Mums who smoke and their baby's birthweight

In Episode 7 of Series 3 of the DIAL Podcast, Rita Pereira from the Erasmus University in Rotterdam and a member of DIAL’s Gene Environment Interplay in the Generation of Health and Education Inequalities(GEIGHEI) project, talks about her research looking at the links between mothers' smoking and their baby's birthweight.

The Interplay between Maternal Smoking and Genes in Offspring Birth Weight is a DIAL Working Paper by Rita Dias Pereira, Cornelius Rietveld and Hans van Kippersluis.

Transcript

In Episode 7 of Series 3 of the DIAL Podcast, Rita Pereira from the Erasmus University in Rotterdam and a member of DIAL’s Gene Environment Interplay in the Generation of Health and Education Inequalities(GEIGHEI) project, talks about her research looking at the links between mothers’ smoking and their baby’s birthweight.

The Interplay between Maternal Smoking and Genes in Offspring Birth Weight is a DIAL Working Paper by Rita Dias Pereira, Cornelius Rietveld and Hans van Kippersluis.

Christine Garrington  0:00  

Welcome to DIAL a podcast where we tune in to evidence on inequality over the life course. In this series, we discuss emerging findings from DIAL research. Our guest today is Rita Pereira from Erasmus University in Rotterdam, she has been looking at the relationships between genes, smoking and birth weight. I started by asking her about the background to the research.

Rita Pereira  0:22  

So this research has been funded by NORFACE grant. And this grant had the idea of studying gene and environment, and particularly how they interact to generate health and education inequalities. So in other words, what this means is, we already know that both genes and environment matter for generating inequalities. But our question was whether there are particularly harmful or beneficial combinations of genes and environments, in determining inequalities. In this setting, we know that maternal smoking during pregnancy is harmful for the baby. And it is reflected in a low birth weight. And we also know that your genetic constituency will determine partially how heavy or light you're going to be born. So the question here is whether there are combinations that are particularly harmful for you.

Christine Garrington  1:26  

Okay, so what was it about mothers who smoke that you specifically wanted to look at and why?

Rita Pereira  1:32  

So we wanted to know whether smoking was particularly harmful for babies, depending on their children. So other possibility was also that some genetic variants protect you against smoking? Right? So in the end, the was not so much about the mothers who smoke, we know that smoking has detrimental effects for the baby. But we wanted to know if the risk of smoking would differ depending on the genes of the offspring.

Christine Garrington  2:01

Right. Okay, so your information came from two great data sources, I wonder if you can tell us something about them? You know, the detail how many people were involved? How those datasets work? and What people are asked that was relevant to your research?

Rita Pereira  2:14  

Our main data set was a data set called ALSPAC, ALSPAC, which is from the UK, and it's a incredible data set is very detailed. And it followed pregnant women and their children that were born in the 90s, in area of risk though. And it's great for us because it has genetic information, both on the mother, on the baby. And it also has biomarkers so there are urine samples, blood samples that were collected from the mother, when they were pregnant. And it has 5000 to mother baby pairs, that are, that have genetic information on both the mother and the baby. And then we wanted to replicate our findings. And for that we use the UK Biobank. So as the name indicates, is also a dataset from the UK, and is a huge data set. So it has 500,000 individuals that are genotyped. But for our study, it has less variables that are relevant. So for example, we don't have maternal genetic information, and we don't have biomarkers during pregnancy.

Christine Garrington  3:28  

So in order to get a clearer picture you needed to get around this issue of simply comparing mothers who smoke with mothers who don't, so what did you do there?

Rita Pereira  3:37 

So endogeneity is a problem that arises if you just compare mothers who smoke with mothers who don't smoke. And the problem is that smoking is likely to be a result of other variables such as stress, social class, education, maybe personality traits. So if you just compare mothers who smoke with mothers who don't smoke, you might attribute a low birth rate to smoke while in reality, there are other factors driving the low birth weight. So our first step was to solve this issue. And to be able to solve this issue, we instrumented the decision to smoke. So an instrument is a variable that causes smoking, but doesn't cause anything else.

Christine Garrington  4:23  

And how did you do that in your research?

Rita Pereira  4:25

Our instrument was a single snip. So a single snip is a single genetic variant that causes an increase in cigarette smoke a day. In this snip is a is located in a nicotine receptor snip, so people think that it might have to do with how people get addicted and how their body simulates nicotine. We use this snip and we also show that it only correlates with birth weight on mothers who smoke so indicating that there is no effect from this snip on birth rates in mothers who don’t smoke, which is what we want. And we also find that it doesn't really correlate with other variables that are correlated with smoking. Such as age at delivery, whether they're married or single, the social class, education. So it seems that they've having snip is somehow random. And then this snip increases the frequency of smoking. So we use that to do with endogeneity.

Christine Garrington  5:18  

So that was the first issue you dealt with, that you need to overcome. You also needed to deal with this whole question of under reporting, didn't you? This question of people not necessarily giving the full picture or telling the full story because they feel they might be judged in some way, especially about something like smoking.

Rita Pereira  5:35

So usually mothers are not very comfortable reporting that they smoked during pregnancy, because it's somehow taboo, or they don't feel comfortable or they feel may be judged. So we were worried that the reporting they're reporting to smoke was incorrect. So this is where ALSPAC is a really great data set, because they collected urine of the mothers when they were pregnant, and they tested for cotinine. And cotinine is a byproduct of nicotine that stays in your system for far longer than nicotine that so it stays in your system for a few days. And that's what usually people test when you want to check if you're a smoker or not. So we had this variable. And then what we found is that while 22 mothers reported that they smoked, actually 31% had enough cotinine in their system to be considered active smokers. So that was the second problem that we needed to solve. And after we solve this issue, we will find that one extra cigarette per day decreases the baby's birth weight by by 20 to 40 grams.

Christine Garrington  6:43  

Right, so there was a there was a sort of a clear link there. And you were able to predict a baby's tendency to be born larger and heavier. How did you get that information? And what did that tell you?

Rita Pereira  6:53 

So we constructed a score, which is based on the the baby's genetic data, I mean, they're not babies now. But they were because then. So based on their genetic data, we can construct a score that gives us the child's predisposition to be born with a low or high birth weight. So just to give you an idea of how predictive this correlates, one standard deviation below or above the mean, means that you're going to be born around 100 to 110 grams lighter.

Christine Garrington  7:25  

Pulling all of this together, what would you say that we learn overall, beyond what most of us know about smoking being bad for our health and the health of our children?

Rita Pereira  7:34  

First and foremost this research highlights how harmful the nicotine effect is on your baby. But besides that, we also want to check whether the risks differ between the babies when we don't find any evidence for meaningful interactions. So this leads us to conclude that the effects seem to be the same for all babies, regardless of their genome. Okay, so it affects every baby apparently, the same more or less the same.

Christine Garrington  8:03  

Do you have any plans then to take this work further?

Rita Pereira  8:06

Yeah, we do, actually. Because now we found this great data set that allows us to have a really cool instrument, and correct for under reporting. So it's definitely in our plans to check whether smoking has other harmful effects, maybe in other outcomes. And besides that, we also wanted to look at gene environmental interactions, maybe in other contexts, such as mental health, education, and so on.

Christine Garrington  8:33  

The Interplay between Maternal Smoking and Genes in Offspring Birth Weight is a DIAL Working Paper by Rita Pereira and colleagues involved in the GEIGEI project. You can find out more about the project at gene-environment.com, and about the wider DIAL programme at dynamicsofinequality.org. Thanks for listening to this episode of our podcast which is presented and produced by Chris Garrington and edited by Elina Kilpi-Jakonen.