This systematic review was registered on PROSPERO (number is 1233026) and was conducted in accordance with the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA 2020) guidelines (12). The review aimed to synthesize both preclinical and clinical evidence on the synergistic and antagonistic interactions between essential micronutrients and biguanides, particularly metformin, and their effects on neurocognitive outcomes in type 2 diabetes mellitus (T2DM). Both human and animal studies were included to capture mechanistic, physiological, and translational perspectives.

A comprehensive literature search was conducted across three electronic databases mainly, PubMed (MEDLINE), Web of Science, and Scopus. The search covered studies published between January 2010 and August 2025. The search strategy combined Medical Subject Headings (MeSH) and free-text terms using Boolean operators to identify studies that examined the combined or individual effects of metformin and micronutrients such as zinc, magnesium, or chromium on cognitive function, memory, or neurobiological mechanisms in T2DM. The PubMed search syntax was structured as follows and adapted appropriately for the other databases: ((“Metformin”[Mesh] OR “metformin”[tiab] OR “biguanides”[Mesh] OR “biguanide”[tiab])) AND ((“Zinc”[Mesh] OR “zinc”[tiab] OR “Zn”[tiab]) OR (“Magnesium”[Mesh] OR “magnesium”[tiab] OR “Mg”[tiab]) OR (“Chromium”[Mesh] OR “chromium”[tiab] OR “Cr”[tiab])) AND ((“Diabetes Mellitus, Type 2”[Mesh] OR “type 2 diabetes mellitus”[tiab] OR “T2DM”[tiab])) AND ((“Cognition”[Mesh] OR “cognitive function”[tiab] OR “learning”[Mesh] OR “memory”[Mesh] OR “executive function”[Mesh] OR “dementia”[Mesh] OR “cognitive decline”[tiab] OR “Alzheimer’s”[tiab])) AND ((“oxidative stress”[Mesh] OR “neuroinflammation”[tiab] OR “inflammatory cytokines”[tiab] OR “AMPK”[tiab] OR “brain insulin resistance”[tiab])) AND ((“animal experiment”[tiab] OR “preclinical study”[tiab] OR “clinical trial”[Publication Type] OR “human study”[tiab])) with filters applied for English language and publication date between 2010 and 2025. Equivalent search strings were customized for Web of Science and Scopus. The retrieved results were exported from each database as comma-separated value (CSV) files for further management and screening.

The CSV files were imported into Rayyan, a web-based systematic review management platform, to facilitate the screening process. Screening was conducted independently by two reviewers following a two-step process with inter-reviewer agreement Cohen’s Kapa of 0.78. The first phase involved title and abstract screening to identify potentially relevant studies, after which 56 records were retained for full-text review and 170 were excluded based on irrelevance, duplication, or inadequate data. Eight conflicts were identified and resolved through discussion, leading to four additional exclusions. The second phase involved full-text screening to confirm eligibility according to predefined inclusion and exclusion criteria. Of the 52 full-text articles reviewed, 40 met the eligibility criteria and were included in the final synthesis. The entire selection process was documented using a PRISMA 2020 flow diagram as depicted in Figure 1, indicating the number of studies identified, screened, excluded, and included, along with reasons for exclusion.

The population comprised adults aged 18 years and above with type 2 diabetes mellitus in clinical studies, and animal models of T2DM in preclinical investigations, including high-fat diet, streptozotocin-induced, or genetically modified models. The intervention included any form of zinc, magnesium, or chromium supplementation administered orally, parenterally, or through dietary enrichment, either alone or in combination with metformin or other biguanides. The comparator consisted of placebo, no supplementation, or standard of care (metformin monotherapy). The primary outcome was cognitive performance measured using validated tools such as the Mini-Mental State Examination (MMSE), Montreal Cognitive Assessment (MoCA), Y-Maze, or Morris Water Maze. Secondary outcomes included oxidative stress markers such as SOD, GSH, and MDA; inflammatory cytokines including IL-6, TNF-α, and CRP; insulin signaling molecules such as AMPK, PI3K/Akt, and GLUT4; and neuroplasticity indicators including BDNF, PSD-95, and SIRT1. Eligible study designs comprised randomized controlled trials, quasi-experimental studies, cohort and case-control designs for clinical data, and controlled animal studies for mechanistic exploration. Only full-text articles published in English with clear methodological detail were included, while reviews, case reports, editorials, and conference abstracts were excluded.

Data extraction was conducted independently by two reviewers (Physiologist and Pharmacologist) using a predesigned Microsoft Excel form. Extracted variables included study title, authors, publication year, country, design, sample size, intervention type and dosage, comparator characteristics, duration, route of administration, key outcomes and mechanistic pathways. Discrepancies between reviewers were resolved through discussion and consensus. The final extraction file served as the foundation for descriptive and thematic synthesis.

Given the heterogeneity of the included studies in design, intervention type, and reported outcomes, data were synthesized descriptively rather than quantitatively. Studies were grouped as clinical or preclinical and analyzed based on micronutrient type, dosage, and duration. The synthesis focused on identifying patterns of synergistic and antagonistic effects, common mechanistic pathways, and consistency of findings across study types.

