Use of Computer Simulations to Understand Female Physiology: Where's the Data?

Physiology is defined in Webster's Dictionary as “a branch of biology that deals with the functions and activities of life or of living matter (as whole body, organs, tissues, or cells) and of the physical and chemical phenomena involved.” As scientists, we operate with the hope that the knowledge we gain through research can be translated into tangible improvements in human health. Researchers use a variety of experimental techniques using human, animal, and cell models to gain physiological knowledge.

Historically, scientists have used mathematical analyses to understand physiology and to test new hypotheses. William Harvey estimated cardiac output and ejection fraction in his original publication De Motu Cordis in 1628. Krogh's computation of oxygen flux in skeletal muscle, and Hodgkin and Huxley's simulation of membrane potential provided early examples of the use of computational methods to advance research. Over the last 50 years, with the development of increasing computational power, mathematical simulations of physiology have been used extensively to understand the human body. There are various types of physiological simulation, such as physics-based deterministic, stochastic, or finite-element models. In developing a simulation, the modeler relies on data in the literature for both the development and the validation of the simulation.

One current area of computer simulation that is receiving serious interest is clinical trials. The development cycle for pharmaceuticals or medical devices is long, which increases the cost of new products. This in turn increases the cost of entry to the market and reduces the number of companies in the market. This cumbersome process stifles innovation and increases the cost to consumers. New techniques are needed to streamline the development of new pharmaceuticals or devices. Computational simulations have recently become extremely important in speeding the process, leading to the term “in silico clinical trials.” Recognizing the value of computer simulation in the drug and device development cycle, the European Commission sponsored the Avicenna Project linking industry, regulators, clinicians, and researchers (1). Their goal is to lay out a road map for the use of in silico clinical trials in the development and regulatory evaluation of medical treatments. The use of computational simulation technology for clinical trials should be useful for predictions in patients of both sexes, across all ages, and for all ethnicities.

In 2001, the Institute of Medicine published a white paper entitled Exploring the Biological Contributions to Human Health: Does Sex Matter? that recommended “Sex, that is, being male or female, is an important basic human variable that should be considered when designing and analyzing studies in all areas and at all levels of biomedical and health-related research” (2). Therefore, research on sex differences will be important in all aspects of medicine to improve healthcare. The Committee also clearly stressed that sex is the classification of organisms as male or female according to the chromosomes and reproductive organs and that gender refers to a “person's self-representation as male or female” (2). However, over the years, physiologists have used and continue to interchange the terms sex and gender; therefore, in the discussion that follows, both terms will be used.

Parameterizing and validating physiological models for clinical utility require data on different sexes, ages, and ethnicities. At the present time (all searches were done on August 10, 2015), this data does not exist. A PubMed search for articles published in all volumes of American Journal of Physiology (AJP) since 1945 resulted in 93,588 articles. However, adding the term gender only yielded 1,481 publications, meaning that ∼3% of the studies considered gender as a variable. Since the year 2000, AJP published 978 papers addressing physiological responses and gender. Since the first paper published in AJP in 1945 addressing gender, 67% of the papers mentioning gender have been published in the last 15 years. If one searches for published papers in AJP focusing on “sex differences,” the outcome is even more concerning, resulting in only 86 papers, with 21 articles in AJP-Regulatory, Integrative and Comparative Physiology. A search of Journal of Applied Physiology yielded ∼30,000 articles, but when combined with “sex difference” yielded 14 articles, whereas the term gender yielded 723 articles.

Female animals have been traditionally excluded from experimental protocols because their hormonal cycles were thought to add variability to studies and, therefore, could confound interpretation of results. In reality, however, most physiological parameters, responses to drugs, etc. are not affected by the stage of menstrual (human) or estrous (animal) cycle in the female at the time the measurements are made. For example, there are many mechanisms responsible for blood pressure control, but women are not given different drugs or different dosages of drugs to control their blood pressure based on whether they are in the follicular phase or the luteal phases of their menstrual cycles.

As scientists, the lack of data on women and female animals in our literature is a glaring omission that should be considered a blemish on our field. What are the adverse outcomes of this lack of gender-based research? That sex differences in physiological systems other than reproduction exist is well known. To ignore how those differences affect half of the world's population charges us as physiologists with neglecting our basic mission: to improve human health through understanding the interplay of physics, biology, and chemistry. By ignoring sex differences, we blind ourselves to confounding variables that, if studied, may deepen our understanding of physiology.

As integrative physiologists focusing on simulation, our sub-area is no better. A PubMed search using the term simulation yields 737 papers that have been published in AJP since 1968 (10). These papers have presented simulations over a wide range of physiology, including renal, metabolism, endocrine, neural, cell, cardiac, and cardiovascular models. The addition of gender in the search resulted in only eight papers in AJP, three focusing on weightlessness, three on growth hormone secretion, one on educational simulation, and one on muscle atrophy. Adding the terms sex difference resulted in only one paper in AJP. A search in Journal of Applied Physiology for the term simulation yielded 496 results, with the addition of gender yielding 5 articles, and the term sex difference yielding 0 articles.

The results of these searches demonstrate that the majority of research in sex differences in physiology has only been done within the last 15 years and that the use of physiological simulation to understand the role of sex on human physiology is severely lacking. Additionally, data from clinical trials do not necessarily provide a breakdown of responses in men and women separately. A cursory review of several completed trails on http://Clinicaltrials.gov presented baseline measures, including participation by either or both sexes. However, outcome measures for both sexes are grouped together, preventing an analysis of potential differences between males and females. Surprisingly, not considering sex differences when reporting these results violates the 1993 NIH Revitalization Act, which required the inclusion of women and minorities in all Phase III trials, as well as analyses of differences between these groups. Using simulation to understand sex differences in physiological mechanisms will be challenging until sufficient data are available to accurately model and validate the responses to a simulation. However, we recently learned that several pharmaceutical companies, such as Bayer, GSK, Lilly, Novartis, and Roche, allow access to clinical trial data (http://clinicalstudydatarequest.com). The data provided include raw datasets, which include the data collected on each patient in the clinical study. This collection of clinical study data could prove to be valuable to the simulation community to address limitations in computer simulation of sex differences.

Our institution has a long history of computer simulation of integrative physiological relationships. However, over the years, we have not focused on understanding gender differences in physiology. We and others need to change this and make every effort to use our simulations to help understand physiological responses between sexes, ages, and ethnicities.