Thanks for sharing, Harrison-
VERY interesting. I found the article behind a paywall, but will share relevant sections to the community via the Digital Media law fair use doctrine.
The authors make a good case against the relative short follow-up which may be very relevant for older participants (in terms of converting to AD,) but fairly useless for the 20 year olds who were followed for 4 years. I’d like to see the numbers re-analyzed only using a subset of older adults who are at greater risk for AD.
However, there are several concerns. Although the increased risk was still present after adjusting for the poten- tial confounds, these confounds, APOE genotype and age, both strongly affect plasma apoE levels. The age range of participants included in this study (20s–80s) with median follow-up of 4 years is both a strength and a weakness, and conversion to dementia may be difficult to assess in large populations over a (relatively) short median follow-up.
Less concerning, they discusses the separate pools of APOE in the periphery and CNS and how that may affect the interpretation of
Rasmussen’s work. Even accounting for the relatively supposedly tiny interplay between the two pools, it remains significant that lower levels of APOE correlate with impaired cognition. I understand that this is a somewhat surprising and unexpected conclusion, but for that reason it certainly deserves further follow-up.
ApoE is synthesized peripherally primarily in the liver, and in the central nervous system (CNS) primarily by glia including astrocytes. The compartments are believed to be distinct, with little or no crosstalk across the blood–brain barrier.5,6 Even the type of lipoprotein particles found in the cerebro- spinal fluid (CSF) are different than the type of particles found in the blood. It is unknown whether or how the lower plasma apoE levels observed to be associated with increased risk of dementia relate to CNS levels. Smaller studies sug- gest that CNS levels are not different between AD and con- trol patients,7,8 and the Alzheimer’s Disease Neuroimaging Initiative data set shows the opposite effect—a relationship between elevated CSF apoE and cognitive decline.9
The authors also raise concerns with regards to the assay methods used which could certainly play a part in distorting results if found to be inaccurate, but the fact that the levels tracked very closely with APOE genotype (E2 exhibiting the highest levels & E4 the lowest) reveals at least a moderate level of precision.
Third, from a technical perspective, there may be concerns about enzyme-linked immunosorbent assay (ELISA) measurements in individuals of different genotypes, because Rasmussen et al show that their ELISA may not detect ApoE2, 3, and 4 protein levels equally well.1
Lastly, the authors ask excellent questions and conclude that further follow-up is warranted.
These concerns aside, taken at face value, the results from Rasmussen et al1 open a new set of fascinating questions about how peripheral ApoE levels are related to dementia. Plasma ApoE levels reflect both synthesis (peripherally, primarily in the liver) and clearance (again, peripherally, primarily in the liver). Is decreased synthesis or increased clearance reflective of underlying physiologi- cal properties that impact the CNS? Could peripheral apoE be part of the complex CNS–peripheral equilib- rium of Ab, with lower plasma apoE levels impacting to some extent CNS levels of Ab or other molecules of importance in CNS function? Or might peripheral apoE levels reflect broader issues related to inflammation, or to cholesterol metabolism, either of which may be relevant in mediating a biological process that impacts likelihood of dementia? Rasmussen et al1 provide an intriguing data set that motivates further exploration of these questions.