Prediabetes is associated with early-emerging and persistent cognitive differences in youth with overweight and obesity

This two-year longitudinal study demonstrates that youth with overweight or obesity who have prediabetes exhibit early-emerging, persistent, and adiposity-independent deficits in IQ, executive function, and processing speed, suggesting that neurocognitive impairment is a feature of metabolic risk that precedes the onset of overt type 2 diabetes.

The Gist

The Big Picture: A "Sugar Glitch" in the Young Brain

Imagine your body's energy system as a massive power grid. Insulin is the key that unlocks the doors to let electricity (sugar/glucose) into your cells to power them up. In prediabetes, the locks are getting rusty. The key doesn't turn as smoothly, so the power grid gets a bit clogged.

For a long time, doctors thought this "rusty lock" problem only messed up the brain in older people, like a slow leak in an old house. But this study asks a scary question: What if this leak starts damaging the house while it's still being built?

The researchers looked at teenagers and young adults with obesity. They split them into two groups:

1. The "Smooth Operators": Kids with obesity, but their insulin keys still work perfectly.
2. The "Rusty Locks": Kids with obesity who have prediabetes (their insulin keys are starting to stick).

The Discovery: The "Rusty Locks" Group is Already Behind

The study found that even before these young people developed full-blown diabetes, their brains were already showing signs of trouble. It wasn't just about being overweight; it was specifically about the sugar metabolism.

Think of the brain like a high-speed computer. The "Rusty Locks" group had a slower processor. Specifically, they struggled with:

• Executive Function: Like a manager trying to organize a chaotic office. They had trouble planning and focusing.
• Psychomotor Speed: Like a runner who is slightly slower off the starting blocks. Their reaction times were sluggish.
• Visuospatial Processing: Like trying to assemble a 3D puzzle while looking at it through a foggy window. They had trouble understanding how things fit together in space.

The "IQ" Twist: Interestingly, the "Rusty Locks" group also had slightly lower overall IQ scores. But even when the researchers accounted for this, the "Rusty Locks" group still had a specific hard time with Spatial Working Memory (remembering where things are in space). It's as if the metabolic issue created a specific "glitch" in their brain's GPS.

The Time Travel Test: Is it Fixable?

The researchers didn't just take a snapshot; they took a movie. They followed these kids for two years.

• The Question: If a kid's sugar levels got better (the lock got less rusty), would their brain catch up?
• The Answer: No.

The cognitive differences were stable. Whether a kid's metabolism got better, worse, or stayed the same over two years, the gap in brain performance between the two groups remained. It's like a runner who starts a race a few steps behind; even if they run hard for two years, they don't magically catch up to the person who started ahead. The brain differences seem to be "locked in" early on.

The Brain Structure: A Construction Site Anomaly

The researchers also looked at the physical brain using MRI scans.

• Normal Development: In healthy brain development, the brain's surface area naturally shrinks a little bit as you get older. Think of this like a city pruning its streets to make them more efficient as the population grows.
• The Anomaly: The "Smooth Operators" (normal glucose) showed this healthy pruning. The "Rusty Locks" (prediabetes) did not. Their brain surface area didn't shrink the way it should have.

The Metaphor: Imagine a forest that is supposed to thin out naturally as trees mature. The "Smooth Operators" are thinning out correctly. The "Rusty Locks" are holding onto too many saplings, cluttering the forest. This suggests that prediabetes might be stalling the brain's natural maturation process, keeping it in a less efficient state.

The "Insulin Spray" Experiment

To see how the brain reacts to insulin, the researchers gave the participants a special spray of insulin through their noses while they were in the MRI scanner.

• The Finding: In the "Smooth Operators," the insulin spray made the brain light up in specific areas (like the parietal sulcus, which handles spatial tasks).
• The Glitch: In the "Rusty Locks" group, the brain didn't respond as strongly to the insulin spray. It's as if the brain's "insulin antenna" was broken. The brain wasn't listening to the signal that says, "Hey, we need to focus and process this information."

