho levels and the increased

risk of CVAs in the present study. To the best of our knowledge, this study is the first to demonstrate a strong association between circulating Klotho level and the long-term incidence of CVA in type 2 diabetes. In addition, the prevalence of microangiopathies was significantly higher in the high Klotho group at the beginning of the study, which may have been because the patients were older and therefore had a longer duration of diabetes. However, the 7-year incidence of the progression or development of microangiopathies was similar between the high and low Klotho groups. A large body of evidence indicates that Klotho may reduce oxidative stress, improve endothelial function, and provide vasoprotection at the cellular level [48,49]. Our study results demonstrated that an increased level of circulating Klotho may exert vasoprotective effects and prevent further pathological changes. As many patients with type 2 diabetes develop vascular complications and suffer from further associated morbidity and mortality, a preventive or therapeutic strategy is important. Our study highlights the clinical impact of circulating Klotho in patients with type 2 diabetes. Furthermore, the long-term longitudinal design also allows for a greater understanding of the more complete dynamic association between circulating Klotho level and vascular complications in these patients. The early recognition of low Klotho levels and timely intervention for these patients may be a potential target for the prevention of vascular complications in patients with type 2 diabetes. Further investigations are needed to elucidate this issue. In spite of the encouraging results observed in this study, several potential limitations should also be considered. First, the study was conducted on patients from a single center, which limits the generalizability of our findings. Second, previous studies have shown that FGF23 and Klotho may have a synergistic effect via a signal transduction pathway to precipitate cardiovascular disease, including CAD, LVH, and vascular endothelial dysfunction [10,35]. In contrast to Klotho, the predictive value of FGF23 in the development of CAD and LVH has been well documented in many clinical scenarios [50]. However, circulating FGF23 level was not measured in this study, and thus we could not ascertain whether the clinical impact of Klotho in type 2 diabetes was dependent on or independent of FGF23. In addition, this study focused on patients with type 2 diabetes, and therefore our results may not be suitable for

88

H.-C. Pan et al. / Atherosclerosis 276 (2018) 83e90

Table 3 Univariate and multivariate analyses of the risk factors for the development of macrovascular complications within 7 years. (A) Univariate analysis to evaluate the impact of the risk factors. (B) Multivariate analysis to evaluate the impact of circulating Klotho level. (A)

Age DM duration Hypertension Male gender Smoking Retinopathy Neuropathy LDL cholesterol HDL cholesterol BUN eGFR UACR Statin Fibrate Klotho (B)

Model Model Model Model Model

1 2 3 4 5

Macrovascular complications

CAD

Odds ratio (95% CI)

P value

Odds ratio (95% CI)

p value

Odds ratio (95% CI)

p value

Odds ratio (95% CI)

p value

1.048 1.001 0.876 3.293 2.314 1.411 1.877 0.993 0.976 1.014 0.981 1.000 0.687 1.137 0.471

0.019 NS (0.734) NS (0.686) 0.003 0.012 NS (0.438) NS (0.186) NS (0.159) NS (0.083) NS (0.473) 0.001 NS (0.338) NS (0.324) NS (0.815) 0.001

1.018 1.004 1.370 3.079 2.842 2.390 1.654 0.996 0.954 1.030 0.979 1.000 1.008 2.308 0.365

NS (0.420) NS (0.169) NS (0.484) 0.013 0.021 NS (0.066) NS (0.393) NS (0.533) 0.040 NS (0.159) 0.007 NS (0.850) NS (0.986) NS (0.128) 0.009

1.085 1.002 1.146 3.532 2.490 0.992 1.150 0.989 0.992 1.019 0.986 1.000 0.649 1.161 0.461

< 0.001 NS (0.531) NS (0.734) 0.005 0.026 NS (0.988) NS (0.809) NS (0.116) NS (0.610) NS (0.354) 0.040 NS (0.972) NS (0.276) NS (0.795) 0.008

0.940 1.001 1.637 1.510 2.694 0.456 1.595 0.988 0.933 1.006 0.997 1.000 0.613 3.684 0.338

NS (0.063) NS (0.756) NS (0.405) NS (0.550) NS (0.097) NS (0.466) NS (0.575) NS (0.231) 0.044 NS (0.865) NS (0.790) NS (0.750) NS (0.478) NS (0.081) 0.043

(1.008e1.090) (0.996e1.006) (0.460e1.666) (1.489e7.283) (1.203e4.452) (0.591e3.372) (0.739e4.768) (0.982e1.003) (0.950e1.003) (0.975e1.055) (0.970e0992) (0.999e1.001) (0.325e1.450) (0.388e3.332) (0.307e0.725)