Mechanistic findings were analyzed thematically to identify shared biological pathways linking micronutrient supplementation and metformin interaction to neurocognitive outcomes. Major mechanistic domains included antioxidant and anti-inflammatory regulation, modulation of insulin signaling and glucose metabolism, enhancement of synaptic plasticity and neurogenesis, and reduction of amyloid and tau pathology. Studies that did not include direct behavioral or neuropsychological testing but evaluated cognition-relevant neurobiological markers or signaling pathways were classified as indirect cognitive inference, whereas studies lacking both cognitive testing and cognition-linked biomarkers were considered non-cognitive assessments. These thematic outcomes were integrated into a conceptual framework illustrating the biological interplay between micronutrients, metformin, and neurocognitive function in T2DM.

Study selection and data extraction were verified independently by two reviewers to ensure accuracy, transparency, and consistency throughout the review process. The use of Rayyan software facilitated blinded screening, conflict identification, and systematic documentation of inclusion and exclusion decisions, which helped minimize potential reviewer bias. The platform enhanced reproducibility by maintaining a transparent record of all decisions and version histories; however, it does not provide a formal scoring system for methodological quality or risk of bias assessment. To maintain rigor, the reproducibility of the search strategy and screening steps was further ensured by PRISMA flow records for audit and validation purposes.

The synthesis comprised 40 studies published between 2013 and 2025, investigating the interactions between micronutrients and metformin in both clinical and preclinical settings. Of these, 27 studies (67.5%) were preclinical, employing Wistar rats, C57BL/6 mice, and STZ- or HFD-induced diabetic models to explore neuroprotective, metabolic, and cognitive outcomes, while 13 studies (32.5%) were clinical, including randomized controlled trials and observational analyses involving adults with type 2 diabetes, prediabetes, zinc deficiency, vascular dementia, or diabetic nephropathy. Chronologically, the research output showed a steady rise over the years: 2013 (1 study; 2.5%), 2016–2017 (2 each; 5%), 2018 (3; 7.5%), 2019–2020 (1 each; 2.5%), 2021–2022 (4 each; 10%), 2023 (10; 25%), and 2024–2025 (6 each; 15%). The peak publication year was 2023, marked by an increase in mechanistic preclinical studies focusing on antioxidant modulation, insulin signaling, and neuroplasticity mechanisms, reflecting growing global interest in nutrient–drug synergy research.

The studies utilized a range of interventions, including pharmacological agents (metformin, vildagliptin, sildenafil, memantine), micronutrients (zinc, magnesium, vitamins D and E, omega-3 fatty acids), phytochemicals (Ocimum gratissimum, resveratrol, tocotrienols, Ajwa seed extract), and environmental interventions such as enrichment therapy as summarized in `Table 1. Study durations ranged from 11 days to 24 weeks, with most clinical trials lasting 8–18 weeks. Geographically, research was distributed across several regions: China (9 studies; 22.5%), India (8; 20%), Egypt (5; 12.5%), Saudi Arabia (3; 7.5%), Malaysia (3; 7.5%), Australia (3; 7.5%), Brazil (2; 5%), and Nigeria, Mexico, Pakistan, Turkey, Iran, Indonesia, Korea, and Canada (1 each; 2.5%).

Overall, Asian countries contributed approximately 67% of all included studies, indicating their strong leadership in micronutrient–metformin interaction research. Across all studies, commonly assessed biochemical and molecular parameters included glycemic indices such as fasting glucose, HbA1c, and insulin sensitivity; oxidative stress biomarkers including superoxide dismutase (SOD), glutathione peroxidase (GPx), and malondialdehyde (MDA); and neuroplasticity markers such as brain-derived neurotrophic factor (BDNF), postsynaptic density protein-95 (PSD-95), and mitogen-activated protein kinase (MAPK). The most frequently implicated molecular pathways included AMPK, SIRT1, PI3K/Akt, and MAPK, which together mediate metabolic regulation, antioxidative defense, and cognitive improvement.

A total of 40 eligible studies were reviewed to determine the prevalence and distribution of micronutrient supplementation in relation to neurocognitive outcomes among individuals or experimental models of type 2 diabetes mellitus (T2DM), as summarized in Figure 2. The figure depicts the overall frequency with which individual micronutrients were investigated across studies, irrespective of whether they were administered as standalone interventions or in combination with antidiabetic pharmacotherapy. Vitamin D was the most frequently investigated micronutrient, accounting for 20% of all reported supplements. Notably, vitamin D was commonly evaluated in combination with metformin, particularly in preclinical and clinical studies examining synergistic effects on insulin signaling, oxidative stress reduction, and neuroplasticity (18, 39, 44). Zinc and iron each represented 16% of reported micronutrients. These trace elements were predominantly studied as standalone interventions or as part of trace-element profiling, rather than consistently combined with metformin, and were associated with modulation of PI3K/Akt signaling, antioxidant defense, and metabolic regulation (15, 27). Sulfate-based compounds accounted for 12% of reported interventions, largely reflecting zinc sulfate and magnesium sulfate formulations used primarily as monotherapy in experimental models (24, 42). Antioxidant vitamins including vitamin E, vitamin B₁₂, vitamin A, and vitamin C— accounted for 24% of all reported micronutrients. These compounds were mostly evaluated as independent supplements or within multinutrient formulations, with only limited studies assessing their concurrent use with metformin (23, 33, 35). Trace minerals such as magnesium, calcium, and copper each appeared in 4% of the included studies. These elements were rarely used as isolated interventions and were more commonly incorporated as adjunctive components within broader micronutrient or antioxidant formulations, or assessed in relation to metabolic or autonomic function rather than direct glucose-lowering effects (24, 47).

