The following is a transcript of an interview by Oriana Kraft with Laura Symul, Postdoctoral Fellow at Stanford University, Department of Statistics.

Laura’s research focuses on fertility, cycle-related symptoms, and drivers of change in vaginal microbiome communities. She uses self-tracked data from mobile phone apps and devices and clinical multi-omics data.

Could you briefly explain the focus of your research?

My research focuses on applying and developing statistical tools and methods for improving our understanding of female health. From a biological perspective, I am especially interested in understanding the impact of hormonal fluctuations driven by the menstrual cycle on different physiological functions in females. This ranges from understanding the interactions between the reproductive cycles and the (vaginal) microbiota to contributing to research on understanding the causes of PMS/PMDD symptoms. I rely on “multi-omics” and self-reported data from digital tracking apps/devices to answer these questions.

Can you tell me more about your research on the relationship between the menstrual cycle and the vaginal microbiota?

I recently got the chance to work with the Vaginal Microbiota Research Consortium (VMRC). The consortium has collected a rich longitudinal dataset from samples of hundreds of pregnant and non-pregnant women. In non-pregnant individuals, we investigated whether the vaginal microbiota composition changed with the menstrual cycle. Specifically, we found that in individuals with diverse (i.e., usually associated with adverse health outcomes) microbiota, temporal fluctuations in the microbiota composition were highly correlated from one cycle to the next. While we still need to understand how hormonal patterns drive these changes, these findings suggest that the menstrual cycle may impact vaginal microbiota composition.

How do you view your research in the context of the personalization of women’s health?

Before answering the question, it is important to define what we mean by “personalized medicine” in contrast to “traditional medicine”. In many aspects, “traditional medicine” is already personalized: not every patient receives the same treatment based on known variants of diseases (e.g., cancers) or risk factors (e.g., hormonal contraception for smoking individuals). We started to talk about personalized medicine when genetic sequencing became so cheap that it became conceivable to sequence every patient. The definition has extended to include sequencing and many of the now available technologies to cheaply and rapidly obtain detailed measurements of many biological factors. So, what distinguishes “personalized medicine” from “traditional medicine” lies in the number of patient variables taken into account for making treatment decisions.

We are still in the infancy of personalized health for many domains when it comes to female health. For example, the choice of hormonal contraception (e.g., the pill formulation) is not based on any biological variable specific to the patient. Usually, women[1] try one pill and change if that one does not work for them. Ideally, we would conduct studies to understand what biological factors (hormonal levels, hormonal sensitivity, etc.) are associated with formulation-specific side effects. This way, women can make better-informed decisions about their contraceptive choices.

We can also think of making progress towards personalized health by refining diagnoses. A concrete example may clarify this idea. PMDD (pre-menstrual dysphoric disorder) is described in the current DSM (DSM-V). The diagnosis is binary: someone either is diagnosed with PMDD or is not diagnosed with PMDD. And there is only one type of PMDD.

However, PMDD diagnosis relies on the patient reporting a set of symptoms (among a list of pre-defined symptoms) during the luteal phase of their cycle. If the patient has a sufficient number of symptoms and their severity is high enough, they are said to meet the criteria for PMDD diagnosis.

To date, very few studies have investigated whether there were subtypes in PMDD (i.e., groups of individuals reporting a specific set of symptoms) and whether these potential subtypes might be associated with the combined severity of symptoms.

Are period tracking apps, currently, the best data sets that researchers have available or what data sets do researchers consider the most valuable in the femtech space?

Period tracking apps and, more generally, female-specific symptom tracking apps are exciting tools for developing personalized female health. First of all, these tools are used by millions of women across the globe. So, with proper control for potential biases due to the specificities of the tracking population and due to the amount of missing data, these tools have already generated the largest datasets of female-specific symptoms and observable body signs (e.g., mucus, menses, etc.) data. Then, these tools can also be deployed in prospective studies in which biological markers are quantified while participants are tracking their cycles and symptoms.

Besides tracking apps, I am very excited by the recent development in remote sensing and remote biochemical assays. Kits for at-home hormonal measurements are an example of a tool that will be extremely powerful in studying the impact of hormonal fluctuations on a series of symptoms.

[1] Here, the term “woman” is used to reflect biological sex, not gender. Specifically, the term “woman” here differentiate individuals with female reproductive systems from others, and identifies fecund individuals since we discuss contraception.


Want to hear more about the vaginal microbiome? Period tracking apps? Personalized medicine ? Menstrual cycle patterns? Then sign up for the 2022 femtech conference  here. 

Leave a Reply

Your email address will not be published. Required fields are marked *