Skip to Content
Medical SciencesMedical Sciences
  • Article
  • Open Access

26 August 2026

24 Pages

Cardiometabolic Risk Factors and Domain-Specific Cognitive Function in Prediabetes and Type 2 Diabetes

,
,
,
,
and
1
Monash Centre for Health Research and Implementation, Faculty of Medicine, Nursing and Health Sciences, Monash University, 43-51 Kanooka Grove, Clayton, VIC 3168, Australia
2
Sheffield Institute for Translational Neuroscience, School of Medicine and Population Health, University of Sheffield, 385a Glossop Rd., Broomhall, Sheffield S10 2HQ, UK
3
NIHR Sheffield Biomedical Research Centre, University of Sheffield and Sheffield Teaching Hospitals NHS Foundation Trust, Glossop Rd, Broomhall, Sheffield S10 2JF, UK
4
Neuroscience Institute, University of Sheffield, Firth Court, Sheffield S10 2TN, UK

Abstract

Background: Cardiometabolic risk factors are increasingly recognised as contributors to cognitive decline, yet their associations with domain-specific cognitive performance in adults with prediabetes or well-controlled type 2 diabetes (T2DM) remain poorly characterised. Methods: This cross-sectional study included 49 adults with prediabetes or well-controlled T2DM (managed by diet or metformin alone; HbA1c < 8%). Cognitive function was assessed using the Cambridge Neuropsychological Test Automated Battery (CANTAB), Trail Making Tests (TMTs) parts A and B, Stroop, and Digit Symbol Substitution Test (DSST). Multivariable linear regression examined associations between cardiometabolic predictors (adiposity [DXA], glycaemic indices [OGTT], and serum lipids) and cognitive outcomes, adjusting for age, BMI, and education. Results: Higher HbA1c was independently associated with increased false alarms on the rapid visual processing task (RVPPFA; β = 0.04 [95% CI 0.01, 0.07]; p = 0.015); fasting and 2 h glucose were not associated with any cognitive measure. Higher triglycerides were associated with lower sustained attention accuracy (RVPA; β = −0.02 [95% CI −0.04, −0.004]; p = 0.016) and lower DSST scores (β = −4.61 [95% CI −8.81, −0.40]; p = 0.033). Higher HDL cholesterol was independently associated with better sustained attention (RVPA; β = 0.08 [95% CI 0.02, 0.15]; p = 0.017). No consistent associations were found between adiposity measures and any cognitive outcome, with no effect modification by sex or diabetes status (all Pinteraction > 0.05). Conclusions: In cognitively healthy adults with prediabetes or well-controlled T2DM, higher HbA1c and triglycerides and lower HDL were associated with poorer domain-specific cognitive performance in the absence of overt cognitive impairment. Larger longitudinal studies are needed to establish whether these associations represent early cognitive changes and to determine whether reducing cardiometabolic risk can preserve cognitive function.

1. Introduction

Population ageing is increasingly recognised as a major driver of the global burden of disease. In 2020, approximately 727.6 million individuals were aged over 65 years, representing 9.33% of the global population [1]. The rate of population ageing is projected to accelerate, with the number of individuals aged ≥60 years expected to double to 2.1 billion by 2050, and those aged ≥80 years to triple between 2020 and 2050 [1,2]. In parallel, the prevalence of cardiometabolic disorders, including obesity, prediabetes, type 2 diabetes (T2DM), and dyslipidaemia, also increases with age. These conditions are now recognised risk factors for cognitive decline and dementia [3,4], with both obesity and diabetes among several modifiable risk factors estimated to contribute to 40% of dementia cases [5]. Impaired cerebral perfusion, insulin resistance, chronic inflammation, and oxidative stress are proposed mechanisms by which cardiometabolic risk contributes to declines in cognitive performance [6,7,8].
Cognitive impairment may manifest as global cognitive dysfunction or as deficits within one or more specific neuropsychological domains, such as memory, pattern recognition, attention, and executive function. Individuals with prediabetes, characterised by dysregulated glucose metabolism below the threshold for T2DM (i.e., fasting plasma glucose of 6.1 to 6.9 mmol/L or 2–h post-load plasma glucose of 7.8 to 11.0 mmol/L, or HbA1c of 5.7% to 6.4%) [9,10], have shown modest impairments in memory and executive function, which may worsen with diabetes progression or if left untreated [11,12]. T2DM is associated with a 19% higher risk of cognitive decline over a 20-year period [13], with ageing compounding this effect [14]. Beyond glucose metabolism, other cardiometabolic risk factors also influence cognitive performance. Obesity, reflected by an increased adiposity, waist-to-hip ratio, and/or body mass index (BMI), has been associated with poorer cognitive function [15]. Elevated triglycerides and/or an unfavourable lipid profile (i.e., dyslipidaemia) are common in T2DM [16,17], and have been implicated in neurodegeneration by promoting atherosclerosis and vascular damage of cerebral arteries [18].
Despite growing recognition of the relationship between cardiometabolic health and cognitive function, the associations between key cardiometabolic risk factors and domain-specific cognitive performance remain unclear, particularly among individuals with prediabetes or well-controlled T2DM who are otherwise cognitively healthy but may exhibit early mild impairments. This is important, as identifying cognitive impairment or dementia at the earliest possible point in the disease trajectory is critical to optimising treatment outcomes and can inform the development of novel metabolic biomarkers for these disease states. A comprehensive approach integrating both cardiometabolic and cognitive measures is therefore needed to characterise these relationships and provide insight into underlying pathophysiological pathways.
To address this, the present exploratory study aims to examine the relationships between body composition, glycaemic measures, and lipid profiles with cognitive performance among individuals with prediabetes or well-controlled T2DM who are otherwise considered cognitively healthy.

