Interning at the FACTS: Research into the Effect of Obesity and Diabetes on DISH
During the summer of 2015 I completed a research based internship at the Forensic Anthropology Center at Texas State (FACTS). My research consisted of a project assessing the correlation between a condition called diffuse idiopathic skeletal hyperostosis (DISH), obesity, and diabetes. FACTS facilitates research like mine by providing a varied array of researchers the tools and resources needed to carry out any number of projects. These tools include a CT scanner, a donated skeletal collection, and a 3D printer not to mention a vast outdoor decomposition facility. Many of these resources I was able to use directly in my research.
For my internship, Megan Veltri (a recent graduate of Texas State University, and FACTS assistant) and I used the Texas State donated skeletal collection housed at the Grady Early Forensic Anthropology Research Lab (GEFARL) to assess the long standing assumption that DISH is caused by obesity and possibly diabetes. To accomplish this, we scored individuals in the skeletal collection for severity of DISH and compared this score to body mass index (BMI) scores and diabetes prevalence.
Based on results from these comparisons, it appears as though DISH is likely not caused by obesity or diabetes. While 55.8% of the overall collection were obese, only 45% of individuals with DISH were obese. Also, only 48.3% of individuals who were obese had DISH. We intend to run correlations but, due to time constraints, such data will be presented at the AAPA. A detailed spread of this data may be seen in Table 1. We will also present data on the progression of DISH in individuals who are obese and diabetic at the AAPA meetings in the spring of 2016.
DISH is, in short, a skeletal anomaly exhibited in the spinal column, expressed by the growth of bony spurs in a flowy, candlewax-like appearance predominantly along the right anterior portion of the lower thoracic vertebrae, though it can progress up or down into the cervical and lumbar vertebrae (Riddle, 2002). The candle-wax appearance may be seen in Figure 1. DISH is actually a fairly common condition, affecting 6-12% of the population and up to 25% of males, and 15% of females over the age of 50 –the vast majority of the donations in the Texas State Skeletal collection are male and over 50, making this collection perfect for this research project (Weinfeld, 1997).
Although DISH causes the fusion of vertebrae, many people with DISH have no inclination that they are affected. This is unlike other spondyloarthropathies individuals with DISH often have no symptoms other than stiffness of the back which is often written off as a symptom of old age. In the most severe cases, DISH may cause sleep apnea, difficulty swallowing, and pinched nerves, all of which may require surgery to remove bone spurs (Mader, 2009).
Because there is very little known about what causes DISH and because DISH is such a complex condition, there were two main questions we needed to ask. First, can it be shown that obesity and diabetes are major causes of DISH? Second, do obesity and/or diabetes play a role in the progression (worsening) of DISH?
There were a number of tasks that proved difficult during our data collection, and although DISH has been described in literature for decades, there is very little information available describing standardized methods for measurement of DISH on dry bone. This meant that many of our methods were developed as we went along, and we had to start over many times after deciding that there was a better way to do things than how we had been doing them. I will therefore describe our methodology below.
To assess whether or not DISH could be caused by obesity or diabetes, I created an Excel spreadsheet showing body mass index (BMI), diabetes diagnosis, and DISH diagnosis for the entire donated collection available at GEFARL (sample size 120). This consisted of calculating BMI for each individual donor using the height and weight information found in their files. We then compared these three factors to assess whether or not they appeared to be linked. This data may be seen in Table1 later in this report.
In order to analyze the progression of DISH and assess the possibility if it worsening with obesity and diabetes, a scoring system was put in place that allowed us to statistically analyze these three factors and their effects on each other.
There is little agreement among scholars as to what exact criteria should be used to diagnose DISH. Riddles’ system did lay forth a method for diagnosing dish, however we felt it was lacking. We decided, therefore to create our own diagnostic characteristics using the available literature, those criteria being: 1. The presence of continuous ossification along the right anterior portion of thoracic vertebrae, spanning at least 3 vertebrae; 2. Retention of disc space between affected vertebrae; and 3. The absence of confounding pathologies or trauma. (Mata, 1998; Resnick, 1975).
