AMED3002 Lab Report 1: Does BMI increase the Presence of Hypetension in Patients?

0.1 SID:520513990

1 Introduction

The Framing Heart Study was conducted in Framingham, Massachusetts in 1948, where it delved into multiple risk factors of cardiovascular disease that included BMI (body mass index), hypertension, blood pressure, lung function, smoking habits and medication use. Initially had 5,209 participants that were observed in 3 different time periods, approxiamately 6 years apart. Hypertension being a established risk factor for cardiovascular disease, researchers have studied the connections with BMI and other associated risk factors. Through analyzing the relationship between BMI and hypertension data across different periods, we aim to give insights possible health implications and management strategies to minimise the effect of hypertension and associated cardiovascular disease.

2 Results

options(warn = -1) 
library(ggplot2)

heart <- read.csv("https://wimr-genomics.vip.sydney.edu.au/AMED3002/data/frmgham.csv")

heart_with_hyp <- subset(heart, PREVHYP == 1)
heart_without_hyp <- subset(heart, PREVHYP == 0)

ggplot() +
  geom_boxplot(data = heart_with_hyp, aes(x = "Hypertensive", y = BMI, fill = "Hypertensive")) +
  geom_boxplot(data = heart_without_hyp, aes(x = "Non-hypertensive", y = BMI, fill = "Non-hypertensive")) +
  geom_point(data = heart, aes(x = factor(PREVHYP), y = BMI), position = position_jitter(width = 0.1)) +
  labs(x = "Hypertension Status", y = "BMI") +
  theme_minimal() +
  scale_fill_manual(values = c("Hypertensive" = "#f7c579", "Non-hypertensive" = "lightgreen")) +
  scale_x_discrete(limits = c("Hypertensive", "Non-hypertensive")) +
  theme(
    axis.line = element_line(color = "black"),
    axis.ticks = element_line(color = "black")
  ) +
  guides(fill = FALSE) +
  geom_hline(yintercept = c(18.5, 24.9), linetype = "dashed", color = "red")

Figure 1: Distribution of BMI across patients with and without hypertension where inbetween the red lines illustrate the range that is considered healthy BMI (2022).

#Summary of patient's BMI with hypertension
summary(heart_with_hyp$BMI)

##    Min. 1st Qu.  Median    Mean 3rd Qu.    Max.    NA's 
##   14.53   24.16   26.55   27.06   29.36   56.80      31

#Summary of patient's BMI without hypertension
summary(heart_without_hyp$BMI)

##    Min. 1st Qu.  Median    Mean 3rd Qu.    Max.    NA's 
##   14.43   22.49   24.60   24.87   26.95   44.27      21

The distribution between BMI and hypertension status shows that there is a difference between the BMI of patients with and without hypertension where patients with a higher BMI (minimum= 14.5, mean= 27.1, maximum= 56.8) than patients with lower BMI (minimum= 22.5, mean= 24.9, maximum= 44.3). The average BMI of patients with hypertension fell into the overweight range (18.5-24.9) patients with hypertension had an average BMI just below the upper bound of the overweight range (18.5-24.9) (2022). Furthermore, there is overlap in boxplots which could express that there is no significant difference between BMI of hypertensive and the non-hypertensive groups.

In figure 1 there is a high number of outliers in both hypertensive and non-hypertensive groups meaning that there is a lot of variability in the data suggesting that there might be a confounding factor or BMI and hypertension have little to no influence on eachother. These outliers skew the means, in the hypertensive group, the higher BMI outliers bring the average up, making it seem like the hypertensive group on average have a higher BMI.

