 All right, well, thank you. Thank you for coming here. So in a good way, this presentation picks up where the last one left off, at least a question at the end there, about quantifying if you do restrict food advertising, what sort of impact it can have. So my paper looks specifically of food advertising. And why food advertising? Well, if you look at the short term studies, the controlled studies, where they have people in laboratory settings, and they run ads or run fake ads, so to speak, it increases consumption of those advertised foods. So like Jennifer was saying at the end, the more foods advertise, parents are going to hear about it, and then there will be that desire to consume it. So when they do those short term studies, they find that obese children consume more of the advertised foods than normal weight. But the important point for this is that all end up over-consuming, end up consuming more of that target food that's being advertised than if it wasn't advertised. So the obese overweight consume more in those studies, slightly more, but the normal weight consume more too. And then the controlled studies, obviously, are the comparison. So with that then, you have a paradox when you look at the data on how much children are watching TV and their weight were BMI. So if you look at the latest NHA, NES data, the National Health and Nutrition Examination Survey, you'll find that there's no statistical difference between hours watched by overweight obese children or normal weight children. And I ran some regressions on it as well, and that effect never comes up. That just TV is linked to overweight when you look at the whole sample. So yeah, sort of a paradox here where you say no food advertising works in the short term. And also you could kind of assume it works that companies are pouring millions, millions of dollars into it, it probably is having some benefit to them as well. So if it doesn't have any, if overweight children aren't watching more TV, then how is that advertising having the impact? Where's the effect coming through? Okay, so that's what my study looks at here. So this last year I've looked a lot, research law about self-regulation of calorie intake and written about it. And a landmark study, at least I think, back in 2004 looked at when they gave access to fast food to lean and overweight adolescents and again, all over consumed, more than you just compared to regular food. But what they found though, when they found that they tracked them over a three day period is that the lean compensated for that and ended up eating less later in that day such that when you took a three day calorie average for the lean, they were normal. Even though on that one day they had the big fast food over consumption, they were able to regulate that consumption over time. Whereas the overweight over obese were not. They ate more on the fast food day, never ate less at dinner. It was a lunch study and then ate the same the next two days such that over a three day period, they had a calorie surplus. So from that you could say that on the next slide here, from that you could kind of assume that overweight obese children have apparently lost that ability to self-regulate calorie intake. That's the assumption I'm making. And then therefore, at least a kind of interesting hypothesis, that if food advertising increases consumption in everyone but then the overweight don't compensate for that and eat less of other foods or eat less foods later, then that's gonna drive overweight obesity to increase the more in that group that they watch television, right? Or see those advertisements, same thing, okay? So any questions here so far? You can jump in if you're, I'm used to being interrupted when I teach, so it's no problem. Some other evidence, just so you can make sure I didn't just make this up. A number of other studies looked at self-regulation of energy intake and this was one of the most interesting ones that I found is that they took a group of overweight obese adults and said, okay, we want you to eat so many fruits and vegetables per day, right? I mean, that's a great thing, eat more fruits and vegetables, you'll be healthier, you'll lose weight. What they found was that they did eat more of those fruits and vegetables, but they actually didn't lose weight, they gained weight because they just added those extra fruits and vegetables onto what they were already eating. So it actually had a negative impact. It's one of the very few studies that's ever shown a negative impact from fruit and vegetable consumption because in this situation, they never said eat less of the other things and more fruits and vegetables, they just said eat another four servings of fruits and vegetables to be healthy. It didn't work, it actually made them worse off because again, no compensation for that extra food, they added it to what they were already eating. So I thought that was a very pivotal study. So then if you take those pieces of information, the theory then would be that the more, if you just censor your sample to only the overweight obese and within that sample only, when you allow for individual variation, the more that they watch TV, aka see food advertisement, the more overweight they should be. So what I'm saying is that food advertising should only have an effect in the subset of people who have lost the ability to self-regulate calorie intake because now it's just gonna tell you to eat more and then you're not gonna eat less, you're just gonna eat more on top of what you normally eat. All right, so then there's a side piece here since this is a paleo conference here, I wanted to talk about does food advertising really serve any biological function? In other words, if we're gonna ban it, then someone might say, well, is it serving any primary function? Not that I can tell, all right? Obviously people did find before food advertisement, and I say again, how do people regulate food intake they seem to do just okay without it. So it'd be hard to argue that food advertisement has any sort of really valuable or biological function to it that is so valuable that if you ban it, the ceiling's gonna cave in, right? It's obviously profitable for companies, we know that, but we're saying from a