Castelblanco et al. Cardiovascular Diabetology (2022) 21:207 Cardiovascular Diabetology https://doi.org/10.1186/s12933-022-01640-3 RESEARCH Open Access Association of α-klotho with subclinical carotid atherosclerosis in subjects with type 1 diabetes mellitus Esmeralda Castelblanco1,2, Marta Hernández3, Nuria Alonso4,5,6, Aina Ribes-Betriu7, Jordi Real2,4, Minerva Granado-Casas4,7,8, Joana Rossell4,7,8, Marina Idalia Rojo-López7,8, Adriana Silvia Dusso1, Josep Julve4,7,8* and Didac Mauricio4,7,8,9* Abstract Background Compelling evidence suggests that the fibroblast growth factor 23 (FGF23) / α-klotho axis is impaired in subjects with diabetes mellitus. We examined the relationship between parameters related to calcium/phosphate homeostasis, including FGF23 and α-klotho, and subclinical carotid atherosclerosis burden in type 1 diabetes mellitus (T1D) subjects. Methods This cross-sectional study involved 226 subjects with T1D and 147 age-, sex- and plaque-matched, non- diabetic (non-T1D) subjects, both with normal renal function. Carotid ultrasound was performed to determine the presence and burden of atheromatous plaques. Concentrations of the intact form of FGF23 and α-klotho were assessed by ELISA. Calcium, phosphate, parathyroid hormone, and vitamin D levels were also determined. Negative binomial regression models were used to examine relationship between parameters studied and subclinical carotid atherosclerosis. Results Only FGF23 was increased in T1D compared with non-diabetic subjects (> 2-fold; p < 0.05). α-klotho was higher in subjects with subclinical carotid atherosclerosis (1.4-fold, p < 0.05). Regression analysis revealed that the log α-klotho concentration was positively associated with the presence of subclinical carotid atherosclerosis both in T1D subjects (incidence rate ratio [IRR]: 1.41; 95% confidence interval [CI], 1.06–1.89; p < 0.05) and in non-T1D subjects (IRR: 1.65; 95% CI, 1.02–2.75; p < 0.05). The models also showed that age, smoking and albuminuria-to-creatinine ratio were positively associated with subclinical carotid atherosclerosis in T1D subjects. Interestingly, sex-related protection against plaque was also revealed in T1D women. Conclusion Higher α-klotho was associated with subclinical carotid atherosclerotic in the absence of kidney dysfunction. This finding also points to a new pathophysiological pathway involved in the development and progression of this complication. Keywords Type 1 diabetes mellitus, Mineral metabolism, Fibroblast growth factor 23, α-klotho *Correspondence: Josep Julve JJulve@santpau.cat Didac Mauricio didacmauricio@gmail.com Full list of author information is available at the end of the article © The Author(s) 2022. Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/ Castelblanco et al. Cardiovascular Diabetology (2022) 21:207 Page 2 of 9 Background without apparent kidney dysfunction [28]. Moreover, Mineral metabolism disturbances and impaired kidney alterations in the concentrations of both FGF23 and function are frequent features in diabetic subjects [1]. α-klotho have been frequently related to disturbances in Adverse cardiovascular outcomes and altered mineral mineral metabolism due to an impaired kidney function metabolism in subjects with impaired insulin signaling [5]. However, analysis of the relationship of both FGF23 have been attributed to kidney dysfunction [2–4]. and α-klotho with atherosclerotic cardiovascular dis- The Fibroblast growth factor 23 (FGF23)/α-klotho axis ease has been lacking in the majority, if not all, reports is heavily impaired in chronic kidney disease (CKD) [5]. [24–27]. Indeed, the relationship, if any, of the main Low concentrations of α-klotho and high concentrations mineral metabolism factors (i.e., calcium, phosphate, of FGF23 generally predict the degree of kidney dysfunc- 25-hydroxyvitamin D (vitamin D), FGF23, a-klotho, tion at early CKD stages, cardiovascular disease, and PTH) with subclinical atherosclerosis burden in T1D death in both subjects with CKD and the general popula- subjects, in the absence of kidney impairment has not tion [5]. Despite the generally accepted inverse relation- been assessed yet. Thus, the intended purpose of this ship between circulating α-klotho and vascular disease sub-study was to uncover novel relationships between in CKD patients [6], it has been refuted by some recent mineral metabolism-related biomarkers (calcium, phos- reports [7, 8]. phate, PTH, vitamin D, FGF23, and α-klotho) and sub- Decreased concentrations of serum α-klotho have also clinical carotid atherosclerosis in T1D subjects who were been associated with increased carotid artery intima- free of renal impairment and previous cardiovascular media thickness (cIMT), a surrogate of subclinical events. carotid atherosclerosis [9], in healthy subjects without impaired kidney function [10]. In this context, α-klotho Methods remained independently associated with metabolic syn- Study design and participants’ selection drome, a condition commonly associated with impaired This study was a cross-sectional, post-hoc analysis of insulin signaling and cardiovascular disease [11], even a previous study designed to assess subclinical carotid after adjusting for age, sex, estimated glomerular filtra- atherosclerosis in subjects with T1D [24]. Selected par- tion rate (eGFR), and overt proteinuria [12]. In addition, ticipants included 226 subjects with T1D, and 147 age-, FGF23 elevations have been recently