Research on the Correlation between Cholesterol Level, Diet, and

Blood Pressure

Jiatong Shen

United World College Changshu China, 215500, Jiangsu, China

Keywords: High Blood Pressure, Cholesterol, Age, Caffeine, Cigarettes, Food.

Abstract: As time progresses, the prevalence of chronic diseases among people has increased, among which is high

blood pressure. High blood pressure is often thought of as being correlated with high cholesterol levels.

Moreover, blood pressure can also be affected by dietary conditions. This paper uses samples from American

elderly aged above 50 between the year 2007-2008 to explore the correlation between cholesterol level, along

with some dietary factors before testing, and blood pressure. A linear regression model is established using R

language to model the relationship between cholesterol levels and blood pressure in the elderly American

population. Also, a logistic regression model is established using R language to model the correspondence

between the cholesterol level and the diagnosis of high blood pressure of the American elderly. This paper

concludes that as age increases, people will have a higher risk of high blood pressure. Generally, the higher

the cholesterol level, the higher the blood pressure. Caffeine and food taken in 30 minutes before the blood

pressure test will slightly increase the blood pressure, while the effect of cigarettes is not stable.

1 INTRODUCTION

High blood pressure is one of the major causes of

premature diseases around the globe. According to

the World Health Organization, unhealthy diets,

physical inactivity, overweight, and consumption of

tobacco and alcohol are all potential factors of high

blood pressure (World Health Organization 2021).

There are many pieces of research concerning the

correlation between cholesterol level, high blood

pressure, and other diseases instead of merely the

correlation between cholesterol level and high blood

pressure. Therefore, this paper chooses to focus on the

correlation between cholesterol level and blood

pressure to fill the gap. This paper investigates the

effect of diet on blood pressure further, as the data

also provides diets from which samples were taken 30

minutes before the blood pressure test.

Cholesterol is a waxy type of fat, or lipid, which

moves throughout people’s bodies in the blood. The

body needs cholesterol to build healthy cells, but high

levels of cholesterol can increase the risk of heart

disease (MAYO Clinic 2021). With high cholesterol,

people can develop fatty deposits in the blood vessels.

Eventually, these deposits grow, making it difficult

for enough blood to flow through the arteries. In

cholesterol level measurements, there are

measurements for the total cholesterol, low-density

lipoprotein cholesterol, and high-density lipoprotein

cholesterol. The total cholesterol is the sum of the

blood’s cholesterol content. Low-density lipoprotein

cholesterol is the kind of cholesterol that can cause

plaques blocking the arteries when there is too much

of it. High-density lipoprotein cholesterol is the kind

of cholesterol that helps to keep away the low-density

lipoprotein cholesterol, making the arteries clean and

unblocked (Cleveland Clinic 2021). This paper uses

the total cholesterol data. The total cholesterol level

is measured during a blood test, and the blood is

drawn from a vein in the arm.

Blood pressure is the force of circulating blood on

the walls of the arteries. Blood pressure is taken using

two measurements: systolic (measured when the heart

beats when blood pressure is at its highest) and

diastolic (measured between heartbeats when blood

pressure is at its lowest). Blood pressure is written

with the systolic blood pressure first followed by the

diastolic blood pressure.

When the arteries become hardened and narrowed

with cholesterol plaque and calcium, the heart must

strain much harder to pump blood through them. As a

result, blood pressure becomes abnormally high

(Cleveland Clinic 2021). Moreover, some external

Shen, J.

Research on the Correlation between Cholesterol Level, Diet, and Blood Pressure.

DOI: 10.5220/0011373400003438

In Proceedings of the 1st International Conference on Health Big Data and Intelligent Healthcare (ICHIH 2022), pages 533-543

ISBN: 978-989-758-596-8

533

dietary factors before the testing can also lead to an

increase in blood pressure. This research explores the

correlation between cholesterol, along with some

dietary factors, and blood pressure of the American

elderly between the year 2007-2008, which can lay

the foundation for comparative research between the

past and the future.

