Candidate code: JPT000

The correlation between soft drink consumption and Type 2 diabetes

IB Biology SL: Internal Assessment

Exploration
Introduction:
The IB Biology course has studied human physiology in the first year, and the topic of diabetes had
interested me tremendously. My grandmother often worries about my sugar intake and the possibility of
getting diabetes, which my mother often dismisses. In China, where I live, unhealthy fast-food consumption
has increased drastically among the youth due to rapid economic development, leading to fastest growing
obesity rates in the world, with 19% of 6 to 17-year-olds and 10.4% of 0 to 6-year-olds being obese in 2020
(“Share of children”). Accordingly, the diabetes rate grew from 4.2% to 9.2% from 2010 to 2021 (“Diabetes
prevalence”). I am worried about the global trend of increasing obesity and diabetes rates, especially caused
by unhealthy food consumption. In this IA, I hope to examine the correlation between Type II diabetes and
soft drink consumption in the 30 wealthiest countries in the world.

Research Question:
What is the correlation between soft drink consumption, as measured through average revenue per capita in
USD, and the incidence of diabetes, calculated in percentage of population between 20 and 79, in the 30 most
wealthy countries, measured through highest GDP per capita, in 2021?

Background Information:
In a human body, blood glucose needs to be kept between 70-130 mg/dL so blood can maintain an
osmotic balance for cells to have enough supply of glucose for cellular respiration (“Insulin, Glucagon”).
Insulin and glucagon, produced by the Islets of Langerhans in the pancreas, control blood glucose
production. Specifically, the release of insulin triggers glucose uptake into muscles and liver cells to convert to
glycogen. Diabetes occurs when a person is unable to produce insulin or use insulin properly, which causes
patients to constantly have high blood sugar, and can cause hyperglycemia, which “over time leads to serious
damage to many of the body's systems, especially the nerves and blood vessels” (WHO). Diabetes is one of
the most preventable yet costly and deadly diseases around the world. Type 2 diabetes, the most common
type of diabetes, is also known as non-insulin-dependent diabetes mellitus, or adult-onset diabetes, can be
caused by lifestyle habits in addition to genetics.
Excess consumption of sugar, which is often found in sodas and other beverages, is found to lead to
obesity. A study done by Bentley et al. finds that in the US, excess sugar consumption has long lasting,
generation-delayed effects on obesity; the hypothesis is supported not only by biological but also
“physiological, historical and economic evidence” (Bentley et al.). An excess amount of sugar can cause spikes
in blood glucose concentrations, which leads to the pancreas releasing more insulin, causing hyperinsulinemia.
Sustained hyperinsulinemia will lead to insulin resistance, one of the causes of Type 2 diabetes (Thomas et
al.). Obesity is also usually linked with diabetes: “obese people are 80 times more likely to develop type 2
diabetes than those with a BMI less than 22” (“Diabetes and obesity”). This is because abdominal fat causes
increased levels of fatty acids and inflammation, which leads to leptin and insulin resistance, which is, as
established, the cause of diabetes (“How does”). This creates a positive relationship between sugar
consumption, obesity, and diabetes. However, the data is currently correlative rather than causal.
A previous study investigating the links between sugar-sweetened beverages and Type 2 diabetes
found that “sugar-sweetened beverage intake was associated with an increased risk of type 2 diabetes” (Wang

1
Candidate code: JPT000

et al.). Another study also found sugar-sweetened soda consumption to be “associated with an increased risk
of diabetes among Mexican women in a magnitude similar to that reported in other populations” (Stern et
al.). Therefore, the trend has been discovered by many scientists.

Hypothesis:
There is a strong positive correlation between soft drink consumption, measured by revenue (USD) per capita
and diabetes rate in a country, measured by percentage of population. This is because high glucose
concentrations lead to high insulin demand, wearing out the pancreas and causing insulin intolerance, which
can likely lead to Type 2 diabetes (MacGill). I will measure its statistical significance by doing the R-squared
test, where a strongly positive correlation would be as close to 1 as possible, preferably over 0.7.

