Dashboard of BMI by Self-reported Diabetes Diagnoisis

Background

The data for this dashboard comes from the National Health and Nutrition Examination Survey (NHANES) for the 2009-2010 and 2011-2012 sample years. NHANES is a continuous, cross-sectional survey that collects detailed health and nutrition information from approximately 5,000 individuals US residents. It is important to note that NHANES over samples some sub populations, like ethnic and racial minorities. This dashboard does not apply survey weights to counteract the oversampling, and thus is not a representative sample of the US. Additionally, since this data set is collected from two NHANES cycles, there are 10,000 observations used in this dashboard before any data cleaning.
The study population for this dashboard consists of U.S. adults aged 18 and older from the 2009-2010 and 2011 to 2012 NHANES cycles. The participants were selected using a stratified multistage probability sampling design. For this dashboard, the focus is on adults with different diabetes status (yes has diabetes, or no does not have diabetes), which is a self reported measure. After filtering out participants less than 18, the sample contains 7,481 observations.

Research Questions

Specifically, my research question is: How does Body Mass Index (BMI) vary between individuals with (either type 1 or type 2) and without diabetes for adults in the U.S. population? Additionally, does BMI differ among adults with diabetes (type 1 or type 2) across different BMI categories? The BMI categories are underweight (BMI < 18.5; blue), normal weight (BMI 18.5–24.9; green), overweight (BMI 25–29.9; yellow), obese (BMI 30–39.9;pink), and extreme obesity (BMI ≥ 40;purple).

Importance

This research is important because it provides preliminary insights into the relationship between BMI and diabetes status in a large, diverse population. Understanding how BMI varies across different diabetes statuses can inform strategies for diabetes prevention, management, and health promotion. Furthermore, examining BMI by category may uncover nuanced patterns between BMI classification and diabetes status, potentially revealing valuable information for healthcare interventions and public health programs.

Figure 1 shows the distribution of Body Mass Index (BMI) by self-reported BMI status. Both the “no diabetes” and “presence of diabetes” groups exhibit a slight rightward skew. The obesity categories are as follows: underweight (BMI < 18.5), normal weight (BMI 18.5–24.9), overweight (BMI 25–29.9), obese (BMI 30–39.9), and extreme obesity (BMI ≥ 40). Let’s filter out extreme obesity values, which will enhance visual clarity in later visualizations.

In Figure 2, we filtered out the extreme obesity category, which contributes to the skew, results in a distribution that is approximately normal. In Figure 1, the median BMI for those without diabetes is approximately 25, while for those with diabetes, the median is to be higher, around 30. These medians remain approximately similar in the Figure 2, even after excluding extreme obesity BMI values.
After confirming the data’s unimodal distribution, Figure 3 provides a clearer comparison of the summary statistics for the “no diabetes” and “presence of diabetes” groups by presenting a boxplot.

Now, let’s break down the graphs we found in Figure 2 and Figure 3 by the various BMI categories. What do you notice?

Discussion

Research Question 1

Figures 2 and 3 reveal that the mean BMI for the “no diabetes” group is 25.3, while the mean BMI for the “presence of diabetes” group is 29.81. This preliminary finding highlights a difference in BMI between individuals with (either type 1 or type 2) diabetes and those without, addressing the first research question: How does Body Mass Index (BMI) vary between individuals with (either type 1 or type 2) and without diabetes in a diverse sample from the U.S. population?

Research Question 2

The second research question—Does BMI differ among individuals with diabetes (type 1 or type 2) across different BMI categories?—is answered in Figure 4. In this graph, BMI and reported diabetes status are further stratified by BMI categories. Figure 4 shows preliminary evidence that individuals with diabetes have higher BMIs across all BMI classifications compared to those without diabetes.

Conclusion

The analysis presented in Figures 1 through 4 provides preliminary insights into how BMI varies between US adults with and without diabetes. Specifically, the findings from Figures 1-3 reveal that US adults with diabetes tend to have a higher mean BMI compared to those without diabetes, suggesting a potential link between higher BMI and the presence of diabetes. In this sample, this trend holds true across all included BMI categories, as seen in Figure 4, supporting the idea that elevated BMI may be associated with an increased likelihood of diabetes, regardless of the specific BMI classification.