Statistics and FSD (part 1 of 2)06/07/2010 at 7:14 am | Posted in Uncategorized | 6 Comments
Tags: experts, female sexual dysfunction, FSD, health, journals, media, medicine, sex, sexual dysfunction, sexual health, statistics
I’ve been meaning to sit down and write these posts for a long time. We interrupt our hopefully-continuing series on BDSM and FSD to bring you statistics, and to look at some problems with the statistics. Although there are a number of studies on sexual dysfunction, for now we’re going to look at two controversial studies, one from the Journal of the American Medical Association and one from Obstetrics and Gynecology. I was hoping to present both of these studies as a unit, but once this post approached the 2,000 word mark, I had to split it up. I will present part 2 shortly.
I notice when I’m reading blogs and books about FSD, some numbers, statistics and claims about the prevalence of sexual health problems keep on popping up. I keep seeing some numbers and themes over and over again.
When I think of sexual dysfunction journals and statistics, my first thoughts are of a study published in the Journal of the American Medial Association in February 1999. It’s a pretty famous journal article and it caused a lot of controversy when first published. You may already be aware of the article yourself; I knew about it before I actually sat down and read it. The article, Sexual Dysfunction in the United States, is available in full text online for free. We can read it together, and we probably should, since, as this page on the Vaginismus Awareness Network points out, a lot of people cite it without having read it. Let’s fix that.
Here’s what I knew about Sexual Dysfunction in the United States before ever siting down and reading the article myself: One of the authors had financial ties to a big drug company, so there was a major conflict of interest with his involvement in an article that talked about sexual dysfunction. And the article took a very broad, sweeping definition of sexual dysfunction. The authors did not factor personal distress into the equation. I knew these things, because other journalists and sexologists like Dr. Tiefer had said as much elsewhere.
That’s what I already knew about the JAMA article. Is that enough? Is that all I need to know about it? Can I just leave it at that and not worry about it? Nah, I can’t just leave this stuff alone. I had to see it for myself. One caveat to keep in mind though, I’m not a statistician. Although I took several statistics classes in college, it’s been a long time since then, and I did my degree is unrelated to the statistics of interest today. So when interpreting numbers and tables, I tend to rely on the conclusions & discussion sections provided by study authors – as well as critiques provided by experts who have read the studies. Fortunately, you can double-check my work since I’ll be providing my sources.
Like I said, a major problem with this study, is that one of the authors, Edward O. Laumann was connected to pharmaceutical giant Pfizer, the maker of erectile dysfunction treatment Viagra. Viagra was already out on the market by the time Sexual Dysfunction in the United States was published, having received FDA approval about one year earlier. Laumann had served as a member of Pfizer’s Scientific Advisory Committee since 1997, and had connections to other big pharma companies like Merck and Eli Lilly, too. That means that right off the bat, Laumann’s involvement in the study was a big conflict of interest. To make matters worse, Laumann’s connection to Pfizer wasn’t explicitly stated at the time of the article’s publication in February. It wasn’t until April 1999 that a correction was printed in JAMA. That doesn’t seem to have hurt Laumann’s career too much however; in doing a quick PubMed search, I see that Laumann is still involved with sexual dysfunction research.
The other problem is that this study may have overestimated the prevalence of sexual dysfunction in the US. This is the study from which we get the statistic that says about 43% of women have sexual dysfunction. So where does that number come from?
The statistics and conclusions in the JAMA article are based on the National Health and Social Life Survey (NHSLS,) a 1992 US study of 1410 men and 1749 women. The NHSLS survey was conducted using in-person interviews. In order to be eligible for the survey, you had to have had at least one sex partner within the previous 12 months, so 139 men and 238 women were excluded due to lack of partnered sex during that time period. Everyone interviewed had to speak English fluently, and most people interviewed were white. Everyone had to be at least 18 years old, and the cutoff age was 59. The JAMA study does not tell me if gender identity was taken into consideration, but as near as I can tell the researchers were interested only in participants’ biological, binary sex. If you’ll look at the tables in the study, you’ll see then that in the end around 1480 womens’ and about 1250 mens’ responses were included in the survey analysis. The math doesn’t quite add up perfectly, and some of the numbers in specific categories of interest in the tables differ a little – 1486 women here, 1477 women there. I don’t know why that is but with numbers that big, little variations may not be significant.
Now unfortunately I do not have the full text of the NHSLS survey yet so I can’t see for myself what questions the interviewers asked participants, and I can’t see for myself the details about how the conclusions were drawn. I have found books that go into detail about this study. If you’re interested, look for The Social Organization of Sexuality: Sexual Practices in the United States for technical details or Sex in America, which was written for a general audience. Personally I think I would go for Sex in America first since it’s the less technical of the two, but that’s a ways down the road since I believe I’ve got quite enough on my plate to keep me writing about FSD for a long time. (So long as I don’t burn out first. Pace yourself, K!)
