Joining Forces to Tackle Obesity, 21-22 January 2002

 

Transcript : The Global Epidemic

 

Speaker

 

Dr William Dietz, Division of Nutrition and Physical Activity, Center for Diseases Control and Prevention, Atlanta, USA

 

ANDREW HILL: Good morning Ladies and Gentlemen. My name’s Andrew Hill. I’m chairing the rest of this morning’s session and this afternoon’s session. You’ve not come here to listen to me, in fact I’ve come here to listen to our next two speakers. You will see that the topic for the remainder of the morning session is ‘The Global Epidemic’, and we have two very prestigious speakers who have, I’d say, a unique view of this global perspective. Our first speaker is Dr Bill Dietz. Bill is Director of the Division of Nutrition and Physical Activity in the Center for Chronic Disease Prevention and Health Promotion in Atlanta. Thank you Bill.

 

WILLIAM DIETZ: Good morning, thank you Andrew. It really is a great pleasure to be here and I’m grateful and honoured that the National Audit Office would invite me to address you on the US experience, the challenge of addressing the environment. I’m particularly grateful to Rob Prideaux of the Audit Office, and Carol Lyons who have made my travel and stay here a very welcome experience. Mark Twain, the famous American writer, wrote a book called A Connecticut Yankee in King Arthur’s Court, and I feel a bit like that today, particularly following the prestigious speakers who opened this conference. But in the United States, as you will see, we have been struggling with this problem now for several years. I hope that our experience in recognising the contribution of the environment and beginning to take some steps to address it, will inform the process here which, in some ways, is more advanced than in the United States, particularly insofar as you’ve already recognised the contribution of community infrastructure and transport policy to physical activity.

 

I would like to start with a series of maps that show obesity trends among US adults. All of these graphs or maps have the same key – there are no data for states in white. States with a prevalence of less than 10% are light blue. This medium blue is a prevalence of 10 to 14%, dark blue is a prevalence of 15 to 20%, and greater than 20% is red. These maps are derived from an annual telephone state-based survey, a very sizeable sample of about 150,000 people per year. Obesity is defined as a body mass index greater than or equal to 30. [Change of slides to coincide with years] So these are data from 1990, 1991, 1992, 1993, 1994, 1995, 1996, 1997, 1998, 1999, 2000. These always show a very rapid progression of this epidemic and, if anything, these are underestimates because they rely on self-reported height and weight. As you know, people tend to overestimate their height and underestimate their weight. The one state which has stayed at a fairly low prevalence is the state of Colorado and there are reasonable questions that can be asked about why that is, and that may eventually lead us towards cause. But these maps, I think, have several lessons associated with them. The first lesson is that changes of this rapidity and magnitude are not caused by shifts in the genetic pool of the population. This is an epidemic that is clearly environmentally driven. The second is that these maps alone have constituted, in my view, one of the most effective communication tools about the prevalence of obesity and have transformed thinking about this as a cosmetic problem to recognising that it’s a very profound health problem. I took the liberty of taking the data that were in the NAO report that looked at the changes in the prevalence of obesity among English women, and compared them to US data derived from three or four major surveys, beginning here with the National Health Examination Survey and moving to the National Health and Nutrition Examination Surveys (NHANES). Beginning in 1980 the prevalence in both countries began to increase rapidly, and I think you can make the argument that the rate of increase in our two countries is parallel. The similarity of these trends raise another important and very interesting question that I think deserves exploration.

 

Another point to be made about the US epidemic is that the more severe forms of obesity are increasing more rapidly than milder forms. Here is Class 3 obesity in the NHANES between 1980 and 1994. There has been a rapid response in the percentage of American adults with a BMI greater than or equal to 40. In contrast, the prevalence of a BMI between 25 and 29.9 has remained relatively constant. So people are moving fairly rapidly through, a BMI of 25 to 29.9 to these more severe forms of obesity, like a BMI > 40, or in to Class 2 obesity (35 to 39.9).

