Lab report 1 - How does blood glucose level differ between sexes and by age

Introduction

The Framingham heart study began in 1948, observing 5209 individuals over 24 years. The data collected provides insights into a range of variables including heart diseases, demographic information and blood testing information (e.g. cholesterol and glucose). Glucose levels in the blood is a key indicator in diabetes diagnoses as the condition is characterised by an inability to reduce glucose levels with insulin.

Understanding the differences in demographics’ glucose levels assists professionals in making informed diagnoses, specific to a patients expected glucose levels. Using the Framingham data, glucose levels can be compared between patients to infer the general range of glucose by age and sex, after removing the readings of diabetic patients from the data set to avoid skewing the results and ensuring that means are relative to individuals without glucose related diseases.

Results

Figure 1

ggplot(heartData_new, aes(x=GLUCOSE, y=AGE)) + geom_hex() + scale_fill_viridis_c(option = "plasma") + labs(title = "Comparison of Blood Glucose and Age", x = "Blood Glucose (mg/dL)", y = "Age")
## Warning: Removed 1353 rows containing non-finite outside the scale range
## (`stat_binhex()`).

The heatmap in figure 1 plots the frequency of blood glucose readings against the age of the participant. A general increase in blood glucose with age is shown, aligning with the current literature concerning age and blood glucose levels. The majority of results fall between 70 - 100 mg/dL, however, the range increases with age. These results suggest blood glucose levels are more variable, and higher, in older individuals despite the mean level remaining relatively steady up to 73 years of age.

Figure 2.

ggplot(heartData_new, aes(y=GLUCOSE, x=sexF, fill = sexF)) + geom_boxplot(outlier.shape = NA) + theme_minimal() + ylim(40,150) + labs(title = "Comparison of blood glucose levels between sexes", x = "Sex", y = "Blood Glucose (mg/dL)")
## Warning: Removed 1378 rows containing non-finite outside the scale range
## (`stat_boxplot()`).

The boxplot in figure 2 compares the mean and intequartile range of blood glucose levels between men and women. The results are insignificantly different between men and women, suggesting levels do not vary greatly by sex. Despite this, range of results were broader in men despite the mean being slightly higher in women, suggesting there may be greater variability in men’s blood glucose levels.

Conclusion

The results suggest that blood glucose levels can differ between demographics, particularly age. From figure 1, it can be observed that blood glucose levels gradually increase with age. Furthermore, a greater variation in blood glucose can be seen as age increases, suggesting that over time the bodies glucose homeostatic mechanisms may worsen, resulting in greater ranges of blood glucose readings observed. The results also indicated little difference between male and female blood glucose readings.