After spending a considerable amount of time searching for an existing scoring system, we decided to edit one produced by Jennifer Riddle. An example of our edited scoring sheet may be seen in Figure 6. To score an individual with DISH using this system, maximum length, width, and thickness measurements must be taken from the site of ossification using sliding calipers. The length, width and thickness measurements are then multiplied together to give a total volume score of the amount of DISH exhibited spinally in any given individual. These scores may then be compared and used to run statistics.
In many individuals, DISH occurred in more than one portion of the spinal column. We therefore measured these separate ossifications in ‘blocks’. A block may be described as a segment of ossification that has an obvious starting and ending point. All blocks are scored individually and then added together. (Note: It is possible for there to be ossification falling in more than one block on one individual vertebrae.) Figure 2 shows how we broke up an individual with three blocks.
In some instances, the length of the ossification was greater than we could measure using calipers. Because of this, it was necessary to break up the length into segments and add them together. This practice adds the opportunity for error, so we developed a method for measuring longer ossifications. If the ossification spans more than 6 contiguous vertebrae, we measured in five vertebrae increments. So if the ossification spans from C-3 to L-5, we took five measurements and added together as seen in Figure 3.
In the event that vertebrae showed obvious ossification that had been broken post-mortem, we used rubber bands to hold the vertebrae together for ease of measurement. We would then measure across the broken area using the method described above to get overall length. This can also be seen in Figure 3.
At times, the width of two parallel blocks overlapped. In these instances, we drew in a line as close to the middle as possible so that the same segment of bone was not measured twice. This line is arbitrary, but necessary as not to overestimate DISH scores. An example of this may be seen in Figure 4.
Because thickness was such a hard measurement to take on hard bone, we decided that radiographs would be a more accurate way to extract these measurements from the bone. Luckily just such technology is readily available at GEFARL. To run radiographs on the fused spinal columns, we mounted the segments to florists foam (used because it only minimally affects the radiograph imaging, and can easily be differentiated from bone on a radiograph) alongside a ruled scale to ensure accuracy. Once complete, we could easily assess width and use it for scoring. It is important to note that when taking the radiographs, the scale must be placed in plane with the area you wish to measure so that there is not any distortion. This meant for us that the scale had to be placed and moved multiple times per scan to ensure accurate placement. Figure 5 shows an example of a radiograph taken for thickness.
Although researchers have assumed for years that DISH is caused by obesity and diabetes, the results of my research do not appear to verify this assumption at this time, at least for the donated collection here at Texas State University. If anything, there appears to be a slight negative correlation between obesity and DISH as well as diabetes and DISH. Although DISH is probably not caused by obesity or diabetes, it could absolutely be made worse by them, which is something that I am still working on determining, the results of which I hope to present at the AAPA meetings next spring.
Mader, Reuven. Et al.
2009 Extraspinal Manifestations of Diffuse Idiopathic Skeletal Hyperostosis. Rheumatology
Mata, Sonia. Et al.
1998 Comprehensive Radiographic Evaluation of Diffuse Idiopathic Skeletal Hyperostosis:
Development and Interater Reliability of a Scoring System. Seminars in Arthritis and
Resnick, Donald. Et al.
1975 Diffuse Idiopathic Skeletal Hyperostosis (DISH): Forestier’s Disease with Extraspinal
Manifestations. Radiology 115:513-524
Riddle, Jennifer E.
2002 Using Diffuse Idiopathic Skeletal Hyperostosis to Determine Age at Death. M.A. Thesis,
Department of Anthropology, University of Nevada, Las Vegas.
Weinfeld, Robert M. Et al.
1997 The prevalence of diffuse idiopathic skeletal hyperostosis (DISH) in two large American
Midwest metropolitan hospital populations. Skeletal Radiology 26(4):222-225