According to Sexual Dysfunction in America, there are seven symptoms of sexual dysfunction. Keep in mind though, the authors did not factor in the degree of symptom severity, and did not factor in how individuals felt about their sexual problems. (Some people have difficulty maintaining an erection or researching reaching orgasm, and are comfortable with that.) The results of this research was based on yes/no responses.
Five sexual problem symptom areas were common to men and women:
- Lack of or low desire for sex
- Arousal difficulties (erection problems for men; lubrication problems for women)
- Difficulty orgasming
- Performance anxiety
- Lack of pleasure during sex.
There were two additional areas that were sex specific:
6. Painful sex (women only)
7. Premature ejaculation (men only)
The article says, “Taken together, these items cover the major problem areas addressed in the Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition” (Laumann et al, 1999). Remember, the DSM-IV and soon-to-be V includes sexual dysfunction. The criteria for sexual dysfunction in the DSM-IV is strongly influenced by the Human Sexual Response Cycle according to Masters & Johnson. It’s interesting that the NHSLS asked questions about anxiety and lack of pleasure during sex, which is a little more comprehensive than the four categories of sexual dysfunction (pain, low libido, low arousal, difficulty orgasming) generally recognized. However, in the Results discussion of the journal article, the authors mostly focused on the DSM’s definitions of sexual dysfunction. The researchers did not inquire about restless genital syndrome (maybe because there was even less research about it back in 1992?)
The article authors came to a number of conclusions. Do read the article when you have a moment if you’re interested in all of them. One social construction criticism of the medical model of sex is that medicine is obsessively genitally-focused, however a lot of the study conclusions seem like they would fit in perfectly fine with a social construction perspective of sexuality. Some examples: The researchers found that higher education was associated with lower rates of sexual dysfunction. But as you probably already know, higher education is also closely tied into class status. And people with higher class privilege have a better chance of getting adequate pay and access to medical and child care. The authors didn’t get into that to much, except to say that “Deterioration in economic position, indexed by falling household income, is generally associated with a modest increase in risk for all categories if sexual dysfunction for women but only erectile dysfunction for men” (Laumann et al, 1999). So what about families that have low income to begin with? Being married was also associated with lower rates of sexual problems – I’m thinking maybe because these people had a long-term experienced sex partner? (And what about same-sex couples, most of whom cannot legally marry in the US?) Women with chronic health problems were more likely to experience pain.Victims of sexual assault and sexual abuse were more likely to report sexual problems, so the effects of abuse can last for a very long time. Fewer black women indicated sexual pain than white women, which is consistent with what I’ve read about vulvodynia before, but it wasn’t a zero response rate.
The authors used a statistics tool, latent class analysis, to come up with the statistics on the prevelance of sexual dysfunction in the US. “These results indicate that the clustering of symptoms according to syndrome can be represented by 4 categories for women as well as for men” Laumann et al, 1999). This is what I meant when I said that in the results section, the authors’ conclusions were very similar to what’s listed in the DSM-IV. Similar, but not quite the same…
For women, the results were:
- 58% no problems
- 22% low sexual desire
- 14% arousal problems
- 7% pain
And for men the results were 70% no problems, 21% premature ejaculation, 5% erectile dysfunction and 5% low sexual desire.
Additionally, tables 1 and 2 presented the statistics for affirmative responses to the questions about sexual anxiety, lack of pleasure during sex and difficulty orgasming for men and women, broken out by various demographics like age and ethnicity. The description for tables 1 and 2 provided on page 540 does not say anything about whether latent class analysis was used. I’m open to correction on this, but I would say to a professor, I think it wasn’t used with tables 1 and 2.
Tables 1 and 2 are a little hard to read; pay attention to the note underneath table 1 that explains, “Percentages are derived from respondents in each category, and the total number represents those who responded to the questions” (Laumann et al, 1999). I found it easier to understand if I went by the total number of responses. For example, the total number of women who said they felt anxious about their sexual performance was about 183 (78 of these women were between the ages of 18-29, 57 between the ages of 30-39, 36 between the ages of 40-49 and 12 between 50-59,) and that’s from a total sample size of 1482 for this question. The reason I emphasized about in the last sentance is that I double-checked my math on this question through some of the other categories like education level and ethnicity and the total number of affirmative responses went from 180-183. I don’t know why there’s a slight variation.