Vitamin D (20%) was the most frequently reported micronutrient, followed by Zinc (16%), Iron (16%), and Sulfate (12%). Antioxidant vitamins E (8%), B12 (8%), A (4%), and C (4%) collectively represented 24% of the total micronutrients. Trace minerals, including Magnesium (4%), Calcium (4%), and Copper (4%), were less frequently reported. Overall, Vitamin D, Zinc, and Iron together accounted for more than 50% of the micronutrient interventions investigated.

Across the 40 studies reviewed, five main categories of compounds were identified with vitamin and mineral formulations, phytochemical extracts, pharmacological agents, nanoparticle-based systems, and mixed vitamin preparations as summarized in Table 2. Vitamin and mineral interventions were the most frequent, representing 35% of all studies. Compounds such as vitamin D₃, zinc sulfate, and magnesium sulfate were commonly used and consistently associated with improved glycemic control, enhanced BDNF expression, and reduced inflammatory cytokines including TNF-α and IL-6. Phytochemical compounds accounted for 27.5% of the studies, with extracts such as Ocimum gratissimum, Nigella sativa, and Tiliacora triandra showing antioxidant, anti-inflammatory, and insulin-sensitizing effects mainly through NF-κB and AMPK pathway modulation.

Pharmacological agents and combination therapies made up 17.5% of the studies, primarily involving metformin, glimepiride, and alpha-lipoic acid, which demonstrated synergistic neuroprotective outcomes when used with vitamins or phytochemicals. Nanoparticle-based formulations comprised 15% of the interventions, including curcumin, zinc oxide, and berberine liposomes, which improved bioavailability and brain delivery efficiency. The remaining 5% involved mixed vitamin and fatty acid formulations such as vitamin E combined with omega-3 fatty acids, showing benefits in neuronal membrane integrity and signaling. Overall, the evidence indicates a growing emphasis on integrated therapeutic strategies that combine micronutrients, phytochemicals, and metabolic drugs to address oxidative stress, inflammation, and cognitive decline in type 2 diabetes mellitus.

Distribution of chemical forms and compounds identified in 40 preclinical and clinical studies investigating neurocognitive outcomes in type 2 diabetes mellitus. Vitamin and mineral formulations represented 35% of all interventions, followed by phytochemical extracts (27.5%), pharmacological agents (17.5%), nanoparticle systems (15%), and mixed vitamin formulations (5%). The predominance of Vitamin D₃, Zinc sulfate, and flavonoid-rich extracts underscores a translational focus on antioxidant and insulin-sensitizing mechanisms.

The analysis of micronutrient dose distribution across studies on neurocognitive outcomes in type 2 diabetes revealed diverse experimental and clinical dosing strategies as summarized in Table 3. Preclinical oral doses were the most frequently applied (30%), typically ranging between 10–400 mg/kg/day, demonstrating optimal glycemic and neuroprotective responses in animal models. High-dose antioxidant and polyphenol interventions accounted for 15% of studies, highlighting compounds such as resveratrol, curcumin, tocotrienols, and alpha-lipoic acid for their potent oxidative stress reducing and anti-inflammatory effects.

Approximately 12.5% of interventions used human-equivalent oral doses within clinically safe ranges (e.g., vitamin D₃, vitamin E, zinc sulfate, and carnosine), showing consistent improvement in metabolic and cognitive parameters. Nanoparticle and liposomal formulations (7.5%) improved bioavailability and produced synergistic outcomes when combined with standard antidiabetic drugs, while trace mineral combinations (7.5%) notably zinc and magnesium enhanced insulin signaling and oxidative stability. Smaller proportions of studies investigated vitamin D derivatives (5%) that activated the AMPK–AKT–GSK3β pathway, botanical extracts (7.5%) with antioxidant and neurocognitive benefits, and diet-based deficiency models (2.5%) designed to evaluate the neurodegenerative effects of chronic vitamin A deprivation.

Distribution and characteristics of micronutrient and bioactive compound doses evaluated in studies on neurocognitive outcomes associated with type 2 diabetes. The analysis indicates that preclinical oral doses (10–400 mg/kg/day) were the most frequently used (30%), followed by high-dose antioxidant and polyphenol interventions (15%), and clinically relevant human oral doses (12.5%). Other dosing strategies included nanoparticle formulations (7.5%), trace mineral combinations (7.5%), vitamin D derivatives (5%), and botanical extracts (7.5%), while diet-based deficiency models (2.5%) were least represented.