Approach to Research Question:

I have chosen to study the linkage between soda consumption and diabetes rate within a country;
however, there are variables not considered that may affect both my independent and dependent variable. I
will use a linear regression for the data to examine if the data points do correlate to a trend of positive
correlation and examine the R2 value to see if the regression is accurate. A value of 0.7-1 would show a strong
positive correlation, as stated in my hypothesis.
However, diet composition, physical activity, education levels, genetic prevalence, sleep rates,
lifestyle, and more of the population in a country could all affect diabetes rate, which means that the results
may not correlate to my hypothesis because of the different cultures in individual countries. Additionally, the
wealth of a country also affects diabetes rates—a previous study by Dagenis et al cites that states with lower
income have the highest diabetes prevalence and countries with higher income have lower diabetes
prevalence, creating an inverse relationship. Factors such as “education, rapid industrialization, urbanization,
international trade with high-income countries (HICs), migration from rural to urban centers, and lifestyle
changes (e.g., calorie-dense diets, reduced physical activity, and tobacco use) may affect diabetes differently in
developed versus developing countries” are all socioeconomic factors that can affect diabetes prevalence.
There is also always variance within a country. For example, populations in cities with only money to
access fast food may have higher diabetes rates than the richer ones who purchase organic food and have a
diet regimen. Therefore, soda consumption may be a surface-level variable studied compared to diet
composition and lifestyle habits within a state, however this data is also simpler and more reliable to analyze
as it could have fewer confounding variables compared to, say, HDI.
Also, there could of course be variability within the data chosen, for example sampling errors or
newly introduced governmental strategies to battle diabetes. Therefore, I must account for random variability
in my data. However, it should be highlighted that since I have chosen the 30 wealthiest countries, there
should be more accurate data as none of them are in states of war, are corrupt, or have other conditions that
would drastically alter data collection.

Variables:
Independent variable: soda/soft drink consumption, measured through average revenue (USD) per capita in
2021
- Soda drink consumption was chosen as the independent variable. This investigation assumes that
with increased consumption, there would be an increased diabetes rate. Soft drink is defined as a
water-based non-alcoholic drink with added sugar (“Soft drinks”).

2
Candidate code: JPT000

- It is found that sugary drinks account for 30% of 4- to 10-year-olds’ daily sugar intake in the UK,
making soft drink consumption a good way to measure sugar consumption, which leads to increased
blood glucose (Public Health England).
- Carbonated and non-carbonated soft drinks as well as energy & sports drink are included in the soft
drink category, encompassing most sugary drinks other than fruit juices.
- Previous research (Wang et al.) has utilized data with consumption of sugar-sweetened drinks to link
it with Type 2 diabetes, and a paper by Goryakin et al. utilized the sales of sodas to examine soda
consumption in different states. This makes the independent variable valuable to examine, as past
research has utilized similar topics.
Dependent variable: proportion of population between 20-79 years with diabetes in 2021
- The effect of obesity on diabetes is known to be generation-delayed according to past research, and
though youth diabetes rates are growing, the lasting effects of excess sugar consumption causing
obesity and diabetes mostly land in those from 20-79 years (Bentley et al.).
- Additionally, most of the data in Diabetes Atlas examine the population with diabetes between these
ages as adults are easier to survey than children.
Controlled variable: 30 wealthiest countries, ranked by GDP per capita in 2021
- It is determined that with a higher GDP per capita, more people have the financial capability to
purchase foods such as sodas which are outside the range of necessity for subsidence. Soft drinks are
more common and more often consumed in wealthier states, according to NationMaster, as 15 out
of 15 countries with highest soda consumption are considered developed states with high income.
- The data coming from developed states will also be more reliable since less developed states may
have less established institutions for research.
- 30 countries were chosen as there would be an appropriate amount of data for an accurate
conclusion while also being focused on similar GDP per capita states.

Data Sources:
Soft drink revenue (USD) per capita 2021 in different countries: Statista
- The data is collected through Statista’s in-house market research, international institutions, national
governmental reports, and various companies. As Statista is a paid service providing market research
used by numerous leading institutions and companies around the world, the data can be considered
as accurate.
Proportion of population between ages of 20-79 with diabetes in a country 2021: Diabetes Atlas
- Diabetes Atlas is a website built by the International Diabetes Foundation that works with over 240
different organizations around the world to tackle the global problem of diabetes. It works with the
United Nations and the World Health Organization, lending it credibility.
Countries ranked by GDP (USD) per capita 2021: Statistics Times
- The data on statistics times is collected from the International Monetary Fund as well as the World
Bank, two important intergovernmental institutions, which lends this data credibility.

