TL neuro

July 2, 2015

CPDD 2015: Sex as a biological variable and pitfalls for behavioral studies

Filed under: Animal Models, CPDD, SABV — mtaffe @ 10:18 am

Cora Lee Wetherington presented on the new NIH initiative on sex as a biological variable (SABV) at the Animals-in-Research forum at the recent CPDD meeting held at the Arizona Biltmore.

Dr. Wetherington has long headed up NIDA’s operations on Sex/Gender differences in substance abuse research and is therefore a key voice on how the NIH will be responding to this new SABV initiative.

Her main point was to outline the elements of the new approach to sex-differences, starting with the Clayton and Collins editorial and leading up to the NOT-OD-15-102 Consideration of Sex as a Biological Variable in NIH-funded Research. This NOT warns NIH applicants that applications submitted after Jan 25, 2015 will need to take new steps in considering SABV.

In the Q&A after Dr. Wetherington’s presentation, Mike Bardo (lab website) asked a deceptively simple question. He wanted to know if people ran their male and female rats in different operant boxes. The point being that male rats might smell something related to the presence of a female in the box just prior to his session, or vice versa. This might have an effect that differed in some way from the presence of a same-sex rat. Bardo indicated that many labs try to have a set of operant boxes that are dedicated to each sex, perhaps even in separate experimental rooms, but intimated that this may become impractical in the new order of SABV after Jan 25, 2015.

Simple question, right? Well, this plays right into the determination of how practical it will be to tell behavioral pharmacology people to start running studies with both sex of rats. The immediate response to the SABV initiative is to say “This will double our experiment size!”. And this comment is rapidly followed by “Wait, by the time we consider female rat cycle, we are not just doubling the experment but possibly increasing our number of groups by 4-fold or more.”.

And indeed several prior questions directed at Dr. Wetherington touched on such issues.

The Bardo question, however, is an even more practical one of experimental throughput. Any given lab is going to be more inflexible in experimental equipment than in the sex of rats it chooses to run from month to month or year to year. Operant boxes are expensive and experimental space to house them is very dear. One cannot just outfit a parallel series of rooms to run the opposite sex. This is going to require compromises. It is going to involve more unknown variability introduced into the design. And it is going to be expensive merely to estimate this variability so as to come up working rules of thumb to address the problem raised by Mike Bardo.

I jumped up to try to emphasize the following points although I am not sure that Wetherington understood what I was driving at.

There are several options available to a lab like mine, if it wants to start running both male and female groups. First, we could do it entirely sequentially. Only females for a 3-4 month interval, then males, etc. This is less than ideal because if we are comparing the sexes, we would like to hold other variables constant. Cohorts run at the approximately the same time has a better chance of doing this. Second, we could run them in different rooms but this has a limit. We only have so many operant box testing rooms and this is typically far too few to accomplish all that we would like to accomplish in a perfect way. (For example, we simply cannot run all of our rats starting at the same exact time of day, relative to their light cycle. Totally impractical.) We have a lot of different projects going on at one time and lining them all up so that males are in one room and females another would slow overall progress. Third, we could keep some boxes within a room for males and some for females. This isn’t much better than separate rooms for operational flexibility purposes and it adds the additional factor of subjects being able to smell (and possible hear the ultrasonic vocalizations of) adjacent animals (the boxes are vented to/from room air). Fourth, we could run the sexes in sequence throughout the day with an extra dose of box cleaning and bedding changing in between the sexes. It adds some work, but preserves a great deal of flexibility.

But really, and getting back to Bardo’s question, we want to know if it matters if we run male and female rats in the same boxes. If it doesn’t, then all of my objections are moot.

So I was trying to point out to Wetherington how much effort it would take for just one of our experimental models to determine if there even were effects of mixed-sex operations on the rats’ behavior that needed to be accommodated. This is at least six months or a year of experiments, running to the tune of tens of thousands of precious direct costs from our NIH grants, to even begin to estimate the kind of variability that would be introduced by running male and female rats in a study within the same limited number of operant boxes. This is methodological work. Not work that can be easily risked by just launching off into the “real” study that is intended to be done.

How tolerant will grant review panels be of proposals that do not address these matters? How tolerant will they be of proposals that do address these matters seriously but plan to burn the first year on such methodological minutia?

How many reviwers will insist on “perfect” design (no females and males run in the same boxes or even housed in the same vivarium room!) regardless of whether it is necessary?

This is just one tiny, tiny, minute methodological question in a restricted subfield of investigation. How many similar questions apply to all of the research funded by the NIH that is poised to come under the SABV dictum seven months from now?

Create a free website or blog at