Sex Matters. As does dedicated grant funding.

"Sex matters. Sex, that is, being female or male, is an important basic human variable that should be considered when designing and analyzing studies in all areas and levels of biomedical and health-related research. "

Quite some time ago Dr. Isis reviewed the complications associated with doing sex comparisons in scientific research.

This is a particular issue that Dr. Isis, as a vascular physiologist and a woman, is painfully aware of and, yet, the difficulties associated with including women in clinical research can be more pragmatic than simple gender discrimination.

I chimed in (reposted) with an observation about the practical realities of scientists engaging in sex-comparison research. I concluded that:

Promoting special funding opportunities are the only way to tip the equation even slightly more favorable to the sex-differences side. The lure of the RFA is enough to persuade the experienced PI to write in the female groups. To convince the new PI that she might just risk it this one time.

Today I noticed (h/t: @KateClancy) a Program Announcement (with Set-aside funds) from the NIH. PAS-10-226 is titled "Advancing Novel Science in Women's Health Research (ANSWHR)".


The funding opportunity announcement is from the NIH Office of Research on Women's Health (ORWH) but a long list of ICs are participating, including the brain ICs of greatest interest to YHN. (The quote at the top of this post comes from the announcement.)
The first point of observation is related to my prior post about grant funding. This is not just a Program Announcement which announces an interest of some part of the NIH to see applications on a general topic over the next three year interval without specific funds being earmarked. It is not just a Request for Applications (RFA) which offers dedicated funds, but only lasts for a single receipt date. The PAS is a hybrid of the PA and the RFA in that it offers dedicated funds, typically for the first Fiscal Year (i.e., all three rounds of receipt dates), and then extends like a regular PA for the additional two years.
This one notes:

It is anticipated that $4 million will be available for FY 2011. The total amount awarded and the number of awards will depend upon the number, quality, and costs of the applications received. Awards issued under this FOA are contingent upon the availability of funds and the submission of a sufficient number of meritorious applications.

The NIH R21 grant mechanism offers up to 2 years of funding with no more than $275,000 in direct costs divided across the two years as the PI desires. There is a caveat that no more than $200,000 can be dedicated to either of the years proposed. So let's say the PAS gets all proposals asking for $200,000 in the first year and apply 56% as our overhead estimate. Dividing $4M by $312K, we arrive at almost 13 awards.
The idea is that the stimulation goes far beyond this, of course. You have all those individuals who prepared applications (perhaps including new preliminary data) which were unsuccessful. Many, especially the near-miss ones, will come back in revised form under the regular PA part of the announcement or even under other funding opportunities which might apply to the project. So the potential is much greater than a mere 13 awards spread across all of the participating ICs.
There are some links in the FOA which may be of interest to the more general audience because they provide the rationale for why we need to do better on studying sex-differences in biomedical research.
You might enjoy reading:
The ORWH report, Agenda for Research on Women's Health for the 21st Century
The 2001 Institute of Medicine (IOM) report "Exploring the Biological Contributions to Human Health, Does Sex Matter?" [National Academy Press version; free for page by page viewing].
As I noted in my original post, the funding of science has a very large influence on what science gets done. Especially for topics that are not considered particularly "hot" or ones that entail additional difficulty above and beyond what might be considered standard approaches. Sadly, studies of the male (in human and in nonhuman research) have a tendency to be the default. This FOA is one positive step to help ameliorate that trend.

9 responses so far

  • whimple says:

    Well, it's a little bizarre the claims of lack of females in clinical research. To get IRB approval for human subjects studies around here you are explicitly required to study men, women and children unless one of these groups can be excluded for key biological reasons, like you're studying prostate cancer and women don't have one and children don't get this kind of cancer. It is insufficient justification that a group be excluded because it would be inconvenient to recruit these subjects, or because "the data will be cleaner" without these subjects. Likewise with race categories. Apparently the non-selection of particular demographic groups from research that may ultimately be of benefit to them contravenes the "Justice" principle of the Belmont Report (http://ohsr.od.nih.gov/guidelines/belmont.html)

  • So let's say the PAS gets all proposals asking for $200,000 in the first year and apply 56% as our overhead estimate. Dividing $4M by $312K, we arrive at almost 13 awards.

    This is almost surely an underestimate, as the overwhelmingly vast majority of R21 applications ask for $150,000 in year one and $125,000 in year two.

  • Dr Becca says:

    @whimple: OK, so you're required to use female subjects, but are you required to do the sex comparisons? I feel like all too often, data gathered from men and women in clinical studies is simply combined. The goal of the PA seems to be not just to include women for "balance" or "fairness" or whatever, but specifically to parse out sex differences in relevant areas of health research.
    Now, will someone promote me to Research Scientist already so I can apply for the darn thing?

  • whimple says:

    @Becca: You do as many comparisons as you can think of until you run out of statistical power. If you're seeing combined data it probably means there was no statistically significant difference detected between males and females, and that they needed the combined sample size to show statistical difference from the control group. As always, the more subjects you enroll, the smaller an effect you can reliably detect.

  • antipodean says:

    If you were actually doing clinical research then you'd get whatever gender balance is in the clinical population.
    Oftentimes that's more women than men. Half the papers I've written recently have been obesity clinics which are 3/4 female so we lack the statistical power to look at men.
    And the first posthoc comparison in anything is for the gender interaction.

  • becca says:

    whimple- or you can devise new methods to improve your statistical power
    (I just went to a thesis defense where she used iTRAQ based proteomics to determine differences between male and female rats in a model of alcohol-induced cardiomyopathy; comparing different runs of iTRAQ in statistically valid ways is a challenge in itself, add to that the amount of data processing to compare *different cellular fractions from single animals* and my brain totes boggled. Good science though.)
    "And the first posthoc comparison in anything is for the gender interaction."
    True. But posthoc comparisons should ALWAYS make any thinking biologist with any knowledge of stats squeamish.

  • DrugMonkey says:

    Ummm... No, becca. Wrongity, wrong, wrong. Posthoc follow up to main effects is standard and appropriate.

  • becca says:

    DrugMonkey- Did I say posthoc follow up to main effects is uncommon or inappropriate? No. I said it should make you squeamish. As in... PAY ATTENTION. As in, are you applying the appropriate corrections for multiple testings?
    There are MANY good motivations for posthoc analysis (not the least of which is to not waste extra animals, if we're talking animal research). There are also many adequate protections to ensure the posthoc comparisons yield reliable conclusions. There are also many posthoc comparisons that aren't analyzed appropriately in the context of multiple tests.

  • I work in the field of breast and ovarian cancer.
    Filling in the "inclusion of women" sections of grant applications is always really, really easy
    (although we do have to justify excluding male breast cancer patients).

Leave a Reply