The analysis shows that metformin was used in various capacities, with moderate preclinical doses (100–300 mg/kg/day) being the most common (27.5%), reflecting its translational relevance to therapeutic levels in humans. A substantial portion of studies (37.5%) did not include metformin, indicating a focus on alternative micronutrient or bioactive compound interventions. High-dose (≥400 mg/kg/day) and low-dose (30–50 mg/kg/day) regimens accounted for smaller fractions (7.5% each), often in combination with antioxidant or nutraceutical agents as summarized in Table 4. In several studies (15%), metformin served as a reference control, emphasizing its role as a benchmark for comparing emerging therapeutic compounds in diabetes-associated cognitive research. Analysis of metformin dosage across studies indicates that synergistic effects with micronutrients and bioactive compounds occur across a broad dose spectrum; however, the most consistent and reproducible synergistic outcomes were observed in preclinical studies using moderate-dose regimens (100–300 mg/kg/day). Low-dose regimens showed preliminary evidence of dose-sparing synergy but were insufficiently represented, while high-dose regimens were largely confined to mechanistic investigations. Due to substantial heterogeneity in study design, outcome measures, and incomplete dose reporting, a single optimal dosage cannot be definitively established. Nevertheless, the evidence supports 100–300 mg/kg/day as the most empirically supported intervention range for synergistic metformin–micronutrient effects.

Distribution and characteristics of metformin doses used across studies investigating neurocognitive outcomes in type 2 diabetes.

Most experiments applied moderate preclinical doses (100–300 mg/kg/day, 27.5%), while 37.5% excluded metformin, focusing on non-pharmacologic interventions. A smaller subset employed reference control use (15%), high-dose testing (7.5%), and low-dose combinations (7.5%), reflecting diverse methodological approaches to assess metformin’s neuroprotective and metabolic efficacy.

Comparators across the analyzed studies predominantly involved standard metabolic control models (35%) typically contrasting normal and diabetic groups to evaluate the impact of interventions on glycemic and neurocognitive indices. Metformin-based comparators (17.5%) were widely employed as positive controls or in combination with micronutrients, indicating its status as a benchmark antidiabetic agent as presented in Table 5. Approximately 15% of studies utilized dual drug or bioactive comparisons, highlighting synergy between nutraceuticals and pharmacological treatments. Placebo and vehicle-based controls (12.5%) provided methodological rigor by minimizing confounding effects, while behavioral and neurocognitive models (10%) reflected growing interest in psychosocial and environmental modifiers of diabetes-related cognitive decline. Clinical stratification controls (7.5%) and a small proportion of positive drug controls (2.5%) contributed to translational validity and comparative benchmarking.

Distribution of comparator and control models used across studies investigating micronutrient and bioactive compound effects on neurocognitive outcomes in type 2 diabetes.

Standard metabolic controls were most prevalent (35%), followed by metformin-based (17.5%) and bioactive compound comparators (15%). Placebo, behavioral, and clinical stratification controls accounted for smaller proportions (12.5%, 10%, and 7.5%, respectively), reflecting diverse experimental and clinical methodologies for validating neuroprotective and metabolic efficacy.

Studies were classified as indirect cognitive inference when cognition was inferred from neurobiological markers without behavioral testing, and as non-cognitive assessment when neither cognitive testing nor cognition-related biomarkers were evaluated as depicted in Figure 3. Direct assessment of learning and memory was reported in 25% (n = 10) of studies, using validated behavioral paradigms such as the Morris Water Maze, Y-maze, Novel Object Recognition, and spatial or recognition memory tasks (26, 29, 36b; 5), all of which evaluated hippocampal-dependent learning or memory outcomes. Global or composite cognitive function was assessed in 15% (n = 6) of studies, primarily in clinical or translational settings using standardized neuropsychological instruments measuring attention, executive function, language, orientation, or overall cognitive impairment (33, 35). Indirect cognitive inference accounted for 12.5% (n = 5) of studies. These investigations did not employ behavioral or neuropsychological tests but evaluated cognition-relevant neurobiological markers or pathways, including hippocampal structural integrity, neurotransmitter modulation, and neuroplasticity-related proteins such as BDNF, PSD-95, SIRT1, or acetylcholinesterase activity (14, 34). Neuropathy-associated neurological outcomes were reported in 7.5% (n = 3) of studies, where sensory or nociceptive function related to diabetic neuropathy was assessed rather than cognition (6). Emotional or affect-related behavioral outcomes, including anxiety- or depression-linked behaviors, were evaluated in 10% (n = 4) of studies (23, 44). 0% (n = 12) of the included studies were classified as non-cognitive assessments. These studies did not include behavioral testing, neuropsychological evaluation, or cognition-linked neurobiological markers and focused exclusively on metabolic, biochemical, cardiovascular, or systemic molecular outcomes (Mohammed H. ElSayed et al., 2023; 13, 16).

