Prevalence and factors linked to renal involvement in prediabetes patients across Europe in the ePREDICE trial Authors: Sebastián Mas-Fontao, Esther Civantos, Nisa Boukichou, Juan A Moreno, Jaakko Tuomilehto, Rafael Gabriel, Jesús Egido Published: 2024 ABSTRACT AbstractThis sub-analysis of the ePREDICE trial, investigated the prevalence and determinants of renal complications, specifically glomerular hyperfiltration, albuminuria, and reduced kidney function, in individuals with prediabetes (PD). The cohort consisted of 967 participants from diverse backgrounds across seven countries. The kidney function was evaluated using the MDRD-4 equation, and the influence of various clinical and demographic factors on renal involvement was assessed by multivariable regression models. Additionally, insulinogenic and disposition indices were examined. Overall, the prevalence of renal abnormalities in this PD cohort was 9.2% (n = 89). Key findings included the detection of hyperfiltration in 20 (2%) individuals, albuminuria in 45 (4.7%), and CKD stage G3a in 29 (3%). Hyperfiltration was inversely correlated with age and height, while albuminuria showed a significant direct association with the disposition index (DI). Age and waist circumference were significantly and directly associated with estimated glomerular filtration rate (eGFR). The ePREDICE study highlights critical factors that affect renal involvement in PD individuals, revealing complex interactions among various parameters. These findings further emphasize the necessity for the search of early kidney abnormalities in people with PD especially in those in older age groups and with a large waist circumference.Keywords: Prediabetes, Chronic kidney disease (CKD), Estimated glomerular filtration rate (eGFR), Albuminuria, BiomarkersSubject terms: Preventive medicine, Kidney diseases, Diabetes complications, Type 2 diabetes Introduction Prediabetes (PD), a precursor to type 2 diabetes (T2D), is characterized by higher-than-normal blood glucose levels that are not yet high enough to be classified as diabetes. It encompasses impaired fasting glucose (IFG) with plasma glucose levels of 6.1 to 6.9 mmol/L and impaired glucose tolerance (IGT), defined by 2-hour post-challenge plasma glucose levels of 7.8 to 11.0 mmol/L in the oral glucose tolerance test (OGTT)1. The progression from PD to T2D is estimated at 4–6% per year for isolated IFG or IGT, and 6–9% per year for combined IFG and IGT2,3. The escalating prevalence of diabetes and its associated complications poses a significant health challenge for contemporary societies and their healthcare systems (https://www.diabetesatlas.org/en/)4. Among the complications that impact morbidity and mortality in individuals with T2D, chronic kidney disease (CKD) is a prominent concern. It accounts for nearly half of all chronic kidney disease (CKD) cases and affects over 4% of the global population5. However, not all individuals with PD develop full-fledged diabetes, nor does every person with T2D progress to CKD. Glucose metabolism disturbances in PD are closely linked to microvascular complications, including kidney impairment6, eye involvement7, and neuropathy8. A meta-analysis of 180,000 individuals found a modest but significant increase in CKD risk among those with PD9, impacting both morbidity and mortality10. A comprehensive study of the prevalence and contributing factors related to renal involvement in individuals with PD is pivotal for early detection, risk stratification, and precisely targeted intervention approaches. In the present study, our primary objective was to investigate the extent of renal dysfunction and delve into the potential factors contributing to the presence of hyperfiltration, albuminuria and the decline of kidney function in a cohort of participants with PD enrolled in the Early Prevention of Diabetes Complications in People with Hyperglycemia (ePREDICE) trial. In pursuit of this objective, we meticulously evaluated a spectrum of anthropometric, physical, metabolic and other biochemical parameters, and lifestyle variables that hold potential implications on renal dysfunction in this population at baseline. Methods Study design, population and clinical endpoints in the ePREDICE trial The ePREDICE trial has been described in detail in a previous article11 (EUDRA-CT 2013-000418-39; ClinicalTrials.gov NCT03222765 registration 19/07/2017). Briefly, ePREDICE is a multicenter, international, independent, randomized, double-blind, placebo-controlled trial. The participants were randomly assigned with equal probability to one of four parallel study arms: placebo, metformin, linagliptin, or a fixed-dose combination of linagliptin/metformin. All participants also received lifestyle intervention involving dietary and physical activity modifications. Detail un participant flow is provided at Gabriel et al.11.The active intervention phase was planned to last for 2 years. EUDRA-CT (ID 2013-000418-39) before the first patient was recruited and after in the ClinicalTrials.gov (ID: NCT03222765). The study recruited men and women aged 45–74 years with the diagnostic of prediabetes from multiple clinical centers in Australia, Austria, Bulgaria, Greece, Kuwait, Poland, Serbia, Spain, and Turkey. The World Health Organization (WHO) criteria for PD were applied, including both IFG and IGT12. The inclusion and exclusion criteria can be found in the study description11. The primary endpoint of the ePREDICE study is defined as the two-year change in the microvascular index, which is a composite measure consisting of three variables: the retinal “Early Treatment of Diabetic Retinopathy Study” (ETDRS) score, the urinary albumin-to-creatinine ratio (UACR), and the sudomotor index for neuropathy measured using the SUDOSCAN® device. Additional endpoints include the incidence of diabetes, quality of life, cognitive function, depressive symptoms, fatty liver index, insulin sensitivity, beta-cell function, and biomarkers of inflammation and endothelial dysfunction in a subsample of participants. Ethical issues The trial protocol was approved by all national Medicine Agencies and local ethic committees of participating centres: Comité Ético de Investigación Clínica del Hospital Universitario La Paz, Madrid, Spain. Approval No. 3850 (01/02/2013); Hellenic Republic Ministry of Health National Ethics Committee, Greece. Approval No. 74/00–01/14 (15/12/2014); The Dean office, Istanbul Medical Faculty Ethical Board for Clinical Trials, Turkey.-pproval No. 2014/495 (10/04/2014); Etickog Odbora. Beograd, Republike Srbije, Serbia. Appoval No. 61/1 (06/04/2013); Ethics Committee of the Faculty of Medicine, University of Belgrade, Serbia. Approval No. 29/III-10(07/03/2013); Komisji Bioetycznej UJ (Uniwersytetu Jagiellonskiego), Poland. Appoval No. BET/185/L/2014(26/06/2014); Executive Officer of Ethics Review Committee (RPAH) & Human Research Ethics Committee (HREC), Australia. Approval No. X13-0046 & HREC/13/RPAH/65 (12/03/2014); комисия по етика “Александровска” ЕАД София, България, Bulgaria. Approval No. KИ-213/18.03.15 (18/03/2015); Die Ethikkommission für das Bundesland Salzburg. “Landeskrankenhaus Salzburg-Universitätsklinikum der PMU, Universitätsklinik für Innere Medizin I”, Austria. Approval No. 415-E/1649/10-2014 (08/07/2014). All study participants gave their written informed consent prior to the participation in the study. The informed consent sheet complied with the applicable regulatory European medical agency requirements, and adheres with the International Good Clinical Practice Guidelines, and the Helsinki Declaration. Baseline renal function assessment Albumin and creatinine assays were carried out in morning spot urine samples at the Fundación Jiménez Díaz Diabetes Laboratory in Madrid, Spain. Serum creatinine quantification was performed at each participating clinical center using standardized measurements to calculate the estimated glomerular filtrate (eGFR) using MDRD-4 (Modification of Diet in Renal Disease equation with 4 variables) formula standardized using the isotope dilution mass spectrometry (IDMS)13. The MDRD equation is based on four variables (age, sex, ethnicity, and serum creatinine)14. According to the eGFR values, people were classified as follows: G1 (> 90 ml/min/1.73 m2); G2 (89 − 60 