Risk factors associated with Cardiovascular Diseases

Introduction

Cardiovascular diseases (CVDs) represent a major global health concern. CVDs are the leading cause of death worldwide causing approximately 17.9 million fatalities per year [1]. Numerous behavioural, social, and demographic factors have been identified as CVD risk factors including; unhealthy eating and drinking habits, cigarette smoking, and drug treamtent of diabetes or hypertension [1,2,3]. As CVDs are a major cause of mortality globally efforts to improve the understanding of known and unknown risk factors for CVDs should continue. The findings generated from research into CVD risk factors may assist in the direction of funding within healthcare institutions, inform appropriate health awareness campaigns, and allow individuals to adjust their lifestyles to minimise their risk of CVD, and allow high risk individuals to be identified sooner. The Framingham Heart Study is a landmark, prospective study of the epidemiology of cardiovascular disease among the population of Framingham, Massachusetts. Since its beginning in 1948 the Framingham heart study has collected an array of clinical and demographic data over a prolonged period of time relating to cardiovascular health [5]. Sex has been identified as a major risk factor for CVD, in adults below 50 the prevalence of CVD has been found to be much higher in males than females [4]. In addition, behavioural factors such as: cigarette smoking, unhealthy eating and exercise behaviours have been associated with an increased risk of CVDs [6,7]

As such, the aim of this report was to analyse some of the risk factors associated with CVDs. Specifically, this report analysed the relationship between sex and CVD prevalence, smoking and CVD occurence, and BMI and CVD occurence.

Results

#Load ggplot and tidyr, read in data 
library(ggplot2)
library(tidyr)
heartData= read.csv("http://www.maths.usyd.edu.au/u/ellisp/AMED3002/data/frmgham.csv")

Q1 - What is the incidence of CVDs in males and females?

#Generate a data frame using the variables: Sex, Period, and CVD 
SPC = data.frame("Sex" = c(heartData$SEX), "Period" = c(heartData$PERIOD), "CVD" = c(heartData$CVD))
#Males - subset this data to observe the males (Sex=1) at Period 1 
MalesCVD = (SPC$CVD[SPC$Sex == '1' & SPC$Period == '1'])
#Calculate the total # males 
TotalMalesPeriod1 = length(MalesCVD) 
#CVD positive = '1', CVD negative ='0', hence summing data = #CVD positive males at period 1 
MalesCVDY = sum(MalesCVD) 
#Number of(CVD Negative) = subtract total males - CVD positive 
MalesCVDN = TotalMalesPeriod1-MalesCVDY

#Females - Subset the initial data frame to observe the females (Sex=2) at Period 1 
FemalesCVD = (SPC$CVD[SPC$Sex == '2'& SPC$Period=='1'])
#Calculate the total number of females which satisfy this criteria 
TotalFemalesPeriod1 = length(FemalesCVD) 
#CVD positive = '1', CVD negative ='0', hence summing data = #CVD positive females at period 1
FemalesCVDY = sum(FemalesCVD) 
#Number of (CVD Negative) = subtract total females - CVD positive
FemalesCVDN = TotalFemalesPeriod1-FemalesCVDY

#Table generation - Place these numbers into a table to facilitate the generation of a figure 
IncidencePeriod1 = rbind("CVD Males" = c(MalesCVDY), "No CVD Males" = c(MalesCVDN), "CVD Females"= c(FemalesCVDY), "No CVD Females"= c(FemalesCVDN))
colnames(IncidencePeriod1) <-c("Incidence") 


#This table was re-arranged to make the figure generation process easier
Table = data.frame(x = c('Yes', 'No', 'Yes', 'No'), Sex = c('Male','Male','Female', 'Female'), Incidence = c(686, 1285, 471, 2019))
colnames(Table) = c("CVD", "Sex", "Incidence") 
rownames(Table) = c("CVDMales", "NoCVDMales", "CVDFemales", "NoCVDFemales") 
Table =as.data.frame(Table)

#Graph generation - Generate a stacked bar graph indicating the incidence of CVD among males and females
library(ggplot2)
Graph <- ggplot() + geom_bar(aes(y= Incidence, x = Sex, fill = CVD), data = Table, stat = "identity")
Graph <- Graph + ggtitle("Figure 1.Incidence of CVD by Sex")
Graph <- Graph + labs(y="Incidence", x="Sex")
Graph

Figure 1 Incidence of CVD by Sex, indicates that CVD prevalence is different between men and women. While in this cohort there were more female than male participants, there was a lower incidence of CVD in women than in men. This figure suggests that sex may be a determining risk factor of the development of CVD. The data presented above indicates that men are at greater risk of developing CVD than women.

