Timecourse of Pattern Separation and Completion Deficits in Alzheimer Disease?

Ooh Ooh Ooh! They Finally Did It!

I am pleased to have encountered the manuscript I will discuss in this blog. A long time collaborator of mine, Ray Kesner, and I have acted somewhat the gadflies to researchers studying behavioral pattern separation in humans to look at their data a new way to see if time played a factor that was previously uncontrolled in their studies. So when I saw a paper that cited Ray’s and my paper (Link) and performed the exact experiments we had been harping on our friends to perform, I spent a fair portion of the afternoon sending the link to everyone I know. This manuscript not only did it, but extended it to study Alzheimer Disease!

The manuscript I will be discussing is “Pattern separation and pattern completion in Alzheimer’s Disease: Evidence of rapid forgetting in amnestic mild cognitive impairment”, by Brandon Ally, Erin Hussey, Phillip Ko, and Robert Molitor from Vanderbilt University. It is currently an Accepted Manuscript in Hippocampus.

What is Being Studied and Why?

Pattern separation is a term used in psychology and behavioral neuroscience to describe how the brain takes very similar information in through the senses and processes it to store each memory so they are as different as possible (this is how you remember two very similar things as different). Pattern completion is a process that allows the brain to remember information if given an incomplete cue (this is why when you catch only a fleeting glimpse of something you have seen before you can clearly remembering the whole thing). If you are a fan of formal models or hardcore math, see the earliest model from David Marr (Here).

A number of tasks have been developed to study pattern separation and pattern completion in humans, and the research originally performed in rat and mouse models have been mostly replicated in humans. The difficulty in studying these processes in humans, however, is that we have oversimplified what we study. The theory Ray and I have been advocating for a while suggests that one cannot just look only at the stimulus being studied during first and subsequent presentations, but must necessarily account for the temporal lag between first and later presentations.

An Innovative Approach

The beauty of this paper was that the Ally and colleagues only had to modify the approach that is widely available in the literature (and available for download Here). Instead of a jittered temporal lag (number of intervening items) between first and later experience with an item they used a fixed lag of 4, 12, or 40 intervening objects.

By making this little change, the authors were able to tease apart abnormalities in pattern separation and pattern completion functions in Alzheimer Disease at a resolution not previously available.

To do this the authors had only to look at the function when the data were plotted across temporal lag. If there is a monotonic increase in performance across lag, that is normal. If this function is flat or drops off, then there is a deficit. This simplicity makes the reported data incredibly compelling.

Take Home Message

The authors demonstrated that Alzheimer Disease presents with not only poor pattern separation, but also poor pattern completion as measures by an inability to perform pattern separation at any temporal lag, but also an inability to retrieve information once encoded. In contrast, individuals with amnestic mild cognitive impairment showed poor pattern separation–thus replicating earlier studies–but intact pattern completion.


So as the authors mentioned in their introduction, Alzheimer Disease related pathology first presents in the CA1 subregion of the hippocampus, the area previously reported in human fMRI studies to be involved in pattern completion. The other areas of the hippocampus are affected later. This completely supports the authors’ interpretation of the data they observed in Alzheimer Disease.

What makes this so important in my eyes is the fact that if one looks at what has been described about hippocampus function in rodents, it is that removing or inactivating CA1 makes it impossible for rats to hold onto information over time gaps or across lots of intervening stimuli. As I see it, I am, as always, unconvinced that in Alzheimer Disease there is a pattern completion for visual information deficit per se, but there is clearly a deficit in temporal processing.


As I see this study, I see a powerful behavioral assay that may be used as a prodromal test in Alzheimer Disease. IF individuals with mild cognitive impairment show impaired pattern separation but intact pattern completion, whereas the Alzheimer Disease patients showed concomitant pattern separation and pattern completion (or temporal processing as I re-interpret their data) deficits, then a simple screen could be imagined to separate MCI from Alzheimer Disease behaviorally to guide potentially targeted treatment.

These types of behavioral tasks can be used to complement the genetic and imaging studies currently under development for predisting Alzheimer Disease progression.


I would love to hear your thoughts on this!

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s