biological, physical perspective, does it do anything? No, and again, you can see how does the channels through works through, a lot of those ads simulate people eating. They've done other studies where they have, again, people who aren't self-regulating their calorie intake, and if they get them to eat with other people, and then they kind of do those stealth experiments where those other people eat more or eat less, then again, it impacts them to eat more or eat less if this kind of person they placed in there is eating a lot or eating a little. The lean, again, they do it a little bit, but then they end up compensating for that. So in a biological sense, if you think about hunter-gatherers in a tribe, if someone's eating more, if you're not hungry, that's not gonna affect you, but if you've lost that ability to regulate calorie intake, it will, it will get you to eat more. So that's what food advertising is, right? Showing the food, showing people eating it, enjoying it, right, it looks great tasting, et cetera, trying to get everyone to eat more of those foods. So my hypothesis specifically, right, is that higher levels of television viewing, which is a proxy for food advertising, is gonna have a positive relationship with Body Mass Index, but again, only in overweight slash obese children. And then a question as to what ads are they seeing, it's pretty obvious if you look at the study and the Yale Rudd Center, Jennifer, they have a lot of data on the amount of advertising points going to children and the types of ads and all that. So you can just see one study I put up, 2008, 49% of advertisements during Saturday morning basically cartoons were about food and actually of those 49%, 91% were for junk food, processed food, so to speak, high in fat, sodium, sugar, whatever, basically processed food. So the majority of those ads are going to process food and as I talked about my abstract, that is predictable from economic theory. A very, very small amount of commercials is directed towards fruits and vegetables. The large majority, over 90% in general in any sample is processed food. So a side question here that might come up, couldn't food advertising change a normal weight child? Into an overweight child over time. I say yes, certainly could, but that's not what I'm studying here. What I'm looking at is only a cross section in that NHS data and saying, okay, at that point in time, are the children watching more TV, are they, do they have a higher body mass index? So certainly this other question is viable, but it would take longitudinal data, a longitudinal study to try to figure that one out. All right, on to the data. So I got the latest data or at least the latest that had all the variables filled in that I needed. So 2009, 2010 data from the CDC. Children's age two to 11. And then what's neat about this is it's individual level data. It's an individual level survey. So they have an ID number basically and then they have a bunch of characteristics for that child and then also their examination data if they've submitted for blood pressure and cardiovascular markers and everything else down the line. So it's pretty comprehensive data set on the individuals in that survey. So again, my sample is only overweight, obese, not the normal weight. And then I group them together and some studies they differentiate and say this much in overweight and this much in obese. I just put them all in one category. I figure the goal is to get people to normal weight. So that's what the public policy really should be focusing on. So the dependent variable, if any of you guys run the regressions in your spare time or a professional time, I guess, is BMI. And then the variation in BMI, the independent variables, you have hours of TV viewing per day. And then physical activity, obviously you want to control for that. So this is the days where a child is at least physically active for 60 minutes. And then the control variables in there, the gender, the race, the income, obviously income is a huge factor. As somebody said, children don't have a lot of disposable income. So that's gonna play a role. And also obese overweight is higher in the lower income group in neighborhoods and that type of situation. There are some other variables I would have liked to have, but there was missing data in the NHS. So I went with what was available. Okay, so the regression results. Hopefully you can see this here. So I try to make it as big as possible. So you get the R squared up top, basically tells you how much of the variation in the dependent variable I'm explaining. It's a little low if you do this sort of stuff at 0.36, but only because I used BMI. I actually ran this with weight as well and the R squared was much, much higher, around 85% or something like that. So that's really more of artifact than just the data and the way it was constructed. But the key thing here is at the bottom. So you have the coefficients. Physical activity is a negative. So that's what you think, right? The more you exercise, the less your BMI is gonna be. TV video is the one I was focused on and then you can see that's positive. So what that variable is saying is for one more hour that a overweight obese child watches television on average, then their BMI should increase by 0.31. So you get some children that watch more, you know, four, five, six, seven hours of TV a day than that's a substantial shift in the BMI. The gender was not statistically significant. Age obviously was, race. Only really African-American was statistically significant and a pretty huge coefficient there. Just in this last 2009, 2010 data, there was a huge jump in African-American youth overweight obesity. Not really sure why, but especially among girls, African-American girls that really skyrocketed. So that's kind of a story for another day. And then that last variable is income-poverty ratio and that's negative and that's what you'd think from the previous things that the closer you are to poverty, right, then, or excuse me, the farther away you are from poverty, right, the BMI is gonna go down. Okay, I think of it as probably the easiest way to think of it. All right, any questions on the results here or anything? All right, so then the prediction from