Religion som instrumentvaiabel?

Denne tar kaka:

Big hits of epidemiology: church makes you fat!:

Time magazine has a headline out yesterday: ‘Why Going to Church Can Make You Fat.’  Here are some excerpts,

Maybe it’s all the church socials, but a new study finds that those who attend religious activities are more likely to gain weight than those who don’t go to church as often . . . researchers at Northwestern University sought to find out how attending religious events affects weight gain over time. They analyzed data from the Coronary Artery Risk Development in Young Adults study, which followed more than 2,400 people aged 20 to 32 for 18 years. Over that time, the scientists reported at an American Heart Association conference, people who went to church or church activities at least once a week were more than twice as likely as people with no religious involvement to become obese.

Is this a great story or what?  Time wasn’t the only media outlet to think so.  Google News lists 66 articles on this today, with MSNBC winning the prize for best headline for ‘Praise the lard? Religion linked to obesity in young adults.’

So what’s going on here?    4th-year medical student Matthew Feinstein is presenting this work at the American Heart Association’s Nutrition, Physical Activity and Metabolism/Cardiovascular Disease Epidemiology and Prevention Scientific Sessions.  There’s no paper yet, so all we have to go on is the press release from Northwestern.  Feinstein and colleagues modeled the likelihood of becoming obese later in life for normal weight young adults as a function of religious participation, controlling for age, race, sex, education, income, and baseline body mass index.  Frequent attendance at religious services was associated with a 50% increase in risk of obesity.

So does this show that going to church make you fat?  Um, no.  Regular attendance at religious services is highly correlated with lots of other variables like region of residence, participation of women in the workforce, number of children, attitudes about science, and others.  I wouldn’t be surprised to find it correlated with dietary customs and degree of comfort exercising in public.  Given all these omitted variables, I think it’s pretty likely that religious attendance is serving as a proxy variable in this model—a variable that doesn’t drive the result itself but stands in for something the researcher doesn’t have.

But this result is so much fun!  And I’m sure it will be good for young Dr. Feinstein’s career.  From a scientific perspective, though, was this worth doing?

(Via The Incidental Economist (Posts).)

Som nevnt, det finnes ikke en artikkel å gå etter i sømmene (ennå) så vi må spekulere litt. Økonometrien i dette tilfellet handlet nok ikke om religion i utgangspunktet, men som i ett ‘moment of clarity’ ble korrelert med andre endogene variabler. Og ja, hvis dette er signifikant, et artig resultat som sikkert tar opp en god dags diskusjon på ett seminar.

Legg igjen en kommentar

Fyll inn i feltene under, eller klikk på et ikon for å logge inn:

WordPress.com-logo

Du kommenterer med bruk av din WordPress.com konto. Logg ut / Endre )

Twitter picture

Du kommenterer med bruk av din Twitter konto. Logg ut / Endre )

Facebookbilde

Du kommenterer med bruk av din Facebook konto. Logg ut / Endre )

Google+ photo

Du kommenterer med bruk av din Google+ konto. Logg ut / Endre )

Kobler til %s

%d bloggers like this: