Loading ...

A study spanning 20 years helps nail down the timing of biomarker changes that occur in the period between normal cognition and a diagnosis of sporadic Alzheimer's disease (AD), something that hasn't previously been extensively investigated in longitudinal studies.

By analyzing cerebral spinal fluid (CSF), as well as cognitive and brain imaging assessments conducted every few years for two decades, researchers were able to plot the course of changing levels of amyloid-beta 42 (Aβ42), phosphorylated tau 181 (p-tau181), and neurofilament light chain (NfL) in adults with AD and mark when those levels began to deviate from those of adults without AD.

Levels of (Aβ42) in CSF and the ratio of Aβ42 to Aβ40 in people who developed AD diverged from those of peers who remained cognitively normal at 18 years and 14 years, respectively, before clinical signs of disease appeared. 

The level of p-tau181 in CSF increased 11 years before disease onset, and NfL levels, a measure of neurodegeneration, increased 9 years before diagnosis. 

These changes were followed by hippocampal atrophy and cognitive decline a few years later. 

The results also show "an apparent accelerated change in concentrations of CSF biomarkers followed by a slowing of this change up to the time of diagnosis," report the authors, led by Jianping Jia, MD, PhD, with the Innovation Center for Neurological Disorders, Xuanwu Hospital, Capital Medical University, Beijing, China. 

The study was published online on February 21, 2024, in The New England Journal of Medicine. 

Time Course of Biomarker Changes

Jia and colleagues conducted a nested case-control study within the China Cognition and Aging Study (COAST). They matched 648 adults who developed AD to 648 who remained cognitively normal. CSF, cognitive, and brain imaging assessments were performed every 2-3 years for a median of about 20 years. 

Within both groups, men slightly outnumbered women. At baseline, CSF biomarker levels, cognitive scores, and hippocampal volumes were similar in the two groups. Adults who developed AD were more likely than their matched controls to be carriers of the APOE ε4 allele (37% vs 20%).

In terms of CSF Aβ42, the level of this biomarker in those who developed AD diverged from the level in controls an estimated 18 years before clinical diagnosis. At that time, the level was lower by a mean 59.13 pg/mL in the AD group. 

A difference in the ratio of CSF Aβ42 to Aβ40 between the two groups appeared an estimated 14 years before the diagnosis of AD (difference in mean values, −0.01 pg/mL). 

Differences between the two groups in CSF p-tau181 and total tau concentrations were apparent roughly 11 and 10 years before diagnosis, respectively. At those times, the mean differences in p-tau181 and total tau concentrations were 7.10 pg/mL and 87.10 pg/mL, respectively.

In terms of NfL, a difference between the groups was observed 9 years before diagnosis, with its trajectory progressively deviating from the concentrations observed in cognitively normal groups at that time, to a final mean difference in NfL of 228.29 pg/mL. 

Bilateral hippocampal volume decreased with age in both groups. However, the decrease began to differ between the two groups 8 years before AD diagnosis, at which time volume was lower by 358.94 mm³ in the AD group compared with the control group.

Average Clinical Dementia Rating–Sum of Boxes (CDR-SB) scores in the AD group began to worsen compared to the control group at about 6 years before diagnosis.

As AD progressed, changes in CSF biomarkers increased before reaching a plateau. 

Important Contribution 

In a linked editorial, Richard Mayeux, MD, Department of Neurology, Columbia University, New York, said the importance of this work "cannot be overstated. Knowledge of the timing of these physiological events is critical to provide clinicians with useful starting points for prevention and therapeutic strategies."

Mayeux said this "remarkable" longitudinal study spanning two decades "not only confirms the hypotheses of previous investigators but extends and validates the sequence of changes" in sporadic AD. 

Mayeux acknowledged that one might consider the finding in this study to be limited owing to the inclusion of only individuals of Han Chinese ancestry. 

However, longitudinal studies of plasma biomarkers in individuals of Asian, European, African, and Hispanic ancestry have shown similar trends in biomarker changes preceding the onset of AD, he noted. 

"Ethnic variation in these biomarkers is known, but that fact does not lessen the effect of the results reported. It merely highlights that similar studies must continue and must be inclusive of other groups," Mayeux concluded.

The study had no commercial funding. Disclosures for authors and editorialist are available at NEJM.org.