Total Drek

Or, the thoughts of several frustrated intellectuals on Sociology, Gaming, Science, Politics, Science Fiction, Religion, and whatever the hell else strikes their fancy. There is absolutely no reason why you should read this blog. None. Seriously. Go hit your back button. It's up in the upper left-hand corner of your browser... it says "Back." Don't say we didn't warn you.

Tuesday, October 28, 2008

Brings new meaning to the term "cup size."

Regular readers of this blog are familiar with my enduring fascination with breasts. More often than not this fascination is not prurient, per se, but rather stems from a recognition of the significance of breasts in our culture.* For example, I once wrote a series of posts dealing with the bizarre attention paid to, and value placed on, natural breasts as opposed to surgically augmented breasts.

I'd feel embarrassed by my occasional posts on the general subject of boobs but, if anything, I am not as interested as some others seem to be. Research has been conducted, for example, with the intent of producing the ideal boob job. Nope, not making that up. On a somewhat less bewildering note, research has also been conducted into the evolution of the breast that points to its origins in the immune system. Pretty neat stuff, actually.

Well, research on breasts hasn't ended and so my attention to them remains strong. Particularly, I recently ran across an article discussing research linking breast size to coffee consumption. Seriously:

Scientists have discovered that drinking just three cups of coffee a day can make women's breasts shrink.

Nearly 300 women were surveyed about their bust measurements and how many cups of coffee they drank in an average day.

According to researchers, three cups a day was enough to start making breasts shrink, with the effects increasing for every cup drunk.

Now, the quantoids out there are noting, as I did, that this research is apparently cross-sectional, meaning that it can't determine causality. So, for example, it could be that increased coffee consumption leads to decreased breast size, it could be that women with smaller breasts just like coffee more, or it could be that some third variable is controlling both breast size and coffee consumption. And in perfect honesty, I don't know which of these options I find less bizarre than the others. And all that assumes this this isn't just a giant load of crap in the first place.

More interesting, however, is what appears buried late in this admittedly short the article:

It's not all bad news for women however as the researchers also found that regular hits of caffeine can help to cut the risk of developing breast cancer.

Scientists said that the effect of coffee is related to its impact on estrogens - the female sex hormones.

Some substances in coffee can change a woman's metabolism so she acquires a better configuration of various estrogens, therefore lowering the potential risk.

But women with bigger breasts that contain more mammary glands are at a higher risk, the scientists added.

And this is an addition I find interesting. You see, what we have here is a substance that is purported to have two particular effects on women. First, it may reduce the size of a body part that is considered central to sexuality in our culture. Second, it may reduce the risk that the aforementioned body part will develop a life threatening disease. Speaking personally, I would think that the latter revelation is more noteworthy, particularly given the absurd prevalence of craptacular merchandise that supposedly contributes to breast cancer research. So, it seems like the headline should read, "Drinking coffee may reduce breast cancer risk," and then somewhere in the article there should be a mention of the additional cosmetic implications. Sadly, however, when you do a google search for the name of one of the study's authors (i.e. Helena Jernstroem) you find an array of reports on the research which all essentially take the same line- coffee consumption shrinks teh boobs:

And I think this says something interesting, and probably not flattering, about our culture. Maybe the measure of a man is hard to take, but it sometimes appears that the measure of a woman is taken across her chest.

* Also I have a prurient interest in boobs.

Labels: , , ,


Blogger Mister Troll said...

I thought it was well-known that breast size was correlated with an increased risk for breast cancer (in women). Yet a quick glance at Teh Google doesn't support this. Hmm. I'll have to do more reading.

But, if true, then there's a confounding correlation in the study you cite.

Slightly off-topic, I've found it interesting that parity is correlated with a reduced risk of breast cancer. Supposedly the interruption of the menstrual cycle reduces the level of the relevant hormones, which decreases risk. Yet I find this baffling because at least some hormone levels go through the roof during pregnancy, which seems contradictory.

If I have time later today, I'll do a quick read of the article you reference and see if I have any insights...

Tuesday, October 28, 2008 11:06:00 AM  
Blogger Drek said...


I'll be curious to hear what you find, if you have time. Despite my love of boobs, this is definitely not my area of professional expertise.

Tuesday, October 28, 2008 12:32:00 PM  
Blogger Mister Troll said...

