Scribd
Upload
62 views15 pages

Obesity Rates vs GDP in Europe & Central Asia

This document investigates the correlation between obesity rates and GDP per capita in Europe and Central Asia, aiming to determine if media portrayals of wealth and obesity are accurate. The analysis reveals a moderate negative correlation, suggesting that as GDP per capita increases, obesity rates tend to decrease, contradicting common media assumptions. The study highlights limitations in the model, including a low R² value and potential biases in country selection, indicating that other factors may also significantly influence obesity rates.

Uploaded by

hongl26
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
62 views15 pages

Obesity Rates vs GDP in Europe & Central Asia

This document investigates the correlation between obesity rates and GDP per capita in Europe and Central Asia, aiming to determine if media portrayals of wealth and obesity are accurate. The analysis reveals a moderate negative correlation, suggesting that as GDP per capita increases, obesity rates tend to decrease, contradicting common media assumptions. The study highlights limitations in the model, including a low R² value and potential biases in country selection, indicating that other factors may also significantly influence obesity rates.

Uploaded by

hongl26
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd

International Baccalaureate Diploma Program

Is there a correlation between a country’s Obesity rate and its GDP


per capita in USD?
Introduction and rationale
In various forms of media, particularly in the animated films I watched as a child, most wealthy
characters were often portrayed as overweight or obese. In fact, the English language even has
the idiom "Fat Cat," which associates wealthy individuals with the image of a "fat" and "cat."

Figure 1. Character Professor Ratigan in the cartoon “The Great Mouse Detective”

However, despite the rapid development of media today, biases in information sources still
persist.

Since GDP per capita (Gross Domestic Product per capita)—the average income of a resident in
a country—partially reflects the wealth of that country's population, and Obesity Rate represents
the percentage of a country's population classified as obese, in this paper, I aim to investigate the
correlation between obesity rates and GDP per capita in countries within the Europe and Central
Asia region, ranging from low- to high-income countries. This is to examine whether media
portrayals of the connection between wealth and obesity hold true.

Data collection
Since most children (under 18 years old) worldwide do not have an official job or salary,
studying the correlation between obesity rates and income in children would be irrelevant. Thus,
the obesity rate data used in this study focuses solely on adults (18+), including both males and
females.

I use the stratified sampling method: I divide the countries in the Europe and Central Asia region
into four main samples based on GDP per capita: Low income (< $14,999), Lower-middle
income ($15,000 - $34,999), Upper-middle income ($35,000 - $60,000), High income (>
$60,001). Each sample includes 8 countries.
The reason I use GDP per capita instead of GNI (Gross National Income) per capita—which is
the standard used by the World Bank to define income levels—is that GNI per capita includes the
income of a country's residents earned outside the country, which does not fully reflect the
domestic economic context.

Moreover, according to the World Bank, low-income countries have a GNI per capita of less
than $1,145, lower-middle-income countries range between $1,146 and $4,515,
upper-middle-income countries range between $4,516 and $14,005, and high-income countries
have a GNI per capita above $14,005. If I were to use this classification, it would introduce
geographical bias in my stratified samples, as countries within the same income level tend to
cluster in specific regions. For example, low-income countries are mostly concentrated in
Sub-Saharan Africa and East Asia & Pacific, while upper-middle-income countries are primarily
found in Europe and Central Asia. Therefore, to minimize geographic bias, I chose to collect data
only from the Europe & Central Asia region, as this region offers the most diverse range of
income levels.

Figure 2. European & Central Asian region according to World Bank’s definition

Using data from The World Bank and World Obesity Federation - Global Obesity Observatory, I
collected and organized the figures into the tables below.
Country GDP per capita ($) Obesity rate (%)

Ukraine 5069 24.41

Armenia 8053 25.06

Bosnia & Herzegovina 8638 22.41

Belarus 7829 20.13

Montenegro 12221 21.35

Albania 8575 22.27

Serbia 12281 26.44

North Macedonia 8624 29.11


Figure 3. Low-income selected countries

Country GDP per capita ($) Obesity rate (%)

Poland 22056 28.28

Bulgaria 15885 24.08

Hungary 22141 36.43

Latvia 22502 25.72

Czech Republic 31591 26.85

Slovak Republic 24491 27.57

Portugal 27331 21.97

Romania 18404 38.34


Figure 4. Low-middle-income selected countries
Country GDP per capita ($) Obesity rate (%)

