Adding Females to Animal Studies

Mar 25 2015 Published by under Uncategorized

NIH held a workshop on adding female subjects to animal studies. It caused discussion of varying degrees of concern. Here is the link to the NIH report.

The intro:

Biological sex, being female or male, is a fundamental characteristic of biological systems. Sex is a fundamental variable in preclinical biomedical research that underlies drug development research, clinical trials, and preven-tion approaches. Although the biomedical community has made major progress in human studies—women now account for roughly half of the participants in NIH-funded clinical trials—there has not been a similar pattern in preclinical research. Animal studies have focused primarily on males, and investigators studying cell models have often ignored the sex of the individual from which the cells were obtained. For the most part, considering sex as a biological variable has been a blind spot in biomedical research, leaving critical gaps in our knowledge.


Four sessions:

  • Session I addressed the basic concept of including female and male subjects in studies.
  • Session II explored the impact of including or not including sex as a basic biological variable.
  • Session III visited specific and practical methods to integrate sex as a variable in research plans and projects.
  • Session IV discussed how to introduce a "sex matters" culture across multiple disciplines.

I've not digested this yet. I will, and will blog on it as I get excited, irritated or fall asleep.

10 responses so far

  • Drugmonkey says:

    Just effin do it. ( my advice)

  • B. Kiddo says:

    what the...? Can hardly believe such basics are not par for the course.

  • The New PI says:

    Since repeating behavioral studies in females is now a requirement, we started doing behavior on males and females at the same time (well, staggered, but from the same litters). Apart from the enormous expense, which the NIH will have to consider in awarding animal costs, preliminary data shows that they are quite different....which could all just be because females are in estrus and very variable. We'll see.

    I was talking about it to someone at a meeting and they said, you can say you'll do males and females, do the males and then do females later...if you have time. Not sure how that is going to fly.

  • Dr Becca says:

    On study section recently, there were several A1s that had added females to their Research Plan. What they had not added was any change in power analysis, experimental design, or budget. Nobody (besides me) on the panel seemed to think this was a problem, and it blew my mind. Imagine any other scenario in which someone proposes at least a doubling of their subjects and experimental groups with no other considerations or adjustments--is this a solid plan?

    New PI, did you actually observe more variability in the females' behavior--i.e. did you run a Bartlett's test of equal variance?

  • Brain says:

    There was an interesting meta-analysis examining variability of behavioral, neurobiological, and molecular traits in males vs. females. It turns out variability in these measures is similar in males and females. Although male C57s don't have ovarian cycles they do fight a lot in the home cage (which is probably why single housing is anxiolytic for C57s).

  • Former Technician says:

    All this hubbub is causing a different response in immunology studies, where we usually only included females. Our bigger issue was when a post-doc ordered a different strain of mice and the results were opposite. (C56 vs BalbC) She ordered the different strain as controls for a transgenic model. Caused quite the uproar in the lab.

    • JustaTech says:

      I was thinking the same thing; in immunology we almost always only used females (makes housing easier). When one of the scientists did have to use male mice, because sometimes that's what you get in a breeding colony, you'd think I had told them they had to use snakes.
      "But I only want females!"
      "Sorry, all I have are males."
      "When will you have females?"
      "If you're really lucky, 9 weeks."
      "Two months?! I can't wait that long!"

      • potnia theron says:

        In my work, I have (mostly) used females. I work with large animals, and females are just easier to work with. Argh, I have to add males now.

  • Ola says:

    As I've iterated at several other blogs discussing this issue, if NIH expects people to do this, then they have to expect to pay for it. Simply demanding double (or more) the number of animal subjects in a study, without a concomitant increase in the budget, is completely unreasonable. Our mouse costs on one grant are about $20k/yr (male only). Doubling that would mean firing a tech, or having no money for any supplies.

    The other thing which is completely unclear from the NIH announcements on this to date, is whether they want incorporation of both sexes as a whole, or whether they specifically want you to examine sex as an independent variable. There's a BIG difference between those two.... say I need 10 WT and 10 KO mice to see a difference in males only. Adding females would be expected to add some noise, so let's say I need 12 of each, if still comparing WT to KO with no independent breakdown of the sexes. For independence we'd need 4 x groups of 10, but also the group numbers would be higher because now we're comparing 4 groups instead of 2, so t-tests are invalid and ANOVA would be required, and a higher N to meet significance. So, that'd be 12 WT male, 12 WT female, 12 KO male, 12 KO female.

    Going from 20 mice to 24, versus going to 48, is a BIG frickin' difference!

    If Franky boy wants the fancy stuff, he gots to pay.

Leave a Reply