that is that if you ban food advertising, again, to adolescents, I'm not gonna get into the adults and the legal issues. Then watching television should have the same effect as any other sedentary activity in children. In other words, if you take the food out out so you negate its power and it's the same as just, you know, watching a movie or maybe, you know, reading a book or something like that. And actually sedentary activities have kind of a mixed finding in the research literature about are they really driving obesity overweight or not? A lot of studies find that they aren't having any effect. In fact, if you read, there was a New York Times article, maybe a month, two months ago, that talked about how actually reading among adults is related to lower weight. So you're just sitting there reading but they weren't sure if that was a marker of other things, other family variables or things like that. But at any rate, what you do is you, you know, if you negate the food ads, then TV is just another sedentary activity that shouldn't be a big driver of overweight obesity. Because again, we're controlling for physical activity. So what I did then was I reversed the effect of the TV. So I took that coefficient that I got from the regression and then I applied it at an individual level. So if someone was watching seven hours of TV per day, then their BMI was rolled back seven times .31. If they're only watching two hours of TV a day, then seven times .31, or sorry, two times .31 to roll it back. So I did this to get really specific accurate estimates rather than just put them all in one pool and say, here's the average four hours of TV and here's the coefficient and multiplied. That's gonna give you a much less accurate estimate than I could get with this individual survey level data. So it's pretty cool, I think, that you can do it on an individual level and get each person's predicted new BMI if all those food ads are taken out. And we'll see in the end here, I think it leads to better estimates than in the past. So I reversed that impact, recalculated BMI for all the subjects in the sample and then found that in the overall sample, if you negate that food advertising in the TV, then it would shift 5.8% of individuals' children from the overweight, obese category to the normal weight category. And then I broke it out by gender here as well, 4.4 and 6.2. So this is the change in total percent of overweight, obesity in children, which is roughly in this 2009 data, like 33%. So it would take it from 33 to 28 or 27 and change something like that. So it's a fairly big shift. As Jennifer said, it's not gonna completely eliminate it, obviously, because the foods are still there and everything, but it is a big chunk of the reason for that big increase in overweight obesity. So to compare the model I ran versus some other studies, Bolton is an older study that predicted the impact of food advertising and then its estimates on young boys and girls was minus 3.9 and minus 4.0. The Delphi was a survey of experts in their estimates as to if you banned it, it was like 33 or so experts from a couple years back as to what they thought if you banned food advertising, what would you get? You know, I'm not sure how technical their estimates were. Maybe they just got a phone call and made up a number or something, I don't know. I'm sure some of them were highly technical, but I'm not sure about the other ones. It seems they were getting kind of high compared to what I would think even without looking. So then my model shows minus 4.4, minus 6.2, a little higher than that Bolton case, but a little lower than that expert panel. And the conclusion then is that food advertising does negatively impact children who are overweight or obese. That's what that regression showed is it was a statistically significant coefficient on that TV video variable between that and BMI. And then also you could see the higher levels of exposure to that was then positively related to the BMI such that if you banned all advertising to children, it would reduce the overall prevalence of overweight obesity by 5.8%. All right, so a questionnaire, yeah. Yes, I totally understand your question, but not the way I set it up in my opinion, only because that original regression that got that coefficient was hours of TV and BMI. So yes, I mean, it could be the case in that hour TV, only 20 minutes or ads. And that tends to be relatively standard. So whatever that one hour of TV was kind of including implicitly the ads, then that's what led to the BMI shift. So it's kind of like I'm assuming that hour of TV that caused it and then you negate it is the same. So in other words, the 20 minutes adds in, the 20 minutes adds out, out of the hour. If you put the hour in and the hour out, it's the same thing. Again, for your point to be accurate, you'd have to show that the number of ads per TV hour was variable over time or over people or something like that, but generally it's the same. Yes, no. No, because what I'm saying is that the TV containing the ads is what's driving the increase in BMI. If you just say, surf the internet for an hour, that shouldn't have any impact on the BMI. There's nothing inflammatory, so to speak, in there. So it is because it's in the TV that the TV is driving the over consumption. But I definitely understand what you're saying. Oh, perhaps I missed it. Did you have a control group with people watching TV and no ads? Is that how you're determining that it was ads that had the effect? No, what I said was the way back here at the beginning, that was at the beginning? Yeah, I ran regressions separately on this, that normal weight and overweight obese children actually watch similar amounts of TV. So it's not that lean people are just not watching TV at all. They're watching TV, but it's having apparently no impact on them, because they're remaining lean. At least it's not impacting their weight. So if you have, roughly it was like boys was like 2.2 hours. So both normal weight boys and overweight boys are watching the same amount of TV. So from that you can't say that TV's doing anything just based on that. So I censored the sample to focus only on overweight obese. Thank you. All right. Okay, thank you. Appreciate it.