Triangulating evidence from observational and Mendelian randomization studies of ketone bodies for cognitive performance

The human brain primarily utilizes glucose for energy, but ketone bodies (KBs) can also cross the blood-brain barrier and serve as an alternative energy source, particularly when glucose availability is reduced [1]. 3-β-hydroxybutyrate (BOHBUT), a major KB, has been implicated in cognitive enhancement [2], potentially through various mechanisms including improved brain energy supply during low glucose states [3], neurotransmitter regulation [4, 5], reduction of oxidative stress [6], and protection against pathogenic proteins like β-amyloid (Aβ) and phosphorylated tau [7, 8], implicated in neurodegenerative diseases. Preclinical and some clinical studies suggest potential cognitive benefits of BOHBUT, including improved memory in type 2 diabetes patients [9], older adults [10], individuals with mild cognitive impairment [11], and those with mild-to-moderate Alzheimer's disease (AD) [8, 12]. These studies also indicate a potential neuroprotective effect of KB treatment [9, 11, 12]. Experimental evidence in mice shows BOHBUT promoting hippocampal brain-derived neurotrophic factor (BDNF) expression [4, 5], a protein vital for neuronal survival, neurotransmitter regulation, synaptic plasticity, and memory formation [13, 14]. BOHBUT may enhance learning and memory via improved neuronal mitochondrial energy [15]. However, the evidence of BOHBUT's impact on cognitive performance and neuroprotection in humans is inconclusive. This study aims to address this knowledge gap through a triangulation approach, combining observational and Mendelian randomization (MR) studies to establish a more robust understanding of the causal relationship between BOHBUT and cognitive function. The study will investigate the relationship between BOHBUT and cognitive performance, and also examine its potential role in AD, a condition strongly associated with cognitive impairment. Acetoacetate (ACACE), a precursor of BOHBUT, will also be considered in the analysis, although it is a less reliable marker of KBs due to its instability [18].

Existing literature presents a mixed picture regarding the impact of ketone bodies on cognitive function. Some studies have demonstrated a positive effect of BOHBUT supplementation on cognitive memory in various populations. However, inconsistencies across studies and a lack of large-scale, rigorously designed investigations have limited the ability to draw definitive conclusions about the causal role of BOHBUT in improving cognitive performance or preventing/delaying cognitive decline. This research gap highlights the need for a comprehensive analysis that integrates multiple lines of evidence to assess the true nature of the association and potential causality between BOHBUT and cognitive function and neurodegenerative diseases. The review of existing literature shows that the existing studies have methodological limitations including small sample sizes, variability in study designs, and potential confounding factors. The triangulation approach using both observational and MR studies helps to mitigate some of these limitations by providing complementary evidence and reducing bias from confounding.

This study employed a triangulation strategy, combining observational and Mendelian randomization (MR) analyses to investigate the association and causality between ketone bodies (KBs) and cognitive function. **Observational Study:** The Whitehall II study (WHII), comprising 5506 participants aged ≥45 years, served as the dataset. Circulating BOHBUT and ACACE levels were measured using nuclear magnetic resonance (NMR) spectroscopy. Three cognitive function scores (word memory, verbal fluency, verbal meaning) were assessed. A principal component analysis generated a "general" cognitive performance score. Multiple linear regression analyses examined the associations between categorized (high vs. normal) and continuous KB levels and cognitive function scores, adjusting for age, sex, diabetes, smoking, alcohol consumption, waist-to-hip ratio, and occupational position. **Mendelian Randomization (MR) Study:** A two-sample MR analysis was conducted using publicly available summary statistics from genome-wide association studies (GWAS). Exposure data (genetic instruments and summary statistics for BOHBUT and ACACE) were from an in-house KB meta-analysis (45,031 participants). Outcome data for cognitive performance (257,841 participants) and AD (54,162 participants) were sourced from the Social Science Genetic Association Consortium (SSGAC) and the International Genomics of Alzheimer's Project (IGAP), respectively. Both strong (P < 5 × 10⁻⁸) and suggestive (P < 1 × 10⁻⁵) sets of instrumental variables (SNPs) were used. The inverse variance weighted (IVW) method was the primary MR method, with sensitivity analyses using weighted median, weighted mode, MR-Egger regression, and MR-PRESSO to assess robustness. Cis-MR analysis on OXCT1, a gene related to KB catabolism, was also performed. Both the observational and MR analyses accounted for potential confounders through statistical adjustments and sensitivity analyses. The study used established statistical methods and software packages, ensuring rigor and transparency in the analysis.

The observational study revealed a positive association between BOHBUT levels (≥0.32 mmol/L) and general cognitive function (β = 0.26, P = 9.74 × 10⁻³). MR analysis using the strong set of instrumental variables showed a positive causal effect of genetically predicted BOHBUT on cognitive performance (IVW: β = 7.89 × 10⁻², P = 1.03 × 10⁻²; weighted median: β = 8.65 × 10⁻², P = 9.60 × 10⁻³). This effect remained consistent using a suggestive set of SNPs. A nominal protective effect of BOHBUT on AD was also observed (IVW: β = −0.31, OR = 0.74, P = 3.06 × 10⁻²). Cis-MR analysis of OXCT1 provided little evidence of a direct therapeutic effect on cognitive impairment. For ACACE, the observational analysis showed a positive association with general cognitive function and verbal meaning; however, no causal effect on cognitive performance or AD was found in the MR analysis. Sensitivity analyses, including weighted median and weighted mode methods, were generally supportive of the primary findings. The study noted potential confounders, but sensitivity analyses generally supported the robust nature of the findings.

The study's findings provide converging evidence from observational and MR studies supporting a beneficial effect of BOHBUT on cognitive performance. The positive association observed in the observational study is strengthened by the causal inference provided by the MR analysis, reducing the influence of confounding factors and reverse causation. The observed protective effect of BOHBUT against AD, though nominal, lends further support to the hypothesis of a positive impact on cognitive function. This study's findings align with previous research suggesting the cognitive benefits of low-carbohydrate diets and intermittent fasting, potentially through increased KB levels. While this study provides strong evidence of an association and suggests causality between elevated BOHBUT and improved cognitive function, the specific therapeutic levels of KBs require further investigation, particularly in the context of different stages of cognitive decline. The study highlights the need for future RCTs to assess the short-term effects of interventions aimed at elevating KB levels. The non-linear relationship between KB levels and cognitive function warrants further exploration, especially with access to individual patient data that can enable a non-linear MR analysis. The focus on specific cognitive functions assessed in this study also suggests further investigations are required to assess the causal effect of KBs on broader aspects of brain function.

This study, using a triangulation approach combining observational and Mendelian randomization studies, demonstrates an association and provides evidence of a causal effect of increased 3-β-hydroxybutyrate (BOHBUT) on improved cognitive performance and a possible protective effect against Alzheimer's disease. These findings support the hypothesis that higher BOHBUT levels may improve cognitive function and reduce the risk of cognitive decline and AD. Further research, including randomized controlled trials, is needed to confirm these findings and explore the optimal therapeutic range of BOHBUT for cognitive enhancement.

Several limitations should be considered when interpreting the findings. Measurement error in KB levels due to variability in fasting periods prior to blood collection is a potential concern. The limited variance explained by the instrumental SNPs in the MR analysis highlights the need for further discovery of underlying KB-associated SNPs in larger studies. The assumption of a linear relationship between KB levels and cognitive performance may not fully capture potential non-linear effects. Finally, the study's sample predominantly consisted of Caucasian individuals, which may limit the generalizability of the findings to other populations.