Learning and memory were most frequently assessed (25%), followed by global cognitive function (15%) and indirect neurobiochemical inference (12.5%). Emotional-behavioral domains accounted for 10%, neuropathic and neuroendocrine effects 7.5%, while 30% of studies lacked direct testing. Overall, interventions improved hippocampal plasticity, oxidative stability, and insulin signaling, supporting their neuroprotective potential in diabetes-related cognitive decline.

The analysis of 40 studies examining micronutrient and metformin interventions in type 2 diabetes revealed a diverse yet complementary use of behavioral, biochemical, and molecular testing methods to evaluate cognitive and neuroprotective outcomes as presented in Table 6. Behavioral tests such as the Morris Water Maze (MWM), Y-Maze, and Novel Object Recognition (NOR) were frequently employed to measure learning, spatial memory, and recognition ability, accounting for approximately 30% of all reported methods. These tests provided direct evidence of improved hippocampal function, synaptic efficiency, and reduced anxiety-like behaviors following supplementation with vitamin D, vitamin E, zinc, curcumin, and alpha-lipoic acid. Biochemical and molecular assays were the most frequently used category (40%), involving ELISA, Western blot, qRT-PCR, and immunohistochemistry to measure oxidative stress, inflammatory cytokines, and neuroplasticity markers such as BDNF, PSD-95, and SIRT1. These analyses confirmed that the interventions consistently activated the AMPK–AKT–GSK3β signaling pathway, improved antioxidant balance, and attenuated neuroinflammation.

Neuropsychological assessments, including the Mini-Mental State Examination (MMSE) and Montreal Cognitive Assessment (MoCA), represented 15% of methods, providing global cognitive evaluation in human studies. These correlated improvements in attention, executive function, and memory with better glycemic control and metabolic stability. Supporting approaches such as histopathological and imaging analyses (7.5%) confirmed reductions in neuronal degeneration and amyloid burden, while computational bioinformatics tools (5%) like molecular docking and KEGG pathway mapping provided mechanistic insight into receptor-ligand interactions and molecular pathways. A smaller fraction (2.5%) employed metabolic and physiological correlates such as OGTT and HRV to link systemic glucose regulation to neural protection.

Behavioral cognitive assays (30%) and biochemical/molecular tests (40%) were the most commonly applied, followed by neuropsychological assessments (15%), histopathology (7.5%), computational modeling (5%), and metabolic correlates (2.5%). Collectively, these methods demonstrated complementary evidence of cognitive enhancement, oxidative balance, and neuroplastic restoration in diabetes-related cognitive impairment.

Across the reviewed studies, micronutrient and metformin co-interventions produced consistent improvements in fasting plasma glucose (FPG), insulin sensitivity, and glucose tolerance as depicted by Figure 4. Most interventions demonstrated significant reductions in FPG, HbA1c, and HOMA-IR, alongside improvements in oral glucose tolerance test (OGTT) performance and lipid metabolism profiles. Approximately 75% of studies reported a measurable reduction in fasting glucose or glycated hemoglobin, with vitamin D, zinc, magnesium, alpha-lipoic acid, and polyphenolic compounds (e.g., resveratrol, curcumin, and tocotrienols) producing the most pronounced glycemic benefits. Vitamin D₃ supplementation at doses of 500–1000 IU/kg/day significantly decreased fasting glucose and insulin resistance, while zinc sulfate and magnesium sulfate enhanced both glucose tolerance and lipid regulation. In animal models, oral micronutrient doses ranging between 10–400 mg/kg/day and metformin doses between 100–500 mg/kg/day consistently normalized glycemic indices.

In clinical contexts, moderate-dose supplementation (e.g., vitamin E 400 mg/day, zinc 15 mg/day, carnosine 2 g/day) yielded 5–15% reductions in FPG and HbA1c, often in synergy with metformin or glimepiride therapy. These improvements were associated with decreased α-amylase and α-glucosidase activity, enhanced pancreatic morphology, and reduced hyperlipidemia, confirming systemic metabolic restoration. Mechanistically, improved glycemic control correlated with activation of AMPK–AKT–GSK3β pathways, increased GLP-1 secretion, and attenuation of oxidative and inflammatory stressors (e.g., IL-6, TNF-α, and CRP). Several studies also reported concurrent normalization of homocysteine, ferritin, and vitamin B12 levels, suggesting that micronutrient repletion supports broader metabolic homeostasis beyond glucose regulation. However, a minority of studies (10%) reported no significant changes in FPG or HbA1c, typically in computational or non-metabolic models, or in cases where supplementation targeted primarily neuroprotective endpoints. Despite this, the majority demonstrated synergistic effects between micronutrients and metformin, indicating a potentiation of metformin’s insulin-sensitizing and glucose-lowering actions through antioxidant and anti-inflammatory mechanisms.