2. Materials and Methods

2.1. Participants

This study is an exploratory cross-sectional analysis of baseline data from a double-blind, randomised, placebo-controlled trial that examined the impact of 14 weeks of carnosine supplementation on cardiometabolic outcomes in individuals with prediabetes or type 2 diabetes. The detailed methodology for the trial is described in the published protocol [19], and the primary findings are reported elsewhere [20]. In brief, eligible participants were aged 18–70 years and were diagnosed with prediabetes or type 2 diabetes through an oral glucose tolerance test (OGTT) and HbA1c, following WHO 2019 criteria [21]. Only those managing their condition through diet and/or with metformin only were eligible for inclusion, and all participants were cognitively healthy based on self-reported history. Exclusion criteria included HbA1c levels exceeding 8% at screening; a BMI over 40 kg/m2; significant weight fluctuations (greater than 5 kg in the previous six months); smoking; excessive alcohol consumption (>4 standard drinks per week for men or >2 for women) [22]; renal failure (estimated glomerular filtration rate < 30 mL/min); active cancer within the previous five years; acute inflammation or infection; the use of anti-inflammatory medications or any medications or dietary supplements known to affect cardiometabolic or cognitive outcomes; and the presence of cardiovascular, endocrine, respiratory, haematological, gastrointestinal, or central nervous system disorders, including cognitive and psychiatric conditions. Participants with T2DM were required to have been on stable treatment for at least three months prior to enrolment and were advised not to change their treatment during the study. Pregnant or lactating women were also excluded.
Participants were recruited through advertisements in newspapers, radio, and posters. Prospective participants underwent an initial screening via a telephone-based questionnaire, and those meeting preliminary eligibility criteria were invited for an in-person screening visit at the Clinical Trials Centre within the Monash Health Translational Research Precinct in Melbourne, Australia.

2.2. Outcome Measures

At the screening visit, a medical review and a 75 g oral glucose tolerance test (OGTT) were performed after an overnight fast to obtain fasting and post-load blood glucose levels, as well as HbA1c, to confirm prediabetes or diabetes status based on WHO (2019) guidelines [21]. Additional fasting blood samples were collected and analysed at Monash Health Pathology (Melbourne, Australia), which comprised a standard biochemical panel (renal and liver function, full blood count) to exclude other underlying conditions and for measurement of serum lipids by commercial immunoassays to characterise cardiometabolic risks. Demographic characteristics including age and level of education (primary, secondary, tertiary or above) were collected via self-reported answers in the CANTAB application. Anthropometric measurements included the BMI, calculated from weight (kg)/height2 (m2); the waist-to-hip ratio, derived from the waist circumference/hip circumference when measured during the initial medical review; and the total body fat (%), measured by dual-energy X-ray absorptiometry (DXA).
Cognitive function was assessed using four cognitive instruments administered in both written and digital formats. These included the Cambridge Neuropsychological Test Automated Battery (CANTAB), the Trail Making Tests (A&B), and the Stroop test, which were performed using a digital tablet (iPad, iOS v.10, Apple Inc., Melbourne, Australia), as well as the Digit Symbol Substitution Test, which was administered on paper. Descriptions of these assessments and their relevant domains and score interpretations are described in Table 1. Assessments took place during working hours (9 a.m.–5 p.m.) in the same location and in the same sequence, although they were not conducted at the same time of day. Before starting each test, participants received a brief explanation of the procedure. To minimise external distractions, the tests were conducted in a quiet, well-lit, and controlled environment, with only the participant and researcher present. Participants were not required to fast before testing.
Table 1. Description of cognitive test procedures and interpretation of scores.

2.3. Ethics

All participants provided written informed consent prior to undergoing baseline screening procedures. The trial was registered on clinicaltrials.gov (NCT02917928; 28 September 2016) and carried out according to the published protocol [19] and Standard Protocol Items: Recommendations for Interventional Trials (SPIRIT) guidelines [23], and was reported in line with STROBE guidelines [24]. Ethical approval was granted by the Monash Health Human Research Ethics Committee (Ref: 16061A; 14 April 2016), and all aspects of the trial adhered to the principles of the Declaration of Helsinki [25].

2.4. Statistical Analysis

Statistical analyses were performed using JASP (v0.95.4; University of Amsterdam, Amsterdam, The Netherlands) and SPSS (v26, IBM Corp, Armonk, New York, NY, USA). Normality of cognitive variables was assessed through visual inspection of histograms, along with the Shapiro–Wilk and Kolmogorov–Smirnov tests to evaluate skewness and kurtosis. Descriptive statistics were presented as means with standard deviations (SDs) for normally distributed continuous data or as medians with interquartile ranges (IQRs) for non-normally distributed continuous data, and as frequencies and proportions for categorical data. For the CANTAB, key variables were analysed following the guidelines provided in the CANTAB handbook [26].
Simple and multivariable linear regression models were used to explore relationships between cardiometabolic parameters as predictors (anthropometric, glycaemic, and lipid measures) and cognitive scores as outcomes. Four separate models were fitted for each metabolic predictor; Model 1 (unadjusted); Model 2, adjusted for age; Model 3, adjusted for age and body mass index (BMI); and Model 4, adjusted for age, BMI, and level of education. Covariates were limited to age, BMI, and level of education given the sample size, as additional adjustment would risk model overfitting. Where significant associations were identified in these regression models, potential effect modification by sex and diabetes status was assessed by incorporating interaction terms into the fully adjusted models. Sensitivity analyses were performed with additional adjustment for metformin use to examine changes in the results. A two-sided p-value < 0.05 was considered statistically significant. To aid interpretation of effect magnitudes, standardised estimates were derived for significant associations in the fully adjusted models by multiplying each regression coefficient with the standard deviation of the cardiometabolic exposure and dividing by the SD of the corresponding cognitive outcome to express each association per one standard deviation of exposure in standard deviation units of the outcome. Across the fully adjusted models, 11 cardiometabolic predictors were examined in relation to 17 cognitive outcomes. Given the post hoc exploratory nature of the study, this secondary analysis was not powered for cognitive outcomes, and there were no corrections for multiple testing. Significant findings should therefore be interpreted as nominal and hypothesis-generating rather than confirmatory, requiring validation in larger adequately powered cohorts.

3. Results

3.1. Sample Characteristics

The demographic, anthropometric, clinical, biochemical, and cognitive characteristics of participants are presented in Table 2. Of the 50 recruited participants, one participant with normoglycaemia was excluded, leaving 49 for inclusion in the final analysis. Approximately two-thirds of the cohort were males (n = 33; 67%), with 23 participants (47%) classified as having T2DM and 26 (53%) with prediabetes per WHO (2019) criteria [21]. Nearly half of the participants were aged 50 years and over (47%; n = 23). The majority were overweight (BMI ≥ 25–29.9 kg/m2; 53%, n = 26) or obese (BMI ≥ 30 kg/m2; 39%, n = 19), with only four participants within the normal BMI range. Participants were considered cognitively healthy based on self-reports and medical histories.
Table 2. Participant characteristics overall and by diabetes status.

