Biomarkers of cellular senescence predict risk of mild cognitive impairment: Results from the lifestyle interventions for elders (LIFE) study.

Objectives: Cellular senescence, characterized by a marked and multifactorial senescence-associated secretory phenotype (SASP), is a potential unifying mechanism of aging and chronic disease. Most studies of the SASP have focused on frailty and other functional outcomes. Senescent cells have been detected in the brains of patients with Alzheimer's disease, but few studies have examined associations between plasma SASP markers and cognition. The objective of this study was to examine the cross-sectional and longitudinal associations between plasma SASP markers and mild cognitive impairment among older adults at high risk of mobility disability.

Design: The Lifestyle Interventions for Elders (LIFE) study was a randomized controlled trial of a group-based physical activity program compared to a "successful aging" health education program to assess effects on major mobility disability that was conducted from February 2010 to December 2013.

Setting: Recruitment occurred at eight centers in the United States.

Participants: We included 1,373 participants enrolled in the study with baseline measures of 27 biomarkers of cellular senescence and adjudication of mild cognitive impairment (MCI) and dementia at baseline and 24-month follow-up. At baseline, participants were aged 70-80, sedentary, and at high risk of mobility disability.

Measurements: A neuropsychological assessment was administered at baseline and 24 months post-randomization. At both timepoints, a clinical adjudication committee determined whether individuals had a diagnosis of cognitively normal, MCI, or dementia; individuals with dementia at baseline were excluded. The concentrations of 26 of the 27 plasma proteins identified as components of the SASP were measured with commercially available Luminex xMAP multiplex magnetic bead-based immunoassays analyzed on the MAGPIX System while 1 protein (Activin A) was measured using an enzyme-linked immunosorbent assay.

Results: Logistic regression models were used to examine the associations of each senescence biomarker, in quartiles, with baseline or incident MCI. Models stratified by clinical site and adjusted for intervention assignment, age, gender, race, and education. Among 1,373 participants, 117 (8.5%) were diagnosed with MCI at baseline. Increasing quartiles of myeloperoxidase (MPO) was associated with higher odds of MCI compared to quartile 1 (Q2: OR = 1.34, 95% CI: 0.74-2.45; Q3: OR = 1.43, 95% CI: 0.80-2.59; Q4: OR = 1.79, 95% CI: 1.02-3.22). Additionally, matrix metalloproteinase 1 (MMP1) quartiles 2-4 had lower odds of MCI compared to quartile 1 (Q2: OR = 0.61, 95% CI: 0.35-1.02; Q3: OR = 0.58, 95% CI: 0.33-0.98; Q4: OR = 0.64, 95% CI: 0.37-1.08). Of the 1,256 cognitively unimpaired participants at baseline, 141 (11.2%) were diagnosed with incident MCI or dementia at the 24-month follow-up. Compared to quartile 1, increasing baseline quartiles of MPO (Q2: OR = 1.10, 95% CI: 0.63-1.92; Q3: OR = 1.36, 95% CI: 0.80-2.33; Q4: OR = 1.92, 95% CI: 1.16-3.25) and matrix metalloproteinase 7 (MMP7, Q2: OR = 0.88, 95% CI: 0.47-1.62; Q3: OR = 1.46, 95% CI: 0.85-2.55; Q4: OR = 2.14, 95% CI: 1.28-3.65) were associated with increased odds of MCI or dementia at 24 months.

Conclusions: Among older adults at high risk of mobility disability, high plasma MPO was cross-sectionally and, along with MMP7, longitudinally associated with increased odds of MCI and dementia. In contrast, high MMP1 was cross-sectionally associated with reduced odds of MCI.