ml/min/1.73m2) and G3a (59 − 45 ml/min/1.73m2). No participant had eGFR below 45 ml/min/1.73m2. According to the albuminuria/ creatinine ratio (ACR) participants were classified as: A1 (< 30 mg/g); A2 (30–300 mg/g) and A3 (> 300 mg/g), according to the KDIGO (Kidney Disease Improving Global Outcomes) guidelines (see Table 1)13. Table 1Assessment of kidney function in the study population, according with the KDIGO guidelines for CKD evaluation and management. The MDRD-4 equation was employed for the GFR estimation, and the percentages for each category are presented. Predominantly, the prediabetic cohort exhibited hyperfiltration/normal renal function (G1 or G2 + A1), highlighted in green. Individuals with diminished eGFR (G3a) or the presence of albumin in urine (A2 or A3), or both concurrently, are depicted in yellow and orange, respectively. We next explored primary determinants that may contribute to renal involvement (hyperfiltration, albuminuria and hypofiltration) (Table 2). For that reason, we compared people with or without renal abnormalities as a whole by sex. Surprisingly, there were no statistically significant differences in BMI and waist circumference between those with and without renal involvement. No marked differences were noted among various components of the glucose metabolism, except in the disposition index (DI). Significant disparities between individuals with or without renal involvement were observed in height (p = 0.038) and hepatic gamma-glutamyl transferase (GGT) (p = 0.005). A trend for the height discrepancy was discernible among women (158 cm vs. 160 cm, p = 0.058), whereas no such variation was noted among men. Conversely, when we compare individuals with renal impairment versus those without renal impairment GGT showed a significant inverse association in women (16 vs. 22, p < 0.001), while no association was seen in men (28 vs. 29). Assessment of kidney function in the study population, according with the KDIGO guidelines for CKD evaluation and management. The MDRD-4 equation was employed for the GFR estimation, and the percentages for each category are presented. Predominantly, the prediabetic cohort exhibited hyperfiltration/normal renal function (G1 or G2 + A1), highlighted in green. Individuals with diminished eGFR (G3a) or the presence of albumin in urine (A2 or A3), or both concurrently, are depicted in yellow and orange, respectively. We next explored primary determinants that may contribute to renal involvement (hyperfiltration, albuminuria and hypofiltration) (Table 2). For that reason, we compared people with or without renal abnormalities as a whole by sex. Surprisingly, there were no statistically significant differences in BMI and waist circumference between those with and without renal involvement. No marked differences were noted among various components of the glucose metabolism, except in the disposition index (DI). Significant disparities between individuals with or without renal involvement were observed in height (p = 0.038) and hepatic gamma-glutamyl transferase (GGT) (p = 0.005). A trend for the height discrepancy was discernible among women (158 cm vs. 160 cm, p = 0.058), whereas no such variation was noted among men. Conversely, when we compare individuals with renal impairment versus those without renal impairment GGT showed a significant inverse association in women (16 vs. 22, p < 0.001), while no association was seen in men (28 vs. 29). Anthropometric, clinical, biochemical, and metabolic variables at baseline Several anthropometric variables, including weight, height, waist circumference, and hip circumference, were collected as part of the study. Heart rate and blood pressure values were also recorded. The biochemical profile analysed in all participants included plasma glucose, serum insulin, and C-peptide levels measured at five points (0, 30, 60, 90 and 120 min) during the oral glucose tolerance test (OGTT). Other measures included glycated hemoglobin (HbA1c), serum lipid profile (total cholesterol [T-C], high-density lipoprotein [HDL-C], triglycerides [TG], and estimated low-density lipoprotein [LDL-C]), liver function markers (aspartate aminotransferase [AST], alanine aminotransferase [ALT], gamma-glutamyl transferase [GGT], alkaline phosphatase [ALP], and amylase), and kidney function markers (serum creatinine, albumin, and uric acid). Various formulae were applied to assess insulin resistance and beta-cell function based on glucose, insulin, and C-peptide data, including the Homeostatic Model Assessment of Insulin Resistance (HOMA-IR), Quantitative Insulin Sensitivity Check Index (QUICKI), Matsuda Index, Oral Glucose Insulin Sensitivity (OGIS), insulinogenic index, and disposition index. The metabolic syndrome was defined using the National Cholesterol Education Program’s Adult Treatment Panel III report (ATP III) criteria15 and liver involvement was assessed using the fatty liver index (FLI) formula16. Statistical analysis Data were collected and stored in a database developed using the OpenClinica platform (www.openclinica.com). Continuous variables are presented as median and interquartile range (IQR), and comparisons were performed using the Wilcoxon signed-rank test for non-parametric data. Normality of data distribution was assessed using the Shapiro-Wilk or Kolmogorov-Smirnov test, depending on the number of observations. Categorical variables are expressed as percentages and compared using the Pearson chi-squared test. Two-tailed P-values were reported, with P < 0.05 considered statistically significant. Multinomial logistic regression models were used to calculate odds ratios (ORs) and 95% confidence intervals (CIs), using normalized z-scores for quantitative variables. All statistical analyses were conducted using R software (v4.0.2) and R Studio (v1.4), employing libraries such as ggplot2, jmv, tidyverse, clinicopath, gtsummary, and finalfit. Results Prevalence of renal involvement In the examination of different features of kidney dysfunction in people with PD three situations were considered: hyperfiltration, albuminuria, and decline in kidney function (Table 1). The presence of renal involvement was considered when at least one of the subsequent conditions was met: (1) an elevation in the eGFR beyond 130 ml/min/1.73 m2 (98th percentile within our population); (2) an ACR surpassing 30 mg/g; or (3) an eGFR below 60 ml/min/1.73 m2. Utilizing these criteria, in total 89 individuals (9.2%) were identified as having renal involvement. Firstly, we executed a separate analysis focusing on individuals with renal hyperfiltration, considering that this condition commonly represents an early and reversible stage of kidney damage preceding the appearance of albuminuria and decline in renal function17. Twenty (2%) individuals surpassed this threshold and were defined as having renal hyperfiltration. The presence of albuminuria was detected in 45 (4.65%) participants, and renal failure (CKD stage G3a: 45–59 ml/min/1.73 m2) in 29 (3%) individuals, while 878 (95%) had normal filtration values ranging between 60 and 129 ml/min/1.73 m2. Two participants showed anomalous glomerular hyperfiltration (GF) and albuminuria and three hypofiltration and albuminuria concurrently. Table 2Descriptive and comparative statistics in baseline demographic and plasma biochemical parameters in individuals without or with renal involvement defined as the presence of hypofiltration < 60 mL/min/1.73 m2, hyperfiltration > 130 mL/min/1.73 m2, and the presence of albuminuria (30 mg albumin per gram of creatinine)VariablesRenal involvement (men)Renal involvement (woman)No renal involvement (men)No renal Involvement (woman)MenWoman p-valueTotal p-valuep-valuePhysical and antropometric data Sex0.34 Men33 (100%)372 (100%) Women56 (100%)506 (100%) Age (yr)61 (54, 65)58 (50, 64)58 (52, 64)59 (53, 64)0.240.350.96 Body mass index29.5 (27.4, 33.0)31.4 (27.0, 36.1)29.5 (27.1, 32.5)30.5 (27.1, 34.5)0.850.290.33 Height (cm)1.73 (1.69, 1.77)1.58 (1.54, 1.61)1.73 (1.69, 1.78)1.60 (1.55, 1.65)0.640.0580.038* Waist circunference (cm)105 (100, 110)100 (92, 108)104 (98, 113)99 (91, 108)0.880.590.72 Systolic blood pressure (mmHg)142 (124, 151)131 (123, 144)133 (123, 143)131 (120, 142)0.110.560.21 Diastolic blood pressure (mmHg)84 (80, 89)84 (77, 89)82 (76, 89)81 (76, 