Oleanolic Acid‐Enriched Olive Oil Attenuates Kidney Function Decline in Patients With Type 2 Diabetes: Results From the OLTRAD Randomised Controlled Trial

ABSTRACT

Aims

Chronic kidney disease (CKD) is a major complication of type 2 diabetes mellitus (T2DM), driven in part by oxidative stress and the accumulation of advanced glycation end products (AGEs). Oleanolic acid (OA), a natural triterpene present in olive‐derived matrices, may modulate AGE metabolism. In this prespecified, exploratory secondary analysis of the OLTRAD trial, we evaluated whether long‐term intake of OA‐enriched olive oil attenuates kidney function decline in patients with T2DM.

Methods

In the OLTRAD randomised controlled trial, 97 patients with T2DM were assigned to consume OA‐enriched olive oil (600 mg OA/kg) or control olive oil for 12 months, alongside standard therapy. Kidney outcomes included estimated glomerular filtration rate (eGFR) and urinary albumin‐to‐creatinine ratio (uACR). Circulating AGEs (methylglyoxal [MG], carboxymethyllysine [CML]), enzymes involved in AGE metabolism (aldose reductase [AR], glyoxalase‐I [GloxI]), and oxidative stress markers (ROS and MDA) were assessed.

Results

OA‐enriched olive oil attenuated eGFR decline relative to control, an effect restricted to participants with mildly impaired eGFR at baseline (60–90 mL/min/1.73 m²; time × group interaction p = 0.010), with no significant interaction among participants with normal baseline eGFR. OA‐enriched oil also prevented the rise in uACR observed with control oil. Mechanistically, OA reduced AR concentration and increased GloxI activity, resulting in lower circulating MG, whereas CML remained unchanged. OA‐enriched oil further reduced ROS and MDA levels, while MDA increased with control oil.

Conclusions

In this exploratory, hypothesis‐generating secondary analysis, long‐term consumption of OA‐enriched olive oil was associated with attenuated kidney function decline in patients with T2DM, an effect limited to those with baseline mildly impaired eGFR. These changes were associated with modulation of AGE‐related pathways and reduced oxidative stress. As the trial was powered for glycemic control rather than renal outcomes, larger, adequately powered studies are needed to confirm these findings before their clinical relevance can be established.

Trial Registration: Clinicaltrials.gov identifier: NCT06030544; https://clinicaltrials.gov/study/NCT06030544

1. Introduction

Chronic kidney disease (CKD) is a major complication of type 2 diabetes mellitus (T2DM) and often remains undetected until advanced stages, increasing the risk of kidney replacement therapy [1]. Its progression is accelerated by comorbidities such as hypertension, obesity, and persistent oxidative and inflammatory stress, even under optimal glycemic and blood pressure control. CKD is also associated with increased cardiovascular risk, mortality, and a substantial burden on healthcare systems, resulting in a significant impairment of patients' quality of life [2]. Identifying modifiable factors contributing to kidney function decline is therefore essential for early prevention.

Reduced estimated glomerular filtration rate (eGFR) and increased urinary albumin‐to‐creatinine ratio (uACR) are key indicators of kidney damage. Elevated uACR is associated with accelerated eGFR decline and progression to end‐stage renal disease [3]. eGFR is the gold‐standard marker of kidney function, declining with age and more rapidly in T2DM [4]. Accordingly, strategies aimed at lowering uACR and preserving eGFR may provide meaningful kidney protection in T2DM.

Advanced glycation end products (AGEs) are key contributors to CKD and its cardiovascular complications. AGEs are irreversible compounds formed through non‐enzymatic reactions between reducing sugars and proteins, lipids, or nucleic acids. Their accumulation is enhanced by chronic inflammation, hyperlipidaemia and oxidative stress, hallmarks of T2DM and CKD [5, 6]. Elevated circulating AGEs are associated with faster kidney function decline and increased cardiovascular risk [7]. In this context, the mediterranean diet, characterised by low dietary AGE content and high antioxidant intake, reduces circulating AGEs and slows atherosclerosis progression [8]. Lowering AGE burden may therefore contribute to kidney function preservation in T2DM, highlighting the potential role of dietary interventions in mitigating AGE‐related kidney damage [9, 10, 11].

Among the bioactive compounds of the Mediterranean dietary pattern, oleanolic acid (OA) (3β‐hydroxy‐olea‐12‐en‐28‐oic acid) is a pentacyclic triterpene widely distributed in the plant kingdom. Although OA has been identified in several fruits and aromatic plants, including grapes, apples, figs and rosemary, olive‐derived products (olive leaves, fruit skin and olive oil) represent its major dietary source in Mediterranean populations. OA has shown antioxidant, anti‐inflammatory and antidiabetic properties [12, 13, 14] that may contribute to kidney and cardiovascular protection in T2DM [15, 16]. In fact, experimental and translational studies have reported beneficial effects of OA and related olive triterpenes on inflammatory responses, vascular function, obesity‐related complications and diabetic kidney injury, supporting their potential use as nutraceuticals and components of functional foods [17, 18, 19, 20] (REF). OA also inhibits key enzymes of the polyol pathway, including aldose reductase (AR) and enhances glyoxalase‐I (GloxI) activity, reducing methylglyoxal (MG) levels [21] and AGE formation [22]. These mechanisms are associated with improved glycemic control, reduced renal AGE deposition and lower urinary glycated albumin in experimental diabetes models [15]. However, clinical evidence in humans remains limited [23, 24].

Therefore, this study aimed to evaluate whether modulation of AGE metabolism through OA, delivered as an OA‐enriched olive oil, can reduce circulating AGE levels and preserve kidney function in patients with T2DM. We further assessed its effects on key metabolic pathways involved in AGE formation and detoxification to explore its potential as a nutritional strategy in this high‐risk population.