3.2. Associations Between Cognitive and Cardiometabolic Measures

3.2.1. Adiposity

The body fat percentage and waist-to-hip ratio were not associated with any cognitive measure across the CANTAB, Trail Making Tests (TMTs), Stroop, or Digit Symbol Substitution Test (DSST) in unadjusted or multivariable models (Table 3, Table 4 and Table 5). BMI was similarly not associated with any cognitive outcome except TMT-B’s completion time, where a higher BMI was associated with faster completion in the fully adjusted model (β = −2.51 [95% CI −4.76, −0.25]; p = 0.03) but not in unadjusted or partially adjusted models (Table 4). Sensitivity analyses with further adjustment for metformin use did not alter these findings, and no significant interaction effects with sex or diabetes status were observed (Pinteraction > 0.08; Table 4).
Table 3. Univariable and multivariable associations between cardiometabolic variables and selected CANTAB domains.
Table 4. Univariable and multivariable associations between cardiometabolic variables and Trail Making Tests.
Table 5. Univariable and multivariable associations between cardiometabolic variables with Stroop and Digit Symbol Substitution Test outcomes.

3.2.2. Glycaemic Measures

Higher HbA1c was associated with a greater probability of false alarms on the CANTAB rapid visual processing task (RVPPFA) in the unadjusted model (β = 0.03 [95% CI 0.003, 0.07]; p = 0.03) (Table 3). This association remained significant after adjusting for age (p = 0.011), BMI (p = 0.014), and level of education (β = 0.04 [95% CI 0.01, 0.07]; p = 0.015). No significant interaction effects with sex or diabetes status were observed for this association (Pinteraction >0.26; Table 3). Fasting and 2–h post-load glucose were not associated with any cognitive measures in univariable or multivariable analyses (Table 3, Table 4 and Table 5; Supplemental Table S1). Additional adjustment for metformin use did not alter results for any of these glycaemic parameters.

3.2.3. Lipid Profile

Higher triglyceride concentrations were associated with lower sustained attention accuracy on the CANTAB rapid visual processing task (RVP A prime [RVPA]; β = −0.02 [95% CI −0.04, −0.01]; p = 0.008; Table 3). This remained significant after adjustment for age (p = 0.012), BMI (p = 0.014), and level of education (β = −0.02 [95% CI −0.04, −0.004]; p = 0.016), with no effect modification by sex or diabetes status (Table 3). Higher triglycerides were also associated with slower processing speed as measured by RVP median latency (RVPMDL) in age-adjusted and BMI-adjusted models (β = 27.19 [95% CI 1.86, 52.52]; p = 0.036), although they were attenuated in the fully adjusted model (p = 0.057). There was a significant interaction with diabetes status for this association (β = 74.10 [95% CI 24.55, 123.66]; Pinteraction = 0.004; Table 3).
In addition to CANTAB domains, higher triglycerides were associated with a higher TMT B/A ratio in the unadjusted (β = 0.42 [95% CI 0.01, 0.83]; p = 0.045), age-adjusted (p = 0.045), and age- and BMI-adjusted models (p = 0.043), but this did not reach statistical significance in the fully adjusted model (p = 0.050), and there were no interactions with sex or diabetes status (Pinteraction > 0.42; Table 4). Higher triglycerides were also associated with lower DSST scores across all adjusted models (all p ≤ 0.03), including in the fully adjusted model (β = −4.61 [95% CI −8.81, −0.40]; p = 0.033), with no interactions with sex or diabetes status (Pinteraction > 0.31; Table 5).
Higher concentrations of high-density lipoprotein cholesterol (HDL-C) were associated with better sustained attention (RVPA; β = 0.08 [95% CI 0.02, 0.15]; p = 0.010) in CANTAB, which remained significant across all multivariable models (all p < 0.02), with no significant interactions with sex or diabetes status (Pinteraction > 0.54; Table 3). Other lipid measures, including total cholesterol, low-density lipoprotein cholesterol (LDL-C), and LDL:HDL ratio, were not associated with any cognitive measures in both univariable and multivariable analyses (Table 3, Table 4 and Table 5; Supplemental Table S1). The results for all lipid measures were not altered in sensitivity analyses with additional adjustment for metformin use.
To aid interpretation, the effect magnitudes of significant associations are presented in Table 6, expressed as per 1 SD of cardiometabolic exposure relative to 1 SD of the cognitive outcome. All associations ranged between −0.36 SD and + 0.4 SD (Table 6).
Table 6. Standardised effect magnitudes for significant cardiometabolic–cognitive associations.

