Diabetic Mapping

Diabetes is a chronic illness that affects the beta cells of the pancreas. There are two common types: Type 1 and Type 2.

Type one diabetes is usually found in children and young adults. It is usually diagnosed before the age of thirty. These are insulin dependent diabetics because their beta cells quit producing insulin. Type two diabetes is usually diagnosed later in life. This diabetic is typically overweight. Usually this type can be controlled by diet and oral medication. In this type of diabetes, the beta cells do produce insulin, but for some reason or another, the cells cannot absorb the insulin that is produced.

This project is designed to map the number of known diabetic cases through out the United States. I emphasize the term known because it is believed that there are many people that have diabetes and do not even know it. This project uses the total number of known cases from 1999 by state and divides that number by the total population by state for 1999. The information was found at the Center of Disease Control (CDC). I took this data and divided into three categories to compare the difference. The categories were equal interval (Figure 1), natural breaks (Figure 2), and quantiles (Figure 3). There is a small range of the data with 1.9% of the population of Alaska being the low end, and Mississippi being the high end with 4.1% of the population.

The first map design was equal interval. According to Slocum, "equal intervals are a method of data classification in which each class occupies an equal portion of the number line." (Slocum 1999) In Figure 1, the data is mapped in equal interval. However, the data is a little skewed because initially it put the data in the same class. I added a tenth of a percent to the last interval to prevent the same number of classes in the same legend. The legend is broken at 1.9% to 2.5%, 2.6% to 3.0%, 3.1% to 3.6%, and 3.7% to 4.1% (Figure 1). This map shows very well that the highest concentration of diabetes cases is found in the South, particularly in Louisiana, Mississippi, and Arkansas. This map is good because it does not create any gaps in the legend. It is also very easily done with a calculator. I believe that this map is balanced and that it is well designed. It is very easy to determine which states fall into which category. It contains most of the map elements except for scale. People who like to see where the heaviest and lowest concentrations of diabetic are located would prefer this map.

The second classification technique that I chose was natural breaks (Figure 2). According to Slocum, natural breaks are "a method of data classification in which a graphical plot of the data (such as a histogram) is examined to determine natural grouping of the data." (Slocum 1999) The data broke very closely to the same as equal area. The data broke at 1.9% to 2.4%, 2.5% to 2.9%, 3.0% to 3.4%, and 3.5% to 4.1% (Figure 2). There was no more than two tenths of a percent difference anywhere in the legend. However, given the close range of the data, it may be a huge difference after all. This map did have a larger upper class than equal area or quantiles. Its range was from 3.5% to 4.1%. This is a great map for anyone that would like to concentrate on knowing where the greatest breaks are in the data. There is an interesting distribution according to this map. It appears that the highest number of diabetes cases border the Mississippi river. There must be something in the water!

The last method of data classification chosen was quantiles (Figure 3). According to Slocum, "quantiles are a method of data classification in which an equal number of observations is placed in each class." (Slocum 1999) This map is a little skewed because Alaska has the lowest number of diabetic cases with 1.9%. This caused the lowest category to be one observation higher than it would be if the calculations where done by hand. The data range was 1.9% to 2.6% for class one, 2.7 to 3.1 for class two, 3.2 to 3.6 for class three, and 3.7 to 4.1 for class four (Figure 3). This map would be used if a percentile of the data were desired. For example, the top 25% of diabetic cases and where their data range is easily discerned using this map.

All three of these maps have shown an interesting phenomenon. Diabetes seems to be concentrated east of the Mississippi river, but it is also slowly working its way west. All three of the maps show that the majority of diabetic cases are centered in the South. This could be because of the rich southern diet that everyone down here is accustoms too. Even though the highest percentages of diabetics are found in the South, these maps prove that diabetes is an equal opportunity disease. This is evident by such a small difference from the highest range (4.1%) to the smallest range (1.9%). There is another interesting occurrence. In all three maps, Georgia has less diabetic cases than most all other states in their region. They must be doing something right in Georgia!

In all seriousness, diabetes is a disease that does not discriminate on where someone lives. These maps testify to that. Anyone can get it! I know because I have it! It is interesting to know which states have the most reported diabetes cases. As stated earlier, this data could be skewed because of the huge number of people who have diabetes and do not know it!

Works Cited

Slocum, Terry A. 1999.  Thematic Cartography and Visualization.  New Jersey: Prentice Hall. 

Figure 1

Figure 2

Figure 3