The result show that most interventions demonstrated significant glycemic improvement, with reductions in fasting glucose and HbA1c observed in about 75% of cases. Vitamin D, zinc, magnesium, and alpha-lipoic acid supplementation, either alone or combined with metformin (100–500 mg/kg/day), consistently enhanced glucose tolerance, improved insulin sensitivity, and restored metabolic efficiency. These effects were mediated through AMPK–AKT pathway activation and reduction of oxidative and inflammatory stress.

Majority, 62.5% of studies reported significant reductions in fasting plasma glucose and HbA1c, indicating improved glycemic control following micronutrient or combined micronutrient metformin therapy as presented in Table 7. These interventions, particularly those involving vitamin D, zinc, curcumin, resveratrol, tocotrienols, and alpha-lipoic acid, effectively enhanced glucose tolerance and pancreatic function. About 20% of studies showed improved insulin sensitivity, reflected by reduced fasting insulin and HOMA-IR levels and increased GLP-1 activity, largely mediated through AMPK–AKT–GSK3β signaling activation. Approximately 15% demonstrated improved lipid metabolism, including lower total cholesterol, triglycerides, and LDL-C, with higher HDL-C, confirming systemic metabolic recovery. Moderate or partial glycemic improvement was seen in 7.5%, often in short-duration or small-sample studies, while 10% either showed no measurable effect or did not report glycemic indices. Adverse or neutral glycemic responses occurred in 5%, mostly in studies involving nutrient deficiencies or insufficient treatment dosing. Another 5% showed normalization of metabolic cofactors such as zinc, ferritin, and hemoglobin, indicating improved metabolic balance.

The study showed that significant reductions in fasting plasma glucose and HbA1c were found in 62.5% of the studies, while 20% demonstrated improved insulin sensitivity and 15% reported enhanced lipid metabolism. Micronutrient supplementation, particularly with vitamin D, zinc, and polyphenols, potentiated metformin’s glucose-lowering effects through antioxidant, anti-inflammatory, and insulin-sensitizing mechanisms, highlighting its therapeutic value in improving glycemic regulation and metabolic balance in type 2 diabetes.

Improvement in insulin sensitivity was observed in 60% of the studies, characterized by reductions in HOMA-IR as presented in Table 8, restoration of β-cell function, and enhanced glucose-stimulated insulin secretion. These studies consistently demonstrated increased GLUT4 expression and PI3K/AKT pathway activation, reflecting enhanced insulin signaling at both peripheral and neuronal levels. Approximately 20% of the studies showed qualitative or inferred improvements in insulin response based on reduced oxidative stress and inflammatory load, even without direct HOMA-IR quantification. In several preclinical models, particularly those combining vitamin D, zinc, and polyphenol-based compounds, normalization of insulin secretion and increased glucose uptake in skeletal muscle were reported.

A smaller proportion, around 10%, presented mixed or neutral findings, often due to short experimental duration or limited sample sizes. Conversely, 5% of studies reported impaired insulin sensitivity associated with high-fat or vitamin-deficient diets, highlighting the detrimental metabolic effects of nutrient imbalance. Micronutrient and bioactive compound supplementation especially vitamin D₃, zinc sulfate, curcumin nanoparticles, and alpha-lipoic acid enhanced insulin sensitivity through activation of the AMPK–IRS–AKT–GSK3β signaling axis, reduction of insulin resistance, and protection of β-cell integrity.

The study showed that 60% of interventions improved insulin sensitivity through enhanced HOMA-IR indices and insulin signaling, while 20% exhibited qualitative metabolic benefits. 10% demonstrated neutral effects, and 5% reported impaired insulin responsiveness under nutrient-deficient or high-fat conditions. Micronutrient supplementation especially with vitamin D₃, zinc, curcumin, and alpha-lipoic acid was most effective in restoring insulin sensitivity via AMPK–IRS–AKT–GSK3β pathway activation and β-cell protection.

Oxidative stress modulation emerged as a central biological mechanism through which micronutrient and bioactive compound supplementation supported neurocognitive and metabolic outcomes in type 2 diabetes, as summarized in Table 9. Across the reviewed studies, antioxidant effects were primarily demonstrated through directional and quantitative changes in redox biomarkers, rather than uniform statistical reporting. Superoxide dismutase (SOD) activity increased in 67.5% of studies, followed by elevations in reduced glutathione (GSH) in 62.5%, catalase (CAT) in 57.5%, and glutathione peroxidase (GPx) in 52.5%, indicating widespread restoration of endogenous antioxidant defense systems (14, 23, 29). Markers of oxidative damage showed consistent quantitative reductions. Malondialdehyde (MDA), the most frequently assessed lipid peroxidation marker, decreased in 70% of studies, with reported reductions ranging from absolute concentration decreases (e.g., 14.03 to 6.12 nmol/mL) to percentage declines exceeding 50% in dose-response models (39, 43). Similarly, reactive oxygen species (ROS) and thiobarbituric acid reactive substances (TBARS) declined in 45% and 25% of studies, respectively, reflecting attenuation of cellular oxidative burden across diverse experimental models (29, 41). At the molecular signaling level, redox-linked inflammatory mediators were less frequently measured but demonstrated consistent directional suppression. Inducible nitric oxide synthase (iNOS), thioredoxin-interacting protein (TXNIP), and nuclear factor-κB (NF-κB) were downregulated in 15%, 10%, and 20% of studies, respectively, supporting mechanistic cross-talk between oxidative stress reduction and inflammatory pathway inhibition (6, 21). Interventions incorporating vitamin D₃, zinc, curcumin (including nanoparticle formulations), alpha-lipoic acid, and Nigella sativa oil demonstrated the strongest antioxidant profiles. These agents consistently increased antioxidant enzyme activity, restored glutathione balance, and reduced lipid and protein oxidation, thereby limiting oxidative stress–driven neurodegeneration and insulin resistance (29, 39).