Austria 56033 18.81

Finland 52925 21.97

Belgium 54700 20.51

Italy 39003 17.97

United Kingdom 49463 26.94

France 44690 10.00

Germany 54343 20.99

Sweden 55516 15.60


Figure 5. Upper-middle-income selected countries

Country GDP per capita ($) Obesity rate (%)

Netherlands 64572 15.05

Ireland 103887 29.27

Luxembourg 128678 18.89

Denmark 68453 15.70

Monaco 256580 14.00

Norway 87925 19.53

Iceland 79637 21.76

Switzerland 99564 12.47

Figure 6. High-income selected countries


Graph Analysis
Since obesity rates range from 1% to 100%, while GDP per capita ranges from $5,069 to
$256,580, graphing them on Desmos or GeoGebra would cause the data to cluster, making it
difficult to identify trends. Therefore, I use Python and the Matplotlib library to generate Figure
7 using this code:
[Link]
k

Figure 7. Graphed GDP per capita and corresponding Obesity rate using Python and Matplotlib

Looking at Figure 7, most data points are clustered within the 15% - 30% obesity rate range. At
the same time, we can predict that there is a moderate negative correlation between GDP per
capita and Obesity rate, as countries with higher GDP per capita tend to have lower obesity rates,
and vice versa. In other words, obesity rates may decrease linearly or exponentially as GDP per
capita increases and may increase linearly or exponentially as GDP per capita decreases.

Based on this prediction, the most suitable models to represent the relationship between GDP per
capita and obesity rate are: Linear, Logarithmic, Power and Exponential Model.
I input the collected data into lists on the TI-84 Plus CE: Obesity rate (%) in the first column
(corresponding to the horizontal axis); GDP per capita ($) in the second column (corresponding
to the vertical axis).

Figure 8. Lists of GDP per capita ($) and corresponding Obesity rates

Using the entered data, I apply the Regression function on the TI-84 Plus CE to generate
equations.

Linear Model ​
General formula của Linear Equation:
𝑦=𝑎×𝑥+𝑏

By navigating to “CALC”, selecting “LinReg(ax+b)”, and entering the two lists of GDP per
capita and obesity rate, I obtain the following Linear Regression equation:
Figure 9. Linear Regression on TI-84 Plus CE

Using the calculated constants rounded to three decimal places for greater accuracy, I substitute
them into the generalized form to derive the Linear equation.

𝑦= − 3569. 688 × 𝑥 + 128793. 972

To visually examine how well the model fits the data, I graph this Linear equation in Figure 7,
resulting in the following graph:
Figure 10. Graphed Linear regression

I repeat this process for the Logarithmic, Exponential, and Power models.

Logarithmic Model
General form of Logarithmic equation:
𝑦 = 𝑎 + 𝑏 × 𝑙𝑛(𝑥)
Figure 11. Logarithmic Regression on TI-84 Plus CE

Logarithmic Modelling Equation:


𝑦 = 295989. 001 − 80517. 599 × 𝑙𝑛(𝑥)

​
Figure 12. Graphed Logarithmic Regression

Exponential Model
General formula of Exponential equation:
𝑥
𝑦 = 𝑎 ×𝑏

Calculated Exponential Regression:

Figure 13. Exponential Regression on TI-84 Plus CE

Exponential Modelling Equation:


𝑥
𝑦 = 152977. 053 × 0. 932
Figure 14. Graphed Exponential Regression

Power Model
Generalized form of Power Equation:
𝑏
𝑦 = 𝑎 ×𝑥

Calculated Power Regression:

Figure 15. Power Regression on TI-84 Plus CE


Power Modelling Equation:
−1.605
𝑦 = 4377277. 33 × 𝑥

Figure 16. Graphed Power Regression

Examining the Linear, Exponential, Logarithmic, and Power models, they all have Pearson's
correlation coefficient ranging from approximately -0.450 to -0.483, indicating a moderate
negative correlation between GDP per capita and obesity rate.
Among these models, the Power model has the highest R² value (≈0.233), meaning it fits
approximately 23.33% of the collected data points, making it the best-fit model among the four.
However, the R² value remains relatively low, with no significant difference compared to the
other models.
Applying the Power model, we observe that obesity rates decrease exponentially as GDP per
capita increases and increase exponentially as GDP per capita decreases. This result contradicts
the media’s assumption of a positive correlation between wealth and obesity.
Conclusion
Based on the analysis and collected data, I conclude a finding that contradicts the typical media
portrayal of a positive correlation between obesity and wealth: There is a moderate negative
correlation between a country's obesity rate and GDP per capita.
The best-fit model in this context is the Power model, which suggests an exponential increase or
decrease in obesity rates as GDP per capita fluctuates.

