COGS Research Meeting 2020: Talks


Sex/Gender and Cardiovascular Disease

Louise Pilote (McGill University)
January 8, 2020

Thank you so much, while I was setting this up, I was listening to Jill and Cara and I was reminded of a book I read when I was in med school, Thomas Kuhn’s book on The Structure of Scientific Revolution, and he talks about paradigm shift, and I just looked it up and I want to read to you the definition; paradigm shifts open up new approaches to understanding what scientists would have never considered valid before. So, do you think we’re in a paradigm shift here with COGS? Yes!

[audience responds] Yes!

It’s an honor for me to be here today, to address you on work that we’ve been doing and it’s a lightning talk, so it’s going to be brief, but it’s going to show a little bit about how we try to put together sex and gender to explain sex differences in cardiovascular diseases that have been observed many times before.

This idea of sex and gender cardiovascular diseases came up within the teamwork- it was a genesis team, which was an interdisciplinary team, and Canadian team, multidisciplinary, was funded by CIHR and heart and stroke. And among many works that we did, we had noticed that all of the cardiovascular times have been over time decreasing, which was really good. There were actually sex differences in prognosis, which was dependent on age. And here we show that, on the left men and women, if we look at the case fatality rate after heart attack, we found that, depending on the age group, women tended to have a higher case fatality rate than men. And this was more pronounced in fact in young women. So, we’re wondering, why is it that young women with heart attacks die more than young men? It came to, why is this sex difference existing?

So, we formed this study called the genesis study- genesis practice study, which is a cohort study of patients who are less than 55 years of age, who have sustained myocardial infarction. And this was a multi-center study, with 24 centres across Canada.

And I have to say that when I started that felt like “We’re going to find this one gene, we’re going to find this one factor that’s going to explain everything”, in fact I was going on a sabbatical to understand genetic epidemiology and “we’re going to collect blood and we’re going to do DNA and we’re going to find the gene”. I have to say that that was my hypothesis, that in terms of explaining why young women, who should be protected through hormones from having cardiovascular diseases.



So we built a conceptual framework, like good scientists, and we said that there would be sex and gender related determinants in young men and women who would relate to presentation, symptoms, severity, and anatomy, and they would be biological factors, traditional risk factors, and psychosocial factors, more gender related factors that would explain the difference in terms of mortality, quality of life, that characterization, and psychosocial response.

So, we were kind of doing sex and gender, trying to identify factors that could explain the difference. But we were kind of stuck at this point, as a root, as to “what’s gender? How are we going to measure gender?”

So our genesis team invited Joyce Johnson to come and explain to us “what is gender”, and then she explained it to us, and then we said “Oh, gender has many different factors, I am an epidemiologist, I have to measure variables, how am I going to get to measuring sex and gender”. So we understood at that time with Joyce, that sex is a biological variable, that there is a [inaudible], and the same thing with gender, there are different possibilities of gender, regulations, rules, it’s still gender and gender identity that make up who the person is in society, and then would this factor effect health, and then do they explain why young women die more from heart attacks than men?

And then we came across the CHRI women’s health research network that explains gender roles, identities and relations in more elaborate terms. So at this point we said “Okay, how are we going to measure gender in our cohort”. So, we actually came together, a group of us from our genesis team, and said we’re going to find validated instruments, questionnaires, that represent these different aspects, and then we’re going to make sure that in the literature, in the report, to show that there is actually a set of differences, there’s actual a difference between men and women in terms of the prevalence of these factors, or the impact that these factors or various outcomes that were measured in the studies, and we were like setting our hazard issue at 1.5 between sexes.

So, this is a bit how we went about how we went about designing a questionnaire that would measure these differences of gender. So, we came up with gender roles, different factors, these are the factors that we had in our questionnaire for gender roles, and then we did the same for gender identity. Here, we came across gender identity, we came across the BEM masculinity and BEM 80 score, and we found at the time, and it’s kind of true now except for this measurement, BEM was the only one, but in fact BEM measures personality traits, and that’s only one aspect of gender, which we put under gender identity, and then stressed the rules and consonants and measurement traits, which are more markers that are related to more the personal attributes, personality attributes of a person.

And then we had measures that measured gender relations, these are some of the factors, and again we wanted to not have to validate our questions, so we were seeking from the literature, already validated instruments. And internalized gender is the factors that we included in our measure.

So, at this point we did measure all of that, and then this was many many variables, right? All of these questions. So, we said, “let’s see if some of these are inter-correlated, let’s see if we can reduce the number of variables”. We ran a very early machine learning type of analysis, the principle components analysis, which had the data tell us which are the variables that are really explaining uniquely variables.



So then we came up with these variables, and then we said “does that really measure gender” and at this point we said “okay, this is the process measure of gender that we have is sex” so we ran a logistic regression with sex as a dependent variable, to achieve score. And this was a score from 0-100, with 0 person having more masculine characteristics, to 100, where the person had more female characteristics, and this shows the distribution in women, showing that you have your sex, but the gender is on a continuous variable in a spectrum.

And then this was the solution of characteristics, and so we said, “do these effect sex or gender, effects your risk of having risk factors” and then the outcome. So, here are the first 9, we will just look at the association between gender and hypertension, diabetes, and so on, and we find that these associated.

On the second line, we look at sex down the line, on the third line we look at gender after adjusting for sex, and we see that gender actually is still an independent predictor of acquiring risk factors for heart disease, but sex on the fourth line is not so much, implying that when you account for gender, you account for most of the sex effects in terms of explaining the difference.

And let me finish by looking at the outcomes of these patients after one year after their infarc, and then if we look at their pattern of gender, we find that if you were on the third line, that was more associated to masculine gender in society, we found that your risk of having a recurring ECS was lower than if you were of a feminine gender, and that was under the Canadian sex association.

So, this is kind of showing that there are sex differences, in fact we didn’t find any sex differences in terms of the outcome, but if you break it down by gender, you find that there’s an increased risk of heart risk and ECS.

And then the question of breaking down versus the score, and here if we breaking down some of those gender variable factors, we found that the responsibility for housework was associated with a lower chance of receiving invasive cardiac procedure after an ECS, and again, this was independent of whether you’re a man or women.

So, lesson learned, we found that gender matters when it comes to heart disease, but since then, there’s been a lot question “Is gender? What is it? Is it really just social determinants?” the intersectionality that we’ve been talking about, “should we measure gender according to the population, age, and culture” “why did we needed to have a score” and at the end of the day, we had a lot of questions, “why gender? Why not just the factors, why do we have to call that gender?”



COGS Research Meeting 2020






Supporting Organizations