All animal experiments were performed in adherence to international guidelines on animal experimentation and were ethically approved by the Institutional Animal Use and Care Committee at the University of Eastern Africa, Baraton. All animals were treated humanely throughout the experiment and had unrestricted access to standard rodent pellets and water.

The bark of Vernoniaauriculifera was harvested from Eldoret, Uasin Gishu County, Kenya (0.5206˚N, 35.2698˚E). authenticated by a botanist at the University of Eldoret, and a voucher specimen deposited in the University of Eldoret Herbarium, voucher number; UOE/BR/2025/15. The bark was air-dried away from direct sunlight. The dried plant material was ground into a fine powder. Sequential cold maceration was performed for 72 hours at a bark-to-solvent ratio of 1:10 (w/v) for each solvent (methanol, ethanol, and distilled water). The extracts were filtered, evaporated to dryness under reduced pressure using a rotary evaporator, and the percentage extraction yields calculated (methanol 18.4%; ethanol:14.7%; aqueous: 10.2%). The dried extracts were stored at 4˚C until further use. For invivo dosing, extracts were dissolved in distilled water and administered orally by gavage at a volume of 10 mL/kg body weight.

Qualitative phytochemical analyses for secondary metabolites—including flavonoids, alkaloids, tannins, glycosides, terpenoids, steroids, saponins, and anthraquinones—were performed on all three extracts using established conventional methods [10][11].

The ferric reducing antioxidant power (FRAP) assay was evaluated spectrophotometrically at 700 nm across five concentrations (100 - 500 μg/mL), with ascorbic acid as the reference standard. Hydrogen peroxide (H₂O₂) scavenging activity was determined at 230 nm at three concentrations (20, 40, and 62.5 μg/mL).

Twenty-five male Wistar rats (body weight 180 - 220 g) were randomly assigned to five groups of five animals each using simple randomization (random number table): Group I—Normal Control (NC); Group II—Diabetic Control (DC); Group III—Diabetic + 200 mg/kg body weight methanol extract (DL); Group IV—Diabetic + 400 mg/kg body weight methanol extract (DH); and Group V—Diabetic + 100 mg/kg body weight metformin (MET). Diabetes was induced by a single intraperitoneal injection of STZ (60 mg/kg body weight) dissolved in 0.1 M citrate buffer (pH 4.5). Seventy-two hours after STZ injection, tail-vein blood glucose was measured following an overnight fast (12 hours); rats with fasting blood glucose ≥ 200 mg/dL were confirmed diabetic and included in the study. All 25 rats completed the study; no animals were excluded after induction. Treatment began on Day 0 (the day after diabetes confirmation) and continued for 21 days (Weeks 1 - 3 corresponding to Days 1 - 7, 8 - 14, and 15 - 21 respectively; Week 4 measurements were taken at Day 21). The sample size of five per group was based on prior similar studies and was consistent with ethical minimisation guidelines; a formal power analysis was not performed.

Blood glucose concentrations were measured weekly using a glucometer. Body weights were recorded on Days 0, 7, 14, and 21. On Day 22 (the day following treatment completion), all animals were euthanised by lethal anaesthetic injection and blood samples were collected by cardiac puncture. Serum was separated by centrifugation and used for the determination of lipid profile (total cholesterol, triglycerides, HDL-C, LDL-C), liver function tests (GGT, AST, ALT, ALP, total protein, albumin, bilirubin), and kidney function tests (urea, creatinine) using standard enzymatic colorimetric reagents. Hepatic tissue malondialdehyde (MDA) content, an index of lipid peroxidation, was determined using the thiobarbituric acid reactive substances (TBARS) assay [12].

Data are expressed as mean ± standard error of the mean (SEM). Statistical comparisons were made using one-way ANOVA followed by Tukey’s post hoc test. A p-value of <0.05 was considered statistically significant. Analyses were performed using Microsoft Excel and R software. One-way ANOVA was applied independently at each time point for the weekly glucose and body weight data; while a repeated-measures ANOVA would be more appropriate for within-subject longitudinal comparisons, the one-way approach was used here due to the exploratory nature of the study. Future studies should employ repeated-measures ANOVA or mixed-effects models to better account for the correlated structure of repeated observations from the same animals.

Qualitative phytochemical analysis revealed that the methanol extract contained the greatest diversity of secondary metabolites—flavonoids, alkaloids, tannins, glycosides, terpenoids, anthraquinones, and saponins—but lacked steroids. The aqueous extract was devoid of glycosides and terpenoids but contained steroids. The ethanol extract profile closely resembled that of the methanol extract, also lacking steroids (Table 1).

Table 1. Phytochemical composition of Vernoniaauriculifera bark extracts.

+: Present; −: Absent.

3.2.1. Ferric Reducing Power

All three extracts demonstrated dose-dependent ferric reducing activity across concentrations of 100 - 500 μg/mL (Table 2). The methanol extract exhibited reducing capacity (OD = 1.716 - 1.764) closely comparable to ascorbic acid (OD = 1.727 - 1.805), followed by the ethanol extract (OD = 1.660 - 1.752).