The Takeaway: Why This Matters

This study is a wake-up call. It tells us that metabolic health is brain health, and this connection starts much earlier than we thought.

1. It's not just "being fat": The brain issues are linked specifically to the sugar problem, not just the weight.
2. It happens early: The damage starts in the "prediabetes" stage, long before a diabetes diagnosis.
3. It's persistent: These brain differences don't just disappear if you fix your diet for a few months. The brain seems to have adapted to the metabolic stress in a way that is hard to reverse quickly.

The Bottom Line: If we want to protect the future intelligence and mental sharpness of our youth, we can't just wait until they are old and sick. We need to fix the "rusty locks" (metabolic health) while the brain is still under construction.

Technical Summary

1. Problem Statement

Type 2 diabetes (T2D) is a well-established risk factor for cognitive decline and dementia in older adults. However, it remains unclear whether these deficits are caused by:

• Direct metabolic effects of chronic hyperglycemia and insulin resistance.
• Secondary consequences of comorbidities (e.g., cardiovascular disease) that accumulate with age.
• Shared risk factors like obesity and poor diet.
• Reverse causality (poor cognition leading to metabolic dysfunction).

Most existing literature focuses on older adults, leaving a gap in understanding how metabolic dysfunction impacts the developing brain during critical neurodevelopmental windows. Specifically, it is unknown if cognitive and structural brain differences emerge at the prediabetes stage in youth, independent of adiposity, and whether these differences are stable or reversible over time.

2. Methodology

Study Design and Population:

• Design: A 2-year longitudinal observational study.
• Participants: Adolescents (ages 8–20) with overweight/obesity (BMI >85th percentile) recruited from the Pathogenesis of Youth Onset Diabetes (PYOD) study.
• Sample Size: 69 participants completed baseline assessments; 42 completed the 2-year follow-up (attrition due to COVID-19 and contraindications like dental braces).
• Grouping: Participants were classified at baseline into preT2D+ (prediabetes, $n=22$ at baseline) and preT2D- (normal glucose control, $n=44$ at baseline) based on OGTT, fasting glucose, and HbA1c criteria.

Data Collection:

• Metabolic Phenotyping: Oral Glucose Tolerance Test (OGTT), fasting insulin/glucose, HbA1c, HOMA-IR, and Matsuda Index (peripheral insulin sensitivity). Body composition (BMI, fat mass, visceral/subcutaneous fat) was measured via BIA and MRI.
• Cognitive Assessment: Comprehensive neuropsychological testing including:
  • IQ: WISC-V (under 16) or WAIS-IV (16+).
  • Domains: Executive function (Spatial Working Memory, Stop Signal Task, Digit Span), Psychomotor speed (Reaction Time, Coding, Trail Making Test), Visuospatial processing (Block Design, VSPLOT), and Memory.
• Neuroimaging:
  • Structural MRI: T1-weighted scans processed with FreeSurfer to measure cortical thickness, surface area, and hippocampal volume.
  • Functional MRI (rsfMRI): Resting-state fMRI performed before and after Intranasal Insulin (INI) administration (1 IU/kg, max 60 IU) vs. Placebo.
  • Imaging Analyses:
    • fALFF: Fractional Amplitude of Low Frequency Fluctuations to assess resting-state brain activity changes in response to insulin.
    • Connectivity: Seed-based connectivity analysis using the caudate (dopaminergic hub) to assess corticostriatal circuitry.

Statistical Analysis:

• Linear regressions and Linear Mixed Effects Models (LMM) were used to assess group differences and longitudinal changes.
• Covariates included age, sex, and BMI z-score (to control for adiposity).
• Multiple comparisons were corrected using the Benjamini-Hochberg procedure (BHC).