88)0.640.430.38 Antihypertensives (Y)4 (30%)30 (50%)154 (40%)184 (36%)0.650.04Glycemic data Fasting glucose (mmol)6.40 (6.00, 6.70)6.40 (6.20, 6.62)6.40 (6.20, 6.73)6.30 (6.10, 6.70)0.430.670.82 Matsuda index2.38 (1.71, 3.56)2.51 (1.80, 3.79)2.56 (1.98, 3.51)2.69 (1.96, 3.74)0.720.650.57 HOMA-IR3.23 (2.14, 4.60)3.24 (2.20, 4.60)3.17 (2.19, 4.39)3.05 (2.29, 4.25)0.930.740.75 Insulinogenic index0.49 (0.27, 0.70)0.55 (0.34, 0.86)0.57 (0.37, 0.90)0.63 (0.39, 1.00)0.10.280.075 Disposition index1.29 (0.71, 1.68)1.52 (0.88, 2.09)1.55 (0.98, 2.21)1.61 (1.05, 2.45)0.0580.280.056 HbA1c5.90 (5.52, 6.03)5.80 (5.60, 6.10)0.520.740.19 Prediabetes criteria12 (36%)101 (27%)0.10.310.71 IGT11 (33%)13 (23%)149 (40%)161 (32%) IFG10 (30%)21 (38%)122 (33%)189 (37%) IGT + IFG5.90 (5.60, 6.20)22 (39%)5.80 (5.50, 6.00)156 (31%)Lipid profile Total cholesterol (mmol)5.02 (4.68, 5.74)5.21 (4.50, 6.02)5.12 (4.53, 5.71)5.35 (4.66, 6.12)0.720.50.8 Triglycerides (mmol)1.57 (0.99, 2.15)1.29 (0.96, 1.86)1.32 (0.95, 1.95)1.32 (1.01, 1.76)0.110.990.31 HDL (mmol)1.12 (1.05, 1.29)1.34 (1.19, 1.57)1.19 (1.03, 1.42)1.40 (1.20, 1.63)0.230.30.18 LDL (mmol)3.28 (2.74, 3.80)3.14 (2.54, 3.96)3.24 (2.69, 3.75)3.29 (2.68, 3.90)0.830.540.73Hepatic profile ALT (SGPT) (UI)28 (21, 38)18 (15, 26)26 (21, 34)21 (17, 28)0.310.120.4 AST (SGOT) (UI)26 (17, 31)17 (15, 24)22 (18, 26)19 (16, 24)0.110.350.97 AST/ALT Ratio0.84 (0.64, 1.03)1.00 (0.70, 1.17)0.82 (0.64, 1.00)0.94 (0.72, 1.12)0.70.720.56 GammaGT (UI)28 (21, 46)16 (14, 24)29 (22, 44)22 (16, 34) > 0.99 < 0.001*0.005* Fatty liver index79 (53, 88)68 (30, 89)73 (50, 89)67 (41, 84)0.450.980.75Others Urate (mg/dL)6.80 (5.60, 7.53)5.06 (4.28, 6.21)6.20 (5.40, 7.10)5.00 (4.20, 5.90)0.120.750.58 C-reactive protein (ng/mL)0.27 (0.08, 0.38)0.32 (0.10, 0.72)0.17 (0.05, 0.42)0.26 (0.10, 0.61)0.370.550.25Statistics presented are median (interquartile range, IQR) or n (%). Quantitative variables are compared using Wilcoxon rank sum test while for qualitative date Pearson’s Chi-squared test are employed. OGTT: oral glucose tolerance test; HbA1c: Hemoglobin A1c; HOMA-IR: Homeostatic Model Assessment of Insulin Resistance; HDL: high-density lipoprotein; LDL: low-density lipoprotein; AST (SGPT): Aspartate Aminotransferase, also known as serum glutamic pyruvic transaminase; ALT (SGOT): Alanine Aminotransferase, also known as serum glutamic oxaloacetic transaminase; GGT: Gamma-Glutamyltransferase. Descriptive and comparative statistics in baseline demographic and plasma biochemical parameters in individuals without or with renal involvement defined as the presence of hypofiltration < 60 mL/min/1.73 m2, hyperfiltration > 130 mL/min/1.73 m2, and the presence of albuminuria (30 mg albumin per gram of creatinine) Statistics presented are median (interquartile range, IQR) or n (%). Quantitative variables are compared using Wilcoxon rank sum test while for qualitative date Pearson’s Chi-squared test are employed. OGTT: oral glucose tolerance test; HbA1c: Hemoglobin A1c; HOMA-IR: Homeostatic Model Assessment of Insulin Resistance; HDL: high-density lipoprotein; LDL: low-density lipoprotein; AST (SGPT): Aspartate Aminotransferase, also known as serum glutamic pyruvic transaminase; ALT (SGOT): Alanine Aminotransferase, also known as serum glutamic oxaloacetic transaminase; GGT: Gamma-Glutamyltransferase. When people with renal abnormalities were stratified by the modality of kidney dysfunction (hyperfiltration, albuminuria and hypofiltration) a number of interesting observations appeared. Association of clinical, biochemical and metabolic variables with hyperfiltration Defining the threshold for hyperfiltration presents a challenge, and in this study, we established a threshold of > 130 ml/min/1.73 m2 based on a consensus. This intentionally stringent threshold aimed at having a high specificity, i.e., minimizing false positives at the expense of true positive cases. Employing this criterion, we identified 20 (2%) individuals with hyperfiltration (3 men and 17 women), The hyperfiltration group comprised 85% women, in contrast to 57% in the normal filtration group (p = 0.013). The odds ratio was 4.23 (95% CI: 1.41–18.21, p = 0.022), suggesting a significantly higher proportion of women in the hyperfiltration group compared to the normofiltration group. Furthermore, hyperfiltrating individuals were younger, the mean age was 54 years (SD = 6.6) for the hyperfiltration group and 58.3 years (SD = 7.6) for the normofiltration group, resulting in a Cohen’s d of 0.60, indicating a moderate effect size and a meaningful difference in age between the groups (p = 0.011), the comparison also exhibit a notable difference in height, primarily attributed to excess of women (p < 0.001*)(Table 3). Table 3Descriptive and comparative statistics of subjects with hyperfiltration (> 130 ml/min/1.73m2) using a calibrated serum creatinine MDRD-4 eGFR measurement versus normal (G1 and G2 stages)VariablesHyperfiltration (> 130) (Males)Hyperfiltration (> 130) (Females)Normofiltration (Males)Normofiltration (Females)Males p-valueFemales p-valueTotal p-valuePhysical and antropometric data Sex0.01* Male3 (%)392 (%) Female17 (%)526 (%) Age (yr)52 (48.7, 57.7)64 (58, 66)58 (52, 64)58 (52, 64)0.410.02*0.01* Body mass index (kg/m2)29.4 (26.1, 38.4)32.0 (27.9, 33.8)29.5 (27.3, 32.5)30.6 (27.1, 34.6)0.350.680.89 Height (cm)1.72 (1.71, 1.74)1.55 (1.53, 1.58)1.73 (1.69, 1.78)1.60 (1.55, 1.65)0.68 < 0.01* < 0.01* Waist circunference (cm)100 (86, 106)100 (85.7, 106.3)104 (98, 113)99 (91.3, 108)0.650.720.33 Systolic blood pressure (mmHg)128 (124, 135)128 (124, 132)133 (124, 144)131 (120, 142)0.450.510.12 Diastolic blood pressure (mmHg)83 (71, 87)84 (71, 87)83 (76, 89)82 (76, 88)0.680.750.56 Antihypertensives (Y)1 (33%)7 (41%)154 (39%)194 (37%)0.820.720.87Glycemic data Fasting glucose (mmol)6.3 (4.5, 6.5)6.4 (6.2, 6.6)6.40 (6.20, 6.73)6.30 (6.10, 6.70)0.320.740.84 OGTT glucose 60 min (mmol)11.7 (10.5, 16)11.2 (9.7, 11.8)11.50 (9.90, 12.80)10.8 (9.3, 12.3)0.440.820.75 OGTT glucose 90 min (mmol)10 (9.9, 11.5)9.4 (7.9, 10.5)9.95 (8.30, 11.50)9.4 (7.8, 10.97)0.120.620.9 Fasting insulin (UI)6.1 (6.1, 6.1)7.7 (5.7, 15.5)11.3 (7.8, 15.3)11.1 (8.3, 15.1)0.210.420.05* Matsuda index5 (5, 5.02)2.2 (1.5, 4.3)2.55 (1.93, 3.50)3.1 (2.3, 4.3)0.170.130.07 HOMA-IR1.8 (1.75, 1.8)2.2 (2.3, 4.3)3.22 (2.20, 4.48)3.06 (2.30, 4.27)0.230.070.04* Insulinogenic index0.1 (0.1, 0.12)0.6 (0.4, 0.8)0.56 (0.36, 0.90)0.63 (0.39, 1.00)0.130.690.6 Disposition index0.59 (0.59, 0.6)1.6 (1.4, 3.1)1.50 (0.96, 2.18)1.61 (1.04, 2.38)0.180.50.6 HbA1c5.8 (5.5, 5.9)5.8 (5.4, 5.9)5.8 (5.5, 6.0)5.8 (5.6, 6.1)0.770.370.38 Prediabetes criteria0.320.960.93 IGT1 (30%)5 (29%)109 (28%)165 (31%) IFG0 (0%)7 (41%)156 (40%)199 (38%) IGT + IFG2 (60%)5 (29%)127 (32%)162 (31%)Lipid profile Total cholesterol (mmol)5.3 (4.6, 5.55)5.9 (4.6, 6.6)5.12 (4.53, 5.72)5.34 (4.66, 6.13)0.970.310.26 Triglycerides (mmol)1.6 (1.5, 1.67)1.3 (0.92, 1.64)1.32 (0.95, 1.98)1.33 (1.02, 1.77)0.420.60.91 HDL (mmol)1.06 (0.92, 1.3)1.3 (1.2, 1.5)1.19 (1.03, 1.41)1.38 (1.19, 1.62)0.570.860.63 LDL (mmol)3.5(2.79, 3.8)3.9 (2.7, 4.5)3.25 (2.70, 3.76)3.29 (2.68, 3.91)0.810.210.18Hepatic profile ALT (SGPT) (U/L)18 (16.3, 34.7)17 (14.3, 25)26 (21, 34)21 (17, 28)0.350.140.02* AST (SGOT) (U/L)22 (17.8, 25.5)16.5 (5.9, 22.8)22 (18, 26.6)19 (16, 24)0.850.080.06* AST/ALT Ratio0.82 (0.64, 1.12)1.00 (0.2, 1.16)0.83 (0.63, 1.00)0.94 (0.72, 1.13)0.360.690.75 GammaGT (U/L)23 (14.7, 37.2)18 (15.3, 23.7)29.5 (22, 45)22 (16, 34)0.360.120.02* Fatty Liver Index47.4 (36.3, 85)54.1 (23.1, 92.2)73.8 (50.5, 89.1)66.8 (41, 83.9)0.450.70.39Others Urate (mg/dL)4.5 (4.1, 4.6)4.4 (3.9, 5.2)6.20 (5.40, 7.12)5.02 (4.20, 5.90)0.01*0.05*< 0.01* C-reactive protein (ng/mL)0.3 (0.3, 0.33)0.5 (0.08, 1.2)0.17 (0.06, 0.42)0.26 (0.10, 0.61)0.520.140.05*People with eGFR of less than 60 ml/min/1.73m2 were excluded. Statistics presented as median (IQR) or n (%). Quantitative variables are compared using Wilcoxon rank sum test while for qualitative date Pearson’s Chi-squared test are employed. OGTT: oral glucose tolerance test; HbA1c: Hemoglobin A1c; HOMA-IR: Homeostatic Model Assessment of Insulin Resistance; HDL: high-density lipoprotein; LDL: low-density lipoprotein; AST (SGPT): Aspartate Aminotransferase, also known as serum glutamic pyruvic transaminase; ALT (SGOT): Alanine