2. Methods

2.1. Design and Study Population

The current work was conducted within the framework of the OLTRAD (oleanolic acid as therapeutic adjuvant for the treatment of type 2 diabetes mellitus) study (ClinicalTrials.gov NCT06030544), a prospective, randomised, parallel‐group, double‐blind, controlled clinical trial. Based on an intention‐to‐treat analysis, the study included 97 patients with T2DM. The primary objective of the OLTRAD study was to evaluate whether the regular consumption of an OA‐enriched olive oil enhances the therapeutic effects of standard antidiabetic pharmacological treatment in patients with T2DM, compared to a control oil without triterpene enrichment over a 12‐month period. Kidney‐related outcomes were prespecified as secondary endpoints of the study.

Participants were recruited between March 2022 and March 2024 at the Endocrinology and Nutrition Service of the Virgen del Rocío University Hospital (Seville, Spain). Eligible participants were adults (18–75 years) with BMI 25–39.9 kg/m² and T2DM diagnosed ≥ 6 months, on stable metformin therapy (≥ 850 mg/day) either as monotherapy or in combination with other antidiabetic agents (excluding pioglitazone and sulfonylureas). Key inclusion criteria included HbA1c < 9% and stable treatment, while major exclusions comprised severe CKD, significant liver disease, recent cardiovascular events, and conditions affecting study adherence. Full inclusion and exclusion criteria are detailed in Table S1.

2.2. Ethical Considerations

The study protocol was approved by the Research Ethics Committee of the Virgen del Rocío University Hospital (October 29, 2020) and the CSIC Ethics Committee (Code 142/2020; February 12, 2021). All participants provided written informed consent. The OLTRAD study was conducted in accordance with the Declaration of Helsinki and the International Council for Harmonization Good Clinical Practice guidelines (https://www.ich.org/).

2.3. Randomization, Intervention and Compliance Assessment

Participants were randomly allocated in a 1:1 ratio to the study groups by using computer‐generated sequences (Figure S1). Randomization was stratified by sex and age (cutoff: 50 years). Only the principal investigator, who was responsible for the study, was aware of each participant's assigned intervention group. All other members of the intervention team remained blinded throughout the study.

Each patient was randomly assigned, in addition to standard antidiabetic pharmacological therapy, to one of two intervention arms: (a) a control olive oil, consisting of a commercially available blend of virgin and refined olive oils with a low content of minor bioactive components and (b) a functional olive oil enriched with OA (600 mg OA/kg). OA‐enriched oil was formulated using the control oil as a base, to which crystallised OA (≥ 95% purity), extracted from olive leaves, was added following the protocol described in European patent EP18382510. Both oils were centrally prepared and provided to participants by the study team in identical containers with colour‐coded labelling to ensure double blinding. The average composition of both the control and OA‐enriched olive oil used in this study has been previously reported by García‐González et al. [25]. The OA‐enriched olive oil contained ≈600 mg OA/kg oil, corresponding to an estimated daily OA intake of approximately 30 mg, whereas the control olive oil contained only trace amounts of this triterpene (Table S2).

Participants were instructed to consume approximately 55 mL/day of the assigned olive oil throughout the intervention period. The caloric contribution from olive oil intake was identical in both intervention groups, making it unlikely that differences in energy intake contributed to the observed outcomes.

Participants received the assigned oil free of charge during scheduled study visits, with a supply sufficient to cover the subsequent three‐month period. They were followed for 12 months, with follow‐up visits conducted on a quarterly basis according to the study protocol. Participants were asked to return empty containers at each visit, which were recorded as an indirect measure of adherence.

Adherence to intervention was evaluated using both subjective and objective measures. Self‐reported compliance was assessed using the Haynes–Sackett adherence test, a validated tool for measuring therapeutic compliance in clinical practice [26]. In addition, participants were instructed to return all olive oil containers at each follow‐up visit, and the number of empty bottles was recorded as an indirect measure of adherence. To further validate compliance, serum concentrations of OA were quantified in fasting blood samples obtained from participants in both intervention and control groups at baseline and after 12 months of follow‐up.

Serum OA concentrations in patient samples collected at baseline and after 12 months of follow‐up were quantified using the method previously described in our earlier publication [27].

2.4. Dietary Assessment

Dietary intake was evaluated at baseline and after 12 months using a validated food frequency questionnaire (FFQ) for the Spanish population. In addition, 24‐h dietary recalls were collected during follow‐up visits to further characterise dietary intake. Information on lifestyle factors, including physical activity, alcohol consumption and smoking status, was also recorded. Participants were encouraged to maintain their dietary habits and physical activity levels throughout the trial, without imposing calorie restrictions. No significant differences in dietary intake or lifestyle factors were observed between intervention groups at baseline or during follow‐up (data not shown).

2.5. Anthropometric Measurements and Laboratory Tests

At baseline and after 12 months of follow‐up, participants underwent a comprehensive clinical evaluation including measurement of anthropometric parameters (BMI, height and waist circumference) and SBP and DBP. A structured interview was also conducted to record gastrointestinal symptoms or other adverse events, as well as to collect information on lifestyle habits, medical history and current medication use.

Fasting blood samples were obtained from the median cubital vein using sterile vacuum collection tubes at both time points (baseline and 12 months). Additionally, participants provided a first‐morning urine sample in sterile containers. All blood and urine samples were processed and analysed at the Clinical Biochemistry and Analysis Laboratory of the Virgen del Rocío University Hospital following standard operating procedures. Blood was separated into serum or plasma through centrifugation within an hour of collection (1500 × g for 20 min at 20°C or 1500 × g for 15 min at 4°C, respectively).

Additionally, three blinded physicians from the trial team retrospectively reviewed participants' electronic medical records to identify any serious health events (vascular, hepatic, renal, oncological, or neurodegenerative) occurring from recruitment to 12 months after the study ended.

2.6. Adverse Events Assessment

Adverse events were systematically assessed during the intervention period using both face‐to‐face medical interviews at each on‐site visit, as well as digitised medical records (‘DIRAYA Navegador’ and ‘DIRAYA Estación Clínica’ softwares). Event severity was graded according to the Common Terminology Criteria for Adverse Events (CTCAE version 5.0). Safety outcomes, including adverse events, gastrointestinal symptoms, liver‐related events, treatment‐related adverse events, withdrawals due to adverse effects and deaths, were recorded and summarised by treatment group.