3. Key Contributions

• Early Emergence: Demonstrates that neurocognitive deficits associated with metabolic dysfunction are present at the prediabetes stage in youth, prior to the onset of overt T2D and chronic comorbidities.
• Independence from Adiposity: Shows that these cognitive and structural differences persist even after statistically controlling for BMI and body fat percentage, suggesting metabolic dysfunction itself drives the impairment.
• Stability: Provides longitudinal evidence that these cognitive deficits are stable over a 2-year period and do not fluctuate significantly with short-term changes in glucose tolerance or adiposity.
• Neurodevelopmental Deviation: Identifies that youth with prediabetes fail to exhibit the expected developmental reduction in cortical surface area seen in healthy controls, suggesting altered neurodevelopmental trajectories.
• Mechanistic Insight: Links peripheral insulin resistance to altered central insulin sensitivity in specific brain regions (intraparietal sulcus) and suggests a role for insulin-dopamine interactions in corticostriatal circuits.

4. Key Results

Cross-Sectional Findings (Baseline):

• Cognition: The preT2D+ group had significantly lower Full Scale IQ (mean difference ~7 points) and deficits in executive function (Spatial Working Memory), psychomotor speed (Reaction Time), and visuospatial processing (Block Design, VSPLOT) compared to preT2D-.
  • Note: When IQ was added as a covariate, most domain-specific deficits were attenuated, except for Spatial Working Memory, suggesting a broad global cognitive shift rather than isolated domain deficits.
• Metabolic Correlates: Reduced peripheral insulin sensitivity (Matsuda Index) was negatively associated with processing speed (Reaction Time, Trail Making Test A) across the entire sample.
• Brain Structure: PreT2D+ youth showed a trend toward smaller total cortical surface area, though this did not survive strict multiple comparison correction at baseline.

Longitudinal Findings (2-Year Follow-up):

• Cognitive Stability: The cognitive gap between preT2D+ and preT2D- groups remained stable over 2 years. Changes in metabolic status (improvement or worsening of prediabetes) did not correlate with changes in cognitive performance magnitude.
• Structural Divergence: A significant Group × Time interaction was found for Total Cortical Surface Area.
  • PreT2D- youth showed the expected developmental reduction in surface area.
  • PreT2D+ youth did not show this reduction, indicating a deviation from typical neurodevelopmental pruning.

Functional Imaging (Exploratory):

• Central Insulin Sensitivity: Higher peripheral insulin sensitivity (Matsuda Index) was positively correlated with increased fALFF (brain activity) in the left intraparietal sulcus in response to intranasal insulin.
• Cognition-Brain Links:
  • Insulin sensitivity in the right medial prefrontal cortex was inversely associated with reaction time (better sensitivity = faster speed).
  • Caudate Connectivity: Greater insulin-induced connectivity between the caudate and the intraparietal sulcus/precuneus was associated with fewer errors in Spatial Working Memory. This supports the hypothesis that insulin action on dopaminergic circuits influences cognitive function.

5. Significance and Implications

• Clinical: The findings suggest that neurocognitive screening and metabolic intervention should begin earlier, potentially at the prediabetes stage in youth, to prevent long-term cognitive decline.
• Theoretical: The study challenges the view that cognitive decline in diabetes is solely a result of long-term vascular damage or aging. Instead, it posits that metabolic dysfunction disrupts neurodevelopmental trajectories early in life.
• Mechanism: The link between peripheral insulin resistance, central insulin sensitivity (specifically in the intraparietal sulcus), and dopaminergic signaling (caudate connectivity) provides a potential biological pathway for how metabolic risk translates to cognitive impairment.
• Limitations: The study acknowledges a reduced sample size due to the pandemic and motion artifacts in fMRI, which limited the power of some functional analyses. The lack of a healthy-weight control group means the specific interaction between obesity and prediabetes cannot be fully disentangled.

Conclusion: Prediabetes in youth with overweight/obesity is associated with a broad, stable, and early-emerging neurocognitive disadvantage that is independent of adiposity and linked to altered neurodevelopmental trajectories and central insulin sensitivity.