Materials:
- A laptop
- Excel
- Internet access to sites (Statista, Diabetes Atlas, Statistics Times)

3
Candidate code: JPT000

Methodology:
1. Go on Statistics Times to find the ranking of the top 30 countries with the highest GDP per capita in
2021 (https://statisticstimes.com/economy/projected-world-gdp-capita-ranking.php). Write down
the names of these states in an Excel document in a single column (A2-A31), named “Country.”
2. Looking at the first country in the document, access Statista. Go to Consumer Markets à Non-
alcoholic drinks à soft drinks. Then select the country from the Excel sheet and look it up in the
Regions section. For example, here is soft drinks for Burkina Faso:
https://www.statista.com/outlook/cmo/non-alcoholic-drinks/soft-drinks/burkina-faso. On the
website, scroll to the second bar graph in the Revenues section, which says “Average revenue per
capita.” Document the average revenue for 2021 in the column to the right of the “Country” column
(B2-B31), naming it “Soft Drink Revenue” in B1.
3. Repeat step 2 for the next 29 countries.
4. Access Diabetes Atlas (https://diabetesatlas.org/data/en/). Look at the country written in the first
column A2 and look it up in the “Country or territory” section. Find the age-adjusted comparative
prevalence of diabetes, then write down the diabetes rate in a new column (C2-C31), named
“Diabetes rate,” in the same row as the country. It will be shown as a percentage of the population,
e.g., United States, 10.8% of population from 20 to 79.
5. Repeat step 4 with the next 29 countries.
6. Now, you should have a list of countries from A2-A31, numbers of soda revenue from B2-B31, and
diabetes rates from C2-C31. Label A32 as Q1, A33 as Q3, A34 as IQR. In B32, under the numbers
of the soda revenue per capita in the countries, write down “=QUARTILE (B2:B31,1)”. This should
give you a number that is Q1, which is the first 25% of data. In B33, write “=QUARTILE
(B2:B31,3)”. This will give you Q3, 75%.
7. Repeat step 6 with the C column, but instead of B2:B31, write C2:C31, and vice versa with the Q3.
8. In B34, write down “=B33-B32”. This will give you the interquartile range, or IQR, which is the
middle 50% of the data. Repeat in C column in C34 with “=C33-C32”.
9. Label A35 “lower outlier” and A36 “upper outlier.” In B35, write down “=B32-(1.5*B34)”. This
means Q1-(1.5*IQR), which is the formula for the lower outlier of the data set. In B36, write
“=B33+(1.5*B34)”. This will give you the upper outlier of the data set.
10. Highlight any numbers in the B column that is either under the lower outlier or above the upper
outlier. Then highlight the same row for A and C column. For example, Qatar has a diabetes rate of
19.5%, which is above the upper outlier of 13.4875%. So, I highlighted the entire row of data in
Qatar.
11. Repeat steps 8-9 with the C column, but instead of B32, write C32 and vice versa.
12. Now, you should have several rows of data highlighted. Either remove the highlighted data or move
it to another column. This data will not be considered in our calculations since they contain outliers
from the data set.
13. Select the data for the diabetes rate and soda consumption for each country (B2-B31 to C2-C31) and
go to the insert graph function. Create a scatter chart, making the X-axis the B column (independent
variable, soda revenue) and the Y-axis the C column (dependent variable, diabetes rate). In the Chart
Elements section, select Trendline à Linear Trendline. Add R2 and R value, which should be some
of the options for chart element.

4
Candidate code: JPT000

Safety, ethical, or environmental issues:

There should be little safety or environmental issues for this investigation, as it is all done on the computer so
it should have little impacts on personal safety or the surrounding environment. However, we must note that
we cannot be certain if the data collection process has been completely ethical or if they are completely
accurate. Although the data sources, which are Statista, Our World in Data, International Diabetes
Federation, and Statistics Times, are known to be credible as they are used by many international institutions,
one must continually consider the sources in which we find our data by keeping an eye out for new
information about how this data is collected.