Aminotransferase, also known as serum glutamic oxaloacetic transaminase; GGT: Gamma-Glutamyltransferase. Descriptive and comparative statistics of subjects with hyperfiltration (> 130 ml/min/1.73m2) using a calibrated serum creatinine MDRD-4 eGFR measurement versus normal (G1 and G2 stages) People with eGFR of less than 60 ml/min/1.73m2 were excluded. Statistics presented as median (IQR) or n (%). Quantitative variables are compared using Wilcoxon rank sum test while for qualitative date Pearson’s Chi-squared test are employed. OGTT: oral glucose tolerance test; HbA1c: Hemoglobin A1c; HOMA-IR: Homeostatic Model Assessment of Insulin Resistance; HDL: high-density lipoprotein; LDL: low-density lipoprotein; AST (SGPT): Aspartate Aminotransferase, also known as serum glutamic pyruvic transaminase; ALT (SGOT): Alanine Aminotransferase, also known as serum glutamic oxaloacetic transaminase; GGT: Gamma-Glutamyltransferase. For glycemic parameters between hyperfiltrating individuals versus those with normal filtration, we observed difference in fasting plasma insulin (7 vs. 11 ; p = 0.047) and HOMA-IR (2.94 vs. 3.14; p = 0.04), though a statistical significance was not reached for the Matsuda Index (3.32 vs. 2.59; p = 0.065). The hyperfiltration group displayed lower GGT values compared with those with normal filtration (18 vs. 26 UI/L; p = 0.016). Also, levels of other hepatic enzymes such as ALT (18 vs. 23 U/L, p = 0.024) and AST (17 vs. 20 U/L, p = 0.064) were lower in the hyperfiltration group. Another distinguishing factor between the two groups was platelet count (254000 vs. 238000 platelets per microliter of blood, p = 0.043). In terms of renal parameters, people with hyperfiltration had lower serum creatinine and uric acid (p < 0.001). Additionally, an association between hyperfiltration and albuminuria was noted (4 vs. 3 mg/g, p = 0.033). HsCRP was numerically higher in individuals with hyperfiltration (0.52 GF vs. 0.21 Non-GF; p = 0.054). Multivariable regression analysis indicated significant associations between hyperfiltration and Matsuda index, hepatic AST, and hsCRP (Fig. 1). Sex-specific variations were also evident. Odds ratios for statistically significant variables of MDRD-4 G1/G2 vs. hyperfiltating individuals comparison. Expresses as OR (95% CI). Quantitative variables are expressed as z-scores. AST: Aspartate Aminotransferase, GGT: Gamma-Glutamyltransferase. Association of clinical, biochemical and metabolic variables with albuminuria There were some sex-specific variations in the study parameters. HDL cholesterol (p = 0.05) had a significant association with albuminuria, particularly in women. In women, GGT had a significant difference between people with albuminuria vs. non-albuminuria (p = 0.017) (Table 4). Table 4Descriptive and comparative statistics between people with albuminuria versus those without.VariablesAlbuminuria maleAlbuminuria femaleNo Alb maleNo Alb femaleMale p-valueFemale p-valueTotal p-valuePhysical and antropometric data Sex0.51 Male21 (100%)384 (100%) Female24 (100%)538 (100%) Age (yr)61 (52, 65)55 (47, 64)58 (52, 64)59 (52, 64)0.440.210.69 Body mass index (kg/m2)29.8 (27.9, 33.0)30.3 (27.0, 35.9)29.4 (27.0, 32.5)30.7 (27.1, 34.5)0.410.750.57 Waist circunference (cm)108 (100, 110)100 (92, 109)104 (98, 113)99 (92, 108)0.710.770.46 Systolic blood pressure (mmHg)146 (127, 151)133 (122, 144)133 (123, 143)131 (120, 142)0.0280.520.055 Diastolic blood pressure (mmHg)84 (80, 92)82 (76, 90)82 (76, 89)82 (76, 88)0.380.860.43 Antihypertensives (Y)6 (30%)10 (41.6%)154 (40%)204 (38%)0.350.730.13Glycemic data Fasting glucose (mmol)6.30 (6.00, 6.80)6.40 (6.20, 6.60)6.40 (6.20, 6.70)6.30 (6.10, 6.70)0.350.920.58 OGTT Glucose 60 min (mmol)12.30 (10.90, 13.70)11.20 (9.38, 12.12)11.50 (9.97, 12.80)10.85 (9.30, 12.30)0.130.680.16 OGTT Glucose 90 min (mmol)11.40 (7.90, 12.40)9.55 (7.65, 10.75)9.90 (8.40, 11.50)9.40 (7.90, 11.00)0.10.550.48 Fasting Insulin (UI)13 (9, 18)13 (9, 19)11 (8, 15)11 (8, 15)0.110.30.062 Matsuda Index1.98 (1.52, 2.86)2.54 (1.77, 3.65)2.59 (1.99, 3.52)2.65 (1.93, 3.78)0.022*0.770.074 HOMA-IR3.83 (2.56, 5.21)3.71 (2.57, 5.47)3.16 (2.18, 4.35)3.05 (2.28, 4.25)0.130.260.062 Insulinogenic index0.52 (0.28, 0.71)0.58 (0.27, 0.87)0.56 (0.36, 0.90)0.63 (0.39, 1.00)0.320.250.11 Disposition_Index1.18 (0.80, 1.46)1.42 (0.99, 1.91)1.55 (0.98, 2.20)1.61 (1.04, 2.46)0.031*0.230.018* HbA1c5.90 (5.80, 6.20)5.90 (5.42, 6.03)5.80 (5.50, 6.00)5.80 (5.60, 6.10)0.034*0.830.21 Prediabetes criteria0.560.250.84 IGT8 (38%)4 (17%)105 (27%)170 (32%) IFG7 (33%)12 (50%)153 (40%)198 (37%) IGT + IFG6 (29%)8 (33%)126 (33%)170 (32%)Lipid profile Total cholesterol (mmol)5.02 (4.73, 5.92)5.24 (4.62, 6.24)5.12 (4.53, 5.69)5.33 (4.63, 6.08)0.380.880.69 Triglycerides (mmol)1.49 (0.96, 2.17)1.48 (1.12, 1.95)1.32 (0.96, 1.95)1.32 (0.98, 1.75)0.390.10.074 HDL (mmol)1.24 (1.09, 1.34)1.28 (1.00, 1.48)1.17 (1.03, 1.40)1.40 (1.20, 1.63)0.760.016*0.05* LDL (mmol)3.28 (2.79, 3.80)3.28 (2.56, 4.19)3.24 (2.69, 3.75)3.29 (2.66, 3.89)0.710.950.81Hepatic profile ALT (SGPT) (U/L)32 (28, 45)18 (15, 26)26 (21, 33)21 (17, 28)0.018*0.380.33 AST (SGOT) (U/L)28 (18, 31)17 (16, 25)22 (18, 26)19 (16, 24)0.058 > 0.990.23 AST/ALT Ratio0.83 (0.58, 0.98)1.04 (0.83, 1.29)0.83 (0.64, 1.00)0.94 (0.72, 1.12)0.520.160.64 Gamma-Glutaryltrnasferase (U/L)33 (23, 84)16 (14, 24)29 (22, 43)22 (16, 34)0.20.017*0.46 Fatty liver index79 (72, 88)70 (40, 89)73 (49, 89)67 (40, 84)0.150.830.23Renal profile eGFR (MDRD-4 SCr Calib)86 (75, 95)87 (83, 101)88 (76, 99)84 (73, 94)0.740.160.34 Urate (mg/dL)7.20 (6.20, 7.60)5.06 (4.18, 5.55)6.20 (5.37, 7.10)5.00 (4.20, 5.90)0.016*0.980.12 Creatinine (mg/dL)80 (73, 88)62 (54, 65)80 (71, 88)63 (58, 71)0.740.180.91Others C-reactive protein (ng/mL)0.29 (0.10, 0.38)0.42 (0.16, 0.73)0.17 (0.05, 0.42)0.26 (0.10, 0.61)0.180.120.061Statistics presented as median (IQR) or n (%). Quantitative variables are compared using Wilcoxon U test while for qualitative data Pearson test are employed. OGTT: oral glucose tolerance test; HbA1c: Hemoglobin A1c; HOMA-IR: Homeostatic Model Assessment of Insulin Resistance; HDL: high-density lipoprotein; LDL: low-density lipoprotein; AST (SGPT): Aspartate Aminotransferase, also known as serum glutamic pyruvic transaminase; ALT (SGOT): Alanine Aminotransferase, also known as serum glutamic oxaloacetic transaminase; GGT: Gamma-Glutamyltransferase. Descriptive and comparative statistics between people with albuminuria versus those without. Statistics presented as median (IQR) or n (%). Quantitative variables are compared using Wilcoxon U test while for qualitative data Pearson test are employed. OGTT: oral glucose tolerance test; HbA1c: Hemoglobin A1c; HOMA-IR: Homeostatic Model Assessment of Insulin Resistance; HDL: high-density lipoprotein; LDL: low-density lipoprotein; AST (SGPT): Aspartate Aminotransferase, also known as serum glutamic pyruvic transaminase; ALT (SGOT): Alanine Aminotransferase, also known as serum glutamic oxaloacetic transaminase; GGT: Gamma-Glutamyltransferase. The DI showed a significant association with albuminuria; individuals with albuminuria displayed a DI of 1.3, whereas in those without albuminuria it was 1.58 (p = 0.018). Of interest, glycemic abnormalities, as well as other related indices such as the Matsuda Index (p = 0.022) and HbA1c (p = 0.034), were exclusively observed in males. While the DI stood as the lone statistically significant variable, a comprehensive multivariable analysis was conducted on other variables that approached significance. The aim was to explore whether these variables could influence the disposition index (DI), but it was not the case (Fig. 2). Odds ratios (95% CI) for major variables between people with albuminuria versus those without. Quantitative variables are expressed as z-scores. HDL: High-Density Lipoprotein. Association of clinical, biochemical and metabolic variables with hypofiltration (G3a) To elucidate disparities between individuals with normal eGFR values (60–129 ml/min/1.73 m2) and those with eGFR < 60 ml/min/1.73 m2, categorized as stage G3a (hypofiltration) according to KDIGO guidelines, exclusions were applied for individuals with filtration rates exceeding 130 ml/min/1.73 m2 to facilitate a proper analysis. Among the factors (Table 5) influencing a reduced renal function within our cohort, age