Q2 - Is there evidence that smoking influences CVD occurence?

#Generate a table containing the variables: current cigarette smoking status and CVD presence  
SmokingT = table(heartData$CURSMOKE, heartData$CVD)
#Perform a Chi Squared test to evaluate if there is an association between smoking and CVD. 
#H0 = there is no relationship between smoking and CVD, the two variables are independent (p > 0.05). Ha = there is a relationship between smoking and CVD, the two variables are not independent (p < 0.05). 
chisq.test(SmokingT) 
## 
##  Pearson's Chi-squared test with Yates' continuity correction
## 
## data:  SmokingT
## X-squared = 5.9981, df = 1, p-value = 0.01432
#Calculate the odds ratio of smoking as opossed to non-smoking and the occurrence or non-occurrence of CVD
library(Rfast)
## Warning: package 'Rfast' was built under R version 3.6.3
## Loading required package: Rcpp
## Loading required package: RcppZiggurat
## Warning: package 'RcppZiggurat' was built under R version 3.6.3
odds.ratio(SmokingT)
## $res
## odds ratio    p-value 
## 1.11245009 0.01350343 
## 
## $ci
## [1] 1.022256 1.210602

The analysis indicates that smoking increases an individual’s risk of developing a CVD, and this association was found to be statistically significant. Smoking and CVD occurence are not independent variables as indicated by the Chi-squared test (p<0.05). As P<0.05 the null hypothesis is rejected, smoking is associated with CVD occurence. Smoking is a possible risk factor for CVD occurrence as indicated by the odds ratio of the occurence or non-occurence of CVD, with or without cigarette use (odds ratio >1, p-value <0.05). These relationships are statistically significant and indicate that cigratte usage is a possible risk factor for the development of CVD.

Q3 - Is BMI associated with CVD occurence?

#Generate a data frame containing the variables "BMI" and "CVD"
BMI = data.frame("BMI" = c(heartData$BMI), "CVD" = c(heartData$CVD)) 
#Generate a boxplot inicating the distribution of BMIs in the CVD cohort and the non-CVD cohort 
ggplot(BMI, aes(x= factor(CVD), y=BMI, fill= factor(CVD))) + geom_boxplot() + ggtitle("Figure 2.BMI and CVD") + labs(y="BMI(Body Mass Index)", x= "CVD occurence") + labs(fill = "CVD") + scale_fill_discrete(name = "CVD", labels= c("No CVD", "CVD")) 
## Warning: Removed 52 rows containing non-finite values (stat_boxplot).

#Calculate the quartile scores, median, and mean of the BMI distribution represented in the boxplot  
median(BMI$BMI[BMI$CVD=='1'])
## [1] NA
#Generate the quartile and median scores of the BMI distribution 
BMICVD = data.frame("BMICVD" = c(heartData$BMI[heartData$CVD=='1']))
summary(BMICVD)
##      BMICVD     
##  Min.   :15.16  
##  1st Qu.:23.86  
##  Median :26.28  
##  Mean   :26.65  
##  3rd Qu.:28.76  
##  Max.   :55.31  
##  NA's   :20
BMINoCVD = data.frame("BMINoCVD" =c(heartData$BMI[heartData$CVD=='0']))
summary(BMINoCVD)
##     BMINoCVD    
##  Min.   :14.43  
##  1st Qu.:22.91  
##  Median :25.18  
##  Mean   :25.62  
##  3rd Qu.:27.82  
##  Max.   :56.80  
##  NA's   :32

The analysis indicates that BMI may be associated with an individual’s risk of CVD however, insufficient analyses have been performed at this point to conclusively state the nature of the association between BMI and CVD occurence. Figure 2 BMI and CVD, indicates that in both the cohort with CVD and the cohort without CVD there is a large distribution across the upper 25% of data. In addition, Figure 2 does indicate that the median BMI of the cohort with CVD is greater than that without CVD. This is supported by the quartile summaries, in the cohort with CVD the median BMI was higher (26.28) than than of the cohort without CVD (25.18). Thus, the data does indicate that the median BMI is larger in those individuals with CVD, however, we can not conclude a relationship of statistical signficance from the data and analysis performed.

