Two days after the holidays are over, and I'm still taking care of unfinished business from last year. Still, the study I'm about to discuss is making the rounds of the blogosphere, and because it's about breast cancer risk I felt the need to weigh in. This is particularly true, given some of the representations of this study that are popping up in the press and in the blogosphere, particularly among right wing bloggers. Let's start with a BBC news story about the study:
Women who exercise by doing the housework can reduce their risk of breast cancer, a study suggests.
The research on more than 200,000 women from nine European countries found doing household chores was far more cancer protective than playing sport.
Dusting, mopping and vacuuming was also better than having a physical job.
The women in the Cancer Research UK-funded study spent an average of 16 to 17 hours a week cooking, cleaning and doing the washing.
Experts have long known that physical exercise can reduce the risk of breast cancer, probably through hormonal and metabolic changes.
But it has been less clear how much and what types of exercise are necessary for this risk reduction.
And much of past work has examined the link between exercise and breast cancer in post-menopausal women only.
The latest study looked at both pre- and post-menopausal women and a range of activities, including work, leisure and housework.
All forms of physical activity combined reduced the breast cancer risk in post-menopausal women, but had no obvious effect in pre-menopausal women.
Out of all of the activities, only housework significantly reduced the risk of both pre- and post-menopausal women getting the disease.
Housework cut breast cancer risk by 30% among the pre-menopausal women and 20% among the post-menopausal women.
I'm betting that you can guess what sorts of reactions to this study are popping up. Unfortunately, as is frequently the case, few seem to have actually read the study itself, and most ignore the cautionary statements of the lead investigator or any hint of nuance in their rush to use this study to justify their ideological biases. For example, it didn't take long at all for our old "friend" Vox Day to leap in and make a fool of himself:
Perhaps it is the specific form of light exercise that housework offers. More likely, women who live the sort of traditional lives that involve running a household and doing housework are less likely to do the sort of things that spark breast cancer, unlike your average post-feminist skank with a career and a gym membership.
Because dosing yourself with estrogen daily and having abortions can't possibly be related to anything bad, like breast cancer, oh no. That's just the Patriarchy's mysogynistic scientists trying to keep women oppressed, barefoot and doing the dishes, a proper feminist science wouldn't be allowed to publish anything that a woman doesn't want to hear.
I honestly don't know how Vox jumped from this study to a rant about estrogen replacement therapy and abortions, but it's not surprising that Vox would misread this study as an indictment of feminism, given that he has the charming view that women have "fascistic" tendencies, a view that he uses to justify his view that women should not be allowed to vote. Oddly enough, he was half making sense with regards to how one potential interpretation of this study is that moderate housework could be a marker for another factor that is the real protective factor against breast cancer. Unfortunately, he couldn't use that Mensa mind that he bills himself as having and keep himself from launching into a tirade about "skanks" and abortion--which is, to say, it was just Vox being his usual sexist self.Sadly, not surprisingly, Vox isn't the only one jumping on this study as some sort of indictment of feminism. There's plenty of misogyny around looking for an excuse to attack feminism. For example, we have this charming post from a guy going by the 'nym of Heretic (you know, when I see a 'nym like "Heretic," I consider it a pretty good indication that the guy thinks way more of himself and his opinions than he and his opinions in fact warrant, and this "Heretic" fits that indication perfectly):
Betty Friedan, in The Feminine Mystique, claimed that, "the women who 'adjust' as housewives, who grow up wanting to be 'just a housewife,' are in as much danger as the millions who walked to their own death in the concentration camps...". Reference
We now see that the opposite is the case. Being a housewife and mother is in fact better for women's health. It is being a feminist that kills you.
Of course, where "Heretic" is coming from can be clearly seen from the little blurb under the title of his blog Heretical Sex:
Ideological movements are generally movements of the gullible led by the manipulative, and feminism is no exception. The feminist movement can be regarded as a taxpayer-funded religious social club run by lesbians, where women are encouraged to gather in secret for the purpose of hating men.
Charming, isn't he? But even Heretic is not alone. How about a guy who calls himself The Eternal Bachelor:
No wonder cases of breast cancer have been rising in recent decades; women don't do housework anymore. It's soooo sexist (unlike expecting a man to be the main breadwinner; that's not sexist at all, no ma'am.)
So next time some women hassle you to donate money for a charity that battles Breast Cancer, go one better than giving them cash; give them a big pile of ironing to do, or ask them to clean up your desk.