associated with sex- and plaque-matched, non-diabetic subjects with increased cardiovascular risk in the absence of apparent complete data of calcium, phosphate, PTH, and vitamin clinical kidney disease [13–15]. D levels, and a plasma sample available to assess FGF23 The relationship between altered mineral metabolism and α-klotho concentrations. components, including FGF23/α-klotho, and athero- The inclusion criteria were age > 18 years old, with T1D sclerosis in subjects with type 1 diabetes mellitus (T1D) duration of at least 1 year, with no evidence of previous without previous cardiovascular events and kidney cardiovascular disease, defined as any form of clinical impairment is supported by different lines of evidence. coronary heart disease, stroke, or peripheral vascular dis- First, decreased vitamin D levels in subjects with T1D ease (this included the diagnosis of diabetic foot disease), has been associated with worse cardiovascular outcomes and normal renal function (eGFR > 60 mL/min/1.73 m2; [16] and coronary artery calcification [17]. Second, circu- albuminuria-to-creatinine ratio (ACR) < 30 mg/g). Sub- lating vitamin D and FGF23 levels showed, respectively, jects with established CKD (defined as eGFR < 60 mL/ negative, and positive associations with arterial stiffness min/1.73 m2; albuminuria-to-creatinine ratio (ACR) in T1D subjects [18]. Third, elevations in parathyroid ≥30 mg/g), prior treatment with vitamin D and/or cal- hormone (PTH), which together with FGF23 keep serum cium supplements, and previous clinical signs of cardio- concentrations of phosphate under tight control [19, 20], vascular disease were excluded. For the non-T1D group, have also been related to the development of cardiovas- inclusion and exclusion criteria were the same except for cular disease [21]. Furthermore, PTH concentrations the T1D diagnostic variables, thus, these subjects had have been positively associated with arterial stiffness fasting glucose and HbA1c values below 100 mg/dL and in T1D subjects, even after adjustment for additional 5.7%, respectively. Clinical records from subjects were risk factors, including eGFR [22]. Finally, recent stud- thoroughly reviewed in addition to an anamnestic evalu- ies have shown that lower levels of α-klotho in subjects ation and physical examination. The study was conducted with T1D are associated with increased values of cIMT in accordance with the tenets of the Declaration of Hel- in the absence of apparent signs of renal dysfunction as sinki, and approval was obtained from the ethics com- revealed by normal values of creatinine clearance [23]. mittees of both participating centers. Written informed Our research group has recently found that the calcium consent was provided by all participants. phosphate product was related to subclinical carotid atherosclerosis in subjects with type 2 diabetes mellitus Castelblanco et al. Cardiovascular Diabetology (2022) 21:207 Page 3 of 9 Clinical and laboratory procedures number (percentage), unless otherwise indicated. Quali- Demographic and clinical data were obtained from tative variables were summarized using absolute or rel- each participant. Weight, height, and waist circumfer- ative frequencies. Bivariate analysis consisting of the ence were measured using standard clinical procedures. chi-squared or McNemar tests were used to compare Subjects were classified as having hypertension or dys- qualitative variables and the t-test or Mann–Whitney U lipidemia when they were receiving antihypertensive test were used to compare quantitative variables using or lipid-lowering treatment, respectively. Biochemical compareGroups R package [29]. The association between variables were determined in fasting blood and urine α-klotho (in subjects with T1D and controls stratified) samples. All biochemical analyses were performed using and subclinical carotid atherosclerosis (expressed as standard laboratory methods. Basic blood and urine number of atherosclerotic plaques detected) was assessed biochemistry were determined using commercial kits using fitted negative binomial regression models. Nega- adapted to a Hitachi Modular DDPP analyzer (Roche tive binomial regression is similar to regular multiple Diagnostics, Indianapolis, USA). HbA1c was measured regression, except for the fact that the dependent (Y) using an HPLC Variant II Turbo (Bio-RAD, Hercules, variable is an observed count. Thus, the possible values USA). Intact PTH was measured in an Elecsys E170 ana- of Y are the non-negative integers: 0, 1, 2, 3, and so on lyzer (Roche Diagnostics, Indianapolis, USA) by elec- (in this analysis, the number of atherosclerotic plaques) trochemiluminescence immunoassay. The eGFR was [30]. Negative binomial regression is a generalization of calculated using the Chronic Kidney Disease Epidemi- Poisson’s regression that loosens the restrictive assump- ology Collaboration equation [25]. Plasma concentra- tion that the variance is equal to the mean made by the tions of the intact form of FGF23 (TECOmedical AG, Poisson model; thus, we used this approach to achieve Sissach, Switzerland) and α-klotho (Cusabio Biotech more robust estimations [31]. The models were per- Co. Ltd., Wuhan, China) were measured by an enzyme- formed using the glm.nb function from MASS R Package linked immunosorbent assay. Serum vitamin D con- [32]. Clinically relevant variables, as potential confusion centrations were measured using an Architect i2000SR