2 METHOD

2.1 Dataset

This paper uses the datasets from NHANES 2007-

2008, including the demographics data, total

cholesterol data, and data of blood pressure

(NHANES 2021).

The demographics data includes information that

was collected using the Sample Person and Family

Demographics questionnaires (NHANES 2021). This

paper specifically extracts the variables of SEQN

(Respondent sequence number) and RIDAGEYR

(Current age of the respondent).

The cholesterol data were collected from blood

specimens. The blood specimens were processed,

stored, and shipped to the University of Minnesota,

Minneapolis, MN for analysis. Vials containing the

blood sample were stored under 30°C until they were

shipped to the University of Minnesota for testing.

There were changes to equipment, lab method, and

lab site (NHANES 2021). This paper specifically

extracts the variables of LBXTC (Total cholesterol).

The blood pressure data were obtained from three

consecutive blood pressure measurements including

systolic and diastolic pressure, which are extracted

from the original table as BPXSY1(Systolic 1),

BPXSY2 (Systolic 2), BPXSY3(Systolic 3), BPXDI1

(Diastolic 1), BPXDI2 (Diastolic 2), BPXDI3

(Diastolic 3). Moreover, there include the

methodological measurements in the data of blood

pressure, including the variables of BPQ150A(Had

food in the past 30 minutes?), BPQ150B(Had alcohol

in the past 30 minutes?), BPQ150C(Had coffee in the

past 30 minutes?), and BPQ150D(Had cigarettes in

the past 30 minutes?).

2.2 Data Preprocessing

In the data preprocessing phase, the three tables for

demographics data, cholesterol data, and blood

pressure data are integrated into one big table. First,

variables are selected, and NA values are omitted.

Next, the mean of the systolic blood pressure as well

as the mean of the diastolic blood pressure are

calculated based on the data of blood pressure from

the three times’ measurements. Then, data are filtered,

and only the ones of which the sample’s age is greater

or equal to 50 are kept. After that, the remaining data

are categorized into different age groups for further

research, including “50-60”, “60-70”, “70-80”, and

“80+”. Finally, a new variable called “Diagnosis” is

created: Samples whose blood pressure is lower than

or equal to 120/80mmHg is labeled as 0, meaning that

this sample is diagnosed as normal, while those

whose blood pressure is higher than 120/80mmHg is

labeled as 1, meaning that this sample is diagnosed as

abnormal or with high blood pressure. Diagnosis

standards are determined based on a conclusion by

the American Heart Association, which indicates that

120/80mmHg and below is the normal range (Heart,

1 Oct. 2021).

2.3 Data Analysis

2.3.1

Data Summary

To analyze the data, the summary for the combined

table is obtained

Figure 1: Summary for the Combined Table.

ICHIH 2022 - International Conference on Health Big Data and Intelligent Healthcare

534

Figure 2: Summary for the Combined Table.

The summary for the combined table shows that

the age group with the most population in this

research is those aged between 60-70, and the age

group with the least population is those aged 80+. The

mean and median of the first blood pressure, whether

systolic or diastolic, is the largest among the three

blood pressure tests. In addition, the means and

medians of systolic blood pressure of people above

50 years old in this research are all above the normal

range, while the ones for diastolic blood pressure are

all within the normal range. Furthermore, the normal

range of total cholesterol is 125 to 200mg/dL, so the

mean of total cholesterol of the tested population

slightly exceeds normal (MedlinePlus, 20 Oct. 2021).

2.3.2 Correlation between Age and Mean

Blood Pressure

Since we want to explore the correlation between age

and mean blood pressure, boxplots regarding the age

and the mean systolic and diastolic blood pressure are

drawn.

Figure 3: Age and Mean Systolic Blood Pressure.

Research on the Correlation between Cholesterol Level, Diet, and Blood Pressure

535

Figure 4: Age and Mean Diastolic Blood Pressure.

These two boxplots demonstrate that as age

increases, the mean systolic blood pressure generally

increases, and the mean diastolic blood pressure

generally decreases. From these trends, we can

conclude that there exist correlations between age and

the mean systolic and diastolic blood pressure.

