0:04

There's a study taken that we've covered for

0:06

many years on the health report from the

0:08

United States called the Nurses'

0:10

Health Study, where they followed thousands of

0:13

female nurses for many, many years with

0:16

a lot of knowledge about their health, their

0:18

well-being, what illnesses they've

0:21

had and so on. And

0:23

it's been a rich source of information

0:26

over the years on women's

0:28

health. A fascinating and quite troubling

0:30

paper was released from the study,

0:32

though, recently, which has looked at

0:34

the sexuality, in other words, the sexual preferences of

0:37

women in this study, following them through

0:39

to look at what effect it may

0:41

have on these women's lifespan. And

0:44

it wasn't necessarily good news. Here's

0:47

Sarah Mcketta, who's one of the researchers in the

0:49

study. She's based at Harvard School of Public Health.

0:52

There's an extremely robust literature that

0:54

spans multiple decades that has shown

0:56

their health disparities depending on sexual

0:58

orientation. So we know that people

1:00

who identify as lesbian, gay or

1:02

bisexual have higher risks of chronic

1:04

disease, of adverse mental health, of

1:06

adverse behavioral health like smoking and

1:08

drinking. And this is due

1:10

to the chronic experiences of stress and

1:12

discrimination to being a minoritized identity. But

1:15

we didn't know a whole lot about mortality

1:17

differences. We'd only had a couple

1:19

of studies where we were really able to identify

1:22

people who were lesbian, gay or bisexual and follow

1:24

them for a long enough time that we could

1:26

see a difference. So you

1:28

used this long running study at

1:30

Harvard of women who were nurses.

1:32

What did you find? One

1:34

of the things that the Nurses Health Study did was

1:36

that very early on in the survey. So the survey

1:38

started in the 1980s. In

1:41

1995, they asked people about their sexual

1:43

orientation, which compared to other surveys is

1:45

extraordinarily early. So we've had 30 years

1:47

of following these women. Just

1:49

before you go on, looking at your data.

1:52

The declaration rates of their sexuality

1:54

were low. You got 1%, maybe

1:56

0.8%. very

2:00

low rates. How valid was this

2:02

declaration of their sexuality? I actually

2:05

love that you asked this because this is something we were

2:07

thinking about very deeply. Part of this is that it was

2:09

1995 in the United States.

2:12

We've since surveyed these people multiple times over the

2:14

past year. I think the last time that we surveyed them

2:16

on their sexual orientation identity was, I want to say it was

2:18

like 2017 or 2019. So we've

2:21

gotten some more information about this sample

2:23

that more people might be concealing their

2:25

sexual orientation. Maybe they're not responding. Maybe

2:27

they're saying they're heterosexual. So

2:29

if there is a premature mortality

2:31

associated with being LGBTQI

2:34

and they were in the

2:36

heterosexual group, they may have

2:38

pulled down the mortality rate

2:40

of the heterosexual group and

2:43

therefore you don't see the difference. This is really

2:45

what you're saying. Exactly. Yeah. So if you

2:47

were thinking about it like a risk ratio or

2:49

something, you would think that maybe that you wouldn't

2:51

see as pronounced of a difference. So

2:54

potentially then, given that

2:56

you did find a significant difference, albeit in

2:58

low rates, low numbers, it actually

3:00

could be bigger because you have this effect.

3:03

That's the startling thing. So I expected the

3:05

effects, and I'll tell you what the effects are.

3:08

Yeah, by the way, yeah, yeah. Sorry, I'm more distracted. Again,

3:11

we were sort of feeling like there were reasons

3:14

why our results might be biased towards

3:16

a null finding or they might be conservative. But

3:19

just to tell you what we found, people who were

3:21

identified as lesbian or gay died 26% sooner

3:24

than people who identified as heterosexual. Because

3:26

then when we looked within the

3:28

lesbian and gay category to look

3:31

at lesbian women and bisexual women

3:33

separately among lesbian women compared to

3:35

heterosexual women, they died 20% earlier.

3:37

And among bisexual women, they died 37% earlier. So there was

3:40

a difference in

3:42

magnitude depending on how people identified. This

3:44

is a bunch of women who were working as nurses, so

3:47

they were healthy enough to work. They all have the same

3:49

job, so there's not going to be a lot of variability

3:51

with regard to sort of socioeconomic indicators or

3:53

statuses that would be kind of on

3:55

the call for a pathway to worse

3:57

health. It's an extraordinarily non-Hispanic white sample.