CVA

(0.974e1.064) (0.998e1.009) (0.568e3.304) (1.324e10.391) (1.169e6.911) (0.944e6.050) (0.521e5.253) (0.982e1.010) (0.912e0.998) (0.988e1.074) (0.964e0.994) (1.000e1.000) (0.419e2.426) (0.786e6.778) (0.170e0.781)

LVH development

(1.037e1.135) (0.997e1.007) (0.522 2.517) (1.450e8.603) (1.115e5.561) (0.382e2.580) (0.371e3.558) (0.976e1.003) (0.962e1.023) (0.979e1.062) (0.973e0.999) (1.000e1.000) (0.298e1.412) (0.375e3.594) (0.260e0.816)

Macrovascular complications

CAD

Odds ratio (95% CI)

p value

Odds ratio (95% CI)

p value

Odds ratio (95% CI)

p value

Odds ratio (95% CI)

p value

0.471 0.354 0.402 0.397 0.394

0.001 < 0.001 0.001 0.001 0.001

0.365 0.368 0.374 0.364 0.260

0.009 0.015 0.028 0.024 0.009

0.461 0.262 0.260 0.248 0.246

0.008 < 0.001 0.005 0.003 0.006

0.338 0.386 0.582 0.594 0.278

0.043 NS (0.090) NS (0.373) NS (0.391) NS (0.056)

(0.307e0.725) (0.210e0.596) (0.238e0.679) (0.227e0.696) (0.224e0.690)

CVA

(0.881e1.003) (0.993e1.010) (0.513e5.230) (0.391e5.834) (0.837e8.666) (0.055e3.772) (0.312e8.145) (0.968e1.008) (0.872e0.998) (0.938e1.079) (0.979e1.016) (1.000e1.001) (0.159e2.370) (0.850e15.966) (0.118e0.968)

(0.170e0.781) (0.165e0.821) (0.156e0.898) (0.151e0.875) (0.095e0.711)

LVH development

(0.260e0.816) (0.125e0.552) (0.102e0.659) (0.098e0.630) (0.090e0.672)

(0.118e0.968) (0.128e1.160) (0.177e1.917) (0.181e1.953) (0.075e1.034)

Macrovascular complications included the development or progression of CAD, CVAs, and PAOD. CAD, coronary artery disease; CVA, cerebrovascular accident; LVH, left ventricle hypertrophy; DM, diabetes mellitus; LDL, low-density lipoprotein; HDL. High-density lipoprotein; BUN, blood urea nitrogen; eGFR, estimated glomerular filtration rate; UACR, urine albumin-to-creatinine ratio; PAOD, peripheral artery occlusive disease. Model 1: Uncorrected; Model 2: Corrected for age and sex; Model 3: Corrected for age, sex, hypertension, smoking, LDL, HDL and statin usage; Model 4: Corrected for age. Sex, hypertension, smoking, LDL, HDL, statin usage and eGFR; Model 5: Corrected for all variables in univariate analysis. Values in bold are statistically significant (p value < 0.05).

Table 4 Correlation between the baseline Klotho level and 7-year diabetes outcomes (Spearman rank correlation coefficients: r).

Klotho CAD CVA LVH development PAOD development/progression Retinopathy development/progression Neuropathy development/progression

Klotho

CAD event

CVA event

LVH PAOD development/ development progression

Retinopathy development/ progression

Neuropathy development/ progression

e 0.230** 0.221** 0.185* 0.030 0.056 0.040

0.230** e 0.207** 0.296** 0.039 0.126 0.046

0.221** 0.207** e 0.006 0.015 0.082 0.068

0.185* 0.296** 0.006 e 0.050 0.053 0.035

0.056 0.126 0.082 0.053 0.233 e 0.072

0.040 0.046 0.068 0.035 0.076 0.072 e

0.030 0.039 0.015 0.050 e 0.233 0.076

CAD, coronary artery disease; CVA, cerebrovascular accident; LVH, left ventricle hypertrophy; PAOD, peripheral artery occlusive disease. *p < 0.05. **p < 0.001.