2.7. Evaluation of Kidney Function

Kidney function was evaluated by determining the serum creatinine (sCr)‐based eGFR, calculated using the CKD‐Epi (chronic kidney disease epidemiology collaboration) equation [28].

In order to analyse the influence of eGFR at baseline on the evolution of kidney function after intervention, participants were stratified into three categories based on their initial eGFR values: (a) normal eGFR (≥ 90 mL/min/1.73 m², n = 53), (b) mildly impaired eGFR (60–< 90 mL/min/1.73 m², n = 34) and (c) severely impaired eGFR (< 60 mL/min/1.73 m², n = 2) [9, 29]. For subgroup analyses, patients with severely impaired eGFR were excluded due to the limited sample size, which would preclude statistically meaningful comparisons (Figure S1).

Several serum and urinary markers were also evaluated at both baseline and after 12 months of intervention. Serum urea, uric acid, total protein and albumin concentrations were quantified using standardised methods [30]. We also evaluated the uACR, both at baseline and after the 12 months follow‐up period. This ratio was computed by dividing the concentration of urinary albumin by that of creatinine, in mg/g. Using this method, which is based on a spot urine test, produces results which are comparable to those obtained using a 24 h urine collection [31].

2.8. Assessment of Circulating AGE Levels and Enzymatic Markers of AGE Metabolism

Circulating MG and Nε‐carboxymethyllysine (CML) levels were measured in serum samples using competitive ELISA kits (OxiSelect methylglyoxal competitive ELISA kit and OxiSelect N‐epsilon (carboxymethyl) lysine competitive ELISA kit, Cell Biolabs Inc., San Diego, CA, USA), following the manufacturer's instructions. These well‐validated competitive ELISAs are an enzyme immunoassay developed for rapid detection and quantitation of MG‐H1 (methylglyoxal‐hydro‐imidazolone) and CML protein adducts, respectively [32, 33]. The values indicate the accumulation of AGEs in the body and are reflective of stable protein‐ or peptide‐associated MG and CML rather than the free compounds. Concentrations of aldose reductase (AR), an enzyme that initiates the polyol pathway by converting glucose into sorbitol, which is further metabolised to fructose, and this accumulation promotes the formation of AGEs, were measured using ELISA kit (CUSABIO, Wuhan, China). The glyoxalase system, particularly GloxI, is an important regulator of the control of intracellular AGEs, mainly MG [34, 35]. Activity of GloxI was determined using ELISA kit (Assay Genie, Dublin, Ireland). All the samples were assayed in duplicate.

2.9. Evaluation of Oxidative Stress Parameters

Reactive oxygen species (ROS) levels were assessed by measuring the fluorescence of the 2′,7′‐dichlorodihydrofluorescein diacetate (DCFH‐DA) probe (Sigma, D6883‐50MG), following a previously described method based on the Fenton reaction [36]. Briefly, serum samples were incubated with HEPES buffer containing DCFH₂ ferrous sulfate (FeSO₄) and fluorescence was measured at 485/538 nm after 12 min. Quantification was performed using H₂O₂ calibration curves. All measurements were performed in triplicates.

Serum malondialdehyde (MDA) was determined using a thiobarbituric acid reactive substances (TBARS) assay adapted from Ohkawa et al. [37]. After protein precipitation and reaction with thiobarbituric acid, absorbance was measured at 532 nm. MDA levels were calculated using calibration curves generated with 1,1,3,3‐tetramethoxypropane. All measurements were performed in triplicate.

2.10. Statistical Analysis

The sample size of the OLTRAD trial was determined based on the primary endpoint of the study, which was glycemic control (HbA1c), as specified in the trial registration (ClinicalTrials.gov NCT06030544). The calculation was based on the hypothesis of at least a 7% reduction in HbA1c levels in the intervention group compared with the control group. Assuming a one‐sided hypothesis test, a significance level of α = 0.01 and a statistical power of 95%, the estimated sample size required was 44 participants per group. This number was increased to 50 participants per group to account for potential dropouts and losses to follow‐up.

The Kolmogorov–Smirnov test was used to assess the normality of quantitative variables. Continuous variables that significantly deviated from the assumption of normality were transformed accordingly. Continuous variables are presented as mean ± standard deviation/error, while categorical variables are presented as proportions. Changes (Δ) were calculated as the difference between post‐intervention and baseline values. Categorical variables were compared using chi‐square tests. Within‐group and between‐group changes were assessed using paired and unpaired t‐tests, respectively. To evaluate data variation according to intervention group and time (baseline to 12 months), repeated‐measures ANOVA was conducted, followed by post hoc multiple comparisons analysis using the Bonferroni correction. Analyses were adjusted for potential confounders or effect modifiers (age, sex, BMI, smoking status and pharmacological treatment—use of angiotensin‐converting enzyme inhibitors [ACEIs], angiotensin II receptor blockers [ARBs], GLP1‐Ras, sodium‐glucose cotransporter‐2 inhibitors [SGLT2is], diuretics and metformin dose), except when the variable of interest already incorporated one or more of these covariates in its definition, in which case adjustments were not applied to avoid redundancy. Analyses of kidney outcomes were performed as an available‐case analysis, restricted to participants with available follow‐up data, without imputation of missing values. Statistical analyses were performed using SPSS version 29 (IBM SPSS Statistics, NY, USA).

3. Results

3.1. Baseline Characteristics of the Study Population

Of the 97 participants initially included in the study, 95 underwent baseline kidney function assessment. At the 12‐month follow‐up, 89 participants completed kidney function evaluation. Renal outcomes were analysed using an available‐case approach: a total of 87 participants were included, after excluding eight individuals with missing follow‐up eGFR data and two with severely impaired kidney function at baseline (eGFR < 30 mL/min/1.73 m²) (Figure S1).

Baseline anthropometric and biochemical characteristics of the patients who underwent the baseline study of kidney function are presented in Table 1. No significant differences were observed between randomised dietary groups in terms of baseline parameters. Furthermore, the baseline characteristics of patients who did not complete the 12‐month follow‐up kidney function assessment were comparable to those who completed it, except for serum creatinine concentrations, which were higher in patients that did not complete the study (Table S3).