covariables, further α-klotho, were included into 3 mod- analyzer (Abbott Diagnostics, Lake Forest, USA) by a els, as a confirmatory analysis. The association measures chemiluminescent microparticle immunoassay. Vitamin were estimated and were expressed by incidence rate D deficiency was defined as a concentration below 20 ratios (IRR) and their 95% confidence interval (95% CI). ng/mL, and vitamin D insufficiency as values between The statistical analysis was carried out using R 4.1.0 [33] 20 and < 30 ng/mL, in accordance with the Endocrine (https://www.R-project.org/). A p-value less than 0.05 Society guidelines [26]. Dietary intake of vitamin D and was considered statistically significant. calcium was evaluated using a 101-food-item semiquan- titative food frequency validated questionnaire (available Results at http://bibliodieta.umh.es/files/2011/07/CFA101.pdf ) Demographic, clinical, and basic biochemical data of [27]. subjects with and without T1D are shown in Table 1. Because the participants in this sub-study were matched Carotid ultrasonography imaging for age, sex, and atherosclerotic plaque, these variables Subclinical atherosclerosis was identified by non-invasive did not differ between both groups of subjects. The num- image analysis (carotid ultrasound) as the presence and ber of atherosclerotic plaques did not differ between number of atherosclerotic plaques in the carotid arteries. groups. All carotid ultrasound protocols were performed by the Adiposity surrogates (i.e., BMI and waist circumfer- same researcher who was blinded to the characteristics of ence) did not significantly differ among both groups. the study subjects. Ultrasonography imaging of the com- Systolic blood pressure, but not diastolic blood pressure, mon and internal carotid arteries was performed using a was significantly elevated in the T1D group (127 mm Siemens Sequoia 512 and a 15 MHz linear array probe. Hg) compared with non-diabetic subjects (121 mm Hg); The presence and number of plaques was identified using accordingly, the percentage of subjects with hyperten- 2D-mode and color Doppler imaging, and these tech- sion was significantly increased in the T1D group (24.3%) niques were similarly used in longitudinal and transverse compared with the non-T1D group (6.1%). planes. A detailed description of the ultrasound proce- The percentage of subjects with T1D treated for dys- dures used in this study has been recently described [24, lipidemia was higher (38.1%) than that observed in 28]. non-T1D subjects (10.2%) (Table 1). In line with taking lipid-lowering medications, the serum concentrations of Statistical analysis total triglycerides, total cholesterol and LDL cholesterol Quantitative variables were presented as mean and stan- were significantly reduced in the subjects with T1D. dard deviation, median and interquartile intervals or Castelblanco et al. Cardiovascular Diabetology (2022) 21:207 Page 4 of 9 Table 1 Demographic and biochemical characteristics and Table 2 Mineral metabolism parameters in non-diabetic and atherosclerosis-related parameters of the subjects included in T1D subjects the study Non-T1D T1D p-value N Variables Non-T1D T1D p-value N N = 147 N = 226 Clinical data N = 147 N = 226 Calcium (mg/dL) 9.41 (0.35) 9.28 (0.38) 0.001 373 Age 43.8 (10.9) 44.6 (10.8) 0.514 373 Phosphate (mg/dL) 3.56 (0.60) 3.54 (0.63) 0.733 373 Sex (women) 84 (57.1%) 120 0.509 373 Vitamin D (ng/mL) 21.6 (8.2) 21.0 (11.4) 0.570 373 (53.1%) PTH (pmol/L) 4.84 (1.70) 4.69 (1.67) 0.396 373 Diabetes duration 21.4 (10.7) 226 FGF23 (pg/mL) 34.9 (19.3) 72.8 0.001 373 (years) (167.0) Ethnicity (Caucasian) 145 (98.6%) 224 0.648 373 α-klotho (ng/mL) 0.19 (0.21) 0.20 (0.27) 0.515 373 (99.1%) Data were presented as mean (standard deviation). FGF23, fibroblastic growth Tobacco exposure 0.773 371 factor 23; PTH, parathyroid hormone; Vitamin D, 25-hydroxyvitamin D Former smoker 37 (25.5%) 53 (23.5%) Smoker 41 (28.3%) 60 (26.5%) were significantly elevated (∼2-fold, p < 0.05) in subjects Hypertension 9 (6.1%) 55 (24.3%) < 0.001 373 with T1D. Neither the levels of PTH or α-klotho, nor Dyslipidemia 15 (10.2%) 86 (38.1%) < 0.001 373 vitamin D differed between groups. Systolic blood pressure 121 (14.2) 127 (18.1) < 0.001 366 The sub-analysis by sex revealed that α-klotho was sig- (mm Hg) nificantly increased in women compared with men irre- Diastolic blood pressure 74.4 (9.6) 74.7 (10.0) 0.771 366 spective of whether they were in the T1D or non-T1D (mm Hg) group. In contrast, FGF23 did not differ between sexes BMI (kg/m2) 25.0 (4.0) 25.8 (4.0) 0.071 373 Waist circumference 88.6 (12.0) 89.0 (12.3) 0.796 357 (Supplementary Table 1). (cm) The subjects with subclinical carotid atherosclerosis Diabetic Retinopathy < 0.001 373 were older (age mean: +10.3 years old) and more obese Healthy 147 (100%) 0 (0%) (waist circumference mean: +5.6 cm) than those with- Diabetic Retinopathy 89 (39.4%) out plaque(s) (Table 3). Although the values of eGFR and No Diabetic 137 ACR were worse in subjects with plaques, they fell into Retinopathy (60.6%) the clinically normal range. Mean systolic blood pres- Biochemistry sure was higher (+ 9 mm Hg) in subjects with plaques HbA1c (%) 5.34 (0.31) 7.64 (1.01) < 0.001 372 than in those without. In line with this, a higher propor- Triglycerides (mg/dL) 95.8 (56.8) 75.7 (37.5) < 0.001 373 tion of subjects with plaques were receiving medication Total cholesterol (mg/ 195 (37) 183 (32) < 0.001 373 for hypertension