DOI: 10.1002/alz.076015 ISSN: 1552-5260

Allostatic load profiles of stress to predict Alzheimer’s disease: an accessible opportunity to identify risk at early stages

Juliana Nery Souza‐Talarico, Yelena Perkhounkova, Maria Hein
  • Psychiatry and Mental health
  • Cellular and Molecular Neuroscience
  • Geriatrics and Gerontology
  • Neurology (clinical)
  • Developmental Neuroscience
  • Health Policy
  • Epidemiology



Alzheimer’s disease (AD) can start decades before symptoms appear, yet detection at preclinical stages remains a challenge due to a lack of screening tools that are both accurate and accessible. The multi‐system Allostatic Load index (AL), a stress‐related marker of cumulative biological risk, is associated with β‐amyloid (Aβ), tau proteins, and hippocampal atrophy, potentially linking AL to AD. If this is correct, a low‐cost and readily accessible peripheral blood AL‐based screening tool can offer an opportunity to identify individuals at risk for AD in primary care facilities. Here, we determined the AL profiles that predict the conversion risk from mild cognitive impairment (MCI) to AD by 36 months before diagnosis and associated with hippocampus volume (HV) reduction.


Using the Alzheimer’s Disease Neuroimaging Initiative (ADNI) database, we analyzed longitudinal data from 385 older adults with MCI at baseline, 12, and 36 months. Blood neuroendocrine (cortisol, LH, prolactin, testosterone) and immunologic (C‐reactive protein,fibrinogen,IL‐13,IL‐16,IL‐18,IL‐3,IL‐6,MCP‐3,MCP4,TNF‐alpha,insulin‐like growth factor) stress‐related biomarkers obtained at baseline and 12 months were transformed into AL indexes to generate the following profiles: 1)Decreasing AL: high at baseline and decrease over 12‐months; 2)Increasing AL: low at baseline, increase; and 3)Sustained High AL: high at baseline, remain high or increase. The ADNI consensus diagnosis determined the AD conversion and brain MRI assessed HV. Models were adjusted for socioeconomic and APOE4 covariates.


41.6% of participants with Sustained High AL converted to AD, while 27.9% with Decreasing profiles remained MCI (p = 0.01). Increasing AL showed higher conversion risk than Decreasing profile (OR: 3.96). Sustained High AL was associated with faster HV reduction (p = 0.04). AL and blood Aβ models were compared for prediction accuracy.


The odds of converting to AD were almost four times higher among participants with Sustained High AL profiles with faster hippocampus reduction than Decreasing AL profiles, suggesting that elevated AL profiles might be a risk factor and a potential mechanism explaining MCI conversion to AD. These findings set the foundation for exploring accessible bio‐behavioral screening tools to sign risk in vulnerable individuals in primary care facilities and refer them to specialized clinics for timely intervention.

More from our Archive