The effect of fasting status on Alzheimer’s disease plasma biomarkers AB40, AB42, GFAP, NfL, CD‐14, and YKL‐40
Erin Livia Vasquez, Tiffany F Kautz, Pia Kivisäkk, Claudia L. Satizabal, Rebecca Bernal, Sudha Seshadri, Carlos Cruchaga, Stephanie Yiallourou, Mitzi M. Gonzales, Jayandra Jung Himali, Matthew P. Pase- Psychiatry and Mental health
- Cellular and Molecular Neuroscience
- Geriatrics and Gerontology
- Neurology (clinical)
- Developmental Neuroscience
- Health Policy
- Epidemiology
Abstract
Background
Prior research has shown that Alzheimer’s disease (AD) blood biomarkers may be altered by fasting status, but this is poorly studied.1 For accurate diagnosis, it is essential to have controlled protocols. To examine the effect of fasting on AD biomarkers, we compared levels of amyloid‐beta (AB) 40, AB‐42, neurofilament light (NfL), glial fibrillary acidic protein (GFAP), YKL‐40, and CD‐14 in fasting and non‐fasting samples.
Method
MarkVCID consortium2 plasma (n = 10) was used to determine the effect of fasting for AB‐40, AB‐42, GFAP, and NfL using the Quanterix Neuro 4‐plex E kit. Blood samples were collected in the morning and again 2‐hours post‐breakfast.
Knight Alzheimer’s disease research center (ADRC) plasma (n = 16) was used to determine the effect of fasting on YKL‐40 (MSD) and CD‐14 (ELISA, R&D system). Blood collections occurred within a 1‐year span.
Biomarker results were normalized as required and paired t‐tests or Wilcoxon signed‐rank tests were used to compare values between fasting and non‐fasting results. ICC was used to correlate fasting and non‐fasting results.
Result
There were significant differences between fasting and non‐fasting levels for NfL (p = 0.0137), GFAP (p = 0.0137), and AB42:40 (p = 0.0020), but no significant changes for AB42 (p = 0.4316) and AB40 (p = 0.1815) (Figure 1). Strong correlations between fasting and non‐fasting results were observed for all four biomarkers (ICC = 0.873‐0.964) (Figure 3).
There was no significant difference between fasting and non‐fasting concentrations for YKL‐40 (p = 0.86) or CD‐14 (p = 0.43) (Figure 2). There was poor correlations between fasting and non‐fasting YKL‐40 (ICC = 0.472) and CD‐14 (ICC = 0.254) (Figure 3). This was not due to the length of time between the two blood draws (YKL‐40, r2 = 0.067; CD‐14, r2 = 0.024) or the collection time of day (YKL‐40, r2 = 0.076; CD‐14, r2 = 0.001) (Figure 1).
Conclusion
Although significant changes in NfL, GFAP, and AB40:42 were observed, these values correlated well, suggesting that fasting might not alter diagnostic categorization. Additionally, while no significant differences were measured for YKL‐40 or CD‐14, there was poor correlation between fasting and non‐fasting values that were not related to the length of time between the draws or the time the draw occurred. We are working to determine the variable(s) responsible for these results.