4. Discussion

This study explored associations between cardiometabolic risk factors and domain-specific neurocognitive performance in cognitively healthy adults with prediabetes or well-controlled T2DM. Our findings suggest that higher HbA1c was associated with a higher false alarm rate on a sustained attention task, and that lipid dysregulation, particularly higher triglycerides and lower HDL-C concentrations, was associated with poorer sustained attention, and processing speed. These associations were independent of age, BMI, and education. Notably, these were observed in the absence of overt cognitive impairment and under relatively well-controlled metabolic conditions, raising the possibility that subtle cognitive differences may be detectable early in cardiometabolic disease using targeted neuropsychological assessments. Effect magnitudes were modest (−0.36 SD to +0.41 SD), suggesting variations within normal cognitive performance, rather than clinically significant cognitive impairment.
Adiposity measures, including body fat percentage, waist-to-hip ratio, and BMI, were not consistently associated with cognitive outcomes in this study. The literature on adiposity and cognition is inconsistent. Some studies have reported no associations between BMI and cognitive measures in cognitively healthy older adults [27], while others have reported detrimental effects on brain volumes in T2DM [28]. Sex-specific associations have also been reported [29,30] but could not be explored here given the unequal sex distribution and modest sample size. Interestingly, Xing et al. [31] found that BMI was not associated with DSST or Stroop performance in T2DM, consistent with our findings, though unexpected positive associations with memory scores were observed. These echo similarly counterintuitive reports that moderately higher BMI may protect against grey matter loss [32] and cognitive decline [33] in later life. In our study, the isolated association between a higher BMI and faster TMT-B completion, present only in the fully adjusted model, may reflect the same paradox, though its model dependency and small sample size warrant cautious interpretation. Collectively, these findings suggest that the adiposity–cognition relationship is complex and requires further investigation in adequately powered studies with balanced sex representation.
Higher HbA1c was associated with increased false alarm responses on the Rapid Visual Processing task (RVPPFA), reflecting poorer inhibitory control, independent of age, BMI, and education. This is broadly consistent with evidence that chronic glycaemic exposure in prediabetes or T2DM is associated with cognitive decline [34,35,36,37], particularly in attention and executive function. However, prior studies largely relied on global composite scores or brief screening instruments and typically included individuals with more severe disease on complex medication regimens, introducing confounding by indication [38,39]. Here, these associations were detectable in a cognitively healthy cohort managed with diet or metformin alone, with HbA1c below 8%, indicating that this association was observable at lower levels of glycaemic exposure than have typically been demonstrated in prior work. This points toward glycaemia-related attentional dysregulation potentially emerging earlier than previously recognised, an intriguing hypothesis that warrants formal testing. The specificity to inhibitory control is also of interest; false alarm responses reflect frontoparietal network function [40], suggesting that this network may be particularly vulnerable to early glycaemic dysregulation. Notably, fasting and 2 h post-load glucose were not associated with any cognitive measure, while HbA1c remained significant across all models, pointing to a cumulative rather than acute glycaemic burden as the more relevant exposure. Indeed, glycaemic variability has been shown to damage neurons through glucose transporter mismatches [41], a cumulative process integrated over time by HbA1c but not captured by isolated glucose measurements, consistent with our findings.
Other mechanisms linking chronic hyperglycaemia and cognitive dysfunction include oxidative stress, endothelial dysfunction, and reduced cerebral perfusion, which directly contribute to neuronal injury [4,39]. Accumulation of advanced glycation end-products (AGEs), which bind to their receptor (RAGE), further perpetuates oxidative and cellular damage associated with up to three-fold increased odds of mild cognitive impairment (MCI) [42,43,44,45]. In parallel, cerebral insulin resistance impairs long-term potentiation of synaptic plasticity, a key process in memory formation and learning [8], by reducing long-term potentiation and brain-derived neurotrophic factor (BDNF) production [46,47]. Hyperglycaemia may also promote amyloid pathology through increased amyloid precursor protein expression [48]. While these mechanisms offer plausible explanations, the key pathways in this population require further investigation, ideally using combined neuroimaging, cognitive testing and biomarker-based approaches.
Lipid dysregulation was the most prominent cardiometabolic risk factor associated with cognitive performance in this study. Independent of age, BMI, and education, higher triglycerides were associated with lower accuracy on the sustained attention task (RVPA) and poorer processing speed (DSST) across all adjusted models, whereas higher HDL-C was associated with better sustained attention across all models. These findings are broadly consistent with prior evidence linking higher triglycerides to poorer executive function [49] and higher HDL-C to better cognitive performance in older adults [50]. Studies in T2DM cohorts have been mixed; some have reported associations between the triglyceride–glucose index with MCI, or between HDL-C and attention and memory [51], while others have reported no independent lipid–MCI associations [52,53] or report associations only for total and LDL-C [54]. These inconsistencies likely reflect differences in T2DM severity and medication use (e.g., statins), and a common reliance on global cognitive assessment tools such as the Montreal Cognitive Assessment (MoCA) [51,54], which may lack sensitivity to detect domain-specific associations. Here, we extend the literature by demonstrating that triglyceride and HDL-C associations with sustained attention, and processing speed may be detectable in a well-controlled, early-stage cardiometabolic cohort—a population largely overlooked in prior work—though these findings remain exploratory and require replication in future studies. Mechanistically, our findings mirror the contrasting roles of triglycerides and HDL-C in neurovascular health: triglyceride-rich lipoproteins promote blood–brain barrier dysfunction, endothelial injury, and amyloid dysmetabolism [55], whereas HDL-C exerts anti-inflammatory, antioxidative, and endothelial-protective effects that support cerebrovascular integrity [56]. These processes may impair or preserve cerebral perfusion and neural efficiency within frontoparietal networks underlying attention and processing speed [40], providing a neurovascular basis for the domain-specific associations observed here. These relationships may be further modulated by genetic susceptibility; apolipoprotein E4 (ApoE4), the strongest genetic risk factor for sporadic Alzheimer’s disease [57], regulates lipid transport within the central nervous system and has been linked to impaired lipid metabolism, neuroinflammation, and cognitive deficits across multiple domains [57,58]. The precise lipid-driven neurovascular mechanisms underlying cognitive risk, and their interactions with genetic factors, if any, await further study.
A key strength of this study is the use of a comprehensive cognitive assessment battery that combines domain-specific CANTAB measures with traditional neuropsychological tests (Stroop, TMTs, and DSST). This enabled the detection of subtle cognitive differences unlikely to be captured by global screening tools. Cardiometabolic risk was characterised using validated measures, including DXA-derived body composition, OGTT-based glycaemic measures and a full lipid panel. The inclusion of key covariates, including age, BMI, and education, strengthened the analyses, with education in particular as an important but less consistently accounted for factor in prior studies. Restricting inclusion to individuals managed by diet or metformin alone minimised confounding by disease severity and eligibility criteria of the parent trial excluded cardiovascular, renal, psychiatric, and central nervous system disease, as well as any medication or supplement affecting cardiometabolic or cognitive outcomes; therefore, no participant was taking antihypertensive or lipid-lowering medications. Blood pressure and depressive symptoms in this cohort were within normal ranges as reported previously [59,60], allowing early cardiometabolic–cognitive associations to be delineated more clearly than in previous studies of populations with more advanced disease.
Limitations of the study include the modest sample size, which restricted statistical power and precluded subgroup analyses, whereby sex and diabetes status could only be examined as interactions for exploratory purposes. With no correction for multiple testing and, given the large number of cognitive outcomes tested, there is an increased susceptibility for type I error; thus, individual findings should be interpreted with caution. The cross-sectional study design precluded causal inference and directionality, and reverse causation cannot be excluded. Although key confounders known to influence cognition (i.e., sex, level of education, and diabetes status) were incorporated in the analysis, the modest sample size precluded adjustment for a wider range of variables due to risk of overfitting. Thus, residual confounding by cardiometabolic (e.g., biomarkers), lifestyle (e.g., diet, sleep) and demographic factors (e.g., ethnicity categories) cannot be ruled out. Data on additional variables, including socioeconomic status or diabetes duration, were not measured in the parent trial, and a healthy control group was not included; aspects which should be considered in future studies. Cognitive health was established on the basis of self-reports and medical histories rather than cognitive screening, and could be subject to misclassification errors. The ApoE genotype was not assessed and may represent an unmeasured confounder or effect modifier, given its established role in lipid metabolism and cognitive decline. Finally, the cohort was predominantly male and highly selected, with strict inclusion criteria (HbA1c < 8%, well-controlled T2DM with diet or metformin only, BMI < 40 kg/m2 and absence of comorbidities). While the cohort’s cognitively healthy and metabolically well-controlled profile strengthens internal validity, it limits generalisability to other populations, including more diverse female populations, and those with more advanced or poorly controlled disease, complex polypharmacy, or established cognitive impairment. Given these constraints, the present findings should be considered hypothesis-generating rather than confirmatory, and are intended to inform the design of larger adequately powered longitudinal studies.