2.3.3 Correlation between Cholesterol Level

and Mean Blood Pressure

To investigate the correlation between cholesterol

level and the mean blood pressure, scatterplots with a

line of the best fit are drawn regarding cholesterol

level and the mean systolic and diastolic blood

pressure.

Figure 5: Cholesterol Level and Mean Systolic Blood Pressure.

ICHIH 2022 - International Conference on Health Big Data and Intelligent Healthcare

536

Figure 6: Cholesterol Level and Mean Diastolic Blood Pressure.

The plots illustrate that as cholesterol amount in

the body increases, mean systolic blood pressure as

well as mean diastolic blood pressure slightly

increase with a weak positive linear relationship.

2.3.4

Caffeine and 1st Blood Pressure

Boxplots regarding whether the sample ingested

caffeine 30 minutes before the blood pressure

measurements and their implications on the blood

pressure firstly measured, also known as 1st blood

pressure, are drawn.

Figure 7: Caffeine and 1st Systolic Blood Pressure.

Research on the Correlation between Cholesterol Level, Diet, and Blood Pressure

537

Figure 8: Caffeine and 1st Diastolic Blood Pressure.

The samples who drank coffee in the past 30

minutes had a slightly higher 1st blood pressure.

Here, 1st blood pressure is chosen because it is more

easily influenced by diet, as there is a shorter time gap

between testing and diet compared with the other 2

test results.

2.3.5

Cigarettes and 1st Blood Pressure

Boxplots regarding whether the sample ingested

cigarettes 30 minutes before the blood pressure

measurements are drawn.

Figure 9: Cigarettes and 1st Systolic Blood Pressure.

ICHIH 2022 - International Conference on Health Big Data and Intelligent Healthcare

538

Figure 10: Cigarettes and 1st Diastolic Blood Pressure.

The samples that had cigarettes in the past 30

minutes had a slightly lower 1st systolic blood

pressure and a slightly higher 1st diastolic blood

pressure.

2.3.6

Food and 1st Blood Pressure

Boxplots regarding whether the sample ingested food

30 minutes before the blood pressure measurements

are drawn.

Figure 11: Food and 1st Systolic Blood Pressure.

Research on the Correlation between Cholesterol Level, Diet, and Blood Pressure

539

Figure 12: Food and 1st Diastolic Blood Pressure.

The samples that had food in the past 30 minutes

had a slightly higher 1st blood pressure.

2.3.7

Correlation Plot

A correlation plot of all the numerical variables in the

table is drawn, where the larger the dot, the higher the

correlation.

The correlation between cholesterol and blood

pressure is not very significant from the plot.

However, the correlation between blood pressure

tests each time is quite high, which shows the stability

or consistency of a person’s blood pressure

Figure 13: Correlation Plot for Numerical Variables.

ICHIH 2022 - International Conference on Health Big Data and Intelligent Healthcare

540

2.3.8

Linear Regression

The linear regression model models the relationship

between the raw cholesterol data and the mean blood

pressure, whether it is systolic or diastolic

Figure 14: Linear Regression Model Summary.

The first plot is the summary of the linear

regression model for cholesterol and means systolic

blood pressure. Residuals are essentially the

difference between the actual observed response

values and the response values that the model

predicted. The values of the residuals vary. The

coefficient standard error measures the average

amount that the coefficient estimates vary from the

actual average value of our response variable. In the

plot above, the coefficient standard error is relatively

small. The t values are relatively far away from zero

and are large relative to the standard error, which

could indicate a relationship between cholesterol

level and blood pressure exists. The values of Pr(>|t|)

are smaller than 0.05, which indicates that it is

unlikely we will observe a relationship between the

predictor (cholesterol level) and response (blood

pressure) variables due to chance (FELIPE REGO, 23

Oct. 2015).

2.3.9

Logistic Regression

The logistic regression model models the relationship

between the raw cholesterol data and the diagnosis.

