Friday, July 30, 2010

Socio-economic predictors of breastfeeding in Ireland

When a child is born the first investment decision a parent makes is whether to breastfed the child. This usually occurs within the first hour. Of course decisions taken beforehand, like diet, also have an important role. Since breastfeeding is generally considered to have important physical and other benefits for both child and mother, understanding who is more likely to breastfeed (& ultimately why) is important. This is particularly the case in Ireland where breastfeeding has a relatively low incidence.
It might be conjectured that social class has some role to play. High SES mothers may have access to better information (or be better able to process the information) as well as support networks. They have less of a need to return to work. On the other hand, the opportunity cost of their time may be higher.
The graph below shows the proportion of children breastfed by socio-economic class (where there are two adults in the household, it is the highest of the two classes).

There is a very striking gradient with children from manual backgrounds being much less likely to be breastfed.
Of course, one can't be sure what the reason for this link is. Class is correlated with education, income and other factors. Looking at the gradient with respect to the primary carer's education one finds:

Again there is a strong pattern with children of more educated mothers being much likely to be breastfed. What happens if we put both factors together in a model? Well both play an independent role although education seems to be the dominant factor. Interesting, household income does not have a significant role to play given social class & maternal education.
One can find other predictive factors: younger mothers and those from rural areas are less likely to breastfed. But these factors seem to have a fairly small role compared to social class and education.
If we look at mother's behaviour during pregnancy for clues as to whether they breastfed one comes across an interesting anomaly. Those who smoked during pregnancy are about 9% less likely to breastfeed, other things being equal, whereas those who consumed alcohol are about 3% more likely.


David Madden said...

Interesting correlations, Kevin. My guess, from my own work on smoking and other health behaviours, and from personal observations regarding child-rearing, is that it all comes back to human capital investment and those unobservables which are correlated with such investments. Breastfeeding can be difficult, mothers are asked to start breastfeeding at a time when they are typically physically (and often emotionally as well) shattered. And there is an easy alternative with formula milk. It is an investment which takes a lot of self-discipline and will-power, so we would expect that it would be correlated with other human capital investments, like education and not smoking. The correlation with drinking is also perhaps not so surprising. Do you know how much the mothers drank? Also do you know if mothers who drank during pregnancy then abstained during breast-feeding? The dose-response relationship between drinking and health is non-montonic (unlike smoking), so it is quite possible that moderate drinking, while not quite a health investment, would not be viewed as being as injurious to health as smoking (while writing this I can't help being reminded of a memorable scene from "The Snapper"!).

Kevin Denny said...

There is information on the volume of drinking but I haven't looked at it: only a blog post! The answer to the second question is No I think, though I think you might know whether they currently (8 or so years later) drink.
I agree about education but one might think of smoking & drinking as proxies for high discount rates. The education effect may not only be due to the psychological factors you mention, it could be due to access to information or support. I don't know much about the situation in Ireland - as far as I know there is not much support for breastfeeding in the hospitals in general: they don't have the resources.
In the US there are lactation consultants who's job it is to help mothers. But my guess is that middle class moms have, better support. Some hospitals also adopt "baby friendly" practices specifically: getting it right before discharge is important. From the UK data that we have been working with there is a big drop-off in breastfeeding after the first week.