What a card. No wonder he remains an eternal bachelor.
Of course, none of these bloggers seem to have actually read the study (or at least the abstract, which is understandable to most lay people). But where would the fun be in that? It's far easier simply to use press reports of the study as a means of confirming your prejudicies. As Erica Barnett put it:
The funniest/saddest/most ridiculous thing about this study "proving" that housework "cuts the risk of breast cancer" by 20 to 30 percent: The researchers' actual conclusion, buried in the 12th paragraph of this BBC story, was that "moderate physical activity" of any kind--walking, riding a bike, masturbating--reduces the risk of breast cancer. So what did the BBC and other media focus on? What else--domestic servitude.
Even more interestingly, as pointed out on Our Bodies Our Blog, a similar study looking at risk factors for endometrial cancer concluded the same thing (that moderate exercise is protective) about two years ago and was spun the same way. Not surprisingly, anti-feminist bloggers like Vox, Heretic, and Eternal Bachelor were all too eager to take such spin to ever more crude and obnoxious levels.
Me being what I am, that's the first thing I thought of doing: Reading the actual study (which is an e-pub ahead of print version and has not yet appeared in the print version of the journal). So, let's start with the abstract:
There is convincing evidence for a decreased risk of breast cancer with increased physical activity. Uncertainties remain, however, about the role of different types of physical activity on breast cancer risk and the potential effect modification for these associations. We used data from 218,169 premenopausal and postmenopausal women from nine European countries, ages 20 to 80 years at study entry into the European Prospective Investigation into Cancer and Nutrition. Hazard ratios (HR) from multivariate Cox regression models were calculated using metabolic equivalent value-based physical activity variables categorized in quartiles, adjusted for age, study center, education, body mass index, smoking, alcohol use, age at menarche, age at first pregnancy, parity, current oral contraceptive use, and hormone replacement therapy use. The physical activity assessment included recreational, household, and occupational activities. A total physical activity index was estimated based on cross-tabulation of these separate types of activity. During 6.4 years of followup, 3,423 incident invasive breast cancers were identified. Overall, increasing total physical activity was associated with a reduction in breast cancer risk among postmenopausal women (Ptrend = 0.06). Specifically, household activity was associated with a significantly reduced risk in postmenopausal (HR, 0.81; 95% confidence interval, 0.70-0.93, highest
versus the lowest quartile; Ptrend = 0.001) and premenopausal (HR, 0.71; 95% confidence interval, 0.55-0.90, highest versus lowest quartile; Ptrend = 0.003) women. Occupational activity and recreational activity were not significantly related to breast cancer risk in both premenopausal and postmenopausal women. This study provides additional evidence for a protective effect of physical activity on breast cancer risk. (Cancer Epidemiol Biomarkers Prev 2007;16(1):OF1-7)
Basically, this is the classic example of a study in which multiple correlations were observed that, save for one (housework), approached, but did not achieve, statistical significance. Right in the abstract, it states that total physical activity was associated with decreased risk of breast cancer; it just missed the "magic" p value of 0.05 or below. Only one type of physical activity did achieve "statistical significance," and that was moderate forms of housework. Given the magnitude of the study (218,169 women from nine different countries), the difficulty standardizing what, exactly, is meant by "light," "moderate," and "heavy" activity, plus the sheer logistical difficulties involved in carrying out such a massive study, this is not a particularly surprising result. To overcome some of these difficulties, the authors used a standardized measurement of physical activity that assigns a MET value (the ratio of work metabolic rate to a standard metabolic rate of 1.0) to each activity. Unfortunately, even with such care getting reliable results out of an epidemiological study like this that looks at difficult to measure risk factors is a very dicey affair indeed, as the authors acknowledge in the paper:
This study has some limitations that need to be considered when reviewing these results. Although the data collection was standardized across the nine countries included in this analysis, data were only available on past year physical activity (24), and thus the effect of physical activity in different time periods of life on breast cancer risk could not be examined. In addition, there were no data available on the frequency, duration, and specific intensities of occupational activity; hence, only categories of occupational activity were recorded. Furthermore, there were few study participants who were categorized in manual and heavy manual occupations, thereby limiting the assessment of the effect of
intense occupational activity on breast cancer risk. Finally, some misclassification of physical activity levels is likely in this study thereby introducing nondifferential misclassification bias that would have biased the results towards the null.
"Biased towards the null," by the way, means that such misclassification, if not biased in one direction or the other, would have made it less likely that any statistically significant correlations would be found, because of the random noise introduced into the raw data by misclassification. It's also looking at a snapshot of how much physical activity these women did in the most recent year. This estimate may or may not correlate with their activity over the years prior, when physical activity would most likely have exerted any beneficial effect.
Although I'm not as "unconvinced" as Medpundit, I do agree that there are problems with this study and that it probably does not mean that housework in and of itself protects against breast cancer any more than other moderate activity. However, I do still have to note that certain aspects of the data seem inconsistent. One observation that leapt out at me was that a 28% higher fraction of the "noncases" (women who didn't develop breast cancer) were nulliparous (they never had children). This is inconsistent with numerous previous studies that demonstrate that nulliparity is a risk factor for the development of breast cancer. Similarly, the fraction of noncases whose age at first pregnancy was under 20 was 36% higher than among cases, again not entirely consistent with decades worth of studies that show that a young age at first pregnancy reduces the risk of breast cancer. True, taken as a whole, there was a tendency towards a younger age at first pregnancy, but the authors never stated whether any of these differences between case and noncase groups was statistically significnat; indeed, other than the trend towards a lower age at first pregnancy, these differences weren't mentioned at all. On the other hand, as one would expect, the cases tended to be older by a few years and in the postmenopausal women, 53% more cases used hormone replacement therapy, consistent with HRT as a (now) known risk factor for breast cancer.
In other words, the characteristics of the group seem to be all over the map, which makes me wonder about how valid the correlation that they found is in actuality. I wouldn't go so far as one wag did and suggest that, because housework was the most common activity among the women in the study that the results are like finding a correlation among breast cancers and wearing bras and trying to blame the bras, but that's a good way of putting the results in perspective. That's why my perspective, all can be concluded from this study is that there may be some sort of "sweet spot" for physical activity as far as producing a protective effect against breast cancer, with moderate physical activity being best. It is also possible that, as the authors speculate, that more regular moderate physical activity may be better than less frequent but more strenuous activity as far as protection from cancer. It is not clear whether one, both, or neither of these hypotheses explain this study's results. What is clear, however, is that all too many men are so hostile to feminism and women's taking on roles outside of that of wife and mother that they'll leap at anything to justify their bigotry, even to the point of using gruel as thin as the results of this study.
By "dosing yourself with estrogen daily" I think Vox Day means not HRT but birth control pills. Of course, this study didn't address that.
Orac wrote "I wouldn't go so far as one wag did and suggest that ... the results are like finding a correlation among breast cancers and wearing bras and trying to blame the bras ..."
Apparently, in 1995 a couple wrote a nutty book on the bra/breast cancer link:
I honestly don't know how Vox jumped from this study to a rant about estrogen replacement therapy and abortions, but it's not surprising that Vox would misread this study as an indictment of feminism, given that he has the charming view that women have "fascistic" tendencies, a view that he uses to justify his view that women should be allowed to vote.
I daresay Vox was ranting more about oral contraception than HRT; can't have women in control of their own reproduction, you know.
Also, I'm assuming you meant to say he was justifying his view that women should not be allowed to vote.
Orac, you're a one-man public service campaign. I got an e-mail about this yesterday night and meant to try to get the study to read this morning, because I knew there had to be flaws in the reporting. As somebody who works very hard to actually digest and understand studies before putting them out in the popular press, I'm stunned at how often I see articles that, for example, mistake coincidence for causality, and get the point of a study totally wrong.
I agree this was thin gruel and it is funny how it's being used to promote women as homemakers and the importance of being thin!
I agree with Medpundit. When I reviewed this data dredge study on Sunday for my blog, not only were none of their correlations tenable for these types of studies, but playing devil's advocate, they really didn't find a correlation with physical activity at all! :)
The real question is whether putting the toilet seat down protects against prostate cancer.
I think it's best not to formulate an opinion about research like this, other than questioning why anyone thought that something meaningful could be learned from it.
We might as well do a study on whether doing epidemiologic studies protects from one sort of delusion or another (looks like maybe it causes them). Unless we can have some contextual understanding of the results, it ends up just being the answer to a trivia question at most
Joseph, I almost sprayed my coffee all over my desk. Great comment.