TABLE 1.

Baseline anthropometric and metabolic characteristics of the participants by randomised study groups. ^a

	Total population (n = 95)	OA‐enriched olive oil (n = 46)	Control olive oil (n = 49)
Anthropometric parameters
Age (years)	59.7 ± 9.5	60.4 ± 8.8	59.1 ± 10.2
Sex (men/women)	58/39	28/19	30/20
Weight (kg)	87.6 ± 14.0	87.2 ± 16.1	87.9 ± 12.8
BMI (kg/m2)	30.5 ± 3.7	30.2 ± 4.0	30.6 ± 3.5
Waist circumference (mm)	108.5 ± 9.7	108.2 ± 9.6	108.8 ± 9.8
Hip circumference (mm)	107.3 ± 9.4	107.3 ± 10.3	107.2 ± 8.6
DBP (mm Hg)	84.8 ± 9.8	84.1 ± 10.6	85.4 ± 9.0
SBP (mm Hg)	133.8 ± 16.9	133.9 ± 17.3	133.7 ± 16.8
Serum markers of lipids and glucose metabolisms and inflammation
LDL‐cholesterol (mg/dL)	91.0 ± 32.4	96.6 ± 31.6	85.8 ± 32.6
HDL‐cholesterol (mg/dL)	51.5 ± 14.1	50.8 ± 11.8	52.3 ± 16.0
Total cholesterol (mg/dL)	167.0 ± 37.1	171.3 ± 36.0	162.8 ± 37.9
Triglycerides (mg/dL)	128.7 ± 63.5	125.1 ± 69.7	132.1 ± 57.6
Fasting glucose (mg/dL)	107.9 ± 40.2	104.1 ± 29.8	111.5 ± 48.0
Fasting insulin (mU/L)	10.8 ± 7.9	9.03 ± 4.16	10.8 ± 7.5
HbA1c (%)	6.32 ± 0.80	6.27 ± 0.86	6.37 ± 0.75
HOMA‐IR	2.83 ± 2.00	2.49 ± 1.53	3.14 ± 2.34
HOMA‐β	135 ± 164	134 ± 118	136 ± 201
hs‐CRP (mg/dL)	2.65 ± 4.49	3.11 ± 6.08	2.21 ± 1.91
Alcohol and tobacco habits
Active alcohol consumption (n, %)	9 (9.3%)	5 (10.6%)	4 (8.0%)
Current smokers (n, %)	19 (19.6%)	9 (19.1%)	10 (20.0%)
Past‐smokers (last 12 months) (n, %)	28 (28.9%)	15 (31.9%)	13 (26.0%)
Non‐smokers (n, %)	50 (51.5%)	25 (53.2%)	25 (50.0%)
Use of medications
ACE inhibitors (n, %)	21 (21.6%)	8 (17.0%)	13 (26.0%)
ARA‐II	33 (34.0%)	13 (27.7%)	20 (40.0%)
Diuretics	37 (38.1%)	21 (44.6%)	16 (32.0%)
Other antihypertensive medications (n, %)	30 (30.9%)	11 (23.4%)	19 (38.0%)
Statins (n, %)	64 (65.9%)	33 (70.2%)	31 (62.0%)
GLP‐1 receptor agonists (n, %)	43 (44.3%)	23 (48.9%)	20 (40.0%)
SGLT2 inhibitors (n, %)	37 (38.1%)	18 (38.3%)	19 (38.0%)
Kidney function parameters
eGFR (mL/min/1.73 m2)	88.8 ± 14.1	88.5 ± 12.3	89.1 ± 15.7
Serum creatinine (mg/dL)	0.83 ± 0.18	0.82 ± 0.17	0.84 ± 0.19
Serum uric acid (mg/dL)	4.83 ± 1.26	5.08 ± 1.19	4.58 ± 1.29
Serum albumin (g/dL)	4.69 ± 0.28	4.71 ± 0.26	4.67 ± 0.29
Total serum protein (g/dL)	7.17 ± 0.43	7.20 ± 0.42	7.15 ± 0.46
Serum urea (mg/dL)	35.2 ± 9.6	32.0 ± 8.1	38.2 ± 9.9
Urine creatinine (mg/dL)	98.3 ± 54.1	102.3 ± 59.0	94.7 ± 49.6
Urine albumin‐to‐creatinine ratio (mg/g)	13.8 ± 22.8	11.4 ± 21.0	16.2 ± 24.4

Note: We used unpaired t‐tests for quantitative variables and chi‐square tests for categorical variables.

Abbreviations: ACE, angiotensin‐converting enzyme; ARA‐II, angiotensin II receptor antagonists; BMI, body mass index; DBP, diastolic blood pressure; eGFR, estimated glomerular filtration rate; GLP‐1, glucagon‐like peptide‐1; HDL, high‐density lipoprotein; HbA1c, glycated haemoglobin; HOMA‐β, homeostatic model assessment of beta‐cell function; HOMA‐IR, homeostatic model assessment of insulin resistance; hs‐CRP, high‐sensitivity C‐reactive protein; LDL, low‐density lipoprotein; OA, oleanolic acid; SBP, systolic blood pressure; SGLT2, sodium‐glucose cotransporter‐2.

^a Values are represented as the mean ± standard deviation.

3.2. Evaluation of Intervention in Clinical, Metabolic and Lipid Profiles

The effects of intervention on anthropometric, metabolic and lipid parameters as well as those related to glucose metabolism, according to study groups, in overall population and according to kidney function, are shown in Tables S4 and S5, respectively.

The trial's primary endpoint, HbA1c, did not differ significantly within or between groups over the 12‐month intervention in both the overall population and according eGFR categories at baseline (Tables S4 and S5).

In the overall population, consumption of OA‐enriched olive oil resulted in a significant reduction in both SBP and DBP compared to baseline (p = 0.010 and p = 0.018, respectively) (Table S4). When patients were stratified by baseline eGFR, OA‐enriched olive oil also significantly reduced SBP and DBP in both those with normal eGFR as well as in those with mildly impaired eGFR (all p < 0.05) (Table S5).