Conclusion

The analysis performed indicates that within the Framinham Heart Study cohort explored, there appears to be an association between the male sex and higher incidence of CVDs and between greater median BMI and CVDs. In addition, it was found that there is a statistically significant association between cigarette smoking and the presence of CVD in an individual. This supports the current understanding of the associations between sex, cigarette usage, and unhealthy eating and exercise habits with CVDs. Hence, health campaigns to raise awareness of the cardiovascular risks associated with smoking and poor eating and exercise habits could be increased as a modification of lifestyle factors could lessen the health burden and global mortality of CVDs.

Bibliography

[1]“Cardiovascular Diseases.” World Health Organization, World Health Organization, 17 May 2017,www.who.int/health-topics/cardiovascular-diseases/#tab=tab_1.

[2]Mosca, Lori, Elizabeth Barrett-Connor, and Nanette Kass Wenger. “Sex/Gender Differences in Cardiovascular Disease Prevention: What a Difference a Decade Makes.” Circulation 124.19 (2011): 2145–2154. Web.

[3]Rodgers, Jennifer L et al. “Cardiovascular Risks Associated with Gender and Aging.” Journal of cardiovascular development and disease 6.2 (2019): n. pag. Web.

[4]“New Data from University of Kentucky Illuminate Findings in Cardiovascular Research (Cardiovascular Risk Factors: Does Sex Matter?).” Obesity, Fitness & Wellness Week (2016): n. pag. Print.

[5] “Cardiovascular Diseases.” World Health Organization, World Health Organization, 17 May 2017, www.who.int/health-topics/cardiovascular-diseases/#tab=tab_1.

[6]Lakier, Jeffrey B. “Smoking and Cardiovascular Disease.” The American Journal of Medicine 93.1 (1992): S8–S12. Web.

[7]Healthyweight.health.gov.au. 2014. Body Mass Index (BMI) | Healthy Weight Guide. [online] Available at: http://healthyweight.health.gov.au/wps/portal/Home/get-started/are-you-a-healthy-weight/!ut/p/a1/lVNNj5swFPwru4ccLRvzZY4E0gS0pO1u0gCX1QPMR0SATZx2019fk0StqhVZYnExmnnvzbwxjnGI4wZ-VgWIqm2g7u-x8frEZs7C1xRvrn4jxPvirL6uf0zp3FfwBsc4Ph64y3M41mLF3wWOlP5n2ohOlDgqOdSiPD384lVRioe0bQRvxITsoGompOACHQTsBc8mBPYcndojAnQloQvp_x4uP6TXHl1aZTgCyCiBzEAKsARpSW6ghJs60rmRE0u1TIUZ54kurT33GUdUsyiAkSAddMmx1AQxRgnKGeNmqoOVpYpUH0n1ZODY5GzOgs5mCu3NMXyHeCZbsmCp0BdTu_Kdub3QzCfJ0GQHz50uXNMKpI_GOP4NwJjlfDLBGXBDYiRnMAdnWGr4pa-xcYLX7-vZs0R_DEP0cXfR3bv4u-1RaH-EtXQfOEEhy4IoUdXkLQ4haY8Ch2nbnfbn4ElYtX17i-1reqSg8HagS153VVOgEpo-0X3BCflX8GLn0M6m9GLnHbn2PwvBgNKhp3a_4nFPuNvtmNp_aL0OyVbvCpuper09qb9X-W7DDvbj4x8OImmU/dl5/d5/L2dBISEvZ0FBIS9nQSEh/ [Accessed 9 March 2020].

[Data] Framinghamheartstudy.org. 2020. Framingham Heart Study. [online] Available at: https://framinghamheartstudy.org/fhs-about/ [Accessed 9 March 2020].