5. Conclusions

In cognitively healthy adults with prediabetes or well-controlled T2DM, glycaemic dysregulation and dyslipidaemia were associated with poorer cognitive performance, independent of age, BMI, and education, and in the absence of overt cognitive impairment. Larger prospective studies are needed to confirm whether these associations reflect early cognitive changes and whether optimising cardiometabolic risk factors can preserve cognitive function in this population.

Supplementary Materials

The following supporting information can be downloaded at: https://www.mdpi.com/article/10.3390/medsci14050520/s1, Table S1: Univariable and multivariable associations between cardiometabolic variables and selected CANTAB measures. Figure S1: Participant timeline: Schedule of enrolment, interventions, and assessments. STROBE Statement—Checklist of items that should be included in reports of cross-sectional studies.

Author Contributions

Conceptualisation, A.M. (Aya Mousa) and B.d.C.; methodology, R.H., A.M. (Aya Mousa) and B.d.C.; software, R.H. and A.M. (Aya Mousa); validation, R.H. and A.M. (Aya Mousa); formal analysis, R.H. and A.M. (Aya Mousa); investigation, B.d.C.; resources, B.d.C.; data curation, R.H.; writing—original draft preparation, R.H. and A.M. (Aya Mousa); writing—review and editing, R.H., S.M.B., A.M. (Arshad Majid), J.F., A.M. (Aya Mousa) and B.d.C.; supervision, A.M. (Aya Mousa) and B.d.C.; project administration, B.d.C. and R.H.; funding acquisition, B.d.C. All authors have read and agreed to the published version of the manuscript.

Funding

R.H. was supported by an Australian Government Research Training Program (RTP) Scholarship. A.M. (Aya Mousa) was supported by a fellowship (grant number: 1161871) from the National Health and Medical Research Council (NHMRC) of Australia. B.d.C was supported by a grant from the CASS (Contributing to Australian Scholarship & Science) Foundation and the Royal Australasian College of Physicians (RACP).

Institutional Review Board Statement

The study was conducted in accordance with the Declaration of Helsinki and approved by the Human Research Ethics Committee of Monash Health (ref: 16061A; 14 April 2016).

Data Availability Statement

Restrictions apply to the availability of some, or all, data generated or analysed during this study to preserve patient confidentiality, or because they were used under licence. The corresponding author will, upon request, detail the restrictions and any conditions under which access to some data may be provided.

Acknowledgments

We acknowledge all the participants of the primary trial for their time and efforts in participation. We acknowledge the funding agencies, including the CASS (Contributing to Australian Scholarship and Science) and Royal Australasian College of Physicians foundations. This is a summary of independent research carried out in conjunction with the NIHR Sheffield Biomedical Research Centre (Translational Neuroscience). The views expressed are those of the authors and not necessarily those of the NHS, the NIHR, or the Department of Health and Social Care (DHSC). Graphical abstract created in BioRender. Hariharan, R. (2026).

Conflicts of Interest

The authors declare no conflicts of interest.