In contrast, control olive oil increased LDL‐cholesterol levels in the overall population (p = 0.020) and reduced DBP in both the total population and in patients with normal eGFR (all p < 0.05). Furthermore, OA‐enriched olive oil reduced fasting insulin levels exclusively in patients with mildly impaired eGFR, whereas control olive oil increased HbA1c levels in this subgroup of patients (all p < 0.05).

3.3. Effect of Intervention on Parameters Related to Kidney Function

We evaluated the impact of the OA‐enriched olive oil, compared to control olive oil, on different parameters of kidney function. In the overall population, consumption of OA‐enriched olive oil maintained unaltered eGFR whereas the control olive oil could not avoid the eGFR decline typically associated with the diabetic CKD progression, with an eGFR decline rate of −3.15 mL/min/1.73 m² from baseline (p = 0.025) (Figure 1A). In the case of patients with normal eGFR values, the intake of both olive oils did not affect this parameter either, and the baseline values were maintained throughout the intervention (Figure 1B). However, a differential effect of OA‐enriched olive oil was evident in patients with mildly impaired eGFR. A significant interaction between intervention group and baseline eGFR category was observed (p = 0.010), indicating that the effect of the intervention differed according to baseline CKD stage. In this subgroup, OA‐enriched olive oil significantly slowed eGFR progression, whereas eGFR was reduced after consumption of the control oil, with a progressive decline rate of −7.50 mL/min/1.73 m² from baseline (p = 0.015) (Figure 1C).

FIGURE 1.

Effect of intervention on eGFR. (A) Overall population (n = 87; 45 in OA‐enriched olive oil group and 42 in control olive oil group), (B) patients with normal eGFR > 90 mL/min/1.73 m² (n = 53; 28 in OA‐enriched olive oil group and 25 in control olive oil group) and (C) patients with mildly impaired eGFR: 60–90 mL/min/1.73 m² (n = 34; 17 in OA‐enriched olive oil group and 17 in control olive oil group). All data are mean ± standard error. Continuous variables were compared using repeated‐measures ANOVA adjusted for age, sex, BMI, smoking status and pharmacological treatments (use of ACEIs, ARBs, GLP1‐RAs, SGLT2is, diuretics and metformin dose). *Significant changes between pre‐ and post‐intervention (p < 0.05). ^#Significant differences between OA‐enriched olive oil versus control olive oil (p < 0.05). ACEIs, angiotensin‐converting enzyme inhibitors; ARBs, angiotensin II receptor blockers; BMI, body mass index; eGFR, estimated glomerular filtration rate; GLP‐1‐RAs, glucagon‐like peptide‐1 receptor agonists; OA, oleanolic acid; SGLT2is, sodium‐glucose co‐transporter 2 inhibitors.

Regarding other parameters related to kidney function, the consumption of both OA‐enriched and control olive oils did not significantly affect serum levels of creatinine, uric acid, urea, total protein or albumin, either in the overall population or when stratified by baseline eGFR (Tables S4 and S5). However, a significant increase in urine creatinine and uACR was observed after intake of control oil compared to baseline, both in the overall population and in patients with normal or mildly impaired baseline eGFR (all p < 0.05). This effect was not observed after consumption of OA‐enriched olive oil (Figure 2A–C).

FIGURE 2.

Effect of intervention on urine albumin‐to‐creatinine ratio. (A) Total population (n = 87; 45 in OA‐enriched olive oil group and 42 in control olive oil group), (B) patients with normal eGFR > 90 mL/min/1.73 m² (n = 53; 28 in OA‐enriched olive oil group and 25 in control olive oil group) and (C) patients with mildly impaired eGFR: 60–90 mL/min/1.73 m² (n = 34; 17 in OA‐enriched olive oil group and 17 in control olive oil group). All data are mean ± standard error. Continuous variables were compared using repeated‐measures ANOVA adjusted for age, sex, BMI, smoking status, and pharmacological treatments (use of ACEIs, ARBs, GLP1‐RAs, SGLT2is, diuretics and metformin dose). *Significant changes between pre‐ and post‐intervention (p < 0.05). ^#Significant differences between OA‐enriched olive oil versus control olive oil (p < 0.05). ACEIs, angiotensin‐converting enzyme inhibitors; ARBs, angiotensin II receptor blockers; BMI, body mass index; eGFR, estimated glomerular filtration rate; GLP‐1‐RAs, glucagon‐like peptide‐1 receptor agonists; OA, oleanolic acid; SGLT2is, sodium‐glucose co‐transporter 2 inhibitors.

3.4. OA‐Enriched Olive Oil Modulates Circulating AGE Levels and AGE Metabolism

To explore the association of AGE accumulation with impaired kidney function, we evaluated circulating AGE levels as well as AR concentration, involved in endogenous AGE synthesis, and GloxI, implicated in the clearance of these compounds, based on patients' baseline eGFR. In patients with normal eGFR, consumption of control olive oil led to an accumulation of circulating AGEs, exemplified by a significant increase in MG and CML levels (Figure 3A,C, respectively) after 12 months of intervention (all p < 0.05). This scenario corresponds to a lower rate of AGE detoxification, reflected by significantly reduced GloxI activity (p = 0.021) (Figure 3G). In contrast, consumption of OA‐enriched olive oil did not alter MG‐H1 and CML levels (Figure 3A,C, respectively). Since AR concentration did not change in these patients throughout the intervention, regardless of the type of olive oil consumed (Figure 3E), the maintenance of circulating CML and MG‐H1 levels in those patients who consumed OA‐enriched olive oil may be explained by increased GloxI activity (p = 0.011) (Figure 3G).

FIGURE 3.