 

The final point about changes in prevalence in the United States is that children are more severely affected than adults. Data from four national surveys show that, among 6 to 11 year olds, and 12 to 17 year olds, the greatest changes occurred between 1980 and 1994. Over this period, obesity in adolescents has trebled, whereas in children it’s doubled. So Obesity is a problem that is widespread within the population. All elements of the population are affected. There are very significant ethnic disparities in the prevalence of obesity. Obesity is increasing more rapidly among African-American females and Mexican-American females than it is among Caucasian females, in both children and adults.

 

The obesity epidemic is followed by another epidemic. These are the same BRSS data showing self-reported diabetes trends - beginning in 1990. [Change of slides to coincide with years] So these are diabetes data from 1990, 1991, 1992, 1993, 1994, 1995, 1996, 1997, 1998, 1999 and 2000. The epidemic of obesity is followed by an epidemic of Type 2 diabetes. The implications for health care costs are extraordinary. In the early 1990’s, among adults 65 years and older, diabetes already accounted for 1/4 of national healthcare costs in our Medicare programme. The lag that you see in the incidence of diabetes, between the onset of obesity and Type 2 diabetes, means that in a relatively short time we are going to be consumed by the cost of both diabetes and the other associated problems with obesity shown here. Principally we’re going to see the effect on cardiovascular disease. The prevalence of cardiovascular disease has declined steadily in the United States, but in the last several years it appears that that decline is plateauing. The implications of the obesity epidemic, followed by the diabetes epidemic, are that these diseases are going to begin to increase. Many of you may have seen our article in last week’s issue of the journal of the American Medical Association which looked at the prevalence of the clustering of these risk factors for cardiovascular disease, which are quite significant and substantial.

Now the other population which is affected are children and, although children rarely have the diseases associated with obesity, the prevalence of Type 2 diabetes is increasing rapidly and, in some locations in the United States, now accounts for half of all new cases of diabetes. I think you’ll hear more about all of these complications tomorrow, but the best indication of the concern that we have are paediatric hospital discharge rates, plotted here over about a 20 year period -- for diabetes, obesity, sleep apnoea and gall-bladder disease. These are underestimates, because in the United States, when one lists obesity as a discharge diagnosis, physicians may not be reimbursed for it. So in this study, we’ve simply looked at discharge rates for obesity-associated diagnoses, with either obesity as a primary or secondary diagnosis. I think you can see here that the frequency of diabetes discharges has increased by almost 60%. And, when one looks at sleep apnoea, gall bladder disease, and even primary obesity diagnoses, although the prevalence of these is quite low, between 0 and 0.5%, the prevalence of obesity and its associated diagnoses has doubled.

Therefore, not only are we facing the impact of these diseases in adults, but there is a wave of obesity moving through the paediatric population, which is going to be associated with many of these same consequences with very substantial implications for morbidity and, for the first time, mortality in young adults and adolescents. The implications for healthcare costs, are quite substantial, and you’ll hear more about healthcare costs from later speakers. The United States costs are similar to those in the UK and in the rest of the world, perhaps even a bit higher.

 

What accounts for this epidemic? I divide the factors related to obesity in to altered that which suggest that physical activity and inactivity captured by television viewing have independent effects on the prevalence of obesity. I’d like to take you through the changes that have occurred in each of these domains over the time period, coincident with the epidemic. Shifts in food practices in the United States have been quite substantial. Fast food consumption now accounts for over 40% of family’s expenditures on food. This is a notable shift over 20 years. When I was a child, which was more than 20 years ago --, going to a restaurant was an exception, something we did once a month as a family. Now it’s common for families to go to restaurants two or three times a weel. These choices are choices that are largely influenced by children, and we’ll come back to this because children in turn are influenced by television, which accounts for the relationship of television viewing to obesity. There’s been a reduction in the frequency of family meals, largely as a consequence of employment of the female workforce, and there are some data which suggest that families that eat meals together have healthier meals, meals that are lower in fat, higher in fruits and vegetables. Families that eat together have less fast food and less fried foods. Among girls, paradoxically, restrained eating and meal skipping have a positive impact on obesity. These are used as weight control measures but, in fact, have the opposite effect. Girls who are themselves restrained eaters or skip meals are more likely to develop obesity than not. Consumption of soft drinks has almost doubled over the period encompassed by the epidemic, and in the United States, soft drinks, which include both sodas and sugared beverages -- sugared juices -- now account for about 6% of the average child’s daily caloric intake. Variety has also increased, and if one thinks about the buffet line like that which exists in the restaurant for this hotel, people tend to over-consume when they’re offered more choices. Many of you know that one of the original models for obesity in rats was the so-called ‘cafeteria diet’, where rats were offered a variety of highly palatable food --and absent, a genetic predisposition nonetheless had rapid weight gain. Exposure to variety may in turn be a factor driving the epidemic.

