MCHB EPI Atlanta Conference
 
December 5 - 7, 2006

Genetic Epidemiology, Genetics, and MCH

JUAN ACUÑA: We’re ready to start our Wednesday’s afternoon plenary session entitled “Genetics, Genomics, Epidemiology at MCH”. Okay, thank you. “May god let you meet many wrong people before you meet the right person so when you do, you will be grateful.” And that is a quote from a Colombian Noble Prize winner Garcia Marquez, a “Hundred Years of Solitude” and he was talking about love but my point is that in epidemiology, we usually or many times fall in love with some variables, and we would die first and let them go second. And that creates an analytical problem, and an analytical problem that is translated into problems and translated later our findings to the program people and the policy makers who make decision based on just alterations or relative risk that do not translate into R-squares that prove that our levels of association are really good.

So we have been, in MCH, an important problem such as pre-term birth. Meeting many of these variables and if you guys have followed up literature investigating the risk factors for preterm birth, you will know that the best models out there in published literature do get R-squares that barely explain up to 15 percent of the whole phenomenon. Which means that we have been really engaged into talking about a relative risk of natural ratio of three, I’m happy with it. Still, with the underlying fact that the level of explanation is mostly in bearer term of our models.

Will we find a variable that explains it all? And the answer is easy, no. Too complex of a problem to just find a smoking gun. So we have to take different approaches or will there be manual variables that will explain the problems that we just consistently and constantly dump into the error term so we can forget about them with our clean conscience type of state of mind? And probably the answer is, “Yeah, we would like to have some of those variables.” So I think that we, as a epidemiologist, should be grateful for what genetics brings to the plate in many of these conditions because a lot of what was before immeasurable that really relates to the problems is within variables that we will be able to measure hopefully soon.

I will not go through explaining background of genetics because it’s really complex and too extensive. But, you will have to accept that it’s a summary from the description of phenotypes and the relationship to genetic sequences and the ability through different mechanisms using the laboratory to really account for those phenotypes of those characteristics and their expression and then the interaction between all those phenomena of those genes and those sequences on how to function and how they express themselves, meaning jumping into the concept of genomic.

If you measure the importance of a topic by the number of cover pages, and that’s how the models do it you know so, genetics must be very important because it has been the front cover of very important magazines. I mean the Vogue of science and at some others. But you see that there are some public reports, private reports, lay magazines there that talk about this important topic and I would like to especially call the attention upon the Newsweek, because if you see in the small print over the title, it has an important article about retirement funds. Has nothing to do with genetics but is very important. So you want get that one as well. So what is the analytical paradigm that we have faced? Well, let’s start by the essence of the weyo two-by-two tapes. So establish a relationship between exposure and an outcome. And this is a specific case based on old reports and there were many of them established in the relationship of oral contraceptives and thrombophilias or thrombotic acids or thrombotic diseases. And you see the relationship for the general population given a very low prevalence of the outcome, which is around two, three per 10,000 doubles when you take oral contraceptives and that’s it. We have been using those figures for decades and everybody knows that oral contraceptives pose a risk for--from embolic diseases.

Okay, what comes next? Well--sorry. How will this relationship change if we are able to plug in the genetic predisposition for those conditions? And it’s very easy; our two by two table will never be the same. It, at least, becomes a two by four table and you see that the same population gets stratified but the amazing change of the risk for very small populations who have the particular allele and have the exposure will determine a humongous risk and that most of that risk was masked by the vast majority of the population who was not exposed or exposed to just the environmental exposure without the gene marker.

So that is some of the immediate consequences. Those are some of the immediate consequences of being able to do these. What would happen if we would have a theoretical outcome related to an exposure where we can analyze just two genes—just three genes with two alleles? Genetics is much more complex especially for complex diseases as you will see. But what we will have is that our two by two table immediately turns into a complex table and that is the minimum one where you factor exposure, not exposure, the outcome, the three different alleles and all the combinations and you have to start with a mindset in at least 16 strata per exposure viral. What if you analyze compounders? So it gets very complex. It gets very complex if, on top of that, you have to plug in to the equation, the costs of alternatives when testing is available. How to measure it? How will the effect of the population will be after this testing or these virals can be introduced in our traditional models? How our compounding relationships going to be dealt with given that we have this information? How to quantify this risk given that the generic characteristics can be measured that will relate to what we will call the genetic risk factor? And what happens when there are more than one, two, three, or more genes related to a particular outcome in the presence of several environmental conditions that will play the same role that we have seen them play in our traditional virals and many more. So as you see, it really becomes a complex issue. That is the reason why we’re having a plenary in this because it really will affect the way we, the epidemiologists, conceptualize data analysis on our outcomes.

So in order to help us response to some of these questions, we will have four presentations, the first one, human genome epidemiology by Dr. Sandy Moore who is the Associate Director for Science at the National Office of Public Health Genomics for the CDC and she will talk about genomics and how will it affect clinical medicine and public health practice beyond the traditional domain of genetic diseases, major gaps that exist in translating gene discoveries into health benefits, and the efforts that must focus in expansion of population-based research on all topics. The second presentation by Dr. Sonja Rasmussen who is the Medical Officer for Disease in National Center of Birth Defects and that will focus on selecting sources for specimens and maximizing participation rates as any of epidemiological study participation rates and how to measure an increase is important that participation rate; human subjects related issues, and choosing genes for analysis and analytical approaches to be used. Dr. Scott Gross will talk to us about quantifying the health benefits of genetic testing in economic evaluations. And he is a Senior Health Economist in the Office of the Director of the National Center of Birth Defects at the CDC. And Scott will talk to us about economic evaluation and how it can show the value for money spent in these genetic testings. Economic evaluation requires estimation of health benefits and how good economic evaluations rely, of course, on sound epidemiology. And our last presentation is going to be about Dr. Siobhan Dolen on preterm birth and how genetics, the genomics, is shown in this very important problem in MCH, as a framework for approaching preterm birth as a common complex disorder. Dr. Siobhan Dolen is an assistant professor in the Department of Obstetrics and Gynecology and Women’s Health at the Albert Einstein College of the Medicine and she will discuss the risk factors and clinical approaches to preterm birth, outline genetics and genomic principles, introduce preterm birth in the new framework as a common complex disorder and propose a framework for a genomic approach to research in preterm birth. And without further delay, I want to invite you all to our first presentation. Thank you.