direct extrapolation to different patient populations. Third, previous studies have suggested that the use of statins and RAAS inhibitors may impact the effects of Klotho [35]. The high prevalence of statin and RAAS inhibitor usage in the study patients may have also affected the results of this study. Nevertheless, 94 patients (56.0%) received statin therapy in our study, and fewer than half of the patients (48.9%) who received statin therapy achieved the treatment goal (LDL < 100 mg/dL). Therefore, the clinical benefits of statin therapy may be underestimated in this study. Fourth, there is lacking of normal levels of circulating Klotho in type 2 diabetic patients in the literature. Our study showed that the patients with lower Klotho levels (below the bottom tertile, 32589.4 pg/mL) had significant higher risk of developing macrovascular complications than those with higher Klotho levels (above the top tertile, 71050.8 pg/mL). However, the cut-off values in this study were determined based on a small sample size, further well-powered

research is needed to more precisely assess cut-off values and achieve high validity. Finally, the predictive accuracy of logistic regression has its own limitations. In summary, this study provides clinical evidence demonstrating an association between low circulating Klotho levels and an increased risk of developing CAD and CVAs in patients with type 2 diabetes. On the basis of these results, circulating Klotho level could be considered to be an early biomarker to predict the incidence of long-term macroangiopathies. Further clinical studies are needed to validate the potential therapeutic role.

Conflicts of interest The authors declared they do not have anything todisclose regarding conflict of interest with respect to this manuscript.

H.-C. Pan et al. / Atherosclerosis 276 (2018) 83e90

Financial support This study was supported by the Chang Gung Memorial Hospital Research Program (grant number, CMRPPG2B0241) and the Ministry of Science and Technology (MOST) of the Republic of China (grant number, MOST 106-2321-B-182-002 and 107-2321-B-182004). Author contributions CCS, HCP, and KMC contributed to the conception, design, and interpretation of data. HCP and KMC contributed to data collection and drafting the manuscript. NIY, CCL, and CCS provided patient information, participated in the design and coordination, and helped draft the manuscript. CCL, CCS, NIY, KMC and the staff of the Community Medicine Research Center provided intellectual content for the work and were involved in editing and revising the manuscript. All authors discussed, contributed to, and approved the final version of the manuscript. Acknowledgments The authors thank the staff of the Community Medicine Research Center of Keelung Chang Gung Memorial Hospital. We also express our sincere gratitude to all staff of the Taiwan Clinical Trial Consortium, TCTC. Appendix A. Supplementary data Supplementary data related to this article can be found at https://doi.org/10.1016/j.atherosclerosis.2018.07.006. References [1] M.J. Fowler, Microvascular and macrovascular complications of diabetes, Clin. Diabetes 26 (2008) 77e82. [2] Standards of Medical Care in Diabetes-2016 Abridged for Primary Care Providers, Clinical diabetes : a publication of the American Diabetes Association 34 (2016) 3e21. [3] C.D. Stehouwer, M.-A. Gall, J.W. Twisk, et al., Increased urinary albumin excretion, endothelial dysfunction, and chronic low-grade inflammation in type 2 diabetes progressive, interrelated, and independently associated with risk of death, Diabetes 51 (2002) 1157e1165. [4] C.G. Schalkwijk, C.D. Stehouwer, Vascular complications in diabetes mellitus: the role of endothelial dysfunction, Clin. Sci. 109 (2005) 143e159. [5] Y. Wang, Z. Sun, Current understanding of klotho, Ageing Research Reviews 8 (2009) 43e51. [6] Y. Saito, T. Nakamura, Y. Ohyama, et al., In vivo klotho gene delivery protects against endothelial dysfunction in multiple risk factor syndrome, Biochem. Biophys. Res. Commun. 276 (2000) 767e772. [7] H. Kurosu, M. Yamamoto, J.D. Clark, et al., Suppression of aging in mice by the hormone Klotho, Science 309 (2005) 1829e1833. [8] D.E. Arking, G. Atzmon, A. Arking, et al., Association between a functional variant of the KLOTHO gene and high-density lipoprotein cholesterol, blood pressure, stroke, and longevity, Circ. Res. 96 (2005) 412e418. [9] J. Bernheim, S. Benchetrit, The potential roles of FGF23 and klotho in the prognosis of renal and cardiovascular diseases, Nephrol. Dial. Transplant. 26 (2011) 2433e2438, https://doi.org/10.1093/ndt/gfr208. Epub 2011 May 4. [10] M. Kuro-o, Klotho, phosphate and FGF-23 in ageing and disturbed mineral metabolism, Nat. Rev. Nephrol. 9 (2013) 650e660. [11] M.C. Izquierdo, M.V. Perez-Gomez, M.D. Sanchez-Nino, et al., Klotho, phosphate and inflammation/ageing in chronic kidney disease, nephrology, dialysis, transplantation, Eur. Ren. Assoc. 27 (2012) iv6-10. ~ o, A.B. Sanz, A. Ortiz, Klotho to treat kidney fibrosis, J. Am. [12] M.D. Sanchez-Nin Soc. Nephrol. 24 (2013) 687e689. [13] M.C. Hu, M. Kuro-o, O.W. Moe, Klotho and chronic kidney disease, in: Phosphate and Vitamin D in Chronic Kidney Disease, Karger Publishers, 2013, pp. 47e63. [14] C. Wu, Q. Wang, C. Lv, et al., The changes of serum sKlotho and NGAL levels and their correlation in type 2 diabetes mellitus patients with different stages of urinary albumin, Diabetes Res. Clin. Pract. 106 (2014) 343e350. [15] V.M. Brandenburg, M.E. Kleber, M.G. Vervloet, et al., Soluble klotho and mortality: the ludwigshafen risk and cardiovascular health study, Atherosclerosis 242 (2015) 483e489.