This finding also correlates with social reality: As the average income of a country's residents
increases, their living standards improve, leading to greater investments in health, healthier diets,
and better nutritional education, ultimately reducing obesity rates.

However, there are major limitations in my model:


Firstly, the generated Power model captures only 23.3% of the collected data points, raising
concerns about its accuracy. This suggests that many other factors—such as culture, local
cuisine, body image stigmas, etc.—may influence a country's obesity rate more than its material
wealth.

Secondly, my country selection method introduces potential biases: While Europe and Central
Asia belong to the same geographical region, they may have significant cultural and lifestyle
differences, leading to large deviations in data points.
Work Cited
World Bank Data:
[Link]
[Link]
-2024-2025

World Obesity Federation - Global Obesity Observatory:


[Link]

Common questions

Powered by AI

Increasing GDP per capita is likely to decrease a country's obesity rates based on the study's results, which show a moderate negative correlation between GDP per capita and obesity rates. Higher GDP per capita suggests better standards of living, which can lead to more investments in health, improved nutritional education, and healthier diets, all of which contribute to reducing obesity rates .

The Power model, with its highest R² value among the models tested, suggests a moderate negative correlation between GDP per capita and obesity rates, contradicting the media's portrayal of wealth being positively correlated with obesity. The model implies that as GDP per capita, indicative of national wealth, increases, obesity rates tend to decrease exponentially, thereby refuting the traditional assumption of wealthy individuals being more obese .

The stratified sampling approach helps minimize geographic bias by focusing exclusively on countries in the Europe and Central Asia region that have diverse income levels. By selecting countries within this region using GDP per capita, the study avoids clustering countries from vastly different geographic regions under the same income categories, which would likely introduce bias due to varying regional characteristics .

The research suggests cultural influences because even though the Power model was the best fit, it only captured 23.3% of the data variance, indicating that factors such as culture, local cuisine, and body image stigmas might influence obesity rates more than economic conditions. The inherent differences in lifestyle across Europe and Central Asia further suggest substantial cultural impacts on obesity, beyond merely economic determinants .

The study infers that obesity is influenced by multi-factorial elements beyond just economic indicators. Despite using GDP per capita, the best-fit Power model only explained 23.3% of the variability in obesity rates, suggesting significant roles played by other variables such as cultural aspects, dietary habits, educational levels, and socio-economic factors. This underscores the complexity and multi-faceted nature of influences on obesity beyond wealth alone .

GDP per capita was preferred over GNI per capita because GDP solely reflects the domestic economic context, while GNI includes income earned by a country's residents outside the country. This preference helps to avoid geographic bias in stratified samples since using GNI could lead to clustering countries in the same income levels from different regions like Sub-Saharan Africa or East Asia. Thus, GDP offers a more domestic-focused economic view, suitable for analysis within Europe and Central Asia .

The Linear, Logarithmic, Exponential, and Power models were considered for representing the relationship between GDP per capita and obesity rates. The Power model was deemed the most suitable as it had the highest R² value of approximately 0.233, suggesting it explained about 23.33% of the data points, indicating a moderate but still the best fit among the models considered .

Using GDP per capita as a sole indicator for obesity rates is limited because it only captures approximately 23.3% of the variability in obesity rates, which means other factors such as culture, cuisine, and body image perceptions significantly influence obesity beyond economic wealth. Additionally, data from an economically diverse region like Europe and Central Asia may have inherent cultural and lifestyle differences that impact obesity, which GDP per capita can't capture .

The study excludes children from the analysis because most children worldwide do not have an official job or salary. Therefore, correlating obesity rates with income for this demographic would not be relevant. The focus is on adults aged 18 and over, whose income levels through GDP per capita better reflect potential economic influences on obesity .

The analysis reveals a moderate negative correlation between a country's GDP per capita and obesity rate, as opposed to the positive correlation often portrayed in media. Media frequently associates wealth with obesity, as seen in idioms like 'Fat Cat.' However, the study's findings suggest that higher GDP per capita relates to lower obesity rates, possibly due to better health investments and nutritional education as living standards improve .

You might also like