The reason I said I think tables 1 and 2 did not use latent class analysis is that, I broke out my calculator and added up the number of affirmative responses to the NHSLS question about “Trouble maintaining or achieving an erection” (Laumann et al, 1999.) Keeping the above paragraph in mind, the survey results said about 127 men reported erectile problems, out of a sample size of 1244. But 127/1244 = about 10%, not 5%… So the math and the category grouping is weird. I tried it again with the question about women’s sexual pain and came up with about 228/1479 = about 15.5%, not 7%. So I think I need help with this LCA stats tool. I’m not getting how we’re going from one stat to the final conclusions about prevalence. It’s like you take the survey result statistics, run the numbers through the LCA machine and come up with different numbers.
This latent class analysis tool bugs me. I mean, look at what the authors say about what it is, and then they are all like, “A more detailed discussion of this method is available on request from the authors” (Laumann et al, 1999) and I’m like – it’s so overly complicated. Could you be any more obtuse? I wonder if it would make any more sense to me if I’d majored in stats.
But in the end, the study authors did not talk about these three responses much anyway.
So the authors conclude that “The total prevalence for sexual dysfunction is higher for women than men (43% to 31%)” (Laumann et al, 1999.) The authors explicitly use the words “Sexual dysfunction” here, not “Sexual problems” or “complications” or something else. I also can’t tell what, if any, steps the people who answered “Yes” took to ameloriate any problems. 20% of women and 10% of men did look for medical advice to address their sexual problems.(Laumann et al, 1999).
But back to that statement about the total prevelance for sexual dysfunction – That’s where a lot of the controversy around sexual dysfunction stems from – the study didn’t factor in personal distress. Sexual dysfunction was arbitrarily declared for anyone who had answered “Yes” to whether they’d experienced one of the sexual problems of interest. That’s a problem – it didn’t leave room for self-identification. The use of language in the study is a point that journalists bring up over and over in order to dismiss FSD as entirely fake. Or to make the point that the numbers of sexual dysfunction are so high in this study but there’s no way it’s really that high. Reading between the lines, the gist I get from journalists is that, sexual dysfunction happens to few enough people so that those few people don’t really matter, as evidenced by journalists devoting very little (if any) column space to the opinions of people who actually have sexual dysfunction, and additionally evidenced by sexologists who crack jokes about people who have sexual dysfunction. For all the debate on female sexual dysfunction, these discussions about it rarely center women with FSD.
There’s something else about the study results. I’m getting mixed up with the stats. 22 + 14 + 7 = 43 and 21 + 5 + 5 = 31, yes. But to come up with that 43% or 31% statistic, each category had to be discrete – not overlapping. But in reality you can have more than one sexual dysfunction problem – they’re not always discrete. Problems can overlap. It is not clear to me whether the researchers factored in overlap or if each category was completely discrete. Plus again, keep in mind, these results came from yes/no responses instead of the degree and/or frequency of symptoms or the distress the symptoms caused.
Note also: There’s no category for lack of orgasm for women. I think that’s an unfortunate omission and I can’t figure out why it was left out. (The cynic in me thinks, “Lack of interest.”) There is no category of pain for men however this may be because, I have read elsewhere, men do report sexual pain much less frequently than women. Less frequently, but still more often than never. The authors didn’t make a category for lack of pleasure during sex, even though the NHSLS asked about it.
However, even with these problems in mind…
I am not yet ready to completely dismiss the results of this study. I am ready to take it with a heaping tablespoon of salt. It’s got flaws, yes, but the data wasn’t made up out of thin air either. The people who were interviewed said “Yes” that they experienced a sexual problem, although they may not have felt it was a big deal. But some difficulties were there from time to time. It really happens, and I’m hesitant to dismiss that. And I cannot dismiss FSD as entirely fake, seeing as I still have it over here. It’s still real. It may not be a dysfunction to all 43% of the respondents who answered in the affirmative that they’d had a sexual problem over the last year, but for some of those respondents… yeah, it really was distressing.
One other interesting feature about this 43% statistic is that I saw a very similar number reported elsewhere, almost a decade later. This is not the last time we’re going to see the number of women reporting sexual difficulties (though not necessarily dysfunctions) somewhere around 43%. There really may be something behind this.
So there’s one big, famous study on female sexual dysfunction for you. It’s “The first population-based assessment of sexual dysfunction in the half-ceutnry since Kinsey et al,” (Laumann et al, 1999) and it treats sexual dysfunction very seriously, but it’s got some problems that need to be addressed. Unfortunately, ultimately the weaknesses in the study aren’t helpful for people who live with sexual dysfunction.
Next week, barring unforeseen circumstances, part 2.