References

  1. United Nations, Department of Economic and Social Affairs, Population Division. World Population Ageing 2020 Highlights: Living Arrangements of Older Persons; ST/ESA/SER.A/451; United Nations: New York, NY, USA, 2020. [Google Scholar]
  2. World Health Organization. Ageing and Health. Available online: https://www.who.int/news-room/fact-sheets/detail/ageing-and-health (accessed on 14 August 2026).
  3. van den Berg, E.; Kloppenborg, R.P.; Kessels, R.P.C.; Kappelle, L.J.; Biessels, G.J. Type 2 Diabetes Mellitus, Hypertension, Dyslipidemia and Obesity: A Systematic Comparison of Their Impact on Cognition. Biochim. Biophys. Acta (BBA)-Mol. Basis Dis. 2009, 1792, 470–481. [Google Scholar] [CrossRef] [Scilit]
  4. Biessels, G.J.; Despa, F. Cognitive decline and dementia in diabetes mellitus: Mechanisms and clinical implications. Nat. Rev. Endocrinol. 2018, 14, 591–604. [Google Scholar] [CrossRef] [Scilit]
  5. Livingston, G.; Huntley, J.; Sommerlad, A.; Ames, D.; Ballard, C.; Banerjee, S.; Brayne, C.; Burns, A.; Cohen-Mansfield, J.; Cooper, C.; et al. Dementia prevention, intervention, and care: 2020 report of the Lancet Commission. Lancet 2020, 396, 413–446, Erratum in Lancet 2023, 402, 1132. https://doi.org/10.1016/S0140-6736(23)02043-3. [Google Scholar] [CrossRef] [Scilit]
  6. Nogueira, G.; Schwanke, C.H.; Closs, V.E.; Luz, B.; Moresco, R.N.; Tatsch, E.; Bochi, G.V.; de Moraes, J.F.; Jacondino, C.B.; Gomes, I.; et al. Inflammation, Oxidation, Caloric Expenditure and Cognitive Impairment in Brazilian Elderly Assisted at Primary Care. Curr. Alzheimer Res. 2016, 13, 1056–1063. [Google Scholar] [CrossRef] [Scilit]
  7. Sugimoto, T.; Nakamura, A.; Kato, T.; Iwata, K.; Saji, N.; Arahata, Y.; Hattori, H.; Bundo, M.; Ito, K.; Niida, S.; et al. Decreased Glucose Metabolism in Medial Prefrontal Areas Is Associated with Nutritional Status in Patients with Prodromal and Early Alzheimer’s Disease. J. Alzheimer’s Dis. 2017, 60, 225–233. [Google Scholar] [CrossRef] [Scilit]
  8. Arnold, S.E.; Arvanitakis, Z.; Macauley-Rambach, S.L.; Koenig, A.M.; Wang, H.-Y.; Ahima, R.S.; Craft, S.; Gandy, S.; Buettner, C.; Stoeckel, L.E.; et al. Brain insulin resistance in type 2 diabetes and Alzheimer disease: Concepts and conundrums. Nat. Rev. Neurol. 2018, 14, 168–181. [Google Scholar] [CrossRef] [Scilit] [PubMed]
  9. American Diabetes Association. Diagnosis and classification of diabetes mellitus. Diabetes Care 2014, 37, S81–S90. [Google Scholar] [CrossRef] [Scilit]
  10. World Health Organization. Definition and Diagnosis of Diabetes Mellitus and Intermediate Hyperglycaemia: Report of a WHO/IDF Consultation; World Health Organization: Geneva, Switzerland, 2006. [Google Scholar]
  11. Dahl, A.; Hassing, L.B.; Fransson, E.I.; Gatz, M.; Reynolds, C.A.; Pedersen, N.L. Body Mass Index Across Midlife and Cognitive Change in Late Life. Int. J. Obes. 2012, 37, 296–302. [Google Scholar] [CrossRef] [Scilit]
  12. Luchsinger, J.A.; Kazemi, E.J.; Sanchez, D.L.; Larkin, M.E.; Valencia, W.M.; Desouza, C.; Carlson, A.L.; Pop-Busui, R.; Seaquist, E.R.; Flórez, H.J.; et al. BMI, Insulin Sensitivity, and Cognition in Early Type 2 Diabetes: The Glycemia Reduction Approaches in Diabetes: A Comparative Effectiveness Study. Obesity 2023, 31, 1812–1824. [Google Scholar] [CrossRef] [Scilit]
  13. Rawlings, A.M.; Sharrett, A.R.; Schneider, A.L.C.; Coresh, J.; Albert, M.; Couper, D.; Griswold, M.; Gottesman, R.F.; Wagenknecht, L.E.; Windham, B.G.; et al. Diabetes in midlife and cognitive change over 20 years: A cohort study. Ann. Intern. Med. 2014, 161, 785–793. [Google Scholar] [CrossRef] [Scilit]
  14. Souza, J.G.; Farias-Itao, D.; Aliberti, M.; Alexandre, T.; Szlejf, C.; Ferri, C.; Lima-Costa, M.; Suemoto, C. Diabetes, Hemoglobin A1c, and Cognitive Performance in Older Adults: Is There Any Impact of Frailty? Evidence From the ELSI-Brazil Study. Braz. J. Med. Biol. Res. 2024, 57, e12939. [Google Scholar] [CrossRef] [Scilit]
  15. Morys, F.; Dadar, M.; Dagher, A. Association Between Midlife Obesity and Its Metabolic Consequences, Cerebrovascular Disease, and Cognitive Decline. J. Clin. Endocrinol. Metab. 2021, 106, e4260–e4274. [Google Scholar] [CrossRef] [Scilit]
  16. Halcox, J.; Misra, A. Type 2 Diabetes Mellitus, Metabolic Syndrome, and Mixed Dyslipidemia: How Similar, How Different, and How to Treat? Metab. Syndr. Relat. Disord. 2015, 13, 1–21. [Google Scholar] [CrossRef] [Scilit]
  17. Athyros, V.G.; Doumas, M.; Imprialos, K.P.; Stavropoulos, K.; Georgianou, E.; Katsimardou, A.; Karagiannis, A. Diabetes and Lipid Metabolism. Hormones 2018, 17, 61–67. [Google Scholar] [CrossRef] [Scilit]
  18. Duong, M.T.; Nasrallah, I.M.; Wolk, D.A.; Chang, C.C.; Chang, T.Y. Cholesterol, Atherosclerosis, and APOE in Vascular Contributions to Cognitive Impairment and Dementia (VCID): Potential Mechanisms and Therapy. Front. Aging Neurosci. 2021, 13, 647990. [Google Scholar] [CrossRef] [Scilit]
  19. Baye, E.; Menon, K.; de Courten, M.P.; Earnest, A.; Cameron, J.; de Courten, B. Does supplementation with carnosine improve cardiometabolic health and cognitive function in patients with pre-diabetes and type 2 diabetes? Study protocol for a randomised, double-blind, placebo-controlled trial. BMJ Open 2017, 7, e017691. [Google Scholar] [CrossRef] [Scilit]