Figure 15: Logistic Regression Model Summary.

Research on the Correlation between Cholesterol Level, Diet, and Blood Pressure

541

Figure 16: Logistic Regression Model Summary

Here is the summary of the logistic regression

model for cholesterol and the diagnosis (Le, James,

2021). The deviance residuals are quite small and

close to zero. The values of Pr(>|t|) are smaller than

0.05, which indicates that it is unlikely we will

observe a relationship between the predictor

(cholesterol level) and response (diagnosis) variables

due to chance.

3 DISCUSSION

Firstly, there can be errors in cholesterol data because

the cholesterol sample can be influenced by changes

in the surrounding environment before testing.

Secondly, the data are from 2007-2008, which is a bit

outdated. But what’s helpful is that this research can

lay the foundation for comparative research later.

Thirdly, the results may be biased because they are

based on samples of the American elderly. Fourthly,

the correlations between variables in this research are

generally weak, which might be because there is

limited data. Expanding samples to young people and

gathering more data in later years might be the

solution.

4 CONCLUSIONS

This paper concludes that as age increases, people

will have a higher risk of high blood pressure.

Generally, the higher the cholesterol level, the higher

the blood pressure. Blood pressure can be affected by

diet before testing -- caffeine and food will slightly

increase the blood pressure, while the effect of

cigarettes is not stable.

As the data available only provide dietary inserts

of caffeine, cigarettes, and food 30 minutes before the

blood test, the correlation between diet and blood

pressure has yet to be more deeply investigated.

Further research may investigate the impact resulting

from other types of diet or the long-term impact of

certain nutrition on blood pressure. Moreover, future

research may investigate how public health policy

after 2007-2008, or any other external factors,

influences the rate of getting high blood pressure.

ACKNOWLEDGEMENTS

I would like to express my gratitude to Professor Ma

Shuangge whose lecture on data analytics lays the

crucial foundation for this research. Also, I would

especially like to thank Yi Jiayi for teaching me the

basics of the R language, without whose teaching this

research will not be realized.

REFERENCES

Cleveland Clinic. my.clevelandclinic.org/

health/articles/11920-cholesterol-numbers-what-do-

they-mean. Accessed 1 Nov. 2021.

Cleveland Clinic. my.clevelandclinic.org/

health/articles/11918-cholesterol-high-cholesterol-

diseases. Accessed 1 Nov. 2021.

FELIPE REGO, 23 Oct. 2015, feliperego.

github.io/blog/2015/10/23/Interpreting-Model-Output-

In-R. Accessed 1 Nov. 2021.

Heart. www.heart.org/en/health-topics/high-blood-

pressure/understanding-blood-pressure-readings.

Accessed 1 Oct. 2021.

ICHIH 2022 - International Conference on Health Big Data and Intelligent Healthcare

542

Le, James. "Logistic Regression in R Tutorial." Datacamp,

10 Apr. 2018, www.datacamp.com/community/

tutorials/logistic-regression-R. Accessed 1 Oct. 2021

MAYO Clinic. www.mayoclinic.org/diseases-

conditions/high-blood-cholesterol/symptoms-

causes/syc-20350800. Accessed 1 Nov. 2021.

MedlinePlus. medlineplus.gov/

cholesterollevelswhatyouneedtoknow.html. Accessed

20 Oct. 2021.

NHANES. wwwn.cdc.gov/Nchs/Nhanes/2007-2008/

DEMO_E.htm. Accessed 1 Oct. 2021.

NHANES. wwwn.cdc.gov/Nchs/Nhanes/2007-2008/

TCHOL_E.htm. Accessed 1 Oct. 2021.

NHANES. wwwn.cdc.gov/Nchs/Nhanes/2007-2008/

BPX_E.htm#BPQ150C. Accessed 1 Oct. 2021.

World Health Organization. 25 Aug. 2021, www.who.int/

news-room/fact-sheets/detail/hypertension. Accessed 1

Nov. 2021.

Research on the Correlation between Cholesterol Level, Diet, and Blood Pressure

543