Cognition as an Outcome Measure in Preclinical Drug Studies

Ooh Ooh Ooh! They Finally Did It!

Very few things make me happier in my scientific life than the feeling like my work is relevant. This morning, a manuscript came out that proved to me that the theories and ideas I have been developing may be taking hold. And, more importantly, that the behavioral tasks I have been designing and advocating are starting to gain traction and be used as outcome measures.

This post is going to talk about my own work (Link) as well as an extension of my work and an application of it to a preclinical treatment study in an article entitled “Lithium treatment alleviates impaired cognition in a mouse model of Fragile X Syndrome” that appears in Genes, Brains, and Behavior by Margaret K. King and Richard S. Jope at the University of Miami.

What is Being Studied and Why?

It has been shown in a number of previous study as well as by anecdotal reports that lithium treatment results in some reduction in behavioral problems in fragile X syndrome and the associated mouse model.

Prior to the report described in this post, unfortunately, the behaviors being evaluated in the fragile x mouse model were very mouse-centric and not really based on any findings in the human population. Accordingly, there has never been a clear interpretation of any fragile X mouse effects that can be intuitively translated into the clinic.

An Innovative Approach

The authors set out to use a battery of cognitive tasks for mice, since Lithium has been shown to improv cognitive function in humans with fragile X syndrome in a formal trial. All other agents have only shown effectiveness for more psychiatric rather than cognitive effects of the mutation.

Unfortunately, the available battery of tasks in rodents has been plagued by inconsistent data being reported. Using the water maze, for example, a number of labs have identified deficits in the fragile X mice, whereas numerous other labs have not. This holds for almost every other behavioral task reported thus far in the literature (thorough review in a review paper of mine).

The authors actually decided to do something I have been waiting on for some time: they used a small battery of behavioral tasks I developed for use in a closely related genetic disorder (Link) and applied it to their model.

The reason this is so exciting for me is that the very rationale behind the battery of tasks I developed was based on the results of standardized testing in fragile X syndrome patients. In other words, this battery was literally developed for use in the fragile X mouse, even though I used it for another population entirely.

Additionally, the authors made use of this battery in exactly the way I had hoped someone would. They used performance of wildtype and fragile X mice on this collection of behavioral paradigm as the primary outcome measure for a treatment study. This suggests that collections of well designed behavioral tests selected specifically to test hypotheses concerning the human patient population can and does in fact result in valid and easily interpretable data in a mouse mode.

Take Home Message

In short, the authors demonstrated that the fragile X mouse shows clear cognitive deficits across four behavioral tasks (I will omit the details here and leave the reader to read the paper for the minutia). These deficits are reduced and performance reaches almost wildtype levels when lithium is administered and maintained at clinically relevant serum levels. Importantly, lithium administration did not help or hinder the performance of the wildtype mice on this battery.

What I find most intriguing in this study is the most intuitive finding in the study. That is, the reduction in behavioral deficits returns when the lithium is discontinued and serum levels return to zero. In fact, the levels fall to exactly mirror performance in untreated fragile X mice. Again, no effects of being on or being taken off lithium were observed in the wildtype littermate control mice.


My interpretation of these data are twofold. My first interpretation is that evaluating cognitive function in the fragile X mouse is possible. Prior to this point the study of cognition in the fragile X mouse was always clouded by uninterpretable effects of contradictory results from collaborating labs. Using more carefully designed tasks, researchers using mouse models to study human disease may actually be able to provide some useful answers. Or at least to provide a valid primary screen for how drugs or treatment strategies may affect cognitive function.

My second interpretation is somewhat more nuanced. Despite developing batteries of behavioral tasks specifically for use as outcome measures, I hesitate to emphasize the efficacy of lithium or any drug. I think the results do clearly show lithium may be useful for assisting fragile X patients overcome their cognitive deficits and function better in an academic setting and life in general, but the results also demonstrate that lithium administration has to be chronic. This is less than optimal since there are widely known side effects of long term lithium administration.


The present task was performed without the benefit of expensive tracking systems and complicated or “sophisticated” techniques. It involved the authors asking a clear question and finding the easiest way to administer a series of well controlled behavioral tests to answer that question.

Overall, I think the authors of this study should be commended, and in fact I am specifically commending them right now. I think that what this study primarily demonstrates is the utility of an expressly behavioral approach to answering some questions in disease-related neuroscience.

I hope that results such as these influence a new generation of neuroscientists to pursue behavioral analysis in a more pure manner. By that I mean they develop a theory based on findings in the human patient population, design tasks that answer simple, straightforward questions that the theory posits. All while clearly testing hypotheses concerning the potential mechanisms underlying the observed deficits.


I would love to hear your thoughts on this!

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

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

Google+ photo

You are commenting using your Google+ 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 )


Connecting to %s