Brain aging shows nonlinear transitions, suggesting a midlife "critical window" for metabolic intervention

Abstract

Understanding the key drivers of brain aging is essential for effective prevention and treatment of neurodegenerative diseases. Here, we integrate human brain and physiological data to investigate underlying mechanisms. Functional MRI analyses across four large datasets (totaling 19,300 participants) show that brain networks not only destabilize throughout the lifetime but do so along a nonlinear trajectory, with consistent temporal "landmarks" of brain aging starting in midlife (40s). Comparison of metabolic, vascular, and inflammatory biomarkers implicate dysregulated glucose homeostasis as the driver mechanism for these transitions. Correlation between the brain's regionally heterogeneous patterns of aging and gene expression further supports these findings, selectively implicating GLUT4 (insulin-dependent glucose transporter) and APOE (lipid transport protein). Notably, MCT2 (a neuronal, but not glial, ketone transporter) emerges as a potential counteracting factor by facilitating neurons' energy uptake independently of insulin. Consistent with these results, an interventional study of 101 participants shows that ketones exhibit robust effects in restabilizing brain networks, maximized from ages 40 to 60, suggesting a midlife "critical window" for early metabolic intervention.

Keywords: aging; brain; fMRI; insulin; neuron.

Fig. 1.

Metabolic changes predominate during the acceleration phase of brain aging, as depicted by nonlinear lifespan trends in brain network instability. (A) The HCP-A functional neuroimaging dataset revealed a sigmoidal trend in the destabilization of brain networks across the lifespan. Curve fitting was utilized to derive landmark age points α (onset of destabilization), I (age of fastest destabilization), and β₁ (age of destabilization plateau). (B) Mean changes for biomarkers representing metabolic health (HbA1c), vascular health [systolic blood pressure (SysBP), diastolic blood pressure (DiaBP)], and inflammatory state [blood levels of C-reactive protein (CRP)] were evaluated across age groups defined with respect to their relationship to the α and I landmarks (N_age<α = 111, N_α≤age<I = 281, $N_{I\le \text{age}<{\beta }_1}=202$). Error bars represent 95% CIs for the mean changes between consecutive age groups, normalized by the variance of each biomarker across the entire age range. The α landmark was associated most strongly with an increase in HbA1c (t = 4.8, P = 4E-6), while the I landmark was associated most strongly with an increase in systolic blood pressure (t = 5.7, P = 3E-8). In contrast, blood CRP, indicative of inflammation, showed no significant changes around either landmark. n.s., not statistically significant, *P ≤ 0.05; **P ≤ 0.01; ***P ≤ 0.001; ****P ≤ 0.00001.

Fig. 2.

Gene expression brain maps highlight neuronal insulin resistance as a driver of brain aging, counteracted by neuronal ketone transport. (A) Diagram illustrating the methodology for establishing similarity between the aging pattern in brain function and gene expression distribution. (B) Color-coded tiles show Spearman correlations between the pattern of age-related changes in ALFF and the spatial distribution of gene expression associated with probable key mechanisms underlying brain aging. Each shown point corresponds to a cortical functional region of interest. Labels indicate expressed proteins rather than the underlying genes. Investigated mechanisms encompassed cellular glucose uptake (genes translated to GLUT1, GLUT3, GLUT4), ketone/lactate uptake (MCT1, MCT2), lipid-transport (APOE), vascular function (NOS1, ACE, ET-1, VEGFA, VEGFB, VEGFR1), inflammation (TNF, TNF receptor 1, IL-1β, IL-6, IL-23A, P2RX7), and housekeeping and cytoskeletal structure as unrelated controls (ACTB, NF-L, GAPDH, PGK1, EEF1A1, RPL13A). The spatial correlations were computed in both UKB and HCP-A datasets. (C) Scatter plots depict associations between brain aging effects and gene expression (log-scaled) for genes that replicated in both the UKB and HCP-A datasets. These included genes encoding for GLUT4, MCT2, and APOE. The functional data shown are from the UKB dataset.

Fig. 3.

D-β-hydroxybutyrate circumvents insulin resistance to reverse brain network destabilization during the accelerated phase of brain aging. (A) Oxidative energy-providing pathways of the brain. Neurons can utilize glucose, ketones, or lactate (via astrocytes) for energy, with distinct transporters facilitating their uptake. D-βHB is one of the primary ketones readily metabolized by neurons. The uptake of D-βHB, along with lactate, relies on insulin-signaling independent monocarboxylate transporters (MCT). Graphics were created with BioRender.com. (B) Experiment design of the metabolic intervention dataset. Each participant was scanned two separate times, time-locked to eliminate diurnal variability, with the D-βHB ketone monoester individually weight-dosed (395 mg/kg). Each individual’s glucose dose was then calorie matched to their D-βHB ketone monoester dose. (C) Baseline (fasting) subtracted effects of the two metabolic interventions involving glucose and D-βHB ketone monoester on brain network instability. D-βHB stabilizes brain networks in age groups 20 to 39 (P = 0.01) and 40 to 59 (P = 0.00003) but not in 60 to 79 (P = 0.4). The administration of calorically matched glucose did not have significant effects. GLC: glucose vs. fasting, D-βHB: D-βHB vs. fasting, Δ: glucose vs. D-βHB. (D) Neuronal metabolism “bends” before it “breaks?” Nonlinear threshold effects of functional network destabilization, hypothesized to result from insulin resistance disruption of neuronal connectivity. n.s., not statistically significant, *P ≤ 0.05; **P ≤ 0.01; ***P ≤ 0.001; ****P ≤ 0.00001.