emerged as a significant contributor. Those in stage G3awere found to be on average 6 years older than people in stages G1/G2 (58 vs. 64 years, p = 0.008). However, this age discrepancy was predominantly influenced by variance among women. Other variables distinguishing the two cohorts included HDL-cholesterol (1.06 vs. 10.09 mmol, p = 0.014), LDL-cholesterol (2.9 vs. 3.3 mmol/L, p = 0.046), and GGT (21 vs. 26 U/L, p = 0.014), all of which were lower in the female G3a group. Table 5Descriptive and comparative statistics from G3a vs. G1 and G2 stages using a calibrated serum creatinine MDRD-4 eGFR measurement.VariablesHypofiltration (G3a) (Males)Hypofiltration (G3a) (Females)Normofiltration (Males)Normofiltration (Females)Males p-valueFemales p-valueTotal p-valuePhysical and antropometric data0.38 Sex Male10 (34%)392 (43%) Female19 (66%)526 (57%) Age (yr)62 (57, 66)64 (58, 66)58 (52, 64)58 (52, 64)0.160.021*0.008* Body mass index (kg/m2)29.5 (26.2, 34.7)32.0 (27.9, 33.8)29.5 (27.3, 32.5)30.5 (27.1, 34.6)0.930.540.46 Height (cm)1.75 (1.70, 1.79)1.60 (1.56, 1.66)1.73 (1.69, 1.78)1.60 (1.55, 1.65)0.390.480.76 Waist circunference (cm)105 (101, 111)100 (94, 109)104 (98, 113)99 (92, 108)0.630.320.47 Systolic blood pressure (mmHg)140 (122, 150)134 (121, 148)133 (124, 144)131 (120, 142)0.630.510.46 Diastolic blood pressure (mmHg)85 (77, 87)85 (76, 90)83 (76, 89)82 (76, 88)0.80.680.65 Antihypertensives (Y)6 (30%)10 (41.6%)154 (40%)204 (38%)0.350.730.13Glycemic data Fasting glucose (mmol)6.60 (6.43, 6.77)6.50 (6.25, 6.70)6.40 (6.20, 6.73)6.30 (6.10, 6.70)0.350.340.21 OGTT glucose 60 min (mmol)11.95 (11.12, 12.67)11.90 (10.80, 13.70)11.50 (9.90, 12.80)10.80 (9.30, 12.30)0.380.029*0.028* OGTT glucose 90 min (mmol)9.95 (8.77, 11.78)11.40 (9.50, 12.45)9.95 (8.30, 11.50)9.40 (7.80, 10.97)0.770.007*0.021* Fasting insulin (UI)8 (7, 11)12 (9, 15)11 (8, 15)11 (8, 15)0.0820.420.69 Matsuda index3.35 (2.61, 5.48)2.30 (1.78, 2.85)2.55 (1.93, 3.50)2.68 (1.94, 3.68)0.049*0.170.97 HOMA-IR2.32 (2.07, 3.18)3.47 (2.70, 4.06)3.22 (2.20, 4.48)3.06 (2.30, 4.27)0.0720.360.71 Insulinogenic index0.51 (0.31, 0.58)0.53 (0.33, 0.92)0.56 (0.36, 0.90)0.63 (0.39, 1.00)0.380.420.27 Disposition index1.69 (0.92, 2.07)1.06 (0.80, 1.96)1.50 (0.96, 2.18)1.61 (1.04, 2.38)0.880.0770.21 HbA1c5.80 (5.60, 5.97)6.00 (5.72, 6.10)5.80 (5.50, 6.00)5.80 (5.60, 6.10)0.960.0880.16 Prediabetes criteria > 0.990.044*0.16 IGT3 (30%)4 (21%)109 (28%)165 (31%) IFG4 (40%)4 (21%)156 (40%)199 (38%) IGT + IFG3 (30%)11 (58%)127 (32%)162 (31%)Lipid profile Total cholesterol (mmol)5.04 (4.40, 5.55)5.05 (4.14, 5.60)5.12 (4.53, 5.72)5.34 (4.66, 6.13)0.740.0530.09 Triglycerides (mmol)1.69 (1.25, 2.17)1.11 (0.92, 1.64)1.32 (0.95, 1.98)1.33 (1.02, 1.77)0.110.290.88 HDL (mmol)1.06 (0.92, 1.09)1.41 (1.25, 1.71)1.19 (1.03, 1.41)1.38 (1.19, 1.62)0.014*0.270.78 LDL (mmol)3.16 (2.79, 3.41)2.86 (2.17, 3.43)3.25 (2.70, 3.76)3.29 (2.68, 3.91)0.790.019*0.046*Hepatic profile ALT (SGPT) (U/L)25 (21, 28)20 (16, 27)26 (21, 34)21 (17, 28)0.380.710.39 AST (SGOT) (U/L)21 (14, 30)22 (16, 24)22 (18, 26)19 (16, 24)0.820.910.76 AST/ALT Ratio0.88 (0.67, 1.12)1.00 (0.68, 1.16)0.83 (0.63, 1.00)0.94 (0.72, 1.13)0.610.670.5 GammaGT (U/L)27 (22, 30)16 (10, 24)30 (22, 45)22 (16, 34)0.150.038*0.014* Fatty Liver Index67 (50, 90)71 (33, 85)74 (51, 89)67 (41, 84)0.890.850.83Others Urate (mg/dL)7.00 (6.03, 8.05)5.90 (4.72, 6.75)6.20 (5.40, 7.12)5.02 (4.20, 5.90)0.220.045*0.082 C-reactive protein (ng/mL)0.16 (0.03, 0.36)0.19 (0.08, 0.36)0.17 (0.06, 0.42)0.26 (0.10, 0.61)0.730.230.32Hyperfiltrating subjects (> 130 ml/min/1.73m2) are excluded statistics presented as median (IQR) or n (%). Quantitative variables are compared using Wilcoxon rank sum test while for qualitative date Pearson’s Chi-squared test are employed. OGTT: oral glucose tolerance test; HbA1c: Hemoglobin A1c; HOMA-IR: Homeostatic Model Assessment of Insulin Resistance; HDL: high-density lipoprotein; LDL: low-density lipoprotein; AST (SGPT): Aspartate Aminotransferase, also known as serum glutamic pyruvic transaminase; ALT (SGOT): Alanine Aminotransferase, also known as serum glutamic oxaloacetic transaminase; GGT: Gamma-Glutamyltransferase. Descriptive and comparative statistics from G3a vs. G1 and G2 stages using a calibrated serum creatinine MDRD-4 eGFR measurement. Hyperfiltrating subjects (> 130 ml/min/1.73m2) are excluded statistics presented as median (IQR) or n (%). Quantitative variables are compared using Wilcoxon rank sum test while for qualitative date Pearson’s Chi-squared test are employed. OGTT: oral glucose tolerance test; HbA1c: Hemoglobin A1c; HOMA-IR: Homeostatic Model Assessment of Insulin Resistance; HDL: high-density lipoprotein; LDL: low-density lipoprotein; AST (SGPT): Aspartate Aminotransferase, also known as serum glutamic pyruvic transaminase; ALT (SGOT): Alanine Aminotransferase, also known as serum glutamic oxaloacetic transaminase; GGT: Gamma-Glutamyltransferase. An additional discerning factor between the hypofiltrating vs. normo-filtrating individuals was the post- challenge glucose. While FPG the G3a group compared with people with normal filtration (NF) showed a notably greater glucose elevation at 60 min (median 12.09 vs. 11.07 mmol/L, p = 0.028). This difference also remained evident at 90 min (10.73 vs. 9.67 mmol/L, p = 0.021). While these differences at 30 and 120 min in the OGTT were numerically higher in the G3a group, it was not statistically significant. No other glycemia-related variables showed significant differences between G3a vs. NF, including HbA1c. A higher proportion of women within the hypofiltration group had combined IGTIFG conditions (58%) compared with those with normal filtration (31%), though this sex-disparity did not achieve statistical significance. Although there was no overall disparity in the use of various drugs, a greater prevalence of antihypertensive drug use was observed in the hypofiltration cohort (62% vs. 38%, p = 0.009), likely attenuating observed blood pressure differences. Figure 3 displays outcomes from the multivariable logistic regression analysis aimed at identifying hypofiltration predictors. This analysis incorporated significant variables from univariate analysis and standardized quantitative data into z-scores for normalization. Odds ratios (95% CI) for major variables between people with hypofiltration versus those without. Quantitative variables are expressed as z-scores. OGTT: Oral Glucose Tolerance Test; HDL: High-Density Lipoprotein; GGT: Gamma-Glutamyltransferase. Discussion The ePREDICE study is a comprehensive investigation involving 967 individuals with PD from 7 countries with diverse dietary and lifestyle habits. One of the aims of the study was to assess the presence and degree of renal and other microvascular damage in this population with impaired glucose metabolism. Several studies have previously demonstrated that prediabetes, and not only frank diabetes, can cause kidney damage. In other words, there is evidence suggesting an increased risk of renal involvement in prediabetic individuals compared with those with normal glucose metabolism. The prevalence of renal involvement in people with PD can vary depending on the population studied, diagnostic criteria used, and other factors. Individuals with PD have a high burden of cardiovascular risk factors, including renal damage. In an analysis of data from NHANES 2011–2014, adults with PD (defined using ADA-FPG or HbA1C) had a high prevalence of hypertension (36.6%), dyslipidemia (51.2%), albuminuria (7.7%), and a reduced eGFR (4.6%)18. These studies indicate that there is a significant prevalence of renal involvement in people with PD, emphasizing the importance of early detection and intervention to prevent progression to overt diabetes and associated complications. A number of studies on PD and renal involvement have mainly been focused on the presence and development of impaired renal function19 or glomerular hyperfiltration status20,21. Other authors observed than PD was independently linked to proteinuria development (OR 1.233; 95% CI 1.170–1.301), but not to eGFR decline (OR 0.981; 95% CI 0.947–1.017)22.To better define renal involvement in the PD population, the ePREDICE study categorized kidney damage into three different components: hyperfiltration, albuminuria and decreased eGFR. At