The study showed that SOD (67.5%) and MDA (70%) were the most consistently modulated oxidative biomarkers, indicating strong antioxidant efficacy of micronutrient and bioactive supplementation. CAT (57.5%), GSH (62.5%), and GPx (52.5%) also increased significantly, while ROS and TBARS declined in nearly half of the studies. Downregulation of iNOS (15%), TXNIP (10%), and NFκB (20%) further confirmed the anti-inflammatory synergy of interventions such as vitamin D₃, zinc, curcumin, alpha-lipoic acid, and Nigella sativa oil, restoring oxidative and inflammatory balance in type 2 diabetes.

Inflammatory modulation emerged as a major mechanism linking micronutrient and bioactive compound supplementation to improved neurocognitive and metabolic function in type 2 diabetes as presented in Table 10. Most studies demonstrated significant suppression of pro-inflammatory cytokines and signaling pathways, emphasizing the integrated anti-inflammatory and neuroprotective roles of vitamins, minerals, and phytochemicals. Interleukin-6 (IL-6) and tumor necrosis factor-alpha (TNF-α) were the most frequently downregulated cytokines, showing decreased levels in 65% and 62.5% of the studies, respectively. Interleukin-1β (IL-1β) reduction was observed in 42.5%, while C-reactive protein (CRP) levels declined in 25%, signifying systemic inflammation control. Anti-inflammatory cytokine IL-10 was elevated in 15%, indicating restored immune balance. Nuclear factor kappa B (NF-κB) suppression in 22.5% and NLRP3 inflammasome inhibition in 12.5% further highlighted the downregulation of pro-inflammatory transcriptional and cellular signaling cascades. Several interventions including vitamin D, zinc, curcumin, alpha-lipoic acid, and Nigella sativa oil demonstrated dual antioxidant and anti-inflammatory effects. These compounds not only reduced cytokine expression but also normalized metabolic hormones such as leptin and adiponectin in 10% of studies, indicating improved insulin sensitivity and energy regulation. Overall, the collective findings demonstrate that micronutrient and bioactive compound interventions effectively inhibits chronic low-grade inflammation characteristic of type 2 diabetes, thereby improving neuronal integrity, cognitive performance, and glycemic homeostasis through suppression of IL-6, TNF-α, and NF-κB–mediated inflammatory pathways.

Key: N= Frequencies, % percentages.

The study showed that IL-6 (65%), TNF-α (62.5%), and IL-1β (42.5%) were the most consistently downregulated inflammatory mediators, confirming the potent anti-inflammatory action of micronutrient and bioactive compound supplementation. CRP reduction in 25% of studies reflected improved systemic inflammation, while IL-10 elevation in 15% indicated restored immune equilibrium. Suppression of NF-κB (22.5%) and NLRP3 (12.5%) pathways demonstrated cross-regulation between oxidative and inflammatory signaling.

Micronutrient and bioactive compound supplementation substantially improved neuroplasticity and synaptic integrity in type 2 diabetes as presented in Table 11. The most frequently enhanced marker, BDNF, was upregulated in 22.5% of the studies, indicating improved neuronal survival and cognitive recovery. Synaptic proteins including PSD-95, synapsin-1, and synaptophysin each showed an increase in 7.5%, signifying better synaptic stability and neurotransmission efficiency. Activation of the PI3K–Akt–CREB signaling pathway in 12.5% of cases confirmed enhanced synaptic remodeling and learning capacity, while SIRT1 elevation in 5% highlighted mitochondrial and epigenetic regulation of neuroprotection. Reduction in Aβ deposition and BACE-1 expression (7.5%) paralleled improved cholinergic function (ACh↑/AChE↓) and neuronal histoarchitecture recovery (15%), reinforcing evidence of combined antioxidant, anti-apoptotic, and synaptogenic effects. Overall, these findings demonstrate that micronutrients such as zinc, vitamin D, vitamin E, and bioactive compounds like curcumin, alpha-lipoic acid, and Nigella sativa restore neuroplasticity by modulating BDNF, CREB, and PI3K–Akt signaling axes.

