Which Came First: The Obesity Or The Related Disease?
A novel study looking at childhood obesity published last week by researchers at UCLA in the journal Academic Pediatrics garnered surprisingly few headlines.
It isn’t new that childhood obesity in the U.S. warrants concern and action or that poor children and children of color have significantly elevated obesity rates. What makes this study unique is that it is the first to examine and discover that obesity parallels a long list of other health conditions in children. Up until now, research has focused on the health problems obese children face later in life as they become adults.
UCLA researchers determined that overall when obese children were compared to normal weight children they “reported poorer health; more disability; a greater tendency toward emotional and behavioral problems; higher rates of grade repetition, missed school days and other school problems; ADHD; conduct disorder; depression; learning disabilities; developmental delays; bone, joint and muscle problems; asthma; allergies; headaches; and ear infections,” according to a press release.
In that same release lead researcher Neal Halfon, MD, MPH, a professor of pediatrics, public health and public policy at UCLA, admitted his own shock at the results: "This study paints a comprehensive picture of childhood obesity, and we were surprised to see just how many conditions were associated with childhood obesity.”
You might have picked up on Dr. Halfon’s use of the word “associated” and not “caused by” when referring to the long list of conditions found in tandem with obesity in the study. There’s good reason for that and "associated" is a key word that should draw the attention of anyone reporting on health.
First, let’s look at what the type of study UCLA conducted can and cannot tell us: The researchers used data from the 2007 National Survey of Children’s Health (NSCH). It boasts a large enough sample size of children (over 43,00) ages 10 to 17 all across the country to enable researchers to safely generalize the results to all U.S. children in that age range.
The survey, conducted by phone, relies on parents reporting the true height and weight of their children. A more accurate, and far more costly way, of getting the children’s measurements would be in a clinical setting.
Because NSCH gives a cross-sectional look at that age group, it is only a snapshot in time. Therefore, it cannot tell us what came first – the obesity or the other conditions the researchers found associated with it.
None of this is meant to knock the NSCH or the UCLA research, but to set the stage for the findings. All of these limitations are duly noted by the researchers in their paper. (For more information about the pluses and minuses of surveys, check out William Heisel’s series in Reporting on Health, Know Your Source, from December 2012.)
Correlation vs. Causality: While this study can show an association, it cannot prove any sort of causality, because it only measures the prevalence of childhood obesity and other conditions or problems (aka comorbidity) at one point in time. This is how the researchers explain it in the study:
The cross-sectional nature of the data limits our ability to determine whether obesity is causing the comorbid condition (ie, a complication), whether the comorbid condition is responsible for obesity, or if both are related to some unmeasured third factor.
So, the three possibilities for why some health problems are more likely to exist in tandem with obesity: 1) obesity contributes to multiple health problems; 2) the multiple health problems contribute to the obesity; 3) both the illnesses and the obesity have a common underlying cause.
Assumptions of causality would have to be made based on already established links by a different sort of study. One example is diabetes -- obesity increases the risk of contracting it.
But most of the associations are not so clean cut. The study gives two examples:
… [S]ome evidence suggests the relationship between obesity and ADHD may be due to the experience of toxic stress in the early years, resulting in alterations in executive function that result in poor impulse control as well as leptin insensitivity, which can contribute to weight gain. The relationship between obesity and the development of asthma is similarly complex, with causal arrows pointing in both directions.
The need for further research: By demonstrating a correlation between childhood obesity and a whole host of other conditions, this UCLA study can guide further research. As the researchers explain in their paper, additional investigations should confirm that the associations they found exist and examine which one came first.
One of the most ambitious examples of a study that has helped answer the “what came first” question for cardiovascular disease is the Framingham Heart Study. It has shed much light on the prevention, causes and treatment of cardiovascular issues. But even something on a smaller scale examining obesity –- what causes it and what it causes -- could help us better understand and tackle this most vexing problem.
Image by Mark Turnauckas via Flickr