  20. Hariharan, R.; Cameron, J.; Menon, K.; Mesinovic, J.; Jansons, P.; Scott, D.; Lu, Z.X.; de Courten, M.; Feehan, J.; de Courten, B. Carnosine supplementation improves glucose control in adults with pre-diabetes and type 2 diabetes: A randomised controlled trial. Nutr. Metab. Cardiovasc. Dis. 2024, 34, 485–496. [Google Scholar] [CrossRef] [Scilit]
  21. World Health Organization. Classification of Diabetes Mellitus; World Health Organization: Geneva, Switzerland, 2019. [Google Scholar]
  22. National Health & Medical Research Council. Australian Guidelines to Reduce Health Risks from Drinking Alcohol; National Health & Medical Research Council: Canberra, Australia, 2020.
  23. Chan, A.-W.; Tetzlaff, J.M.; Altman, D.G.; Laupacis, A.; Gøtzsche, P.C.; Krleža-Jerić, K.; Hróbjartsson, A.; Mann, H.; Dickersin, K.; Berlin, J.A.; et al. SPIRIT 2013 statement: Defining standard protocol items for clinical trials. Ann. Intern. Med. 2013, 158, 200–207. [Google Scholar] [CrossRef] [Scilit] [PubMed]
  24. von Elm, E.; Altman, D.G.; Egger, M.; Pocock, S.J.; Gøtzsche, P.C.; Vandenbroucke, J.P. The Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) statement: Guidelines for reporting observational studies. Lancet 2007, 370, 1453–1457. [Google Scholar] [CrossRef] [Scilit] [PubMed]
  25. World Medical Association. World Medical Association Declaration of Helsinki: Ethical principles for medical research involving human subjects. JAMA 2013, 310, 2191–2194. [Google Scholar] [CrossRef] [Scilit] [PubMed]
  26. Cambridge Cognition. CANTAB Connect Research: Admin Application User Guide; Cambridge Cognition Ltd.: Cambridge, UK, 2021. [Google Scholar]
  27. Sturman, M.T.; de Leon, C.F.M.; Bienias, J.L.; Morris, M.C.; Wilson, R.S.; Evans, D.A. Body mass index and cognitive decline in a biracial community population. Neurology 2008, 70, 360–367. [Google Scholar] [CrossRef] [Scilit]
  28. West, R.K.; Livny, A.; Ravona-Springer, R.; Bendlin, B.B.; Heymann, A.; Leroith, D.; Liu, X.; Lin, H.-M.; Hochner, H.; Friedlander, Y.; et al. Higher BMI is associated with smaller regional brain volume in older adults with type 2 diabetes. Diabetologia 2020, 63, 2446–2451. [Google Scholar] [CrossRef] [Scilit]
  29. Boccara, E.; Golan, S.; Ganmore, I.; Ravona-Springer, R.; Livny, A.; Ben-Meir, M.; Azuri, Y.; Weller, A.; Beeri, M.S. Sex-specific differences in the association between regional adiposity, cognition, and brain pathology in older adults with Type II Diabetes. Alzheimer’s Dement. 2024, 20, e091659. [Google Scholar] [CrossRef] [Scilit]
  30. Boccara, E.; Golan, S.; Ganmore, I.; Ravona-Springer, R.; Ben-Meir, M.; Livny, A.; Inbar, Y.; Azuri, Y.; Weller, A.; Beeri, M.S. Regional Adiposity, Cognition and Brain Volumes in Older Adults with Type 2 Diabetes: Sex-Specific Associations. Alzheimer’s Dement. 2025, 21, e104389. [Google Scholar] [CrossRef] [Scilit]
  31. Xing, Z.; Long, C.; Hu, X.; Chai, X. Obesity is associated with greater cognitive function in patients with type 2 diabetes mellitus. Front. Endocrinol. 2022, 13, 953826. [Google Scholar] [CrossRef] [Scilit]
  32. Gogniat, M.A.; Robinson, T.L.; Mewborn, C.M.; Jean, K.R.; Miller, L.S. Body mass index and its relation to neuropsychological functioning and brain volume in healthy older adults. Behav. Brain Res. 2018, 348, 235–240. [Google Scholar] [CrossRef] [Scilit]
  33. Luchsinger, J.A.; Pan, Q.; Knowler, W.C.; Munshi, M.; Watson, K.; Gadde, K.M.; Schlögl, M.; Carmichael, O.T. Association of Body Mass Index and Cognitive Performance in the Diabetes Prevention Program Outcomes Study. Alzheimer’s Dement. 2025, 21, e098699. [Google Scholar] [CrossRef] [Scilit]
  34. Maan, H.; Meo, S.; Rouq, F.; Meo, I. Impact of Glycated Hemoglobin (HbA1c) on cognitive functions in Type 2 diabetic patients. Eur. Rev. Med. Pharmacol. Sci. 2021, 25, 5978–5985. [Google Scholar] [CrossRef] [Scilit]
  35. Dybjer, E.; Nilsson, P.M.; Engström, G.; Helmer, C.; Nägga, K. Pre-diabetes and diabetes are independently associated with adverse cognitive test results: A cross-sectional, population-based study. BMC Endocr. Disord. 2018, 18, 91. [Google Scholar] [CrossRef] [Scilit]
  36. Zheng, F.; Yan, L.; Yang, Z.; Zhong, B.; Xie, W. HbA1c, diabetes and cognitive decline: The English Longitudinal Study of Ageing. Diabetologia 2018, 61, 839–848. [Google Scholar] [CrossRef] [Scilit]
  37. Casagrande, S.S.; Lee, C.; Stoeckel, L.E.; Menke, A.; Cowie, C.C. Cognitive function among older adults with diabetes and prediabetes, NHANES 2011–2014. Diabetes Res. Clin. Pract. 2021, 178, 108939. [Google Scholar] [CrossRef] [Scilit]
  38. Cheng, G.; Changquan, H.; Deng, H.; Wang, H. Diabetes as a Risk Factor for Dementia and Mild Cognitive Impairment: A Meta-analysis of Longitudinal Studies. Intern. Med. J. 2012, 42, 484–491. [Google Scholar] [CrossRef] [Scilit] [PubMed]
  39. Geijselaers, S.L.C.; Sep, S.J.S.; Stehouwer, C.D.A.; Biessels, G.J. Glucose regulation, cognition, and brain MRI in type 2 diabetes: A systematic review. Lancet Diabetes Endocrinol. 2015, 3, 75–89. [Google Scholar] [CrossRef] [Scilit]
  40. Hung, T.-H.; Chen, V.C.-H.; Chuang, Y.-C.; Hsu, Y.-H.; Wu, W.-C.; Tsai, Y.-H.; McIntyre, R.S.; Weng, J.-C. Investigating the effect of hypertension on vascular cognitive impairment by using the resting-state functional connectome. Sci. Rep. 2024, 14, 4580. [Google Scholar] [CrossRef] [Scilit]