than those without plaque. Finally, total dL) cholesterol was elevated (p = 0.04) in the subjects with LDL cholesterol (mg/dL) 116 (32) 103 (27) < 0.001 371 atherosclerotic plaques, with this being mainly explained HDL cholesterol (mg/ 61.8 (18.9) 64.2 (16.1) 0.200 372 by a concomitant, though marginal (p = 0.076), elevation dL) in LDL cholesterol. More subjects with plaque were med- Creatinine (mg/dL) 0.78 (0.15) 0.77 (0.15) 0.589 373 ACR (mg/g) 6.00 (9.63) 8.24 (15.2) 0.083 373 icated for dyslipidemia than those without plaque (52.9% eGFR (mL/min/1.73 m2) 104 (14) 103 (13) 0.908 373 vs. 19.4%, p < 0.001). Microalbuminuria 8 (5.52%) 12 (5.80%) 1.000 352 Notably, the concentrations of α-klotho were signifi- Atherosclerosis imaging cantly higher (1.4-fold, p = 0.01) in subjects with sub- Atherosclerotic Plaque 0.902 clinical carotid atherosclerosis than in those without One plaque 16 (11%) 28 (12%) (Table 3). No further differences in any other component Multiple plaques 16 (11%) 25 (11%) of mineral metabolism were observed. Data are presented as mean (standard deviation), or n (%), as appropriate. α-klotho was positively associated with subclinical ACR, albuminuria-to-creatinine ratio; BMI, body mass index; eGFR, estimate atherosclerotic disease (IRR: 1.41; 95% CI, 1.06–1.89) glomerular filtration rate based on the CKD-EPI (Chronic Kidney Disease Epidemiology Collaboration) equation; HDL, high-density lipoprotein; LDL, (Table 4), even after additional adjustment for other low-density lipoprotein potential confounders (model 3). α-klotho also showed a positive association with subclinical atherosclerotic The eGFR did not differ between groups (Table 1), with disease (IRR: 1.65; 95% CI, 1.02–2.75; p < 0.05) in a sub- filtrate values over 60. Likewise, the ACR was similar in analysis only considering non-diabetic participants (Sup- both groups. plementary Table 2). Among the different mineral metabolism components Multivariable linear regression analysis also identified analyzed, serum calcium, but not phosphate, was signifi- other independent factors associated with plaque in sub- cantly decreased in subjects with T1D compared with jects with T1D (Table 4). Despite the doubling of FGF23 non-T1D subjects (Table 2). Of note, the levels of FGF23 levels in subjects with T1D (Table 2), this phosphaturic Castelblanco et al. Cardiovascular Diabetology (2022) 21:207 Page 5 of 9 Table 3 Subanalysis of demographic and biochemical with the number of atherosclerotic plaques in non- characteristics, including mineral metabolism-related parameters, T1D subjects (Supplementary Table 2). Notably, ACR in subjects with and without subclinical carotid atherosclerosis remained significantly associated with plaque even in No plaque Plaque p-value N subjects displaying clinically normal kidney function Clinical data N = 288 N = 85 (Table 4, Supplementary Table 2). Age 41.9 (10.3) 52.2 (8.8) < 0.001 373 Another variable positively associated with subclini- Sex (women) 159 (55.2%) 45 (52.9%) 0.806 373 cal carotid atherosclerosis was medication for hyperten- T1D number, (%) 173 (60.1) 53 (62.4) 0.801 373 sion and dyslipidemia though only in non-T1D subjects Ethnicity (Caucasian) 285 (99.0%) 84 (98.8%) 1.000 373 (Table 4, Supplementary Table 2). Tobacco exposure 0.278 371 Former smoker 69 (24.0%) 21 (25.0%) Smoker 73 (25.4%) 28 (33.3%) Discussion Hypertension 28 (9.7%) 36 (42.4%) < 0.001 373 The potential of mineral metabolic disturbances, includ- Dyslipidemia 56 (19.4%) 45 (52.9%) < 0.001 373 ing FGF23, α-klotho, PTH, phosphate, and vitamin D Systolic blood pressure 123 (15.7) 132 (19.1) < 0.001 366 status, as non-traditional biomarkers of atherosclerosis is (mm Hg) still under debate. Since kidney disease positively influ- Diastolic blood pres- 74.4 (9.7) 75.4 (10.2) 0.434 366 ences mineral metabolism and hence cardiovascular risk sure (mm Hg) [34], we specifically analyzed the relationship between BMI (kg/m2) 25.1 (3.9) 26.9 (3.9) < 0.001 373 different mineral metabolism-related biomarkers for Waist circumference 87.6 (11.9) 93.2 (12.0) < 0.001 357 subclinical carotid atherosclerosis in T1D subjects with (cm) physiologically normal values of eGFR and without pre- Diabetic Retinopathy 0.004 373 vious cardiovascular events. The results revealed that Healthy controls 115 (39.9%) 32 (37.6%) α-klotho was associated with subclinical carotid athero- Diabetic Retinopathy 58 (20.1%) 31 (36.5%) sclerosis burden in T1D subjects. Interestingly, α-klotho No Diabetic 115 (39.9%) 22 (25.9) concentrations were also associated with the number of Retinopathy atherosclerotic plaques in subjects without T1D. Our Biochemistry HbA1c (%) 6.67 (1.39) 6.97 (1.35) 0.072 372 data strikingly differed from those from other studies Triglycerides (mg/dL) 81.2 (45.6) 92.0 (51.1) 0.081 373 reporting that lower, rather than higher, concentrations Total cholesterol (mg/ 185 (32) 195 (39) 0.040 373 of serum α-klotho are related to increased atherosclerotic dL) cardiovascular risk [35–38]. LDL cholesterol (mg/ 107 (27) 114 (35) 0.076 371 Serum α-klotho concentrations are positively influ- dL) enced by different intrinsic factors. Consistent with other HDL cholesterol (mg/ 63 (18) 63 (16) 0.912 372 reports [36, 39, 40], our data showed concomitant eleva- dL) tions of α-klotho in females. Although the rationale for Creatinine (mg/dL) 0.77 (0.15) 0.78 (0.15) 0.442 373 such sex-specific differences is poorly defined, they could Albuminuria-to-creati- 