29 Jun Sex Of Research Subjects Plays a Large Role in Outcomes

MedicalResearch.com Interview with:
Dr Natasha Karp PhD
Senior Staff Scientist – Biostatistician
Wellcome Trust Sanger Institute

MedicalResearch.com: What is the background for this study?

Response: There is evidence that the prevalence of disease, the symptoms experienced,  the progression,  and the side effects can be  dependent on sex.  However, women are underrepresented within biomedical research and this can be seen in the reporting, design of the experiments and subsequent statistical analysis.

Examples

• A review of international animal research between 2011 and 2012 found that 22 per cent of studies did not state the sex of the animals, and of those that did, 80 per cent of studies used solely males and only 3 per cent included both males and females (https://www.ncbi.nlm.nih.gov/pubmed/25175501)
• In a study across 10 fields of biology,  80% of the rodents were male  (Beery and Zucker 2011 Neurosci Biobehav Rev).  This rate has not changed in 20 years  (Mazure and Jones 2015 BMC Women’s health)
• Women encounter adverse drug reactions more often than men (odds ratio 1.596 confidence interval: 1.3-1.94)( Zopf et al 2008  Eur J Clinic Pharmacol)
• Example – cardiovascular disease is the number one killer of US women,  and women differ to men in symptoms, outcomes and risk factors, yet only one third of cardiovascular trials subjects are female and only 31% of trials report results by sex (http://www.brighamandwomens.org/Departments_and_Services/womenshealth/ConnorsCenter/Policy/ConnorsReportFINAL.pdf)
• Example – It took 21 years to lower the  dosing guidelines for women for the insomnia drug Zolpidem due to differences in the clearance rate of the drug exposing women to greater health risks (https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4918870/)

As a result of these observations, multiple bodies have called for sex to be considered as a biological variable in preclinical research.  The largest funders of biomedical research (National Institute of Health) made inclusion of sex a requirement of funding with a few well defined exceptions.  There was some push back on this,  particularly that scientists should be trusted and would know when sex was a variable and it was a waste of resources. I felt there was a knowledge gap on the role of sex.  We have had published studies of individual examples where sex had a role but no large scale study.

MedicalResearch.com: What are the main findings?

Response:  In this study we looked to answer how often the outcome would depend on sex in an individual experiment. The value comes from the scale of the study.  We looked across thousands of experiments,  which allows us to assess the prevalence of sex impacting the outcome of the experiments.

The team found that in the standard group of mice – the control mice – their sex had an impact on 56.6 per cent of quantitative traits, such as bone mass, and on 9.9 per cent of qualitative traits, including whether the shape of the head was normal or abnormal. In mice that had a gene switched off – the mutant mice – their sex modified the effect of the mutation in 13.3 per cent of qualitative traits and up to 17.7 per cent of quantitative traits. It is worth noting that the study had limited sensitivity to detect the outcome depending on sex and thus is a conservative estimate of the prevalence.

MedicalResearch.com: What should clinicians and patients take away from your report?

Response: There is a movement in healthcare to personalise healthcare.  Sex is a first port of call we should be considering in order to optimise treatments.

MedicalResearch.com: What recommendations do you have for future research as a result of this study?

Response: Whilst studying both sexes in preclinical research will increase costs the long term benefits will be significant.  

MedicalResearch.com: Is there anything else you would like to add?

Response: It is worth considering why this situation has arisen.  All experiments are a simplification of a complex world to allow us to isolate cause and effect.  We then generalise the results to the broader situation.  Under our ethical obligations (3R principles) we need to reduce the number of animals used to the minimum.   Many years ago,  it was noted in a study that the female animals were more variable and this was tracked to the impact of the hormonal cycle.  It was therefore argued that we should work with males as we would use less animals.

A meta study has since proven that for the majority of experiments females are no more variable than males (Prendergast BJ, Onishi KG, Zucker I (2014) Female mice liberated for inclusion in neuroscience and biomedical research. Neurosci Biobehav Rev 40:1–5.).

Another historic driver, arises from the ban on including women of childbearing potential from trials.  This was lifted in 1993 but has left a historic focus on males and a subsequent generalisation that women are just little men. 

MedicalResearch.com: Thank you for your contribution to the MedicalResearch.com community.

Citation:

Prevalence of sexual dimorphism in mammalian phenotypic traits
Natasha A. Karp, Jeremy Mason Jacqueline K. White
Nature Communications8, Article number: 15475 (2017)
doi:10.1038/ncomms15475

Note: Content is Not intended as medical advice. Please consult your health care provider regarding your specific medical condition and questions.

Last Updated on June 29, 2017 by Marie Benz MD FAAD