Table 2. Ferric reducing power (Absorbance at 700 nm) of V.auriculifera bark extracts.

Values represent mean absorbance at 700 nm.

3.2.2. Hydrogen Peroxide Scavenging Activity

All extracts demonstrated concentration-dependent H₂O₂ scavenging activity (Figure 1Figure 1). At 20 μg/mL, the methanol extract scavenged 60% and the aqueous extract scavenged 45% of H₂O₂, compared to 85% for ascorbic acid. At 40 μg/mL, these values increased to 75%, 60%, and 90%, respectively (P < 0.05), confirming significant antioxidant efficacy.

All diabetic groups exhibited baseline blood glucose concentrations between 277 - 283 mg/dL following STZ induction, confirming successful diabetes induction. Treatment with the 400 mg/kg extract produced the greatest hypoglycaemic effect, reducing blood glucose from 280 ± 10 mg/dL at baseline to 97 ± 7 mg/dL at Week 4 (P < 0.01), which was lower than that achieved with metformin (111 ± 9 mg/dL) (Table 3).

Figure 1. Hydrogen peroxide scavenging activity (%) of methanol, aqueous bark extracts of V.auriculifera, and ascorbic acid at different concentrations.

Table 3. Blood glucose levels (mg/dL) across treatment groups.

*P < 0.01 vs. Diabetic Control. Values expressed as mean ± SEM.

The normal control group showed progressive weight gain (+9%) over the experimental period. All diabetic groups experienced weight loss relative to Day 0: the diabetic control (−13%), the 200 mg/kg extract group (−13%), and the 400 mg/kg extract group (−15%). The metformin-treated group showed the most favourable weight outcome among the diabetic groups (−9%) (Table 4).

In liver tissues, the normal control group exhibited the lowest absorbance (approximately 0.20), whereas diabetic control showed a slight increase (~0.23). Administration of the 200 mg/kg extract raised absorbance to ~0.41, and the 400 mg/kg extract to ~0.69. The metformin group recorded the highest liver tissue absorbance (~0.72). Plasma absorbance values were consistently higher than liver tissue values across all groups. The normal control plasma absorbance was approximately 0.75, which declined in the diabetic control (~0.61). The 200 mg/kg extract restored plasma absorbance to ~0.79, the 400 mg/kg extract to ~0.91, and metformin yielded the highest value (~1.00) (Figure 2Figure 2).

Table 4. Effects of Vernoniaauriculifera bark extract on fasting body weight (FBW) of streptozotocin-induced diabetic rats over 21 days.

Values expressed as mean ± SEM; a = significantly different from normal control (P < 0.05).

Figure 2.Absorbance of liver tissue and blood plasma ferric-reducing capacity across treatment groups in streptozotocin-induced diabetic Wistar rats treated with V.auriculifera bark extract and metformin.

Diabetes induction significantly elevated liver enzyme activities. Treatment with the 200 mg/kg extract produced the greatest reduction in GGT (3.85 ± 0.51 u/L vs. 11.38 ± 0.8 u/L in diabetic control), ALT (107.47 ± 5.58 u/L), and ALP (192.03 ± 14.85 u/L). Albumin levels were most improved in the 200 mg/kg extract group (35.54 ± 1.83 g/L vs. 27.81 ± 1.1 g/L in diabetic control). Serum urea was significantly reduced in the 200 mg/kg extract group and the metformin group; however, urea levels were paradoxically elevated in the 400 mg/kg extract group (Table 5).

Diabetes induction elevated triglycerides (1.37 vs. 0.93 mmol/L in normal control) and LDL-C (0.32 vs. 0.20 mmol/L), and reduced HDL-C (0.61 vs. 0.77 mmol/L). The 400 mg/kg extract normalised triglycerides (1.07 mmol/L) and HDL-C (0.75 mmol/L, approaching normal control values), and reduced LDL-C (0.24 mmol/L). The 200 mg/kg extract unexpectedly elevated triglycerides and total cholesterol (Table 6).

Table 5. Liver and kidney function biomarkers across treatment groups.

DL: Low-dose extract; DH: High-dose extract. Values expressed as mean ± SEM.

Table 6. Serum lipid profile across treatment groups.

*P < 0.05 vs. Diabetic Control; ^†P < 0.05 vs. Diabetic Control. DL: Low-dose; DH: High-dose. Values expressed as mean ± SEM.

Diabetic control rats exhibited the highest hepatic MDA levels (~2.95 μM), reflecting severe oxidative stress. Treatment with the 400 mg/kg extract reduced MDA to ~1.53 μM, slightly below the normal control (~1.65 μM) and metformin-treated values (~1.68 μM), indicating superior in vivo antioxidant activity at this dose (Figure 3Figure 3).

Figure 3.Effects of Vernoniaauriculifera bark extract on lipid peroxidation (MDA levels) in liver tissues of streptozotocin-induced diabetic Wistar rats.