 

Now it’s important to note that none of these food related practices have been linked causally to the epidemic of obesity. In part I think that reflects the types of analyses that we’ve engaged in which have focused on macro-nutrients, that is calories or percent calories from fat or total fat intake, rather than patterns of food consumption. We begin to look more and more at patterns of food consumption, I suspect we will eventually be able to link some of these food practices to development of obesity. For example, there’s been a recent study by an economist that argues that 90% of the shifts in obesity can be explained by the increase in the number of fast food restaurants, the increase in the number of regular service restaurants, and the increase in the price of cigarettes, which has a deterrent effect on smoking which therefore contributes to obesity.

 

These are economic models. From the medical point of view, I don’t think we yet have the evidence that says these are food behaviours on which we should focus to reduce the epidemic of obesity. The reason that many of these food practices are now occurring are complicated but at the heart of this problem is that these choices -- fast food choices, unhealthy choices -- are easy choices. The challenge for us as public health professionals is to make healthful choices easier choices, because in many neighbourhoods in the United States, fast food restaurants are the only restaurants that exist. In many neighbourhoods in the United States, access to fruits and vegetables is limited by cost and availability. There are not big supermarkets in many urban areas. So how we begin to reverse these trends.

 

Portion size may also be a factor driving the epidemic. If one reads Fast Food Nation, the book by Eric Schlosser about the fast food industry, he argues in there that it’s very profitable for companies to increase portion size because the difference between a regular portion and a larger portion, in terms of the cost of production, is quite small, which makes the profit margin quite substantial. So it’s in a company’s interest to increase portion size, although it may not be in the public health interest.

 

The second series of factors relate to decreased physical activity, and I think that’s best demonstrated by the contrast that one sees in modes of travel in the United States compared to Europe. Netherlands is at the top of the list, where 30% of all trips are by bicycle, about 20% are by walking, and 45% are by car. England is here, with 8% of trips by bicycle, 12% by walking, and here that 14% are by public transport. One of the figures this graph does not address is the impact that public transport has on physical activity, because there are reasonable data that indicate that people who use public transport actually are more physically active because you have to walk to the station to catch the train, you have to walk from the train or the underground to your job. Despite the fact that only 14% of people here rely on public transport, that may actually promote physical activity. In contrast, in the United States, only 1% of trips are by bicycle, about 10% are by walking, and 85% or so are by car. This reflects, I think, the community infrastructure of the United States because, in many cities, there has been a rapid growth of suburbs that lack centralised schools and lack centralised shopping facilities. The two statistics, that are most compelling for me is that only 1/3 of children who live within one mile of school, walk to school. When I speak to audiences in the United States, almost all of the adults in those audiences walked to school when they were children, but only about 10% of them have children or grandchildren who walk to school. The other statistic which I think reflects the design of our country is that, although 25% of all the trips we make are less than a mile, 75% of those trips are by car. We lack neighbourhood schools, we lack central shopping facilities and, as a result, we’ve become a car reliant society.

The final area of reduced physical activity has to do with the lack of physical education classes, and I was interested to hear the concerns in this country that have been expressed about the lack of physical education programmes. Physical Education classes also have declined substantially in the United States. In 1991, 42% of all schools offered daily physical education, and in 1999 that had declined to 29% of schools. This is a major concern because, increasingly, given the sprawl of the neighbourhoods or the danger in inner cities schools are one of the last safe places where children can be physically active. The elimination of physical education programmes, I think, undermines the opportunities for physical activity.