The study showed that BDNF expression increased in 22.5% of the studies, with corresponding rises in PSD-95, synapsin-1, and synaptophysin (7.5%), reflecting improved synaptic connectivity and neuronal integrity. Upregulation of PI3K–Akt–CREB and Nrf2 pathways (12.5–7.5%) indicated a coordinated enhancement of antioxidant and neuroplastic mechanisms. Concurrently, reductions in Aβ and BACE-1 expression (7.5%) demonstrated protection against neurodegenerative changes, confirming the synergistic role of micronutrients and phytochemicals in restoring synaptic plasticity and cognitive resilience in type 2 diabetes.

AMPK and insulin signaling enhancement was a key mechanistic target of micronutrient and bioactive compound interventions in type 2 diabetes as presented in Table 12. Activation of AMPK was reported in 25% of the studies, promoting glucose uptake, lipid oxidation, and mitochondrial biogenesis. The PI3K/Akt pathway was the most frequently upregulated signaling axis (27.5%), driving improved insulin sensitivity and neuronal metabolism. Complementary molecular changes included increased GLUT4 expression (10%) and decreased GSK-3β phosphorylation (10%), indicating improved synaptic plasticity and reduced tau aggregation. Inhibition of mTOR (7.5%) and activation of PPARs (7.5%) reflected improved metabolic and autophagic balance. Cross-activation between AMPK and Nrf2–HO-1–CREB signaling (7.5%) provided additional antioxidant and neuroprotective support. Collectively, these results suggest that nutraceuticals such as vitamin D, curcumin, zinc, and alpha-lipoic acid, alongside metformin, enhance metabolic and neuronal resilience through AMPK–PI3K/Akt–GSK3β signaling modulation, resulting in better glycemic control and preserved cognitive integrity.

The study showed that AMPK activation (25%) and PI3K/Akt signaling enhancement (27.5%) were the predominant mechanisms underpinning improved glucose regulation and neuroprotection. GLUT4 upregulation (10%) and GSK-3β inhibition (10%) reinforced enhanced insulin sensitivity and prevention of tau pathology, while suppression of mTOR and NF-κB pathways (5–7.5%) highlighted improved autophagy and anti-inflammatory balance. These findings confirm that combined micronutrient and pharmacologic interventions synergistically restore insulin signaling and neuronal metabolism via AMPK–Akt–GSK3β and Nrf2–CREB cross-talk.

Synergistic interactions were the most prevalent pattern across the reviewed studies, accounting for 77.5% of reported outcomes (Figure 5). These synergistic effects were most frequently documented when metformin was combined with vitamins, minerals, or plant-derived bioactive, resulting in superior metabolic, antioxidant, anti-inflammatory, and neuroprotective outcomes compared with single-agent interventions. Representative examples include vitamin- and antioxidant-based combinations (e.g., vitamin D₃, vitamin E, vitamin B₁₂) that enhanced insulin sensitivity, redox balance, and cognitive markers (29, 39, 44), mineral–metformin interactions involving zinc and magnesium that restored insulin signaling and antioxidant enzyme activity (42, 44), and phytochemical–metformin combinations such as curcumin nanoparticles, alpha-lipoic acid, flavonoid-rich extracts, and Nigella sativa oil that activated AMPK–Akt–BDNF and Nrf2-dependent pathways (14, 29, 43).

Furthermore, when stratified by nutrient class, vitamin–metformin combinations showed the highest prevalence of synergistic effects, accounting for approximately 35–40% of all synergistic interactions, particularly with vitamin D₃ and antioxidant vitamins (29, 39). Mineral–metformin interactions (zinc and magnesium) contributed roughly 20–25%, demonstrating consistent synergistic or additive benefits via improved insulin sensitivity, hematologic balance, and oxidative stability (27, 42). Plant extracts and phytochemicals represented approximately 30% of synergistic outcomes, with compounds such as curcumin, alpha-lipoic acid, resveratrol, Ocimum gratissimum, and Nigella sativa producing robust antioxidant, mitochondrial, and neuroplastic effects when combined with metformin (40, 43). Additive (non-synergistic but complementary) effects were observed in 5% of studies, typically where micronutrients and metformin acted through parallel but mechanistically distinct pathways, such as vitamin D providing stronger anti-inflammatory effects while metformin exerted superior glycemic control (44). In contrast, antagonistic interactions accounted for 10% of outcomes and were largely confined to long-term metformin use associated with vitamin B₁₂ depletion, elevated homocysteine, and increased inflammatory burden, with potential adverse implications for neurocognitive health (35, 37). Finally, single-agent or non-combination interventions comprised 2.5–5% of studies and generally conferred modest metabolic or neuroprotective benefits without synergistic amplification (16, 32).

The study showed that synergistic effects accounted for 77.5% of the interventions, while antagonistic effects were limited to 10%, mostly due to metformin-associated nutrient depletion. Additive interactions (5%) reflected cooperative but independent mechanisms. These results confirm that combined therapies involving metformin and micronutrients notably vitamin D, zinc, and alpha-lipoic acid produce enhanced neuroprotective, anti-inflammatory, and insulin-sensitizing outcomes, supporting the integrative use of nutraceuticals in diabetes-associated cognitive dysfunction.