6.2 (10.8) 11.1 (19.3) 0.027 373 be related to hormonal differences. However, according nine ratio (mg/g) to our data, sex (women) was not negatively associated eGFR (mL/min/1.73 m2) 106 (13) 96 (12) < 0.001 312 with the presence and number of atherosclerotic plaques Microalbuminuria 13 (4.78%) 7 (8.75%) 0.178 352 in T1D subjects. In line with this, the sex ratio did not Mineral metabolism significantly differ between the group of subjects with FGF23 (pg/mL) 56.7 (124.0) 62.0 (157.0) 0.777 373 plaque compared with those without plaque, and thus α-klotho (ng/mL) 0.18 (0.25) 0.26 (0.24) 0.010 373 Calcium (mg/dL) 9.33 (0.35) 9.34 (0.43) 0.941 373 this parameter did not help to explain the observed posi- Phosphate (mg/dL) 3.55 (0.60) 3.53 (0.68) 0.854 373 tive relationship between plasma α-klotho and the pres- Vitamin D (ng/mL) 21.2 (8.7) 21.3 (14.3) 0.971 373 ence of subclinical carotid atherosclerosis. PTH (pmol/L) 4.73 (1.67) 4.82 (1.72) 0.679 373 We want to underline that FGF23 is increased in sub- Data were presented as mean (standard deviation), or n (%). BMI, body mass jects with obesity [41]. In our study, both BMI and waist index; eGFR, estimate glomerular filtration rate based on the CKD-EPI (Chronic circumference were significantly increased in subjects Kidney Disease Epidemiology Collaboration) equation. FGF23, fibroblastic growth factor 23; HDL, high-density lipoprotein; LDL, low-density lipoprotein; with atherosclerotic plaques (Table 3). However, BMI PTH, parathyroid hormone; Vitamin D, 25-hydroxyvitamin D was not associated with carotid atherosclerosis in sub- jects with T1D (Table 4) or in non-diabetic subjects. hormone was not associated with an increased risk (Supplementary Table 2). Unfortunately, we did not have of plaque (Table 4). The models also showed that age, complete data from all study subjects regarding waist tobacco exposure and ACR were associated with the circumference. Nevertheless, as waist is better mea- number of atherosclerotic plaques in T1D subjects sure of abdominal obesity [42], we decided to explore (Table 4). However, only age and ACR were associated this potential association. Therefore, using the subset of Castelblanco et al. Cardiovascular Diabetology (2022) 21:207 Page 6 of 9 Table 4 Binomial negative regression models to identify factors associated with the subclinical carotid atherosclerosis burden in subjects with T1D. Model 1 Model 2 Model 3 Predictors IRR 95% CI IRR 95% CI IRR 95% CI α-klotho (log scale) 1.40 1.06–1.86 1.41 1.06–1.89 1.41 1.06–1.89 * * * Sex (women) 0.57 0.33–0.97 0.54 0.31–0.95 0.55 0.31–0.97 * * Age 1.08 1.05–1.11 1.08 1.05–1.11 1.08 1.05–1.11 ‡ ‡ ‡ sBP 1.00 0.98–1.01 1.00 0.99–1.01 1.00 0.99–1.01 BMI 1.04 0.98–1.11 1.04 0.98–1.11 1.04 0.98–1.11 Smoker 2.39 1.37–4.22 2.28 1.29–4.07 2.29 1.28–4.13 † † † Former smoker 0.96 0.49–1.82 0.98 0.51–1.85 1.00 0.52–1.88 Hypertension 1.71 0.98–3.02 1.65 0.93–2.93 1.65 0.93–2.92 Dyslipidemia 1.53 0.93–2.55 1.62 0.95–2.80 1.60 0.94–2.77 ACR 1.01 1.00-1.02 1.02 1.01–1.03 1.02 1.00-1.03 ‡ † † Calcium 0.93 0.47–1.88 0.95 0.47–1.94 Phosphate 1.15 0.73–1.83 1.13 0.71–1.81 PTH 0.94 0.80–1.10 0.93 0.79–1.10 FGF23 1.00 1.00–1.00 Vitamin D 1.00 0.98–1.01 Observations 225 225 225 R2 Nagelkerke 0.574 0.578 0.580 Association measure expressed by Incidence rate ratio (IRR) and 95% confidence interval (95% CI). ACR, Albuminuria-to-creatinine ratio; BMI, body mass index; FGF23, fibroblastic growth factor 23; PTH, parathyroid hormone; sBP, systolic blood pressure; Vitamin D, 25-hydroxyvitamin D * p < 0.05; † p < 0.01; ‡ p < 0.001 subjects with available data, we performed an additional marginally associated with subclinical carotid athero- analysis including waist instead of BMI in the model sclerosis burden in T1D subjects (Table 4), reaching inci- (data not shown). This analysis did not reveal any differ- dence rate values close to those calculated for α-klotho. ence from the model that included BMI as a measure of Intriguingly, the calculated associations between lipid- adiposity. However, we strongly believe adiposity, espe- lowering or antihypertensive treatments and plaque were cially abdominal obesity measures, should be included in exacerbated in non-T1D subjects (Model 3, Supplemen- future research on this matter. tary Table 2). To directly check the potential interaction Serum α-klotho concentrations are also positively between the lipid-lowering or antihypertensive medica- influenced by extrinsic factors. Statins and angiotensin tions and α-klotho concentrations and its relationship II blockade are positive regulators of the gene expres- with the number of atherosclerotic plaques, we further sion α-klotho [43, 44]. Whether circulating α-klotho may repeated the binomial negative regression analysis but be indirectly augmented by antihypertensive or lipid- this time after adjusting for the intake of such treat- lowering agents needs to be explored in future studies. ments in T1D subjects (Supplementary Table 3). Our Notably, the relative proportion of subjects with subclini- analysis showed that the positive association of α-klotho cal carotid atherosclerosis taking lipid-lowering agents with subclinical carotid atherosclerosis remained signifi- and angiotensin II receptor blockers was significantly cant in this subanalysis, suggesting the participation of increased compared with those without plaque (anti- other mechanisms still not identified to explain such a dyslipidemic medication: 