 

The final factor is increased inactivity. We published data from the late 1960s and compared these data to a survey done in 1990 on the frequency of television viewing. This slide shows youth ages 12 to 17 years, but the data is similar for youth aged 6 to 11 years. In 1967 to 1970 in the National Health examination survey, here indicated by the white bars, the median amount of television the children in the United States were watching about two hours per day. By 1990, the median had increased to 4.8 hours per day, and about 35% of children were watching five or more hours of television a day. Now this isn’t simply a consequence of impaired parenting or reluctant parenting, in many communities it reflects the lack of alternatives for children, after school particularly, or the perception of reduced neighbourhood safety, that makes it impossible for children to play freely out of doors. And I think it focuses us on the need to supply children with alternatives after school if we’re going to increase their levels of physical activity.

In these same surveys, we also showed a linear relationship of television viewing to the prevalence of obesity. This isn’t the severity of obesity, but the prevalence of obesity. Between the National Health Examination Survey and the National Longitudinal Survey of Youth, this relationship became more acute. There was a more virulent, if you will, effect of television viewing on the prevalence of obesity.

 

At the CDC, we’ve begun to think about how one constitutes a national nutrition and physical activity programme, in part to address the epidemic of obesity. However, it’s also been clear to us that the same strategies that impact obesity will also have a major impact on other chronic diseases, such as cardiovascular disease, cancer and diabetes. So, many of the comments that I’m going to make now about a national nutrition and physical activity programme address obesity as well as the primary prevention of other chronic diseases. A strong science base has to be at the heart of public health programmes, and that begins with surveillance studies.

 

Now, we have five major surveillance systems in the United States that have enabled us to begin to examine some of the factors related to obesity, most particularly the changes in prevalence of obesity, like the maps that I showed you from the Behavioural Risk Factor Surveillance System. I will not review any of these in detail, except to emphasise that there are several very critical gaps to surveillance, and I suspect they characterise the UK’s surveillance system as well. The first of these is that we lack information on a critical population, namely youths 6 to 14 years of age, who are not captured in our high school Youth Risk Behaviour System (YRBS), nor are they captured in the Paediatric Nutrition Surveillance System, which is low income children aged 0 to 5. So, because childhood is a critical period for the development of obesity, and because about 1/3 of all adult obesity begins in childhood, this gap is critical.

Secondly, these are all cross-sectional systems, and most of them lack the additional information on associated factors which may influence obesity, such as physical activity levels or diet. If we’re really going to establish targets for intervention, it’s absolutely crucial that we have more longitudinal data that enables us to develop causal linkages between behaviours and their outcomes. So surveillance is useful, but does not capture the essential information necessary to influence state based programmes.

The other issue around surveillance is that it doesn’t tell us how we should talk to the population about overweight or obesity. You heard this morning from Mr Leigh of the difficulty in the use of the term ‘obesity’. We don’t use the term ‘obesity’ when we’re referring to children or teenagers because the image is of a very severely overweight individual that has no counterpart in how people perceive their child’s weight. So surveillance systems are not going help us generate the information necessary to develop a dialogue on this topic. That’s going to require additional periodic surveys.

As I mentioned earlier, causal models that help us link these behaviours with obesity, such as the dietary factors that I showed you, are absolutely essential. The next feature of a strong science base are evidence based strategies. I think there are three strategies that we can initiate today in a very effective way, to begin to control this epidemic. These are the behaviour change strategies which we’re emphasising as part of the comprehensive nutrition and physical activity programme. Only vegetable and fruit consumption, has not yet been shown to reduce obesity, but I think we can make a reasonable argument that television viewing, physical activity and breastfeeding all have a reasonable body of evidence that allow us to implement these as obesity prevention strategies.