Analysis
Raw Data:
Table 1: Soft Drink Revenue and Diabetes Rate for Countries with Top 30 GDP per capita in 2021
Country Soda revenue per capita Diabetes rate (% of population
($USD) (±0.05) between 20 and 79) (±0.05)
Luxembourg 362.0 5.9
Switzerland 404.5 4.6
Ireland 419.4 3.0
Norway 454.8 3.6
Iceland 490.2 5.5
United States 934.0 10.7
Singapore 494.4 11.6
Qatar 169.1 19.5
Denmark 351.8 5.3
Australia 583.1 6.4
Netherlands 453.8 4.5
Sweden 295.5 5.0
Austria 310.5 4.6
Finland 260.9 6.1
Germany 310.4 6.9
Canada 359.9 7.7
Belgium 441.7 3.6
Israel 237.9 8.5
United Arab Emirates 137.6 16.4
United Kingdom 501.1 6.3
New Zealand 407.5 6.2
France 223.4 5.3
Japan 480.9 6.6
Italy 120.1 6.4
Kuwait 138.4 24.9
Korea, Rep. 312.1 6.8
Brunei Darussalam 125.2 11.1
Spain 281.5 10.3
Malta 290.6 8.0
Cyprus 242.2 8.6

5
Candidate code: JPT000

Sample Calculations:
Interquartile Range (IQR):
𝐼𝑄𝑅 = 𝑄3 − 𝑄1
Ex: IQR for soft drink revenue
450.7575 − 246.85 = 203.9075
Lower outlier:
𝑙𝑜𝑤𝑒𝑟 𝑜𝑢𝑡𝑙𝑖𝑒𝑟 = 𝑄1 − (1.5 × 𝐼𝑄𝑅)
Ex: lower outlier for soft drink revenue
246.85 − (1.5 × 203.9075) = −59.01125
Upper outlier:
𝑢𝑝𝑝𝑒𝑟 𝑜𝑢𝑡𝑙𝑖𝑒𝑟 = 𝑄3 + (1.5 × 𝐼𝑄𝑅)
Ex: upper outlier for soft drink revenue
450.7575 + (1.5 × 203.9075) = 756.61875

Processed Data:
Table 2: Data Processing for Soft Drink Revenue and Diabetes Rates
Statistics Soft drink revenue statistics Diabetes rate statistics
Q1 246.85 5.3
Q3 450.7575 8.575
IQR 203.9075 3.275
Lower Outlier -59.01125 0.3875
Upper Outlier 756.61875 13.4875
Thus, data for United States, Qatar, United Arab Emirates, and Kuwait are removed since they contain
outliers for either soft drink revenue or diabetes rates.

6
Candidate code: JPT000

Graph 1: The impact of soft drink consumption on the diabetes

rate in the countries with top 30 GDP per capita
14
diabetes rate (% of population between 20 and 79) (±0.05)

12

10

y = -0.006x + 8.6189
4
R² = 0.1035

0
0 100 200 300 400 500 600 700
soft drink revenue per capita (USD) (±0.005)

Linear regression: 𝑦 = −0.00604358𝑥 + 8.61891

R2 value= 0.1035
R value=-0.3218

Interpretation
The scatter chart in Graph 1 does not show any particular trend, instead the data points seem to be
placed at a more centered place with no particular sense of movement or any obvious positive or negative
relationship. The lack of any obvious observable trend implies that there is little correlation between the soft
drink revenue and diabetes rate of these countries, thus not supporting my hypothesis that there would be a
positive relationship. The chart shows that the obesity rate is around 6% for most of these countries and the
soft drink revenue around $350 per person. Similarly, when observing Table 1, the data show interesting
results, such as the Australia having quite high soft drink revenue ($583.1 per capita), but only 6.4% of Type 2
diabetes, compared to states like Brunei where the diabetes rate is almost twice that (11.1%) but the soft drink
revenue is only $125.2 per capita. Additionally, wealthier countries do not have higher incidence of Type 2
diabetes, as the first five richest states all have less than 6% diabetes rate.
There are multiple anomalous points in the results, which include the United States’ soft drink
revenue of $934.01 per person and Qatar’s diabetes rate of 19.5% of the population, which are shown in
Table 1. In order to remove the outliers that may drastically alter the accuracy of the regression, I calculated
the lower and upper outlier value and removed them from consideration. I will explain the reasoning behind
the presence of these outliers in the evaluation section.