Hey, you linked to a non-authoritative source! No fair making me do my own MedLine searches :-)

... where of course I don't actually find it. Eventually I tracked it down in the British Journal of Cancer (advanced online publication). I could mail the pdf if you like.

For a discussion of the paper (and see links to previous research, which describes a study of coffee and breast cancer risk--I believe you're looking for that?):

The study tries to link coffee consumption and breast size (with implication to breast cancer risk, but this was not studied directly). The study was stratified for oral contraceptive use (makes sense). Users showed no relationship between coffee consumption and breast volume. Non-users showed a relationship between coffee consumption, breast volume, and a *specific* genotype. However, measured hormone levels did not explain the relationship.

Hmm... I'm going to have to think much more carefully about this. Looks like I found my bus-reading material... if I find anything interesting, I'll check back.

(I think I've spotted some intriguing biostatistics issues... as you say, cross-sectional studies have their problems. But I'm not sure that I'm right, so for now I have to keep my mouth shut.)

On a related topic, are you familiar with Ionnidis' paper "Why most published research findings are false"?

Tuesday, October 28, 2008 6:41:00 PM  
Blogger Mister Troll said...

I'm back -- is the party over yet?

I had some time to actually read the research. I'll take back my initial skepticism... I thought I'd spotted some statistical errors, but after I was more careful in my reading I realized the authors did at least attempt to control for the the problems I'd identified.

Still, I'm unimpressed. First, as you pointed out, it's hard (impossible?) to prove causation in a cohort study. Second, the sample sizes are very small (30-40 women each in of two groups with the same relevant genotype), with almost no non-drinkers in the sample. Third, smoking is an obvious possible confounding variable [correlates with coffee drinking], although the authors did correct for current/past smoking (as binary variables, not in terms of pack-years which would be much better). Fourth, I would assume that breast size, breast density, coffee drinking, smoking, said genotype, and other unknown confounding variables could be correlated within families. The entire study (~400 women) included only ~100 families. When the authors corrected for family clustering, their p-values (all hail the magical p-value!) stayed significant but the association was definitely weakened.

Most importantly, the authors don't have a plausible biological mechanism relating coffee consumption and breast size. They point out that coffee induces hormone activity (related to the genotype in question), but they don't know what compounds in the coffee are relevant in this process. (Caffeine? Others?) Getting a plausible biological mechanism is a must. And when they measured hormone levels, they did not find an association between coffee drinking and hormone levels *such that* it would explain the apparent link between coffee consumption and breast size. (Oops!)

And lastly, breast density (according to the authors) is a more important risk factor for breast cancer in women than breast size (which is a risk factor primarily in lean women). The study did not include a mammogram, which is very unfortunate -- so ultimately I don't see how the study at all implies anything for breast cancer risk (which was of course their goal: show coffee reduces breast size, to suggest that coffee reduces breast cancer risk).

As I said, I was unimpressed. I would call this type of research "hypothesis generation". Of course, it generated lots of fascination on the internet (coffee shrinks teh boobs!), as you of course are reflexively aware :-)

And I'll add a footnote as well: none of this is my area of professional expertise. I'm happy to be corrected by anyone who knows more than I do about the topic.

Thursday, October 30, 2008 5:35:00 AM  
Blogger Drek said...

Hey Mister Troll,

Not too late to the party at all! I meant to respond to your comment from yesterday but I got busy for most of the day. You know how it is.

Thanks for finding and reading the relevant article. It definitely sounds like the work is speculative and, more importantly, is probably plagued by confounds in a big way. I can't say that I'm surprised since, if nothing else, controlling for enough environmental factors to make a presumably weak cancer reducing effect from coffee emerge would be prohibitively difficult. This work reminds me a little of that spate of articles a few years back claiming that various teas or wines prevent cancer and/or heart disease. Maybe some do, maybe some don't, but the research was often cross-national and rife with opportunities for unobserved heterogeneity. Still, it's nice to know my skepticism isn't completely baseless.

What's the cite on the Ionnidis paper?

Thursday, October 30, 2008 6:13:00 AM  
Blogger Mister Troll said...


PLoS Medicine 2, 696-701.
JAMA 294, 218-228.


Thursday, October 30, 2008 9:44:00 AM  

Post a Comment

<< Home

Site Meter