 

With respect to television viewing, there are now two clinical studies and two school based studies that demonstrate that reduced inactivity, or reduced television viewing, has an impact on obesity prevalence. This is a clinical study published by Epstein in 1995. Both groups here had identifical diet advice, identical involvement of parents and identical behaviour change strategies, but the group that was reinforced for reducing television had a much more profound weight loss than the group that was reinforced for increasing exercise. Following the intervention, rates of relapse or weight regain were comparable. Interestingly enough, there is now an accumulating body of evidence that suggests that the impact of reduced television viewing on activity. Television is probably not simply a form of inactivity, but also affects food intake. For example, in a study in California suggests that 25% of a child’s food intake occurs while they’re watching television. Therefore reductions in television time alone may reduce the opportunities for food intake.

The second strategy is increased physical activity. Interestingly enough, we don’t know the dose of physical activity necessary to prevent obesity. We know a lot about the impact of physical activity on obesity associated co-morbidities. We know a lot about the dose of physical activity necessary to prevent relapse of obesity following weight loss, but we don’t know about the dose of physical activity necessary to prevent obesity. We have recently been engaged in the development of a chapter for The Guide for Community Preventive Services. This is an evidence based assessment of strategies, in this case to promote physical activity. We have another chapter in progress for the guide that is examining effective obesity prevention strategies. These evidence based strategies represent the strongest approaches that we have. Point of decision prompts to promote physical activity, such as signage saying, "Use the stairs because it will help your heart or it will help your health", have a very positive impact on stair use. School based physical education programmes now have an evidence base that supports their usage as a strategy for physical activity. Enhanced access to areas or parks for physical activity increase physical activity. These are all strategies which we’re now beginning to incorporate into our surveillance systems, because if these are effective strategies, it’s important to know how well they’re being implemented and what impact their implementation has on the prevalence of obesity within populations.

 

The final area is breastfeeding. There are now three very large and well-powered studies that have been published, and two others that are now in press, that demonstrate that there’s an impact of ever breastfeeding as well as the duration of breastfeeding on the prevalence of obesity. Never breastfed children have about a 15 to 25% reduction in the prevalence of subsequent obesity in childhood, whereas children who are breastfed for six months or longer have about a 25 to 30% reduction in the prevalence of obesity.

From the programmatic point of view, it is critical to identify policy and environmental changes which help implement these policies. For example, if we’re interested in having children walk to school it’s critical that schools have a central location, but in many states in the United States there are now zoning regulations that make it impossible to locate schools on anything less than 32 acres of land. Well, the only place you’re going to find 32 acres of free land is on the periphery of communities, so Policies such as this make it difficult to incorporate physical activity in everyday life.

 

In closing, I’d like to draw an analogy with smoking, which I think is a recent example of public health success to which we may be able to look for some guidance. This slide shows the frequency of per capita cigarette consumption. Beginning in the 1900s there was a rapid acceleration of cigarette use that only began to plateau when the first medical reports linking smoking and lung cancer began, about the same time the spread of counter-advertising in the media led to the withdrawal of those television ads, followed by the Surgeon-General’s report that linked smoking to lung concern. But it wasn’t until the non-smokers rights campaigns began that cigarette smoking began to decline, and the rate of decline, at least up until the late 1990s, was not affected by increases in cigarette taxes. I think this example has several lessons for us. My belief is that, in terms of obesity, we’re where smoking was in the 1950’s. We’re increasing the visibility of the problem and beginning to redefine obesity as a medical problem, not a cosmetic problem. Hopefully, that recognition will begin to plateau rates of obesity. But I think the struggle for us is to identify the equivalent of the non-smokers rights campaign. I’m not sure that medical costs and the burden that all of society will bear for the increasing prevalence of obesity and it’s cost, will have the same impact as the non-smokers rights campaign, which began from exposure to a toxic air environment. Nonetheless, I’m encouraged by the attention that this problem has received in the United States, and the attention that it is now receiving in England, as evidenced by this conference and the National Audit Office report. I look forward to hearing from you the experience of the UK related to this epidemic. Thank you.

 

ANDREW HILL: Thanks very much Bill.