DOI: 10.1002/alz.071912 ISSN: 1552-5260

Circulating cell‐free mitochondrial DNA (ccf‐mtDNA) and memory in mild cognitive impairment (MCI) and Alzheimer’s disease (AD)

Bing Xin Song, Erika Beroncal, Ana C. Andreazza, Damien Gallagher, Susan Marzolini, Mark J. Rapoport, Jocelyn Charles, Nathan Herrmann, Paul I. Oh, Alex Kiss, Walter Swardfager, Bradley J. MacIntosh, Tarek K. Rajji, Mehreen Siddiqui, Danielle Vieira, Luc Rivet, Krista L. Lanctôt
  • Psychiatry and Mental health
  • Cellular and Molecular Neuroscience
  • Geriatrics and Gerontology
  • Neurology (clinical)
  • Developmental Neuroscience
  • Health Policy
  • Epidemiology



Circulating cell‐free mitochondrial DNA (ccf‐mtDNA), a measure of mitochondrial dysfunction, is emerging as a potential biomarker for early Alzheimer’s disease (AD). Decreased levels of cerebrospinal fluid ccf‐mtDNA concentrations have been detected in early AD patients. Few studies have investigated peripheral ccf‐mtDNA in early AD or with cognitive measures, much less evaluating the role of physical activity within the relationships. Considering the hallmark of episodic memory impairment in early AD, we examined the association between serum ccf‐mtDNA and memory, while considering potential contributing factors, physical activity level and age.


Participants were diagnosed with mild cognitive impairment (MCI) or mild AD using the Diagnostic and Statistical Manual of Mental Disorders criteria for mild or major neurocognitive disorder. Serum ccf‐mtDNA concentration (copies/uL) was measured using quantitative polymerase chain reaction from circulating cell free DNA purified by spin columns with primer for the MT‐ND4 gene. Memory was assessed using the Word Recall Task from the Alzheimer’s Disease Assessment Scale‐Cog (ADAS‐Cog). Physical activity level was assessed by Leisure‐Time Exercise Questionnaire (LTEQ). Linear regression analyses were used to assess the association between ccf‐mtDNA levels and recall memory with a priori covariates, age and LTEQ score. Ccf‐mtDNA levels were log‐transformed prior to analyses due to significant departure from normality in Shapiro‐Wilk tests.


Of 29 participants [17 (59%) male, 21 (72%) MCI], the mean (SD) age was 74.2 (8.7), years of education was 16.2 (2.2), Montreal Cognitive Assessment (MoCA) total score was 22.1 (3.3), LTEQ score was 36.2 (17.4), and ccf‐mtDNA levels were 2545.8 (2223.6) copies/uL. Higher ccf‐mtDNA was associated with worse recall [F(3,25) = 5.21, R2 = .39, p = .006]. Controlling for LTEQ and age, one log ccf‐mtDNA unit was associated with 1.87 (SE = .71) fewer words recalled (p = .01, f2 = .21).


In this sample, higher serum ccf‐mtDNA concentrations were associated with worse recall, controlling for age and physical activity level. This relationship supports the potential involvement of mitochondrial dysfunction in cognitive functions and neurodegenerative diseases and suggest the need for further research to explore the role of physical activity and age with mitochondrial dysfunction and memory.

More from our Archive