the timing of study entrance (baseline) in the ePREDICE trial the prevalence of renal involvement (1 or 2 out of the 3 components) was 9.2%. Among the participants with renal impairment, 29 individuals had an estimated glomerular filtration rate (eGFR) below 60 ml/min/1.73m2, which corresponds to stage G3a CKD (2.99%). Albuminuria, an important marker of kidney damage, was detected in 45 individuals (4.5%). Among these, 44 individuals had urinary albumin excretion between 30 and 300 mg/g (stage A2), while one individual had urinary albumin excretion of 500 mg/g (stage A3). Three participants had both decreased eGFR (G3a) and albuminuria (A2) concurrently. Twenty individuals surpassed presented an elevation in the eGFR beyond 130 ml/min/1.73 m2 (98th percentile within our population) and were defined as having renal hyperfiltration. Glomerular hyperfiltration, albuminuria and decreased eGFR probably are the result from different clinical and metabolism, and pathophysiological mechanism. While they often coexist, they are associated with various factors, some of which overlap. Glomerular hyperfiltration is an early renal abnormality mostly occurring as a compensatory response to metabolic changes, such as hyperglycemia, leading to increased shear stress on the glomerular capillary walls causing albuminuria. Over the time hyperfiltration can subsequently elicit structural changes in the glomerulus causing further albuminuria and increasing the risk of progression to chronic kidney disease (CKD)17,21,23. Multivariable analysis of the ePREDICE data revealed that two prominent factors influencing eGFR were age and waist circumference. Each year of age increased the odds ratio (OR) for decreased eGFR by 5%. These findings are consistent with previous research by Hallan et al.16. Waist circumference showed a strong association with decreased eGFR, with an OR of 3.26. Glomerular hyperfiltration was inversely associated with age (OR 0.92). It is possible that individuals with factors associated with renal sensitivity to hyperglycemia are more likely to exhibit hyperfiltration at a younger age and progress to hypofiltration at a more advanced age. As a new finding, height was inversely associated with renal function, indicating an increase in hyperfiltrating individuals. It is postulated that this inverse association with height may be due to a bias in the GFR estimation algorithm, which should be further investigated, although height-related increase in post-challenge glucose in the OGTT has also been described24,25. An inverse association was found between eGFR and GGT, showing a 3% decrease per unit change. Liver dysfunction has been linked to diabetes development in observational studies26–28. Elevated liver enzymes, such as ALT29 and GGT30, are associated with diabetes incidence and can serve as surrogate predictive biomarkers31. Previous studies indicate GGT is not only a marker of liver dysfunction but also a key biomarker in prediabetes and metabolic syndrome. Elevated GGT is associated with oxidative stress, contributing to systemic inflammation, insulin resistance, and the progression to diabetes, as well as cardiovascular and renal risks in prediabetes32,33. GGT is also involved in the pathogenesis of metabolic dysfunction-associated steatotic liver disease (MASLD) and related complications, including diabetes-associated chronic vascular conditions34. This aligns with our findings of GGT’s association with decreased eGFR, suggesting it may be an early marker of renal involvement in metabolic dysfunction. Research on GGT and CKD presents conflicting results; some studies indicate lower GGT levels in advanced CKD stages, though mechanisms remain unclear. While higher GGT levels are linked to increased CKD risk35, hemodialysis shows an inverse association36, and a meta-analysis found no significant link between elevated GGT and CKD risk37. The ePREDICE study showed lower GGT levels in stage G3a and hyperfiltrators, but these associations were not statistically significant in multivariable analysis. Concerning the metabolism abnormalities, glycemic control emerges as a crucial factor in the likelihood of developing kidney disease. This study examined parameters such as the insulinogenic index (IGI) and the DI, which were decreased in prediabetic subjects, but did not reach statistical significance. IGI has been associated with the incidence of type 2 diabetes mellitus (T2DM)38,39 and assesses early stages of insulin signalling40. DI, on the other hand, predicts the development of glucose intolerance, capturing impaired insulin secretion and insulin resistance, which may coexist in individuals with PD41–43. Analysing the metabolism abnormalities potentially related to each renal component, our study revealed disparate results. People with renal function impairment (eGFR < 60 ml/min/1.73 m2) showed alterations in 60 and 90 min OGTT glucose measurements but did not affect parameters of insulin resistance or β-cell function. In subjects with hyperfiltration, variations in insulin resistance measures (such as HOMA-IR or Matsuda Index) were observed. However, albuminuria emerged as the most significant contributor, with a 30% reduction in the odds ratio associated with a decrease in DI. Limitations of the study: The study has several limitations, including its cross-sectional design, which limits the ability to infer causality. The reliance on eGFR estimates and UACR may not fully capture the complexity of renal function. Additionally, the stringent hyperfiltration threshold may have led to the underestimation of true hyperfiltration cases. Other limitations of this study is the small sample size of the hyperfiltration group, which impacts the statistical power of our analysis. Although significant differences were observed between groups, including a moderate effect size for age (Cohen’s d = 0.60) and a high odds ratio for gender (OR = 4.23, p = 0.022), the achieved power for gender analysis was 0.74, which is below the recommended threshold of 0.80. This reduced power indicates a higher risk of Type II error, meaning that we may have been less likely to detect true differences, particularly with a small sample size. These limitations suggest that while the observed differences are notable, they must be interpreted with caution. Also, variability in population characteristics across the seven participating countries could also influence the generalizability of the findings. Further longitudinal studies are needed to validate these results and explore the progression of renal involvement in people with PD. Conclusions The ePREDICE study shows that people with PD examined at several centers mainly in Europe depict a notable prevalence (9.2%) of renal disease, indicated by an eGFR outside the normal range (> 130 or < 60 ml/min/1.73 m2) or the presence of albuminuria. The study identifies impaired insulin signaling as a key determinant of renal involvement in the population examined. These findings underscore the importance of targeted screening and early intervention to prevent the progression of CKD. Future research should investigate the mechanisms connecting PD to renal dysfunction and evaluate the effectiveness of therapeutic interventions. Acknowledgements Members of the e-PREDICE Consortium. e-PREDICE Consortium investigators: Steering Committee: Jaakko Tuomilehto, Rafael Gabriel, Jaana Lindström, Jesús Egido, Andrea Natali, José Carlos Pastor, Michael Brainin, Marcus Lind, Luis Silva, Peter Schwartz, Aleksandra Gilis-Januszewska. Safety Committee: Carmen Suárez Fernández: Hospital University La Princesa. Madrid, Spain. Beverly Balkau, INSERM Veifville. France. Matti Uusitupa, Institute of Public Health and Clinical Nutrition. University of Eastern Finland, Kuopio, Finland. Internal Ethic Committee: Julio Romero. Hospital University La Princesa, Madrid, Spain. Lars Ryden. Karolinska Institute, Stockholm, Sweden. External Scientific Advisory Board: Ralph DeFronzo, University of Texas. San Antonio, Texas, USA. Manuel Serrano Ríos. Universidad Complutense, Madrid, Spain. Michael Roden, German Center for Diabetes Research (DZD) Heinrich Heine University Düsseldorf, Düsseldorf, Germany. John Nolan, European Diabetes Forum, European Association for the Study of Diabetes, Trinity College