89

[16] J. Donate-Correa, E. Martin-Nunez, R. Martinez-Sanz, et al., Influence of Klotho gene polymorphisms on vascular gene expression and its relationship to cardiovascular disease, J. Cell Mol. Med. 20 (2016) 128e133. [17] L. Luo, Q. Hao, B. Dong, et al., The Klotho gene G-395A polymorphism and metabolic syndrome in very elderly people, BMC Geriatrics 16 (2016) 46. [18] M.C. Hu, M. Kuro-o, O.W. Moe, Secreted klotho and chronic kidney disease, Adv. Exp. Med. Biol. 728 (2012) 126e157. [19] W. Xia, A. Zhang, Z. Jia, et al., Klotho contributes to pravastatin effect on suppressing IL-6 production in endothelial cells, Mediat. Inflamm. 2016 (2016), 2193210. [20] Y. Yao, Y. Wang, Y. Zhang, et al., Klotho ameliorates oxidized low density lipoprotein (ox-LDL)-induced oxidative stress via regulating LOX-1 and PI3K/ Akt/eNOS pathways, Lipids Health Dis. 16 (2017) 77. [21] J. Van Ark, H.-P. Hammes, M.C. Van Dijk, et al., Circulating alpha-klotho levels are not disturbed in patients with type 2 diabetes with and without macrovascular disease in the absence of nephropathy, Cardiovasc. Diabetol. 12 (2013) 1. [22] E.Y. Lee, S.S. Kim, J.S. Lee, et al., Soluble alpha-klotho as a novel biomarker in the early stage of nephropathy in patients with type 2 diabetes, PLoS One 9 (2014), e102984. [23] N. Keles, B. Dogan, M. Kalcik, et al., Is serum Klotho protective against atherosclerosis in patients with type 1 diabetes mellitus? J. Diabetes Complicat. 30 (2016) 126e132. [24] R.T. George, V.C. Mehra, M.Y. Chen, et al., Myocardial CT perfusion imaging and SPECT for the diagnosis of coronary artery disease: a head-to-head comparison from the CORE320 multicenter diagnostic performance study, Radiology 274 (2015) 626. [25] J.A. Chalela, C.S. Kidwell, L.M. Nentwich, et al., Magnetic resonance imaging and computed tomography in emergency assessment of patients with suspected acute stroke: a prospective comparison, Lancet 369 (2007) 293e298. [26] R.M. Lang, L.P. Badano, V. Mor-Avi, et al., Recommendations for cardiac chamber quantification by echocardiography in adults: an update from the american society of Echocardiography and the european association of cardiovascular imaging, official publication of the American Society of Echocardiography, J. Am. Soc. Echocardiogr. 28 (2015), e14, 1-39. [27] I.T. Lee, C.N. Huang, W.J. Lee, et al., High total-to-HDL cholesterol ratio predicting deterioration of ankle brachial index in Asian type 2 diabetic subjects, Diabetes Res. Clin. Pract. 79 (2008) 419e426. [28] C.P. Wilkinson, F.L. Ferris 3rd, R.E. Klein, et al., Proposed international clinical diabetic retinopathy and diabetic macular edema disease severity scales, Ophthalmology 110 (2003) 1677e1682. [29] P.H. Marathe, H.X. Gao, K.L. Close, American diabetes association standards of medical care in diabetes 2017, J. Diabetes 9 (2017) 320e324. [30] J. Donate-Correa, M. Muros de Fuentes, C. Mora-Fernandez, et al., Pathophysiological implications of fibroblast growth factor-23 and Klotho and their potential role as clinical biomarkers, Clin. Chem. 60 (2014) 933e940. [31] S. Devaraj, B. Syed, A. Chien, et al., Validation of an immunoassay for soluble Klotho protein: decreased levels in diabetes and increased levels in chronic kidney disease, Am. J. Clin. Pathol. 137 (2012) 479e485. [32] M.C. Hu, M. Shi, J. Zhang, et al., Renal production, uptake, and handling of circulating alphaKlotho, J. Am. Soc. Nephrol. : JASN (J. Am. Soc. Nephrol.) 27 (2016) 79e90. [33] Y. Lin, Z. Sun, In vivo pancreatic beta-cell-specific expression of antiaging gene Klotho: a novel approach for preserving beta-cells in type 2 diabetes, Diabetes 64 (2015) 1444e1458. [34] J.F. Navarro-Gonzalez, J. Donate-Correa, M. Muros de Fuentes, et al., Reduced Klotho is associated with the presence and severity of coronary artery disease, Heart 100 (2014) 34e40. [35] M.H. de Borst, M.G. Vervloet, P.M. ter Wee, et al., Cross talk between the reninangiotensin-aldosterone system and vitamin D-FGF-23-klotho in chronic kidney disease, J. Am. Soc. Nephrol. 22 (2011) 1603e1609. [36] Y.L. Wu, J. Xie, S.W. An, et al., Inhibition of TRPC6 channels ameliorates renal fibrosis and contributes to renal protection by soluble klotho, Kidney Int. 91 (2017) 830e841, https://doi.org/10.1016/j.kint.2016.09.039. Epub 2016 Dec 12. [37] R.D. Semba, A.R. Cappola, K. Sun, et al., Plasma klotho and mortality risk in older community-dwelling adults, J. Gerontol. A Biol. Sci. Med. Sci. 66 (2011) 794e800. [38] J. Xie, S.K. Cha, S.W. An, et al., Cardioprotection by Klotho through downregulation of TRPC6 channels in the mouse heart, Nat. Commun. 3 (2012) 1238. [39] M.C. Hu, M. Shi, H.J. Cho, et al., Klotho and phosphate are modulators of pathologic uremic cardiac remodeling, J. Am. Soc. Nephrol. : JASN (J. Am. Soc. Nephrol.) 26 (2015) 1290e1302. [40] J. Chen, X. Zhang, H. Zhang, et al., Indoxyl sulfate enhance the hypermethylation of klotho and promote the process of vascular calcification in chronic kidney disease, Int. J. Biol. Sci. 12 (2016) 1236e1246. [41] D.C. German, I. Khobahy, J. Pastor, et al., Nuclear localization of Klotho in brain: an anti-aging protein, Neurobiol. Aging 33 (2012), 1483.e1425-1430. [42] A. Kosakai, D. Ito, Y. Nihei, et al., Degeneration of mesencephalic dopaminergic neurons in klotho mouse related to vitamin D exposure, Brain Research 1382 (2011) 109e117. [43] I.J. Deary, S.E. Harris, H.C. Fox, et al., KLOTHO genotype and cognitive ability in childhood and old age in the same individuals, Neurosci. Lett. 378 (2005) 22e27.