Effect of intervention on circulating AGE levels and AGE metabolism according to baseline eGFR categories. (A, B) Serum levels of methylglyoxal in patients with normal eGFR > 90 mL/min/1.73 m² and patients with mildly impaired eGFR: 60–90 mL/min/1.73 m², (C, D) serum levels of Nε‐carboxymethyllysine in patients with normal eGFR > 90 mL/min/1.73 m² and patients with mildly impaired eGFR: 60–90 mL/min/1.73 m², (E, F) serum aldose reductase concentration in patients with normal eGFR > 90 mL/min/1.73 m² and patients with mildly impaired eGFR: 60–90 mL/min/1.73 m² and (G, H) serum glyoxalase I concentration in patients with normal eGFR > 90 mL/min/1.73 m² and patients with mildly impaired eGFR: 60–90 mL/min/1.73 m². Normal eGFR > 90 mL/min/1.73 m² (n = 53; 28 in OA‐enriched olive oil group and 25 in control olive oil group). Mildly impaired eGFR: 60–90 mL/min/1.73 m² (n = 34; 17 in OA‐enriched olive oil group and 17 in control olive oil group). All data are mean ± standard error. Continuous variables were compared using repeated‐measures ANOVA adjusted for age, sex, BMI, smoking status and pharmacological treatments (use of ACEIs, ARBs, GLP1‐RAs, SGLT2is, diuretics and metformin dose). *Significant changes between pre‐ and post‐intervention (p < 0.05). ^#Significant differences between OA‐enriched olive oil versus control olive oil (p < 0.05). ACEIs, angiotensin‐converting enzyme inhibitors; ARBs, angiotensin II receptor blockers; BMI, body mass index; eGFR: Estimated glomerular filtration rate; GLP‐1‐RAs, glucagon‐like peptide‐1 receptor agonists; SGLT2is, sodium‐glucose co‐transporter 2 inhibitors.

In patients with mildly impaired eGFR, consumption of OA‐enriched olive oil alleviated AGE formation after 12 months of intervention, with a significant decrease in serum MG‐H1 levels (Figure 3B). This effect of OA may be explained by an inhibition of AR concentration and an increase in GloxI activity (all p < 0.05) (Figure 3F,H, respectively). In contrast, control olive oil intake increased circulating levels of MG‐H1 (Figure 3B) and CML (Figure 3D) and reduced GloxI activity (Figure 3H), without affecting AR levels (all p < 0.05) (Figure 3F).

3.5. Effect of Intervention on Parameters Related to Oxidative Stress

Consumption of OA‐enriched olive oil for 12 months resulted in a significant decrease in circulating ROS levels in both patients with normal eGFR and those with moderately impaired eGFR, compared with baseline (all p < 0.05), whereas intake of control olive oil had no effect. Furthermore, among patients with moderately impaired eGFR, the reduction in ROS in the intervention group was significantly greater than that observed in the control group (p = 0.038).

Circulating MDA levels were lower after 12 months of OA‐enriched olive oil intake than after consumption of control olive oil only in individuals with normal eGFR (p = 0.025) (Table S4).

3.6. Safety Outcomes

Safety outcomes are presented in Table S6 Adverse events were comparable between groups, were predominantly mild to moderate in severity, and no withdrawals due to adverse effects or deaths were reported during the study.

4. Discussion

In this exploratory secondary analysis of the OLTRAD randomised controlled trial, 12‐month consumption of an OA‐enriched olive oil was associated with a slower eGFR decline and prevention of a uACR increase. These effects were particularly evident in patients with mildly impaired kidney function at baseline (eGFR 60–90 mL/min/1.73 m²), a clinically relevant subgroup in which early intervention may be especially effective. No such effect was observed among participants with normal baseline eGFR. Beyond clinical outcomes, OA modulated key pathways of AGE metabolism, reduced circulating AR levels, the rate‐limiting enzyme of the polyol pathway, and increased GloxI activity, a major detoxification system for reactive carbonyl species such as MG. These changes were accompanied by reduced circulating MG, while CML remained stable. This differential response is consistent with the role of MG as a highly reactive AGE precursor involved in renal injury [38], whereas CML reflects longer‐term systemic AGE burden [39]. Together, these findings suggest that modulation of AGE‐related pathways may be involved in the renal effects observed after OA supplementation. As kidney outcomes were prespecified secondary endpoints in a trial powered for glycemic control, these results should be regarded as hypothesis‐generating.

To our knowledge, no controlled clinical trials have evaluated the long‐term consumption of OA‐enriched olive oil as a strategy to modulate AGE metabolism and attenuate kidney function decline in patients with T2DM. Diabetes remains the leading cause of end‐stage renal disease, with CKD characterised by albuminuria and progressive eGFR decline and strongly associated with cardiovascular risk. Accumulating evidence supports a central role for AGE accumulation in the pathogenesis of diabetic kidney disease by promoting structural and functional renal damage [40].

Our results are consistent with large dietary intervention studies, including PREDIMED‐Plus and CORDIOPREV, in which Mediterranean diets rich in extra virgin olive oil were associated with slower eGFR decline, particularly among individuals with mildly impaired baseline kidney function [29, 41]. Importantly, this effect was not observed in low‐fat dietary interventions, highlighting the relevance of dietary fat quality and olive‐derived bioactive compounds. However, it is important to consider that the beneficial effects of olive oil within the Mediterranean diet have been primarily attributed to extra virgin or virgin olive oil, which is rich in minor bioactive compounds. In contrast, the control olive oil used in the present study had a lower content of these compounds due to the refining process, which may explain the lack of beneficial effects observed in this group and does not contradict previous findings. In the present work, OA was delivered as a functional olive oil with demonstrated bioavailability [25], providing a targeted nutritional approach within the Mediterranean dietary pattern. Its antioxidant, anti‐inflammatory and metabolic effects may jointly contribute to the observed renal benefits [15, 16, 42].

Preclinical studies have previously demonstrated nephroprotective effect of OA in experimental models of kidney disease [43], showing improvements in kidney function and oxidative stress markers [23, 44]. However, clinical evidence in humans has been limited. In contrast, synthetic triterpene derivatives such as bardoxolone methyl have shown inconsistent and sometimes adverse effects in CKD patients with T2DM [45, 46]. In the present study, the 12‐month consumption of OA‐enriched olive oil was well tolerated, with a safety profile comparable to that of the control group. Most adverse events were mild to moderate in severity, no withdrawals due to adverse effects were reported, and no deaths occurred during follow‐up, supporting the favourable safety profile of OA supplementation in this population [47].