52.9% vs. 19.4%; antihyper- relationship. tensive medication: 42.4% vs. 9.7%) (Table 3), and hence Some strengths and limitations of our study deserve they were overrepresented in the group of subjects with additional comments. Among the strengths, our study plaque. Thus, our data might be suggesting that α-klotho is the first considering simultaneously evaluation of concentrations could be increased in response to lipid- multiple factors associated with calcium/phosphate lowering or antihypertensive treatments and may indi- homeostasis, including FGF23, α-klotho, PTH, vitamin rectly reflect adherence to medications. However, the D, phosphate, and calcium in subjects with T1D with a binomial negative regression analysis revealed that medi- thorough clinical characterization. Further, some limi- cation for either dyslipidemia or hypertension was only tations should also be acknowledged. First, this study’s Castelblanco et al. Cardiovascular Diabetology (2022) 21:207 Page 7 of 9 Acknowledgements observational and cross-sectional nature does not allow We want to particularly acknowledge the participants, and the IGTP- us to ensure complete control of all the potential (still HUGTP and IRBLleida (B.0000682) Biobanks integrated in the Spanish unknown) confounding factors, especially those related National Biobanks Network of ISCIII (PT17/0015/0045 and PT17/0015/0027, respectively). to the increase of α-klotho in participants with subclini- cal carotid atherosclerosis. Moreover, the study design Author contributions also fails to examine cause-effect relationships or the D.M. and E.C. designed and conceptualized the study. All authors contributed to researched data; E.C., M.H., N.A., A.R-B., J.R., J.J. and D.M. analyzed and temporary association between α-klotho and atheroscle- interpreted data; D.M., J.J., A.S.D. and E.C. contributed to the discussion rotic plaque burden, but allows an estimation of the like- and reviewed the manuscript; J.J. and E.C. wrote the original draft of the lihood of having plaque in those subjects with elevated manuscript; M.H., A.S.D., N.A. and D.M. reviewed/edited the manuscript; D.M. supervised the study; funding acquisition was provided by E.C. and D.M. All α-klotho. Second, the lack of urinary phosphorous levels authors read and approved the manuscript. impeded the assessment of a critical renal response to the FGF23/α-klotho axis responsible for phosphorous reten- Funding This work was funded by Ministerio de Sanidad y Consumo, Instituto de tion and its consequent inflammatory component. High Salud Carlos III (Madrid, Spain) grant PI15/0625 (to DM and EC), FEDER “Una circulating α-klotho without phosphaturia would indicate manera de hacer Europa”, and by CIBER for Diabetes and Associated Metabolic α-klotho dysfunction as occurs with its proatherogenic Diseases (CIBERDEM, CB15/00071, PI. Dr. Didac Mauricio) an initiative from ISCIII, Spain, with co-funding from the European Regional Development impact in these subjects. Third, urinary and circulat- Fund (ERDF). Miguel Servet Type 2 contract (JJ, CPII18/00004; ISCIII). Institut ing concentrations of parameters of mineral metabo- de Recerca de l’Hospital de la Santa Creu i Sant Pau is accredited by the lism are profoundly affected by the menopausal status. Generalitat de Catalunya as Centre de Recerca de Catalunya (CERCA). Unfortunately, the menopausal status was not available Data availability from women of this cohort. Finally, we acknowledge that Proposals relating to the data access should be directed to the corresponding both FGF23 and α-klotho are not currently used in clini- authors. To gain access, data requestors will need to sign a data access agreement. cal practice; however, we believe that further research is needed to gain further insight in the translational and Declarations clinical significance of the relationship between α-klotho and subclinical atherosclerosis in T1D, and the potential Ethics approval and consent to participate application of this protein as a biomarker. The study protocol was approved by the Ethics Committees of University Hospital Arnau de Vilanova (Lleida) (CEIC-1571) and the University Hospital Germans Trias i Pujol (Badalona) (PI-15-147), and all participants were informed Conclusion and gave their consent to participate in the study. Our data suggest that higher α-klotho concentration Consent for publication could be a marker of subclinical carotid atherosclerosis Not applicable. burden, pointing to a new pathophysiological pathway involved in the development and progression of this com- Competing interests The authors declare that they have no competing interests. plication in T1D subjects. Additional population studies on mineral metabolism-related biomarkers are warranted Author details 1 to validate the association and to establish the mecha- Endocrinology, Metabolism and Lipid Research Division, Department of Medicine, Washington University School of Medicine, 63110 St Louis, nism involved between these parameters and subclinical MO, USA carotid atherosclerosis in this population. 