90

H.-C. Pan et al. / Atherosclerosis 276 (2018) 83e90

[44] V. Majumdar, D. Nagaraja, R. Christopher, Association of the functional KL-VS variant of Klotho gene with early-onset ischemic stroke, Biochem. Biophys. Res. Commun. 403 (2010) 412e416. [45] D.B. Dubal, J.S. Yokoyama, L. Zhu, et al., Life extension factor klotho enhances cognition, Cell Reports 7 (2014) 1065e1076. [46] M.G. Pavlatou, A.T. Remaley, P.W. Gold, Klotho: a humeral mediator in CSF and plasma that influences longevity and susceptibility to multiple complex disorders, including depression, Transl. Psychiatry 6 (2016), e876. [47] Y. Kim, J.H. Kim, Y.J. Nam, et al., Klotho is a genetic risk factor for ischemic

stroke caused by cardioembolism in Korean females, Neurosci. Lett. 407 (2006) 189e194. [48] T.B. Drueke, Z.A. Massy, Circulating Klotho levels: clinical relevance and relationship with tissue Klotho expression, Kidney Int. 83 (2013) 13e15. [49] J.J. Liu, S. Liu, N.G. Morgenthaler, et al., Association of plasma soluble alphaklotho with pro-endothelin-1 in patients with type 2 diabetes, Atherosclerosis 233 (2014) 415e418. [50] I. Agarwal, N. Ide, J.H. Ix, et al., Fibroblast growth factor-23 and cardiac structure and function, J. Am. Heart Assoc. 3 (2014), e000584.