7
Candidate code: JPT000

The linear regression provided on Graph 1 creates a negative relationship, as shown by the negative
R value of -0.32, which means that when the independent variable increases (soda revenue is higher), the
dependent variable decreases (lower diabetes rate), which is the opposite of my hypothesis. However, this
linear regression has a very small slope (-0.0006), which shows that there is only very small decrease in
diabetes rate that comes with an increase of soda consumption. It is important to note that this linear
regression does not accurately portray the trend of the data, as shown in the R2 value, which is merely 0.1035.
The closer the R2 value is to 1, the less variance there is and the closer the regression represents the data. This
means that even though there is a regression that shows an inverse relationship, it cannot be relied on as it is
vastly inaccurate.

Evaluation
Conclusion
Unfortunately, the results of this experiment do not support my hypothesis that there is a strong
positive correlation between increased soft drink revenue per capita compared to the diabetes rate in the
country. This is because there are no observable trends in the scatter chart in Graph 1, and the linear
regression provides a negative relationship with a slope of -0.006 and an R value of -0.32. However, this does
not support the idea that there is a negative relationship because the linear regression does not represent the
data very well, as its R2 value is extremely low, at 0.1035, but one would prefer an R2 value at around 0.7-1 to
show a strong relationship. This means that the conclusion to be drawn from this data analysis suggests that
there is no direct correlation between soft drink consumption in a country with the diabetes rates.
The results from this experiment do not match similar experiments performed in the past. Stern et al.
found there to be a strong correlation between sugar-sweetened soda and diabetes. However, they had
adjusted the data for different height and ethnic group, finding Hispanic women with height over 149cm are
most vulnerable to diabetes after increased soft drink consumption. In research done by Wang et al., it is also
found that there is a positive relationship, especially in the USA and Europe, generally regions of high
development.
Despite the results of this experiment, I still believe that there is a positive correlation between soda
consumption and diabetes. Malik et al. that sugar-sweetened beverages “are positively associated with weight
gain and obesity, the well-established risk factors for type 2 diabetes.” Since soft drinks are the most common
source of sugar-sweetened beverage, one can also analyze the effects of sugar on obesity. This is because soda
can “induce rapid spikes in glucose and insulin blood concentrations, contributing to a high glycemic load,”
which leads to hyperinsulinemia, which is an unusually high amount of insulin in one’s blood. This
hyperinsulinemia often occurs before Type 2 diabetes, where hormones become less sensitive to insulin,
enhancing insulin resistance, as “fasting insulin rises from normal glucose tolerance to impaired glucose
tolerance (IGT)” (Thomas et al.). This
Graph 2: Diabetes rates around the world, 2021
insulin resistance causes the body to
become unable to utilize insulin correctly,
leading to Type 2 diabetes.

Source: Diabetes Atlas

8
Candidate code: JPT000

Graph 3: Sugar consumption around the world, 2021

Source: Protectivity Insurance

From Graphs 2 and 3, one can observe a correlation between sugar consumption and diabetes rates,
particularly in South America, Africa, and the Middle East. However, this data is not perfectly correlated since
regions like Russia have high sugar consumption but low diabetes rates. So, although this data does not
strongly support the hypothesis and scientific explanation above, it suggests that there is a correlation to a
moderate extent. However, the possible reasons for the lack of a specific conclusion will be discussed below.