They've actually tried to find that association! ...in a way. :)
But when they tested the hypothesis that unmarried/divorced/widowed men (i.e. the non-toilet seat putter downers) were more likely to to get prostate cancer, they found that married status was not associated with the development of prostate cancer. http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&li…
(It was associated with screening and treatment decisions.)
News of the Weird: Daily mentioned this. Now, NOTW isn't always the best source for combating woo; the point of NOTW is to amuse, so stories are judged largely on comedic value. But the author, Chuck Shepard, had this to say: "Yr Editor smells some severe over- or under-disaggregation of variables."
Of course, it would seem likely that not only breast cancer but cancer in general would be helped by moderate physical activity. What would Vox Day et al have to say about a study which said that regular housework protected against prostate cancer, I wonder?
VD needs to write "Correlation does not equal causation" on the blackboard 100 times before he is let back into the blogsphere. So, when do we get a study addressing the effect of housework on prostate cancer?
Hmmm, bad science reporting picked up by ignorant and self-satisfied online babblers. . . . This sounds very familiar.
If you're interested in how the press expresses, validates and amplifies popular prejudices, you've got a great opportunity. It starts in the "Femail" section of the Daily Mail on 11/28/2006, where Fiona Macrae reviewed Dr. Louann Brizendine's book The Female Brain ("Women talk three times as much as men, says study"). Macrae focused especially on two of Brizendine's quantitative bullet points:
[W]omen talk almost three times as much as men, with the average woman chalking up 20,000 words in a day - 13,000 more than the average man. [...]
Studies have shown that while a man will think about sex every 52 seconds, the subject tends to cross women's minds just once a day.
Now, just for the record, Dr. Brizendine has never done any research on either of these topics, and none of the sources that she cites provide any support for either the talking numbers or the sex-thoughts numbers. And the scientific studies that actually have counted words find that men and women appear to talk about the same amount, on average, with men sometimes a bit ahead; and the one study I've found that counted sexual thoughts reports that the frequency for males was every 12,300 seconds on average, compared to every 19,200 seconds for the females. If you care about the science, you can read about talking here and sexual thoughts here, and more on the science (and pseudo-science) of all sorts of sex differences here. [Hyperlinks in original trimmed to avoid ScienceBlogs filter.]
But this post is not about the science of sex differences — we're talking about the epidemiology of influence. And in that respect, Macrae's Daily Mail article contained two pieces of information of great value to science. First, Macrae misspelled Dr. Brizendine's first name as "Luan" (instead of "Louann"), and second, she cited the book as The Female Mind (instead of The Female Brain). These scribal errors are as good as a fingerprint or a hyperlink, and they will allow future scholars of media influence to track the flow of misinformation from Brizendine via Macrae to all sorts of places around the globe, simply by text search.
The rhetoric of the reactions is fascinating. I've cited some in an earlier post, from reader comments on the Daily Mail site and from fark.com, where the dominant reaction was "Why spend money on studying the obvious?" This is a deliciously ironic reaction, since the numbers were apparently invented without any studies being done, and the generalizations based on them appear to be false, and the only fiscal flow has been from the public to Dr. Brizendine and her publisher, Bertelsmann/Random House/Morgan Road.
That's why my perspective, all can be concluded from this study is that there may be some sort of "sweet spot" for physical activity as far as producing a protective effect against breast cancer, with moderate physical activity being best.
Nooo! Plot the effect sizes against the median MET for each class, using separate colours for the two types of activity. Visually, there's little or no evidence for any difference in effect: the difference is almost certainly an artifact of the analysis (there may be an effect in post-menopausal women, but one woiuld need to do the analysis properly. For pre-menopausal women, the quick and dirty analysis says there's definitely no effect).
What would Vox Day et al have to say about a study which said that regular housework protected against prostate cancer, I wonder?
A lie propagated by feminists who are hell-bent on denying conclusive and incontrovertible scientific results for deleterious purposes of their own. Perhaps.
At the risk of being crass, I have to ask: just how small are those guys' penises?
The other thing that's interesting here is that this is OLD NEWS. I mean, a short report in the British Medical Journal printed an association between populational progesterone concentrations and breast cancer risk in 2001 that indicated that populational lifestyle variation played a role in breast cancer etiology (and that author, Jasienska, has gone on to write quite a lot of good stuff on this topic). And that wasn't the first indication that physical activity and breast cancer were tied. The NIH even has a study section for their grants called "energy balance and breast cancer."
Thank you for this post. I am more fed up with the media right now than I can possibly articulate.