Diabetes‐associated kidney function decline has been estimated at approximately 2 mL/min/1.73 m² per year [47]. Our data suggest that OA may attenuate this decline, limiting eGFR reduction to less than 1 mL/min/1.73 m² over 12 months. This effect may be partially mediated by improvements in cardiometabolic factors, including reductions in blood pressure and insulin levels, particularly in patients with mildly impaired eGFR. The greater decline in eGFR observed among participants with mildly impaired baseline kidney function receiving the control olive oil may reflect the well‐recognised heterogeneity of renal disease progression in T2DM, where a subset of individuals experiences accelerated kidney function loss. Importantly, this observation was accompanied by concordant changes in albuminuria and AGE‐related biomarkers, supporting the overall consistency of the renal findings.

Although all participants had normal baseline uACR (< 30 mg/g), individuals in the control group developed microalbuminuria after 12 months, whereas those receiving OA maintained normal levels. This finding is consistent with a potential protective effect at early stages of kidney dysfunction.

Beyond traditional risk factors, AGE accumulation is increasingly recognised as a contributor to CKD progression. AGEs promote oxidative stress, inflammation and structural renal damage [48]. The inverse relationship between AGE levels and kidney function suggests an important role for renal clearance [49]. In our study, higher baseline AGE levels in patients with mildly impaired eGFR support this bidirectional relationship. These findings are in line with previous work from our group and others showing that modulation of AGE metabolism, particularly reductions in MG levels, is associated with preservation of kidney function in patients with T2DM [8, 9, 10, 50]. OA may reduce AGE burden through inhibition of the polyol pathway and enhancement of detoxification systems [21], which may contribute to reduced AGE accumulation and kidney protection. However, the precise contribution of this pathway to CKD progression and to the renal effects observed with OA requires further investigation [51, 52]. In addition, recent experimental studies suggest that OA may exert renoprotective effects through complementary mechanisms involved in diabetic kidney disease, including the attenuation of podocyte injury [20].

Oxidative stress represents a key link between AGE accumulation and kidney damage. Reactive carbonyl species such as MG promote ROS generation, exacerbating renal injury. The reduction in ROS and modulation of MDA observed with OA further support its role in attenuating oxidative and lipid peroxidation damage [53]. These effects were more evident in early CKD stages, suggesting greater responsiveness to nutritional intervention.

Nevertheless, CKD progression is increasingly recognised as a multifactorial process involving immune–metabolic crosstalk within the renal microenvironment, where interactions among metabolic disturbances, inflammatory pathways, immune cells, and resident renal cells contribute to tissue injury and disease progression [54]. This concept is consistent with emerging systems‐level approaches, which suggest that nutritional interventions may simultaneously affect metabolic, inflammatory, and cellular networks involved in kidney disease progression [55]. In addition, although not evaluated in the present study, the gut–kidney axis may represent another relevant mechanism, as dietary bioactive compounds can modulate gut microbiota composition and metabolite production, thereby influencing host pathways associated with renal health [56, 57]. Future studies incorporating emerging biomarkers of tubular injury, inflammation, and metabolic dysfunction may help to further elucidate the biological pathways underlying the renoprotective effects of OA.

The OA formulation used in this study consisted of a functional olive oil with demonstrated bioavailability [27], along with its safety and efficacy in preventing the progression from prediabetes to T2DM [42]. In line with its pharmacokinetic profile [25], including a short half‐life (t _½≈5.5 h), rapid return to baseline concentrations within 10–16 h after ingestion, and a large volume of distribution (V _d = 48 L), fasting serum OA concentrations are expected to remain close to baseline throughout follow‐up regardless of treatment allocation, reflecting rapid systemic clearance and extensive tissue distribution rather than the absence of intake. Importantly, this pattern suggests that OA fasting serum OA concentrations may underestimate chronic exposure and do not adequately reflect sustained intake. This is consistent with population‐based data showing only modest increases in circulating triterpene levels, typically within the low ng/mL range [25, 58].

This study has several limitations. Its single‐centre design and relatively small sample size may limit generalizability of the findings. Kidney function was assessed using eGFR and single spot uACR rather than gold‐standard methods such as inulin clearance or 24‐h urine collection, although these approaches are impractical in routine clinical settings. In addition, the study was powered to detect differences in the primary endpoint (HbA1c), whereas renal outcomes were prespecified secondary endpoints; therefore, the findings related to kidney function should be interpreted with caution. Analyses of renal outcomes were performed as an available‐case analysis, restricted to participants with available follow‐up data, without imputation of missing values; this approach, rather than a strict intention‐to‐treat framework, may further limit the interpretation of these findings. Furthermore, although all statistical analyses were adjusted for the use of relevant pharmacological treatments, including SGLT2 inhibitors, GLP‐1 receptor agonists, ACEIs, ARBs, and diuretics, potential interactions between OA supplementation and concomitant medications cannot be completely excluded. Although the intervention involved an increased intake of olive oil (~55 mL/day), the caloric contribution from olive oil was identical in both study groups. Moreover, dietary intake was assessed using validated tools and no differences were observed between groups; adherence to specific dietary patterns, such as the Mediterranean diet, was not formally evaluated using a standardised score. Therefore, while the observed effects are unlikely to be explained solely by differences in energy intake, a residual influence of unmeasured dietary factors cannot be entirely ruled out. In addition, although adherence was monitored using validated questionnaires and the return of empty oil containers, and serum OA was measured as an objective biomarker, the rapid pharmacokinetics of OA (short half‐life and non‐cumulative circulating behaviour) combined with reliance on single fasting blood samples precluded a robust quantitative comparison of systemic OA exposure between the intervention and control arms over time. Therefore, these measurements alone do not allow a precise quantification of the difference in systemic OA exposure between arms over time. Self‐reported dietary data remain subject to recall bias. Finally, the study population consisted exclusively of overweight or obese adults with T2DM from southern Spain, which may limit the extrapolation of these findings to other populations.