Dublin, Ireland. Workpackages (WP): WP1: coordinador. EVIDEM CONSULTORES Rafael Gabriel, Jaakko Tuomilehto, Nisa Boukichou, Tania Acosta, Ruy López-Ridaura, Luis Silva, Eliana del Águila, Ana Rosón. WP2. CLINICAL CENTRES: Spain. Primary Care centres of SERMAS (Servicio Madrileño de Salud), Madrid. Tomás Gómez-Gascón, Juan Carlos Abánades Herranz, María Esther Sánchez Carranza, Alicia Rodríguez Blanco, Fernando Villasante Claudios,Consuelo Ugarte Pérez, Belén Peláez Raposo, Beatriz Cáceres Sánchez, Sergio González Gasca; Margarita Herrero Delgado, Isabel García Del Río, Mª Lorena Rodríguez Pérez, Mª Carmen Reyes Madridejos, Mª Carmen Castillo López, Mª Jesús Paloma Huerga González, Ana Alayeto Sánchez, Carmen Pascual Díez, Esperanza Villar Coloma, Tirso Galiano Arroyo, Mª Olga Peña Peña, Mª Elena Pejenaute Labari, Mª Mercedes Rojo Tardón, Mª Teresa Recio García, María Campos López-Carrión, Sara Criado Jorge, Virginia García Campo, Almudena Pazos González, Aranzazu Pérez Medina, Ricardo Benito Fernández, Mercedes Ricote Belinchón, Mª Teresa Sánchez-Villares Rodríguez, Noelia Polo Fernández, Antonio Cabrera Majada, Eva María Revuelta Marinez , Itziar Vázquez Carrión, Mª Nuria Fernández De Cano Martín, Manuel Gutiérrez Cabanas, Soledad Fernández Saavedra, Yolanda Jiménez Aguilar, Mª Ángeles Rodríguez Loarce, Emilia Pedroche Morales, Teresa Nieto Monreal, María Begoña Hernández Olivares, Carmen Domínguez Encinas, Ana Martínez-Cabrera Peláez, Mª Inés Casas Jiménez, Pilar Pérez Egea, Concepción Espariz Campano, Ángeles Brieva García, Mª Azucena Saez-Berlana, Carlos Casanova García, Mª Carmen Belinchón Moya, Mª Dolores Parejo Pablos, Elisa Varona Lahuerta, Esther Labrador Arranz, Mª Ángeles Conde Llorente, Mª Teresa Gómez Martínez, Milagros Velázquez García, Mª Patrocinio Verde González, Mª Rosario Campo Martínez, Mª Rosario Del Álamo Gutiérrez, Mª Victoria Cantera Urcía, Alejandra González Esteban, Laura Rodríguez Cortizo, Sabrina Sosa Alés, Eva Torres Cantero, Idoia Baíllo Peña, Tamara García López, Cristina Calle Domínguez, Inmaculada Peña Sainz, Mª Antonia Minguito Lobos, Consuelo Viamonte Andrés, Francisco Manuel García García, José María Lobos Bejarano, Raquel Juez Pimienta, Emiliana Villares Motino, Elvira Pérez Peñas, Silvia Jiménez, Laura Manuyama Pacaya, Carmen Morales Guevara, Carmen Melero González, Blanca Novella Arribas, Marta Cuevas, Belén Sierra García, Marta Ruiz, Amelia González Gamarra, Rosa Mª Sánchez Alcalde, Belén Peláez Raposo, Ángela Gallego Arenas, Mª Soledad Mayayo Vicente, Javier López González, Manuel Jovino Arango Victoria, Ana María Santos Caballero, Isabel Jimeno, Juana Iribertegui, Ramón María Salgado, María Olga Ortega de Santos, María Gema García, María José LLorénsBalducel, Juan Carlos de la Fuente, Claudia Fernández Illen, Beatriz Manrique Olmedo, Tati Arévalo Gallego, Juan Machuca Gómez, Esther San José Blázquez, Teresa Castellanos Ruiz, Macarena Espejo Saucedo, María Esther Fernández Yedra, Carmen Torres Martínez, Ángel Lindo Torres, Víctor Raúl Montes Pina , khosrowDadbinDadbin, Rosa ArdáMaillo, Inmaculada Parra Álvarez, Justino Flores Ramos, Mª Dolores Vicente de Forondo, Antonio Calvo Cebrian, Yolanda Gines Díaz, Paloma Henares García, Luís La Puente Montoso, Aurora López Gil, Mª Concepción Marcello,Silvia Membrado Gómez, Nieves Puente García, Nuria Rodríguez Pata, Antonio Sánchez Calso, Mª Sánchez Casado, Mª Pilar Saladana Calzo, Carmen Velayos Rodríguez, Consuelo Velaz López, Miguel Ángel Venga Mendía, C. Susana Abad Guijarro, Gema Calderero Castellanos, Esperanza Corral Agüero, Ángeles Fernández Ortega, Carmen García Regidor, Edurne Hernández Sanzano, Pablo Martín Cano, Mª Del Carmen Martínez Coello, Sandrine Miguel Miguel, Mª Jesús Ramos Martín de Argenta, Beatriz Ruescas Aurrecoechea, Esteban González López, Elena Ramos Quirós, David Pérez Manchón, Leticia Pontón, Jon Koldo SagarduiVillamor, Mireia Rey Pérez de Pipaon, Mª Luisa Idarreta Zubiria, Juan Carlos Sánchez Ruiz, Ángela Rodríguez de Cossio, Milagros Merino Pella, Nuria Ruiz Hombrebueno, Rafael Llanes, Yolanda Vicente Prior, Mercedes Picó, Francisco Pérez Durán, Isabel Pérez Botella, Ángeles Cuevas, Francisco Martínez García, Raquel Cobeñas Mateo, María Teresa Rodríguez De Fonseca, Naldi Luz Cerdeña Ocola, Irene Alma Polanco García, María Esther Amez De Castro, Susana Barrios Espinoza, Antonio Guijarro Jiménez, Ana María De La Uz Pardos, Francisco Javier Cabrera Pérez, José Ignacio Torres Jiménez, Francisco Martínez García, María Isabel Vidal De La Riva, María Teresa Rodríguez De Fonseca, Gabriel Barderas Cuevas, Gonzalo Carrillo De Albornoz Martínez Pantoja, Mª Isabel García Romero. Castilla y León. Primary care centers of Ávila. Primary Care centre of Arévalo: Saturio Vega Quiroga; Roberto Aldrich García, Carlos Cañas Ruesgas, Carmen Vian Baron, Josefina Fernández Fernández, Mª Antonia Jiménez Carabias, Laureano López Gay, Mª Pilar Marqués Macías, Almudena Cantalejo Martín, Ana Benito Pérez, Modesta Mulero San José, Vanesa Martín Hernández, Laura Sánchez Domínguez, Rosa Mª García Martín, Victor Manuel Álvarez Zurdo. Primary Care centre of Sotillo de la Adrada: David Álvarez Suárez, Carmen Lázaro del Nogal, Lourdes González López, Mª del Mar Varas Reviejo, Juan Luis Martín Clavo, Mª Isabel Blázquez Blanco,Mª Luisa Ramos González, Guadalupe Rinaldi Català, Montserrat López Ramírez, Vanesa Hernández Blázquez, Vanessa Gutiérrez León, Raquel Pérez Cruz, Josefina Fernández Fernández, Almudena Fernández García, Raquel Alonso Moralejo. Primary care centers of Segovia. Primary Care centre of Carbonero El Mayor: María Soledad Fragua Gil, Virginia Silva Guisasola, Concepción Manrique de la Fuente, Ángeles Lazcoz Fontán, Héctor Aceves Gamarra, Alba Marina Hernández López, Mª Jesús Blanco Ledesma, Alfonso Santos López, Cristina de la Cruz Maeso, Mª del Espíritu Santo Otero Herrero, Cristina Olmos Marinero, Patricia Redondo Arranz, Mónica Álvaro García. Primary Care centre of Segovia III: Luis Gonzálvez López, María Ángeles Raquejo Grado, José Rodríguez Sanz, Juan Manuel de Andrés Rubio, Nuria González Acebes, Joaquina Galán Sánchez, Teresa López Fernández-Quesada, Almudena Sanz Prieto, Carmen Montero Morales, María Dolores Alba Jiménez, Beatriz Ayala Miranda. Castilla-La Mancha. Primary Care centres of CUENCA. Hospital General: Jaime Santiago Aranda Regules, Alba Caterina del Hoyo Herráiz, María Victoria Cantero Ayllón, María José Guillén Izquierdo, María Sandra Ruiz Mora, Ana Peña Cabia, Rosa Sánchez Amo, Mº Josefa Moya López. Cuenca I: Fructuoso Muelas Herráiz, Mº Ángeles Molina Morate, Fernando Salcedo Aguilar. Cuenca II: Nieves Valero Caracena, Beatriz Ortega Noheda, Mº Carmen García González. Cuenca III: Cristina Martínez Martín, Miryam Pardo Villalvilla, Mª Eugenia García Castellanos, María Elena de las Heras Martínez. Primary Care centre of Tarancón: Filomena del Saz Castellanos, Encarnación Palomares Cañada, María Concepción Fraile Jiménez, Pilar Palomar Moreno, Bárbara Martínez Garrido, María Pilar Orgaz Gallego, María José Tricio Armero, Cristina García del Pino Cañadas, Isabel Tierno Aparicio. Málaga. Fundación FIMABIS. Servicio Andaluz de Salud (SAS). Regional University Hospital of Málaga, Biomedical Research Institute of Málaga (IBIMA), Universityof Málaga (UMA), Málaga, Spain. Medicina Interna: Ricardo Gómez-Huelgas, María Dolores López-Carmona, Luis M Pérez-Belmonte, María Rosa Bernal-López, María Teresa Moyano Paris, Paula Moya Rodríguez, Antonio Vargas Candela, Alberto Vilches Pérez, María Isabel Ruiz Moreno, Maite Muñoz Melero, Pilar Gómez Martin. Oftalmología: Jacinto Villalvilla, Álvaro Santos, Antonio Archilla, Carlos Rocha, Silvia Lozano Ruiz. Primary care centers of Málaga. Alameda- Perchel: Francisco Javier Orellana Lozano, Manuel Guarino Nuño, Alhaurín de la Torre: Daniel Martin Castillo, María José Guerra Maldonado, José Rogelio Sánchez Ortiz. Alozaina-Yunquera: David Fernández Bonilla, J A Cortes, Juan Antonio Cordero Cabrera. Antequera: José Antonio Godínez, José Jesús Moreno Jiménez, David Paniagua Urbano. Archidona: Antonio Cansino Osuna, María Del Carmen Rojo Camacho, Celinda Lara Moreno, Ignacio Hinojosa Núñez, Almudena Puga González, Capuchinos: Yolanda Rey, Yolanda Rodríguez Gallego. Carranque: Carmen Aylón Moliner. CartamaEstacion-Cartama pueblo-Pizarra: Francisco Jose Guirado