Living the Scientific Life (Scientist, Interrupted)

This study, carried out by Cancer Research UK, showed that women who did 16-17 hours of housework per week cut their risk of breast cancer by 20% for postmenopausal women and 30% for premenopausal women. Further, it was housework specifically that has this beneficial effect, not other forms of physical activity;

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.

[ .. ]

Out of all of the activities, only housework significantly reduced the risk of both pre- and post-menopausal women getting the disease.

This study sounds fishy to me.

Cited story.
Image source.

Comments

#1 Monado
December 29, 2006

Maybe they’re staying out of the air pollution?
#2 doctorgoo
December 29, 2006

Maybe they’re staying out of the air pollution?

Who? Those who are outside or those who are working with noxious fumes from the cleaning solutions?
#3 parrotslave
December 29, 2006

What is next?

The barefoot and pregnant study?

Or maybe they’ll say that watching football and drinking beer reduces prostate cancer.
#4 Elf Eye
December 29, 2006

“The international authors said their results suggested that moderate forms of physical activity, such as housework, may be more important than less frequent but more intense recreational physical activity in reducing breast cancer risk.”

This doesn’t sound so illogical. Housework involves steady and frequent (possibly daily!) exercise versus athletic workouts, which, while intense as you are doing them, have to be fit in at special times and places. I know it is easier for me to fit in housework than a trip to the gym! By the way, may I recommend the vacuuming workout? It is especially effective if you have two dogs and one cat whose propensity to shed will require multiple passes over each square foot of carpet. If a more vigorous workout is required, follow up by shampooing the carpet. For the cooldown phase, folding laundry is effective. Seriously, if you do your own housework, by the end of the week you may have put in a lot of hours and burned a lot of calories.
#5 Bob O'H
December 29, 2006

The paper is avaiable as a pdf. Observational data being what they are, one should be cautious about inferring causation from correlation. In particular, I wonder if doing a lot of housework is correlated with being a housewife, and hence not having to go to work. It could then be that this is the causative factor (for whatever reason).

I would also be a bit concerned that the data are not balanced (this is observational data: they never are), so there could be partial confounding, e.g. household activity could be correlated with another factor that has an effect. This doesn’t seem to be discussed.

Bob
#6 J-Dog
December 29, 2006

I hate to sound sexist, but some guy had to do this study!

I picture like a 1950-1960′s science guy (Desi or Daren) in a lab coat, telling his stay-at-home wifey (Lucy or the bewitched witch, whatever her name was) all about this! What a scam.

I wonder if it will work at my house? I can’t wait to tell my better half… I mean 3/4s Okay, 7/8′s..

Have a Happy New Year everyone!
#7 Mustafa Mond, FCD
December 29, 2006

So what are the stats on housework vs. prostate cancer?
#8 Bob O'H
December 30, 2006

I’ve just looked at the paper properly. Table 3 is the important one. They measure different activity classes using something called MET, which is calculated in a rather complicated way. Here are the cut-offs between hte categories:

Recreational: 14, 25, 42
Housework: 28, 52, 90

They then tested for trends by regressing against the scores of the categories, so a MET score of 43 in in class 4 for recreational activity, but the same effort scores as class 2 for housework! The effect of this is to reduce the magnitude of the recreational slope.

I did a rough-and-ready analysis, which needed a few short-cute (i.e. don’t be surprised if the real results are slightly different), and I got p-values for the difference in the effects of activity of 0.47 for pre-menopausal women and 0.03 for post-menopausal women. So, there might be an effect in post-menopausal women, but the evidence isn’t strong. A proper analysis is needed.

Overall, there’s an effect of activity, and at best a small effect of housework over recreational work. The difference is that housework entails doing more activity that recreation.

OK, must go for a run, and then do some dusting…

Bob
#9 Bob O'H
December 30, 2006

Went for a run, did some cleaning, and even went shopping. And no sign of any suspicious lumps. So exercise must work.

Bob
#10 Tabor
December 30, 2006

Did they do any correlation to these women who do housework…do they do it for others? Do they get more sex? Just asking?
#11 Orac
December 30, 2006

Yeah, what should greet me when I get back but this study?

I’m gonna have to take a crack at it next week.
#12 Bob O'H
January 1, 2007

Orac – I’ve emailed the authors asking a few questions about the analysis, so we’ll see what they say. I can keep you updated, if you wish.

And a happy new year to you all!

Bob