Sunday, September 9, 2012

The Curious Case of Gay Men in Excellent Health

I'm giving a very short talk (15 mins) at the Gay and Lesbian Medical Association meeting in San Francisco on Saturday September 22nd at 10:15AM in the Commonwealth room on the second floor of the Westin Hotel, 50 3rd Street, downtown. You can just walk in. Please do, it would be lovely to see you.

Here's a rough draft:

Hi. Welcome. Say hello to your neighbors. Swap names. Get friendly.
This morning, I'm going to break out of the usual background-hypotheses-results-conclusions format a bit. I'm going to use a more narrative format - something closer to telling a story. So curl up and gather 'round. This is a story of inconvenient data that wouldn't go away. And a story about what happened when I settled down to listen to what it had to say.

I've broken it up into roughly three sections, the first I'm calling "the knock on the door", the second "explaining the mystery", and the third section will be trying to develop lessons, morals if you will, from the tale.

In 2006, shortly after I moved to California from my beloved Rhode Island, I decided to undertake a project to document health disparities affecting gay men, lesbians and bisexuals. There were already a bunch of studies documenting one or another specific health outcome at a time, I wanted to use a consistent methodology to track a bunch of health outcomes, including the ones we didn't expect to be different, like arthritis.
But a curious thing happened when I looked at the first health outcome in the first dataset: general health in the California Health Interview Survey.
When I graphed it out, I saw the same thing that others had already published on, that gay and bisexual men were more likely to report being in poor health (the red bars at the bottom of the graph) than heterosexual men.
But it's pretty obvious that that's not the dominant trend there - gay men in particular are more likely to report being in "excellent" and "very good" health than either bisexuals or heterosexuals.
And with everything I knew about gay men's health, that just didn't make sense. Between HIV infection, depression & suicidality, eating disorders, smoking, and so on, there's no reason that gay men should be more likely to be in excellent health.

So I did what any decent epidemiologist would do. I wrote it off and didn't give it a second thought.

Until the next year, when the next wave of the California Health Interview Survey came out. I pored over the study, and again the same findings - gay men more likely to report being in poor health (as expected), but also more likely to report being in excellent or very good health than heterosexuals.
At this point, I had to use the next level of defense to dispense with these unexpected findings. Rather than simply ignore them, I assured myself that "general health" is a squishy endpoint - it's very subjective - so who knows what it means anyway.

More datasets were coming on-line that asked questions about sexual orientation and health outcomes, so when I got my grubby mitts into the New York City Community Health Surveys, I started out with the general health question, and men who had sex with men were more likely to report being in excellent health there too. General health might be a squishy endpoint, but I knew that there were a lot of studies showing that it was a very strong predictor of mortality, and tracked very closely with other population health measures, even if it doesn't mean much for a given individual. So I had to up the ante, and pull out the big guns to dismiss the fact that both of these surveys presented such contrary findings. It must be due to sampling bias - the gay men are probably younger - that would explain it. No need to get excited.

But at this point, I definitely had my eye open, and started looking in as many datasets as I could lay my hands on. Like the unemployment survey...
The unemployment survey doesn't ask if you're gay or lesbian, but it does ask who you live with and how you're related - so you can identify men living with male partners and women living with female partners. Not ideal, but the advantage is the huge sample size. The unemployment survey calls about 50,000 people a month. So pulling several years together, I found over a thousand men in male couples. Here again, men in male couples were more likely to report being in excellent health.
To be fair, there were also studies that didn't follow the pattern - the General Social Survey for one, and a study from the Philadelphia area. But the general trend was definitely clear after looking in about a dozen studies, especially in the largest and best-conducted studies.

Explaining the Mystery...
I talked with a lot of friends, relatives, co-workers, strangers at gay health meetings, and tried to figure out what might explain this unexpected finding. Despite HIV infection rates, despite depression, despite smoking and substances misuse, somehow gay men were reporting excellent health more often. What could explain it?
The first set of explanations I looked into was whether it could be explained by demographic differences. Gay men, or at least men who will say they are gay to a stranger on the phone, tend to be younger and more highly educated than the general male population. They are more likely to live in large cities. Those could all be tested.
The second set of explanations was more psychological, and much harder to tease apart with available data. What if "excellent" means something different to gay men than to straight men? Maybe gay men, many of whom have seen death up close and far too personal, have a greater appreciation for their own health, and thus are more likely to report being in excellent health than a straight man in comparable health.
Maybe gay men have grown so accustomed to being told that gay men are sick and unhealthy by society at large, and medicine in particular, but see their own lives as being in better shape than the stereotypical gay man, and thus report being in excellent health to a greater degree than a comparable straight man. I could only get at these indirectly.
The third set of explanations was that maybe, just maybe, gay men were in fact healthier than straight men, that we are doing something different from straight men - doing something healthier than straight men. Like going to the gym more. One explanation that came up here was that gay men were healthier because they had fewer kids. I didn't quite understand the logic of that, but it kept coming up, so I looked into it.