2 Unitat de Suport a la Recerca Barcelona, Institut Universitari d’Investigació en Atenció Primària Jordi Gol i Gurina (IDIAP Jordi Gol), Abbreviations 08007 Barcelona, Spain 3 ACR Albuminuria-to-creatinine ratio Department of Endocrinology & Nutrition, Hospital Universitari Arnau de BMI body mass index Vilanova & Institut d’Investigació Biomédica de Lleida (IRB Lleida), cIMT carotid artery intima-media thickness 25198 Lleida, Spain 4 CKD chronic kidney disease CIBER de Diabetes y Enfermedades Metabólicas Asociadas, Instituto de dBP diastolic blood pressure Salud Carlos III, 28029 Madrid, Spain 5 eGFR estimate glomerular filtration rate Department of Endocrinology & Nutrition, Hospital Universitari Germans FGF23 fibroblastic growth factor 23 Trias i Pujol, 08916 Badalona, Spain 6 HDL high density lipoprotein Department of Medicine, Autonomous University of Barcelona, LDL low density lipoprotein 08916 Barcelona, Spain 7 PTH parathyroid hormone Institut d’Investigació Biomèdica Sant Pau (IIB Sant Pau), sBP systolic blood pressure 08041 Barcelona, Spain 8 Vitamin D 25-hydroxyvitamin D Department of Endocrinology & Nutrition, Hospital de la Santa Creu i Sant Pau, 08041 Barcelona, Spain 9 Faculty of Medicine, University of Vic (UVIC/UCC), 08500 Vic, Spain Supplementary Information The online version contains supplementary material available at https://doi. Received: 23 July 2022 / Accepted: 15 September 2022 org/10.1186/s12933-022-01640-3. Supplementary Material 1 Castelblanco et al. Cardiovascular Diabetology (2022) 21:207 Page 8 of 9 References 21. Folsom AR, Alonso A, Misialek JR, Michos ED, Selvin E, Eckfeldt JH, Coresh 1. Yeung SMH, Bakker SJL, Laverman GD, De Borst MH. Fibroblast growth J, Pankow JS, Lutsey PL. Parathyroid hormone concentration and risk of factor 23 and adverse clinical outcomes in Type 2 Diabetes: a bitter-sweet cardiovascular diseases: the Atherosclerosis risk in communities (ARIC) study. symphony. Curr Diab Rep. 2020;20(10):50. Am Heart J. 2014;168(3):296–302. 2. Kuro OM. The Klotho proteins in health and disease. Nat Rev Nephrol. 22. Zobel EH, Theilade S, von Scholten BJ, Persson F, Tarnow L, Lajer M, Hansen 2019;15(1):27–44. TW, Rossing P. Higher parathyroid hormone level is associated with increased 3. Donate-Correa J, Martin-Nunez E, Gonzalez-Luis A, Ferri CM, Luis-Rodriguez arterial stiffness in type 1 Diabetes. Diabetes Care. 2017;40(3):e32–3. D, Tagua VG, Mora-Fernandez C, Navarro-Gonzalez JF. Pathophysiological 23. Keles N, Dogan B, Kalcik M, Caliskan M, Keles NN, Aksu F, Bulut M, Kostek implications of imbalances in fibroblast growth factor 23 in the development O, Isbilen B, Yilmaz Y, et al. Is serum Klotho protective against atheroscle- of Diabetes. J Clin Med 2021, 10(12). rosis in patients with type 1 diabetes mellitus? J Diabetes Complications. 4. Berezin AE, Berezin AA. Impaired function of fibroblast growth factor 23 / 2016;30(1):126–32. Klotho protein axis in prediabetes and diabetes mellitus: Promising predictor 24. Carbonell M, Castelblanco E, Valldeperas X, Betriu A, Traveset A, Granado- of cardiovascular risk. Diabetes Metab Syndr. 2019;13(4):2549–56. Casas M, Hernandez M, Vazquez F, Martin M, Rubinat E, et al. Diabetic 5. Lu X, Hu MC. Klotho/FGF23 Axis in chronic kidney disease and cardiovascular retinopathy is associated with the presence and burden of subclinical carotid disease. Kidney Dis (Basel). 2017;3(1):15–23. atherosclerosis in type 1 diabetes. Cardiovasc Diabetol. 2018;17(1):66. 6. Kitagawa M, Sugiyama H, Morinaga H, Inoue T, Takiue K, Ogawa A, Yamanari T, 25. Levey AS, Stevens LA, Schmid CH, Zhang YL, Castro AF 3rd, Feldman HI, Kikumoto Y, Uchida HA, Kitamura S, et al. A decreased level of serum soluble Kusek JW, Eggers P, Van Lente F, Greene T, et al. A new equation to estimate Klotho is an independent biomarker associated with arterial stiffness in glomerular filtration rate. Ann Intern Med. 2009;150(9):604–12. patients with chronic kidney disease. PLoS ONE. 2013;8(2):e56695. 26. Holick MF, Binkley NC, Bischoff-Ferrari HA, Gordon CM, Hanley DA, Heaney RP, 7. Seiler S, Rogacev KS, Roth HJ, Shafein P, Emrich I, Neuhaus S, Floege J, Murad MH, Weaver CM. Evaluation, treatment, and prevention of vitamin D Fliser D, Heine GH. Associations of FGF-23 and sKlotho with cardiovascular deficiency: an Endocrine Society clinical practice guideline. J Clin Endocrinol outcomes among patients with CKD stages 2–4. Clin J Am Soc Nephrol. Metab. 2011;96(7):1911–30. 2014;9(6):1049–58. 27. Vioque J, Navarrete-Munoz EM, Gimenez-Monzo D, Garcia-de-la-Hera M, 8. Inci A, Sari F, Olmaz R, Coban M, Dolu S, Sarikaya M, Ellidag HY. Soluble Klotho Granado F, Young IS, Ramon R, Ballester F, Murcia M, Rebagliato M, et al. levels in diabetic nephropathy: relationship with arterial stiffness. Eur Rev Reproducibility and validity of a food frequency questionnaire among preg- Med Pharmacol Sci. 2016;20(15):3230–7. nant women in a Mediterranean area. Nutr J. 2013;12:26. 9. Willeit P, Tschiderer L, Allara E, Reuber K, Seekircher L, Gao L, Liao X, Lonn E, 28. Castelblanco E, Betriu A, Hernandez M, Granado-Casas M, Ortega E, Soldevila Gerstein HC, Yusuf S, et al. Carotid intima-media thickness progression as B, Ramirez-Morros A, Franch-Nadal J, Puig-Domingo M, Fernandez E, et surrogate marker for Cardiovascular risk: Meta-analysis of 119 clinical trials al: Ultrasound Tissue Characterization of Carotid Plaques Differs Between involving 100 667 patients. Circulation. 2020;142(7):621–42. Patients with Type 1 Diabetes and Subjects without Diabetes. J Clin Med 10. Keles N, Caliskan M, Dogan B, Keles NN, Kalcik M, Aksu F, Kostek O, Aung SM, 2019, 8(4). Isbilen B, Oguz A. Low serum level of klotho is an early predictor of athero- 29. Subirana I, Sanz H, Vila J. Building bivariate tables: The comparegroups pack- sclerosis. Tohoku J Exp Med. 2015;237(1):17–23. age for R. J Stat Softw. 