Hidalgo, María Eva Ruiz Coronado, Susana Barea Diañez, Beatriz Navarro Aranda. Casarabonela: M Carmen Arroyo Martínez. Ciudad Jardin: Antonio Baca Osorio, José Mancera, Salvador Ruiz Vera, Idelfonso Martos Cerezuela. Delicias:Fernando López Verde, Mª Carmen Barba Cañete, Cristóbal Gómez Acevedo, Margarita Sánchez Pavón. La Luz: Antonio Oropez Mesa, Antonio Rojas Barrilado. La Roca: Esther Martin Aurioles, Rocío Ramos, Francisco Javier Camino, María Eugenia Valdes, Dolores Bravo Fernández. Limonar: Silvia Hazañas, Amparo Vargas Machuca Benítez, Eva María Taboada Ríos. Puerta Blanca: Antonio Hormigo Pozo, Idelfonsa Martínez Zaragoza. Rincón de la Victoria: Milagrosa Espinar Toledo, María Del Rosario Rosillo Rein, Gloria Inmaculada Mestre Reoyo, María Auxiliadora Naranjo Sánchez, José Ángel Sánchez Ortiz, Mª José González Vega, José Carlos Pérez Sánchez, Antonia Cabra Navarro, Antonio Vivas Molina. San Andrés-El torcal: Antonio Ramírez Ceballos, Francisco Ruiz Solares. Tiro Pichón: Juan José Bedoya Belmonte, Germán Ortega Núñez, María Encarnación Bueno Caro. Trinidad-Jesús cautivo: Santos Agreda. Victoria: María José Bujalance Zafra, Montse Román Cereto, Rafael Ángel Maqueda. Poland. Uniwersytet Jagiellonski, Collegium Medicum, Poland: Aleksandra Gilis-Januszewska, Alicja Hubalewska-Dydejczyk, Beata Piwońska-Solska, Justyna Biegańska, Katarzyna Cybulska, Bernadeta Marcykiewicz, Magdalena Duraczyńska, Anna Cybulska, Joanna Stankiewicz-Góra, Alina Mruk, Grzegorz Młyński, Lucyna Rozpondek, MichałSroka, Maciej Gilis-Januszewski, Edyta Sacha, Adela Justyńska, Magdalena Szopa, BartłomiejMatejko. Greece. National and Kapodistrian University of Athens. Greece: KonstantinosMakrilakis, Stavros Liatis, EvangeliaSiami, ChryssoulaStathi, Katerina Barmpagianni, MariaNikoloudi, MeropiKontogianni, IoannaKechrimpari, AphroditiTsiakou, Melina Karaolia. Alexandra Hospital. Universityof Athens. Greece: AsiminaMitrakou, Georgios Panagopoulos, ParaskeviKontou, Petros Thomakos, Georgios Giagkou, EvangeliaAvgeraki, EiriniMamalaki. Bulgaria. University Multi-Profile Hospital for Active Treatment Alexandrovska EAD. Sofia, Bulgaria: Zdravko Asenov Kamenov, Antoaneta Trifonova Gateva, Yavor Sashov Assyov, Tsvetan Vladimirov Gatev, Vera Nacheva Karamfilova, Iveta Slavyanova Nedeva. Austria. Gemeinnutzige Salzburger Landeskliniken Betriebsgesellschaft. Salzburg, Austria: Bernhard Paulweber, Ludmilla Kedenko, Andrea Undeutsch. Turkey. Istanbul University Istanbul. Turkey: Ilhan Satman, B. Fulya Turker, Ayse K. Uzum, Sakin Tekin, Ramazan Çakmak, Elif T. Bagdemir, Selda G. Celik, Cemile C. Idiz, Halime C. Sackoparan, Zafer Cebeci, Nur Kir, Dilara Karsidag, Yildiz Tutuncu, Aslihan Demirbas, Busra Yildiz. Serbia. Medical System Beograd-MSB. Belgrade, Serbia. Predrag Djordjevic, Margarita Dodevska, Nevenka Raketic, Aleksandar Stamenkovic, Marko Jovic, Fadil Canovic, Ljiljana Milivojevic, Kristina Savic, Ljiljana Savic, Mirjana Sarkic. Faculty of Medicine, University of Belgrade. Serbia. Nebojsa Lalic, Katerina Lalic, Aleksandra Jotic, Jelena Stanarcic, Ljiljana Lukic, Tanja Milicic, Natasa Rajkovic, Marija Macesic, Dijana Risimic, Mladen Bila. Australia. The Univerity of Sydney. Australia: Stephen Colagiuri, Anthony Keech, Kristine Maddock, Andrzej S. Januszewski, Liping Lee, Tegan Picone, Emma Sainsbury, Alison Coenen, Chelsea Hendy, Namson Lau, Tania Markovic, Erica Bessell, Nick Fuller. Kuwait. Dasman Diabetes Research Institute: Jaakko Tuomilehto, Abdullah Alkandari, Abdullah Bennakhi, Monira Alarouj, Mohammad Jalali, Medinella Fernandez, Makka Ali Osman, Jincy Raj, Ala’a Al-Obaid, Hyatt Alsayegh, Najeeba Almatrouk. WP3. Lifestyle Intervention. Terveyden ja Hyvinvoinnin Laitos. Finland: Jaana Lindström, Päivi Valve, Katri Hemiö, Katja Wikström, Esko Levälahti, Pirjo Saastamoinen. WP4. Central laboratories and Biobank. Instituto de Investigación Sanitaria de la Fundación Jiménez Díaz. Madrid, Spain (Central Laboratory, Biobank and Inflammatorybiomarkers): Jesús Egido, Sebastián Mas, Sandra Zazo, Esther Civantos, Rosario de Nicolás, Federico Rojo. Consiglio Nazionale delle Ricerche. Pisa, Italy. (NAFLD laboratory): Amalia Gastaldelli, Fabrizia Carli, Emma Buzzigoli, Melania Gaggini. Fundació Hospital Universitari Vall d’ Hebron, Institut de Recerca. Barcelona, Spain (Retinal biomarkers): Rafael Simó, Cristina Hernández, Marta García-Ramirez. Queen Mary University of London. UK (Genetic laboratory): Graham A Hitman. WP5. Microvascular assessment coordination. Universita di Pisa, Italy. (Endotelial assessment): Andrea Natali, Lucrecia Motta. WP6. Retinal Assessment. Instituto de Oftalmobiología Aplicada (IOBA) Universidad de Valladolid, Spain. (Central Retinal Reading Centre): Maribel López, José Carlos Pastor, Lucía Manzanas, Ignacio Alonso, Verónica Velasco, Laura Mena. e-DIAGNOSTIC Oftalmología. Madrid, Spain. (Retinal e-platform): Diana Bravo, Víctor González Rumayor, Marica D´Angelo, Álex Manau. WP7. Neuropsychological assessment. Department for Clinical Neurosciences and Preventive Medicine, Danube University Krems. Krems, Austria: Michael Brainin, Yvonne Teuschl, Alexandra Dachenhausen, Karl Matz. Instituto de Investigación y Asistencia Psiquiátrica (IAP). Madrid, Spain: Laura Ferrando Bundío. Oivauni Oy. Kuopio,Finland: Henri Tuomilehto, Seppo Silvennoinen. WP8. Statistical analysis. Vastra Gotalands Lans Landsting. Gotenburg, Sweden (Statistical Analysis coordinator): Marcus Lind, Aldina Pivodic, Hans Wedel. Institute of Neuroscience, National Research Council (Consiglio Nazionale delle Ricerche). Pavoda, Italy. (Insulin and C-Peptide modeling): Andrea Mari, Andrea Tura. University of Helsinki. Department of Public Health: Pekka Jousilahti. WP9. Technology Assessment. IMPETO Medical. Paris, France (Sudoscan assessment): Jean-Henri Calvet, GaelleLerise, Alice Vilier. Mezen Bouzaien. AARDEX Group SA. Geneva. Switzerland (MEMs drug´s adherence monitoring): Bernard Vrijens, Rodrigo Paiva, Eric Tousset. WP10. Dissemination and communication: Federation Internationale du Diabete Region Europe Lala Rabemananjara. Abbreviations Type 2 diabetes Prediabetes Estimated glomerular filtration rate Disposition index Author contributions Sebastián Mas-Fontao: Investigation, Formal Analysis, Supervision, Methodology, Writing – Original Draft Esther Civantos: Investigation, Methodology Nisa Boukichou: Conceptualization, Formal AnalysisJuan A Moreno: Writing – Review & Editing Jaakko Tuomilehto: Investigation, Conceptualization, Writing – Review & Editing Rafael Gabriel: Investigation, Conceptualization, Writing – Review & Editing Jesús Egido: Supervision, Writing – Original Draft. Funding All the funding or sources of support received during this specific study: - European Commission, FP7 EC-GA No. 279074 - Merck-Serono, Germany - Boeringher Ingelheim, Germany (IIS Program. Grant number 1218.166) - Instituto de Salud Carlos III, Spain PI11/01653 The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript. The IIS-FJD diabetes research group was funded by Instituto de Salud Carlos III/FEDER (PI20/00487, PI23/00119), Spanish Biomedical Research Centre in diabetes and metabolic diseases (CIBERDEM) and Spanish Society of Nephrology (SEN). Data availability The data that support the findings of this study are available from ePREDICE Study group but restrictions apply to the availability of these data, which were used under license for the current study, and so are not publicly available. Data are however available from the authors upon reasonable request and with permission of ePREDICE Study group (contact person Nisa Boukichou: nisa83_1@hotmail.com). Declarations Competing interests The authors declare no competing interests. Footnotes Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. Contributor Information Sebastián Mas-Fontao, Email: smas@fjd.es. Jesús Egido, Email: jegido@quironsalud.es. References Associated Data Data Availability Statement The data that support the findings of this study are available from ePREDICE Study group but restrictions apply to the availability of these data, which were used under license for the current study, and so are not publicly available. Data are however available from the authors upon reasonable request and with permission of ePREDICE Study group (contact person Nisa Boukichou: nisa83_1@hotmail.com).