  41. Xie, Z.; Guo, J.; Kang, Y.; Ou, X. Mismatch between GLUTs and glucose levels causes neuronal damage during glucose fluctuations. Med. Hypotheses 2018, 112, 37–39. [Google Scholar] [CrossRef] [Scilit]
  42. Luévano-Contreras, C.; Gómez-Ojeda, A.; Macías-Cervantes, M.H.; Garay-Sevilla, M.E. Dietary Advanced Glycation End Products and Cardiometabolic Risk. Curr. Diab Rep. 2017, 17, 63. [Google Scholar] [CrossRef] [Scilit]
  43. Prasad, C.; Davis, K.E.; Imrhan, V.; Juma, S.; Vijayagopal, P. Advanced Glycation End Products and Risks for Chronic Diseases: Intervening Through Lifestyle Modification. Am. J. Lifestyle Med. 2019, 13, 384–404. [Google Scholar] [CrossRef] [Scilit]
  44. Gorska-Ciebiada, M.; Saryusz-Wolska, M.; Borkowska, A.; Ciebiada, M.; Loba, J. C-Reactive Protein, Advanced Glycation End Products, and Their Receptor in Type 2 Diabetic, Elderly Patients with Mild Cognitive Impairment. Front. Aging Neurosci. 2015, 7, 209. [Google Scholar] [CrossRef] [Scilit]
  45. Chen, J.; Mooldijk, S.S.; Licher, S.; Waqas, K.; Ikram, M.K.; Uitterlinden, A.G.; Zillikens, M.C.; Ikram, M.A. Assessment of Advanced Glycation End Products and Receptors and the Risk of Dementia. JAMA Netw. Open 2021, 4, e2033012. [Google Scholar] [CrossRef] [Scilit]
  46. Chen, X.; Huang, Y.; Xiong, X. Mechanisms underlying cognitive impairment and management strategies in type 2 diabetes. Front. Endocrinol. 2025, 16, 1655768. [Google Scholar] [CrossRef] [Scilit] [PubMed]
  47. Spinelli, M.; Fusco, S.; Grassi, C. Brain Insulin Resistance and Hippocampal Plasticity: Mechanisms and Biomarkers of Cognitive Decline. Front. Neurosci. 2019, 13, 788. [Google Scholar] [CrossRef] [Scilit]
  48. Lin, B.; Ginsberg, M.D.; Busto, R. Hyperglycemic but not normoglycemic global ischemia induces marked early intraneuronal expression of β-amyloid precursor protein. Brain Res. 2001, 888, 107–116. [Google Scholar] [CrossRef] [Scilit]
  49. Parthasarathy, V.; Frazier, D.T.; Bettcher, B.M.; Jastrzab, L.; Chao, L.; Reed, B.; Mungas, D.; Weiner, M.; DeCarli, C.; Chui, H.; et al. Triglycerides are negatively correlated with cognitive function in nondemented aging adults. Neuropsychology 2017, 31, 682–688. [Google Scholar] [CrossRef] [Scilit]
  50. Crichton, G.E.; Elias, M.F.; Davey, A.; Sullivan, K.J.; Robbins, M.A. Higher HDL Cholesterol Is Associated with Better Cognitive Function: The Maine-Syracuse Study. J. Int. Neuropsychol. Soc. 2014, 20, 961–970. [Google Scholar] [CrossRef] [Scilit]
  51. Cai, R.; Han, J.; Sun, J.; Huang, R.; Tian, S.; Shen, Y.; Wang, S. Effects of ABCA1 R219K Polymorphism and Serum Lipid Profiles on Mild Cognitive Impairment in Type 2 Diabetes Mellitus. Front. Aging Neurosci. 2017, 9, 257. [Google Scholar] [CrossRef] [Scilit]
  52. Ma, L.; Yuan, Y.-X.; Cheng, F.-J.; Liu, Y.; Wei, Q.; Peng, Y.-F.; Wang, Y. The association between blood lipids and cognitive impairment in type 2 diabetes mellitus. Eur. J. Med. Res. 2024, 29, 1. [Google Scholar] [CrossRef] [Scilit]
  53. Chakraborty, A.; Hegde, S.; Praharaj, S.K.; Prabhu, K.; Patole, C.; Shetty, A.K.; Mayya, S.S.; Acharya, R.V.; Hande, H.M.; Prabhu, M.M.; et al. Age Related Prevalence of Mild Cognitive Impairment in Type 2 Diabetes Mellitus Patients in the Indian Population and Association of Serum Lipids with Cognitive Dysfunction. Front. Endocrinol. 2021, 12, 798652. [Google Scholar] [CrossRef] [Scilit]
  54. Ansari, T.; Sawane, M. Association of Hyperlipidemia and Hyperglycemia with Cognitive Function in Type 2 Diabetes: A Cross-Sectional Analysis. Cureus 2024, 16, e72227. [Google Scholar] [CrossRef] [Scilit]
  55. Dimache, A.M.; Șalaru, D.L.; Sascău, R.; Stătescu, C. The Role of High Triglycerides Level in Predicting Cognitive Impairment: A Review of Current Evidence. Nutrients 2021, 13, 2118. [Google Scholar] [CrossRef] [Scilit]
  56. Denimal, D. Antioxidant and Anti-Inflammatory Functions of High-Density Lipoprotein in Type 1 and Type 2 Diabetes. Antioxidants 2024, 13, 57. [Google Scholar] [CrossRef] [Scilit]
  57. Rajič Bumber, J.; Rački, V.; Mežnarić, S.; Pelčić, G.; Mršić-Pelčić, J. Clinical Significance of APOE4 Genotyping: Potential for Personalized Therapy and Early Diagnosis of Alzheimer’s Disease. J. Clin. Med. 2025, 14, 6047. [Google Scholar] [CrossRef] [Scilit]
  58. Emrani, S.; Arain, H.A.; DeMarshall, C.; Nuriel, T. APOE4 is associated with cognitive and pathological heterogeneity in patients with Alzheimer’s disease: A systematic review. Alzheimer’s Res. Ther. 2020, 12, 141. [Google Scholar] [CrossRef] [Scilit]
  59. Saadati, S.; Cameron, J.; Menon, K.; Hodge, A.; Lu, Z.X.; de Courten, M.; Feehan, J.; de Courten, B. Carnosine Did Not Affect Vascular and Metabolic Outcomes in Patients with Prediabetes and Type 2 Diabetes: A 14-Week Randomized Controlled Trial. Nutrients 2023, 15, 4835. [Google Scholar] [CrossRef] [Scilit]
  60. Kabthymer, R.H.; Feehan, J.; Mousa, A.; Aldini, G.; de Courten, M.; Cameron, J.; de Courten, B. The Effect of Carnosine Supplementation on Depressive Symptoms and Health-Related Quality of Life in Individuals with Prediabetes and Well-Controlled Type 2 Diabetes: A Randomized Controlled Trial. Food Sci. Nutr. 2026, 14, e71514. [Google Scholar] [CrossRef] [Scilit]
Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content.

Article Metrics

Citations

Article Access Statistics

Multiple requests from the same IP address are counted as one view.