Strengths and Weaknesses

A strength in this investigation is the source used. Statista and Diabetes Atlas are some of the most
internationally recognized databases, utilized by IGOs like WHO and businesses like Google have used,
which lends the data considerable credibility. Additionally, as the most developed states were chosen for
investigation, which means that the data collection process should be more accurate as there are more
institutions for such purposes. However, the data collected may have variation as well. Though the random
biological variation is accounted for as these would even out in the data collection process, one can never be
sure of the measurement and systematic errors that came from this database. Since there is a chance that the
diabetes rate may come from self-reporting and the soft drink revenue may come from consumer surveys,
one must consider that these numbers could be exaggerated at both ends, either underestimated or
overestimated. Furthermore, there could be greater healthcare in some countries or educational prevention
that would decrease the chances of diabetes by promoting exercise even with a high soft drink consumption.
A weakness in this investigation is the lack of statistical tests. It is difficult to utilize many different
statistical tests as there is not enough result for the pooled effect estimate. Chi-squared is not used in this
investigation because it measures ecological data that is categorical instead of continuous like the data
presented here. However, a positive aspect is that I had standardized my data by accounting for anomalies
that are outliers by finding the lower and upper outlier for both categories, soda revenue and diabetes rate.
Another weakness is the lack of nuance within the data. For example, there are different contents of
sugar across different drinks, and different countries may have people who buy soft drinks with higher or
lower sugar content. If we are able to acknowledge the different tastes for soft drinks between the different
countries, this could deepen our understanding of the relationship between sugar consumption and diabetes.
A major flaw in the investigation is the complexity of diabetes, an uncontrolled variable which can be
a cause for the anomalies and a major reason why the results from the investigation do not support my
hypothesis. As mentioned in the Approach to RQ section, there are many other variables that can affect
diabetes rate, one of them being the lifestyle of the population in a country. For example, Qatar has city
planning where there is little walking to be done, and with the rapid economic development of the country,

9
Candidate code: JPT000

consumption habits were unable to keep up: people often ate fast food, which causes obesity. However, the
soft drink consumption in the state is quite low compared to the other states, at around $169.12, which
means that soft drinks may not be a common part of the local diet. So, although consumption of soft drinks
in Qatar is comparatively low, the people still consume fats that cause obesity, which increases diabetes rates.
Since developed countries were chosen, governments may be more proactive in tackling issues like diabetes,
for example New York’s infamous soda tax that is aimed to lower obesity and diabetes among youth in the
United States. Details and complexities like these cannot be accounted for, as this experiment measured one
singular aspect of consumption habits that diverges across countries. Also very importantly, there are also
racial and ethnic differences when it comes to vulnerability to insulin resistance. Studies have shown that in
the United States, African Americans and non-Hispanic whites have “greater insulin sensitivity than East
Asians and South Asians,” which increases their risks for obesity and diabetes as they may become unable to
use insulin correctly (Raygor et al.). According to a study done by Diabetes Care, Caucasians have an insulin
sensitivity index of 6.87 while Asian Americans 4.17, making Caucasians more vulnerable to diabetes (Chiu et
al.). The differences between the ethnic makeup and genetic factors have not been studied in this
investigation, and this may be influential as well when it comes to countries in different continents. These are
all factors that may have influenced the conclusion achieved from this database analysis.

Suggestions for improvement:

A potential method for extension is to focus on a few countries (e.x. United States) or a region (e.x. European
Union) to perform a meta-analysis, which is the method utilized by past researchers such as Stern et al. This
would account for variation in lifestyle, as it is assumed that citizens in one country or region would generally
have more similar lifestyles, habits, diets, genetic makeup, education levels on average compared to that of
another country in a different continent. This would allow for a more accurate conclusion to be drawn as
there are less uncontrolled variables that differ from one another, which helps the weakness of the lack of
nuance within the data and lack of account for scientific and social variations between states. That said, there
may still be extreme differences between the habits of, say, different states in the US. Therefore, one could
adjust the data set with known factors that affect diabetes, such as education, sleep levels, BMI, and so on,
with tests like the pooled effect estimate. Additionally, another way to improve upon this investigation is to
use sources like PubMed, SpringerLink, and Elsevier databases, which are scientific journal databases with
paid access. Though the resources used here are already quite credible, using databases specific for scientific
and medical content could perhaps add to the complexity of the data as it would present specific aspects, like
how the data is collected and its accuracy.