The best dataset I had access to that would enable looking into these questions was the BRFSS - the Behavioral Risk Factor Surveillance Study. The BRFSS is a random-digit dial survey of a few hundred thousand people every year, from every state, the District of Columbia, Guam, Puerto Rico and the Virgin Islands. Like the unemployment survey, it asks who you live with and whether you're married or in an unmarried partnership, so it's possible to identify who is in a male couple vs. a mixed-sex couple.
For reasons I really don't have time to go into, you can't look at married same-sex couples in BRFSS, basically because a few straight married couples mis-identify themselves as same-sex, and even a tiny number of mis-identified straight couples mangles the same-sex married group.
So, here the men in male couples are more likely to be in excellent health than men in mixed-sex couples.

The first thing I looked into was whether demographics could explain the difference - the men in male couples were younger on average, and younger people are more lilely to report excellent health. Men in male couples were also much more likely to be highly educated, and social class is a very strong predictor of health.

And in fact when you standardize, so that everyone is weighted as though they have exactly the same age and educational attainment, the difference between men in male couples and men in married mixed-sex couples virtually disappears. But the men in male couples are still more likely to report excellent health than the men who, like them, are in "un-married" couples. Further adjusting for a range of other demographics - like race/ethnicity, income, employment status, veteran status, children in the home, state of residence or urbanicity didn't make any appreciable difference once age and education were accounted for.
So, demographics are part of the story. But men in male couples were still about 20% more likely to be in excellent health than men in mixed-sex un-married couples.

I didn't have a good way to directly test the second set of explanations - maybe "excellent" means something different for gay men than straight men? The best I could do was compare across men whose health was comparable, as best as I could figure it out from the available data, and when you control for asthma, arthritis, cardiovascular disease, diabetes, high blood pressure, high cholesterol, smoking, drinking, exercise, body mass index, and fruit and vegetable intake, the groups get more similar. But, the survey doesn't ask HIV status, and I'm not really sure that your health can really be summarized easily with that short list of factors. In short, I'm hesitant to say that this rule in or out that "excellent" health means something different for gay men than straight men, but there is definitely not strong evidence for that explanation.

Among the third set of explanations, it becomes possible to ask what gay men do differently. One of the few health advantages that has been consistently identified in the literature is that gay men are less likely to be obese than straight men - and the opposite is the case for lesbians, who tend to be heavier than straight women. Insert your own wry comment about "the male gaze" here.
And in fact, when you standardize for body mass index in addition to age and education, the differences between men in male couples and men in mixed-sex couples get very small indeed - about 12% more likely to be in excellent health. So that's pretty strong evidence that gay men are in fact doing something different.
And that something different is healthy. A lot of the gay health world has cast a disparaging eye at gym culture. Body dysmorphic disorder, beauty obsession, all kinds of pathologizing terms. And I don't mean to dismiss those phenomena as not problematic, but I think it's worth taking a step back to recognize that whatever it is that gay men are doing to their bodies differently than straight men, it has, on the whole, made us healthier.

The Moral of the Story
Stepping back to the beginning of the story, I want you to recall that it was very hard for me to recognize that there might be something interesting going on here, because the findings did not fit the dominant paradigm of gay health these days - it wasn't a health disparity, but apparently a health advantage. And minority groups aren't supposed to have health advantages. That's why the 'Hispanic Paradox' is called a 'paradox' instead of common sense.
So, what do we miss when we have trouble seeing health advantages?
Before tackling that, I want to step to the other side of that question and ask what we get from identifying health disparities? Curiously enough, we get a sense of satisfaction - health disparities confirm the idea that minority health is adversely affected by societal hostilities and incivilities. But there is more to it than that- identifying and promoting health disparities is an important modality for raising awareness and garnering resources. Raising awareness in the target community to encourage enhanced screening or whatever, and raising awareness in the eyes of funding agencies gets more resources to address the health disparities.
So seeing, describing, and promoting health advantages would seem to be at cross-purposes. First of all, if we, as a minority, are healthier, that flies in the face of being a minority that deserves special consideration. If we are healthier, then we don't need resources devoted to our health.

But I think that is a misguided interpretation. Like every other minority group, we have a mix of health disparities and health advantages. And in all likelihood, many more health similarities than either of those. So telling only the story of health disparities tells a very partial and inaccurate story about who we are, what we need, and especially what resources we can bring to the table. It is a story that focuses exclusively on the negative. It can be discouraging.
Perhaps paradoxically, it is also a narrative that perfectly fits our most potent political adversaries as well. The American Family Association, Exodus, NARTH, and others happily gobble up every health disparity we identify and claim that it proves that we are "inherently disordered", and thus unworthy of basic dignity, let alone legal protections.
So, if health disparities are a bit of a "Black box", confirming one's preconceptions rather than pointing us towards solutions, what are the alternative approaches to minority health?

First, acknowledge the health similarities and health advantages.
Second, learn what we can about why the health advantages, in particular, exist. Asking "how are you healthy?" disrupts the health disparity narrative. Whatever we learn about what's going right can potentially be leveraged to address the health disparities that we are so concerned about.

Third, let's study causes, not effects. Health disparities are effects. If we think that societal homophobia causes the health disparities, then let's measure the effects of societal homophobia directly.

Fourth, let's intervene in the causes directly, not just raise awareness of the carnage that homophobia has caused.

No comments:

Post a Comment