2014;57(12):1–16. 11. Grundy SM, Cleeman JI, Daniels SR, Donato KA, Eckel RH, Franklin BA, 30. Hilbe JM. Negative Binomial Regression. In: Negative Binomial Regression. Gordon DJ, Krauss RM, Savage PJ, Smith SC. Jr. et al: Diagnosis and man- Cambridge University Press; 2012. agement of the metabolic syndrome: an American Heart association/ 31. Schober P, Vetter TR. Count data in medical research: Poisson regression and National Heart, Lung, and Blood institute scientific statement. Circulation. negative binomial regression. Anesth Analg. 2021;132(5):1378–9. 2005;112(17):2735–52. 32. Venables WN, Ripley BD. Modern applied statistics with S. 4th ed. New York: 12. Kim HJ, Lee J, Chae DW, Lee KB, Sung SA, Yoo TH, Han SH, Ahn C, Oh KH. Springer; 2002. Serum klotho is inversely associated with metabolic syndrome in chronic kid- 33. team Rc. R: A language and environment for statistical computing. In. Vienna; ney disease: results from the KNOW-CKD study. BMC Nephrol. 2019;20(1):119. 2021. 13. Yeung SMH, Binnenmars SH, Gant CM, Navis G, Gansevoort RT, Bakker SJL, 34. Bi X, Yang K, Zhang B, Zhao J. The protective role of Klotho in CKD-associated de Borst MH, Laverman GD. Fibroblast growth factor 23 and mortality in cardiovascular disease. Kidney Dis (Basel). 2020;6(6):395–406. patients with type 2 diabetes and normal or mildly impaired Kidney function. 35. Navarro-Gonzalez JF, Donate-Correa J, Muros de Fuentes M, Perez-Hernandez Diabetes Care. 2019;42(11):2151–3. H, Martinez-Sanz R, Mora-Fernandez C. Reduced Klotho is associated with the 14. Sorensen MH, Bojer AS, Jorgensen NR, Broadbent DA, Plein S, Madsen PL, presence and severity of coronary artery disease. Heart. 2014;100(1):34–40. Gaede P. Fibroblast growth factor-23 is associated with imaging markers of 36. Semba RD, Cappola AR, Sun K, Bandinelli S, Dalal M, Crasto C, Guralnik JM, diabetic cardiomyopathy and anti-diabetic therapeutics. Cardiovasc Diabetol. Ferrucci L. Plasma klotho and cardiovascular disease in adults. J Am Geriatr 2020;19(1):158. Soc. 2011;59(9):1596–601. 15. Tunon J, Fernandez-Fernandez B, Carda R, Pello AM, Cristobal C, Tarin 37. Pan HC, Chou KM, Lee CC, Yang NI, Sun CY. Circulating Klotho levels can N, Acena A, Gonzalez-Casaus ML, Huelmos A, Alonso J, et al. Circulating predict long-term macrovascular outcomes in type 2 diabetic patients. fibroblast growth factor-23 plasma levels predict adverse cardiovascular Atherosclerosis. 2018;276:83–90. outcomes in patients with diabetes mellitus with coronary artery disease. 38. Koga S, Ikeda S, Akashi R, Yonekura T, Kawano H, Maemura K. Serum soluble Diabetes Metab Res Rev. 2016;32(7):685–93. Klotho is inversely related to coronary artery calcification assessed by intra- 16. Joergensen C, Hovind P, Schmedes A, Parving HH, Rossing P. Vitamin D levels, vascular ultrasound in patients with stable coronary artery disease. J Cardiol. microvascular complications, and mortality in type 1 diabetes. Diabetes Care. 2021;77(6):583–9. 2011;34(5):1081–5. 39. Pedersen L, Pedersen SM, Brasen CL, Rasmussen LM. Soluble serum Klotho 17. Young KA, Snell-Bergeon JK, Naik RG, Hokanson JE, Tarullo D, Gottlieb PA, levels in healthy subjects. Comparison of two different immunoassays. Clin Garg SK, Rewers M. Vitamin D deficiency and coronary artery calcification in Biochem. 2013;46(12):1079–83. subjects with type 1 diabetes. Diabetes Care. 2011;34(2):454–8. 40. Gkentzi D, Efthymiadou A, Kritikou D, Chrysis D. Fibroblast growth factor 23 18. Llaurado G, Megia A, Cano A, Gimenez-Palop O, Simon I, Gonzalez-Sastre M, and Klotho serum levels in healthy children. Bone. 2014;66:8–14. Berlanga E, Fernandez-Veledo S, Vendrell J, Gonzalez-Clemente JM. FGF-23/ 41. Hu X, Ma X, Luo Y, Xu Y, Xiong Q, Pan X, Xiao Y, Bao Y, Jia W. Associations of Vitamin D axis in type 1 Diabetes: The potential role of mineral metabolism in serum fibroblast growth factor 23 levels with obesity and visceral fat accu- arterial stiffness. PLoS ONE. 2015;10(10):e0140222. mulation. Clin Nutr. 2018;37(1):223–8. 19. Sun M, Wu X, Yu Y, Wang L, Xie D, Zhang Z, Chen L, Lu A, Zhang G, Li F. 42. Ross R, Neeland IJ, Yamashita S, Shai I, Seidell J, Magni P, Santos RD, Arsenault Disorders of calcium and phosphorus metabolism and the proteomics/ B, Cuevas A, Hu FB, et al. Waist circumference as a vital sign in clinical practice: metabolomics-based research. Front Cell Dev Biol. 2020;8:576110. a consensus statement from the IAS and ICCR working group on Visceral 20. Gutierrez O, Isakova T, Rhee E, Shah A, Holmes J, Collerone G, Juppner H, obesity. Nat Rev Endocrinol. 2020;16(3):177–89. Wolf M. Fibroblast growth factor-23 mitigates hyperphosphatemia but 43. Yoon HE, Lim SW, Piao SG, Song JH, Kim J, Yang CW. Statin upregulates the accentuates calcitriol deficiency in chronic kidney disease. J Am Soc Nephrol. expression of klotho, an anti-aging gene, in experimental cyclosporine 2005;16(7):2205–15. nephropathy. Nephron Exp Nephrol. 2012;120(4):e123–33. Castelblanco et al. Cardiovascular Diabetology (2022) 21:207 Page 9 of 9 44. Yoon HE, Ghee JY, Piao S, Song JH, Han DH, Kim S, Ohashi N, Kobori H, Kuro-o M, Yang CW. Angiotensin II blockade upregulates the expression of Klotho, Publisher’s Note the anti-ageing gene, in an experimental model of chronic cyclosporine Springer Nature remains neutral with regard to jurisdictional claims in nephropathy. Nephrol Dial Transplant. 2011;26(3):800–13. published maps and institutional affiliations.