Works Cited:
Bentley et al. “U.S. obesity as delayed effect of excess sugar.” Economics and Human Biology, Volume 36, January
2020. https://doi.org/10.1016/j.ehb.2019.100818
Chiu, Ken C et al. “Insulin Sensitivity Differs Among Ethnic Groups With a Compensatory Response in
Beta-Cell Function.” Diabetes Care, vol 23, no. 9, Sep 2000.
https://care.diabetesjournals.org/content/diacare/23/9/1353.full-text.pdf
Dagenais, et al. “Variations in Diabetes Prevalence in Low-, Middle-, and High-Income Countries: Results
From the Prospective Urban and Rural Epidemiological Study.” Diabetes Care, 10 May 2016, American
Diabetes Association. https://doi.org/10.2337/dc15-2338
“Diabetes.” World Health Organization. 10 November 2021.https://www.who.int/news-room/fact-
sheets/detail/diabetes
“Diabetes and Obesity.” Diabetes.co.uk. 15 January 2019. https://www.statista.com/outlook/cmo/non-
alcoholic-drinks/soft-drinks/worldwide

10
Candidate code: JPT000

“Diabetes Data Portal.” Diabetes Atlas, International Diabetes Federation. 8 November 2021.
https://diabetesatlas.org/data/en/country/36/ca.html
“Diabetes Prevalence, 2019.” Our World in Data. https://ourworldindata.org/grapher/diabetes-
prevalence?time=2019
Goryakin, Yevgeniy et al. “Soft drink prices, sales, body mass index and diabetes: Evidence from a panel of
low-, middle- and high-income countries.” Food policy vol. 73 (2017): 88-94.
doi:10.1016/j.foodpol.2017.09.002
“How does obesity cause type 2 diabetes?” Vitagene. 20 July. https://vitagene.com/blog/does-obesity-cause-
type-2-diabetes/
“Insulin, Glucagon, and Diabetes.” Kognity. https://app.kognity.com/study/app/biology-sl-fe-2016/hum-
phys-core/hormones-homeo-core/insulin-glucagon-core/
“Lifestyle > Food and drink > Soft drink > Consumption: Countries Compared.” NationMaster.
https://www.nationmaster.com/country-info/stats/Lifestyle/Food-and-drink/Soft-
drink/Consumption
MacGill, Markus. “How soda impacts diabetes risk.” Medical News Today. 18 December 2018.
https://www.medicalnewstoday.com/articles/259604#takeaway
Ndumele, Chiadi E. “Obesity, Sugar and Heart Health.” Johns Hopkins Medicine.
https://www.hopkinsmedicine.org/health/wellness-and-prevention/obesity-sugar-and-heart-health
“PHE urges parents to cut sugary drinks from children’s diets.” Gov.UK. 17 July 2015.
https://www.gov.uk/government/news/phe-urges-parents-to-cut-sugary-drinks-from-childrens-diets
Raygor, Viraj et al. “Impact of race/ethnicity on insulin resistance and hypertriglyceridaemia.” Diabetes &
vascular disease research vol. 16,2 (2019): 153-159. doi:10.1177/1479164118813890
“Share of children and teens with overweight problems in China in 2020, by age group.” Statista. 10 Feburary
2021. https://www.statista.com/statistics/1202241/china-prevalence-of-overweight-among-children-
and-teens/
“Soft Drinks.” Statista. https://www.statista.com/outlook/cmo/non-alcoholic-drinks/soft-drinks/worldwide
Stern, Dalia, et al. “Sugar-Sweetened Soda Consumption Increases Diabetes Risk Among Mexican
Women,” The Journal of Nutrition, Volume 149, Issue 5, May 2019, Pages 795–
803, https://doi.org/10.1093/jn/nxy298
“Sugar Consumption Map.” Protectivity Insurance. https://www.protectivity.com/sugar-consumption-map/
Thomas, Dylan D et al. “Hyperinsulinemia: An Early Indicator of Metabolic Dysfunction.” Journal of the
Endocrine Society vol. 3,9 1727-1747. 24 Jul. 2019, doi:10.1210/js.2019-00065
Wang, Meng et al. “Association between sugar-sweetened beverages and type 2 diabetes: A meta-analysis.”
Journal of diabetes investigation vol. 6,3 (2015): 360-6. doi:10.1111/jdi.12309
“World GDP per capita Ranking/” Statistics Times. 26 August 2020.
https://statisticstimes.com/economy/world-gdp-capita-ranking.php

11