This study has important strengths. The randomised, double‐blind design with two intervention arms and an identical control oil differing only in OA content allowed for an isolated evaluation of OA effects. Moreover, unlike studies relying on skin autofluorescence, we quantified circulating AGEs and enzymes directly involved in AGE metabolism using validated ELISA‐based methods, enabling a more comprehensive assessment of AGE‐related pathways [59, 60].

In summary, long‐term consumption of OA‐enriched olive oil appears safe and, in this exploratory secondary analysis, was associated with attenuated kidney function decline in patients with T2DM, an effect limited to those with early stages of impairment. These effects are accompanied by modulation of AGE‐related pathways and oxidative stress, suggesting a potential association between AGE detoxification and renal protection. This food‐based strategy may represent a promising avenue for further investigation in early renal dysfunction in T2DM; larger, adequately powered multicentre trials with kidney function as a primary endpoint are needed to confirm these hypothesis‐generating findings.

Author Contributions

Conceptualisation: Elena María Yubero‐Serrano and José María Castellano. Methodology: Aída García‐González, Antonio Jesús Martínez‐Ortega, Elena María Yubero‐Serrano, Pedro Pablo García‐Luna and José María Castellano. Investigation: Alejandro Matamoros‐Domínguez, Aída García‐González, Juan Manuel Espinosa‐Cabello, Gisela Pérez‐Muñoz, Manuel Jesús León Aranda, María Del Carmen Roque‐Cuellar, Silvia García‐Rey, Andrés Jiménez‐Sánchez, Antonio Jesús Martínez‐Ortega, Miguel Ángel Mangas‐Cruz, Isabel Cerrillo, Javier S. Perona, Elena María Yubero‐Serrano, Pedro Pablo García‐Luna and José María Castellano. Resources: Alejandro Matamoros‐Domínguez, Aída García‐González, Juan Manuel Espinosa‐Cabello, Gisela Pérez‐Muñoz, María Del Carmen Roque‐Cuellar, Silvia García‐Rey, Andrés Jiménez‐Sánchez, Antonio Jesús Martínez‐Ortega, Miguel Ángel Mangas‐Cruz, Isabel Cerrillo, Javier S. Perona, Elena María Yubero‐Serrano, Pedro Pablo García‐Luna and José María Castellano. Formal analysis: Alejandro Matamoros‐Domínguez, Andrés Jiménez‐Sánchez, Antonio Jesús Martínez‐Ortega, Manuel Jesús León Aranda, Elena María Yubero‐Serrano, Javier S. Perona, Pedro Pablo García‐Luna and José María Castellano. Validation: Elena María Yubero‐Serrano, Javier S. Perona and José María Castellano. Data curating: Elena María Yubero‐Serrano and José María Castellano. Visualisation: Javier S. Perona, Elena María Yubero‐Serrano and José María Castellano. Writing – original draft preparation: Elena María Yubero‐Serrano, Alejandro Matamoros‐Domínguez, Aída García‐González and José María Castellano. Writing – review and editing: Andrés Jiménez‐Sánchez, Antonio Jesús Martínez‐Ortega, Miguel Ángel Mangas‐Cruz, Isabel Cerrillo, Javier S. Perona, Elena María Yubero‐Serrano, Pedro Pablo García‐Luna and José María Castellano. Supervision: Elena María Yubero‐Serrano, Pedro Pablo García‐Luna and José María Castellano. Funding acquisition: José María Castellano. Project administration: Pedro Pablo García‐Luna and José María Castellano. All the authors have read and agreed to the published version of the manuscript.

Funding

This research is part of the R+D+i project PID2019‐107837RB‐I00, funded by the Spanish Ministry of Science and Innovation/Spanish State Research Agency (Agencia Estatal de Investigación, AEI) (MCIN/AEI/10.13039/501100011033/). A.M.‐D. is grateful for funding received from the Andalusian Regional Government/Regional Ministry of University, Research and Innovation and the European Social Fund Plus (ESF+) (DGP_PRED_2024_00736). A.G.‐G. is grateful for funding received from the European Union through the Next Generation EU funds and the Recovery, Transformation and Resilience Plan, and by the Ministry of Universities, in the framework of the Margarita Salas, Maria Zambrano grants for the Requalification of the Spanish University System 2021–2023, organised by the Pablo de Olavide University, Seville.

Conflicts of Interest

The OA‐enriched olive oil used in this study was developed under the European patent EP18382510. J.M.C. is listed as an inventor on this patent. The rights to this patent are held by ACESUR Group (Dos Hermanas, Seville, Spain), which also supplied the commercial olive oil used as the control oil in this study, free of charge. As a named inventor, J.M.C. holds a personal interest in the intellectual property covering the tested product, which constitutes a competing interest in the context of this study. The authors have not received direct financial compensation, licensing revenue, or other commercial benefit related to this patent to date. ACESUR Group had no role in the study design, data collection, analysis, interpretation of the data, or the decision to publish the results. [Correction added on 21 August 2026, after first online publication: All instances of ‘J.M.E.‐C.’ have been changed to ‘J.M.C.’.]

Supporting information

Figure S1: Supporting Information.

Table S1: Inclusion and exclusion criteria.

Table S2: Chemical composition of the olive oils used in the study.

Table S3: Baseline metabolic comparability between patients included and not included in the follow‐up kidney function analysis [1].

Table S4: Effect of intervention on clinical, metabolic characteristics and lipid profile.^†

Table S5: Effect of intervention on clinical and metabolic characteristics (∆ changes produced between post and pre‐intervention) according to kidney function.^†

Table S6: Adverse events by study group in participants included in the kidney‐related analysis [1].

García‐González A., Matamoros‐Domínguez A., Espinosa‐Cabello J. M., et al., “Oleanolic Acid‐Enriched Olive Oil Attenuates Kidney Function Decline in Patients With Type 2 Diabetes: Results From the OLTRAD Randomised Controlled Trial,” Diabetes, Obesity and Metabolism 28, no. 11 (2026): 10195–10209, 10.1111/dom.71206.

Data Availability Statement

The datasets generated during the trial are available in DIGITAL CSIC repository (https://doi.org/10.20350/digitalCSIC/18374) upon reasoned request to the corresponding author.