4:00

And in the United States, of course, mortality is

4:02

very much patterned along racial and ethnic lines. And

4:05

these are people... Which is a way of saying that in

4:07

statistical terms, there weren't many confounders. In other words, there wasn't

4:09

much else that could actually affect the outcome. Exactly.

4:12

These are people with high health literacy. And so

4:14

I would expect that they have a lot more

4:17

things in common than they have things that are

4:19

different. So summarizing then the

4:21

findings is that you found a

4:23

significant increase in premature

4:26

death compared to the larger cohort

4:28

of female nurses in this study.

4:31

It was larger if you were

4:33

bisexual compared to declaring yourself as

4:35

lesbian. And it looks

4:38

preliminarily across all causes of death.

4:41

Let's go back to what's going on here and

4:44

where you go next with this. Because

4:46

we've known this background literature for a really long

4:48

time about health outcomes, and again,

4:50

not across mortality, but across health outcomes, we

4:53

actually expected to find the patterning that we

4:55

found. So we expected that we would see

4:57

a disparity and it would be more

4:59

pronounced among bisexual women. And the

5:01

reason for that is that people know that sexuality is

5:04

a spectrum, but often these

5:06

risks are very discreet and it's

5:08

because of people's roles within the

5:10

queer community. So bisexual women experience

5:13

discrimination not only from

5:15

heterosexual folks, but also from people within

5:17

the queer community. They're often partnered

5:19

with men and so they might have more pressure

5:21

to conceal their identity and across the

5:23

board. Because they're not necessarily welcomed into the

5:25

community. And what you're saying is that

5:27

if you're bisexual, you might just be out on

5:29

a limb more isolated and more stressed. Yeah, you

5:32

might not be able to tap into sort of

5:34

other queer communities which have these sort of protective

5:36

social effects. And so one of the

5:38

things that happens, and we see this, our particular

5:40

study didn't look at this, but in this cohort

5:42

and in other studies across a bunch of national

5:44

surveys, we know that for

5:46

example, bisexual women have the highest

5:48

rates of tobacco use, of alcohol

5:50

use, of anxiety and depression relative

5:53

to not only lesbian women, but of

5:55

course also heterosexual women. And

5:57

so these concealment processes and these experiences of discrimination

5:59

can... lead to these behaviors to cope. And

6:02

that's coupled with healthcare avoidance

6:04

and discrimination. Physicians

6:07

who meet a woman who has a husband

6:09

often won't ask them about their sexual orientation

6:11

identity and that can lead

6:13

to inappropriate screening. It can lead

6:16

to further fears of disclosure. It

6:18

can lead to this really kind of toxic spiral. And

6:20

just to explain that a little bit more, there is

6:23

evidence, for example, that doctors don't

6:25

push cancer screening, for example, in

6:27

lesbian, gay or bisexual women as

6:30

they do most heterosexual women for

6:33

some bizarre reason. I

6:36

don't know what it's like in Australia, but it's definitely like

6:38

that over here. There's this

6:40

perception that they're not at risk and that's

6:42

not borne out by the evidence. And

6:45

so what's going on is that the

6:49

things that affect health, which are these

6:51

systemic experiences of discrimination and stress, they

6:54

affect all kinds of health and they

6:56

affect health behaviors and health experiences that

6:58

are implicated in all of the top

7:00

10 causes of disease. This

7:03

is due to sort of cumulative burden of stress

7:05

rather than just one thing that

7:07

we could put our finger on. And we did do

7:09

a sensitivity analysis, which is a sensitive analysis is when

7:11

we just change our model, lead assumptions and

7:13

see if the results hold. But we did a

7:16

sensitivity analysis and said, what if we only

7:18

looked at this relationship in women who never

7:20

reported smoking? So we sort of removed

7:22

that from the equation and the

7:25

results were the same. So it's not just

7:27

the leading cause of preventable mortality,

7:29

it's everything. It's the sort of

7:31

cumulative experience of these

7:33

stressors on health risk behaviors and on

7:36

the body. Sarah, thank you very much for

7:38

talking to us. Oh, thank you so

7:40

much. It was such a pleasure to talk about this work. Dr.

7:43

Sarah Miketa from Harvard School of Public

7:45

Health. And it goes along with a

7:47

lot of other research which shows that

7:49

gender dysphoria, gender identity issues, as well

7:51

as sometimes sexual preferences can

7:53

make an enormous difference to your health

7:55

and well-being. And at its heart, it's

7:57

not something that's deeply physiological.

8:00

about discrimination which affects your