Association between the atherogenic index of plasma and cognitive impairment

Yiying Li, Yuke Zhang, Xinyu Yang, Yuehong NiScientific 2026 · DOI

This longitudinal study of nearly 3,000 adults reveals a surprising "inverted U-shaped" link between the Atherogenic Index of Plasma (AIP) and cognitive decline. While moderate AIP levels significantly increase the risk of impairment, extremely high levels appear paradoxically protective, challenging traditional linear models of how lipid health impacts the brain.

What Problem Does This Paper Address?

As global populations age, cognitive impairment and dementia have become a primary public health challenge, particularly in China, which currently houses approximately 17 million dementia patients. While metabolic disorders like dyslipidemia (unhealthy lipid levels) are known risk factors, traditional markers like total cholesterol or LDL often show limited predictive power for cognitive decline.

The Atherogenic Index of Plasma (AIP)—calculated as $\log_{10}(\text{Triglycerides} / \text{HDL-C})$—has emerged as a more sensitive indicator of atherosclerosis and overall lipid balance. However, previous research on AIP and cognition has mostly been cross-sectional (capturing only a single point in time) and assumed a simple linear relationship (i.e., "higher AIP is always worse"). This paper seeks to fill that gap by investigating the long-term, non-linear relationship between AIP and cognitive health over a decade.

Methods

The researchers employed several sophisticated statistical techniques to analyze longitudinal data:

  • Cox Proportional Hazards Model: Used to determine the independent risk factors for cognitive impairment over time while adjusting for confounders like age, sex, education, and BMI.
  • Restricted Cubic Splines (RCS): This is the study's methodological highlight. Unlike standard regression, RCS allows for the modeling of complex, non-linear "dose-response" relationships, which is how the authors discovered the "inverted U" curve.
  • Kaplan-Meier Survival Analysis: Used to compare the cumulative "survival" (remaining free of cognitive impairment) across different AIP quartiles.
  • Mini-Mental State Examination (MMSE): Cognitive function was assessed using this standardized scale, with cut-offs adjusted for the participants' educational levels to ensure accuracy.

Datasets

The study utilized data from the China Health and Retirement Longitudinal Study (CHARLS), a high-quality national database.

  • Timeline: 2011 to 2020 (spanning five survey cycles).
  • Participants: 2,971 individuals aged 45 and older who were cognitively healthy at the start of the study.
  • Demographics: Included 1,550 females and 1,421 males.

Key Insights and Ideas

  • The Inverted U-Shape: The most striking finding is that the risk of cognitive impairment does not rise indefinitely with AIP. Instead, it peaks in the middle range and then drops significantly at very high levels.
  • The "Danger Zone": The risk was notably increased when AIP values fell between 0.205 and 0.423.
  • The High-AIP Paradox: In the highest quartile (AIP > 0.64), the risk of cognitive impairment actually decreased. The authors propose an "adaptive protection" hypothesis: extremely high triglycerides might provide necessary energy substrates or specific neuroprotective lipid components (like long-chain polyunsaturated fatty acids) that help maintain neuronal membranes, offsetting the damage from atherosclerosis.
  • Gender and Education Gaps: Women faced a 1.24x higher risk than men, likely due to estrogen loss during menopause affecting the hippocampus. Conversely, higher education was a massive protective factor; those with a high school education or above had only ~10% of the risk compared to illiterate participants.

Key Results

  • Incidence: Over the 10-year follow-up, 40.46% of men and 54.45% of women developed new-onset cognitive impairment.
  • Hazard Ratios (HR):
    • Moderate AIP (T2 group): HR = 1.412 (41% higher risk than the lowest group).
  • Extreme High AIP (T4 group): HR = 0.475 (a ~52% reduction in risk compared to the lowest group).
  • Survival Probability: Kaplan-Meier curves showed that the T4 group had the highest 10-year survival probability (80%), while the T3 group (moderate-high AIP) had the lowest (30%).

Limitations and Future Work

The authors acknowledge several caveats:

  • Observational Nature: While the cohort design establishes a temporal sequence, it cannot strictly prove causality, and reverse causality remains a possibility.
  • Population Specificity: The data is limited to Chinese adults; results may differ in other ethnic or geographic populations with different diets and lifestyles.
  • Unmeasured Confounders: The study did not account for the APOE $\epsilon4$ allele (a major genetic risk factor for Alzheimer's) or specific dietary patterns.
  • Future Directions: The researchers call for multi-omics studies (genomics, lipidomics) and neuroimaging to uncover the exact biological mechanisms behind the "paradoxical" protective effect of high AIP.

Summary updated

Abstract

The Atherogenic Index of Plasma (AIP), a burgeoning composite lipid marker that reflects overall lipid balance, has an incompletely understood longitudinal relationship with cognitive impairment. This study conducted a systematic investigation into the long-term, nonlinear association between AIP and the risk of cognitive impairment among middle-aged and older adults, utilizing longitudinal data from the China Health and Retirement Longitudinal Study (CHARLS) spanning from 2011 to 2020. The study encompassed 2971 participants who were free from cognitive impairment at baseline and were followed for a period of up to 10 years. The cumulative incidence of new-onset cognitive impairment was found to be 40.46% in men and 54.45% in women. A multivariable-adjusted Cox regression analysis revealed that an elevated AIP, particularly within the 25th to 75th percentile range, was an independent risk factor for cognitive impairment (P < 0.05). Further analysis employing restricted cubic splines (RCS) uncovered a significant inverted U-shaped nonlinear association between AIP and cognitive impairment risk (P-nonlinear < 0.001), with a notably increased risk within the AIP range of 0.205 to 0.423. Subgroup analyses, stratified by sex, age, educational level, and BMI, showed consistent trends across various demographic groups. This study suggests that AIP, as a straightforward composite lipid marker, exhibits a significant nonlinear association with the risk of cognitive impairment in middle-aged and older adults. Monitoring AIP levels could assist in the early identification of individuals at high risk, and targeted interventions within specific AIP ranges, such as 0.205–0.423, could have significant public health implications, offering a new potential target for preventive strategies against cognitive decline.
BibTeX
@inproceedings{li2026association,
  author = {Yiying Li and Yuke Zhang and Yu Zhang and Xinyu Yang and Yuehong Ni},
  title = {{Association between the atherogenic index of plasma and cognitive impairment}},
  booktitle = {Scientific},
  year = {2026},
  doi = {10.1038/s41598-026-41335-3},
}
Figure