Figure 1 also shows that the patients are mostly in the overweight range as the means of both groups are higher and just under the healthy weight bound.

options(warn = -1) 

hyp_colors <- c("Hypertensive" = "#f7c579", "Non-hypertensive" = "lightgreen")
heart_with_hyp <- subset(heart, PREVHYP == 1)
heart_without_hyp <- subset(heart, PREVHYP == 0)
period_labels <- c("Baseline", "6 years later", "12 years later")

# Boxplot, over 3 periods for hypertensive and non hypertensive
ggplot() +
  geom_boxplot(data = heart_with_hyp, aes(x = factor(PERIOD, labels = period_labels), y = BMI, fill = "Hypertensive")) +
  geom_boxplot(data = heart_without_hyp, aes(x = factor(PERIOD, labels = period_labels), y = BMI, fill = "Non-hypertensive")) +
  labs(x = "Periods when Patients were Examined", y = "BMI", fill = "Hypertension Status") +
  facet_wrap(~PREVHYP) +
  theme_minimal() +
  scale_fill_manual(values = hyp_colors) +
  theme(
    axis.line = element_line(color = "black"),
    axis.ticks = element_line(color = "black")
  ) +
  guides(fill = guide_legend(title = "Hypertension Status")) +
  geom_hline(yintercept = c(18.5, 24.9), linetype = "dashed", color = "red")

Figure 2: Distribution of BMI across patients with and without hypertension at different period of time where inbetween the red lines illustrate the range that is considered healthy BMI(2022).

Figure 2 depicts the distribution of patient BMI across different periods of time, being from the baseline to 12 years after the baseline was conducted. One notable trend is a slight decrease across the 12 year span in both non-hypertensive and hypertensive groups. For the non hypertensive group the mean BMI decreased from 25 at baseline to 24.8 6 years later and 24.7 12 years later. Similarly, the hypertensive group decreased from baseline of 27.6 to 27 after 6 years and 26.7 12 years later. This trend could suggest decrease in BMI and patients age. Nevertheless, the hypertensive group had higher BMIs in comparison to non-hypertensive group across all examination periods.

options(warn = -1) 

#Violin plot 
ggplot() +
  geom_violin(data = heart_with_hyp, aes(x = factor(PERIOD, labels = period_labels), y = BMI, fill = "Hypertensive")) +
  geom_violin(data = heart_without_hyp, aes(x = factor(PERIOD, labels = period_labels), y = BMI, fill = "Non-hypertensive")) +
  labs(x = "Periods when Patients were Examined", y = "BMI", fill = "Hypertension Status") +
  facet_wrap(~PREVHYP) +
  theme_minimal() +
  scale_fill_manual(values = hyp_colors) +
  theme(
    axis.line = element_line(color = "black"),
    axis.ticks = element_line(color = "black")
  ) +
  guides(fill = guide_legend(title = "Hypertension Status")) +
  geom_hline(yintercept = c(18.5, 24.9), linetype = "dashed", color = "red")

Figure 3: Distribution of BMI across patients with and without hypertension at different period of time where inbetween the red lines illustrate the range that is considered healthy BMI (2022).

Figure 3 illustrates the distribution more visually than Figure 2. It can be seen that there is a consistent diamond shape across hypertensive and non-hypertensive groups indicating that there is even distribution of data on both sides of the mean. The hypertensive group has a higher tail end compared to the non-hypertensive, this is due to the outliers shown in Figure 2.

3 Conclusion

According to the results, patients with hypertension have higher BMI in comparison to patients that are non-hypertensive. The graphs showed a lot of outleirs indicating that the data has high variability and that BMI was either incorrectly recorded or there is a confounding factor that influences the relationship between BMI and hypertension. The outliers skew the means making the data unreliable and not accurate. More testing is needed into different factors that could influence hypetension in individuals. Additionally, in Figure 2 there was a tendency where as the patients aged, the overall BMI lowered for both groups, denoting that increase in age lowers BMI. Overall, the results show that patients with hypertension tend to have higher BMI and patients with lower BMI tend to have no hypertension. But due to the variability of data and the overlap of box plots there isn’t enough evidence to be conclusive about whether BMI increases the risk of hypertension.

4 References

Centers for Disease Control and Prevention. (2022, June 3). Assessing your weight. Centers for Disease Control and Prevention. https://www.cdc.gov/healthyweight/assessing/index.html

Data Used: Framingham Heart Study Longitudinal Data Documentation