Sugar substitute 'sparking global diabetes epidemic'

"Syrup found in biscuits, ice cream and energy drinks fuelling diabetes on a 'global scale'," reports the Daily Mail, highlighting that countries that use large amounts of fructose corn syrup have diabetes rates "20% higher" than countries where it is less common.

This report comes from an ecological study looking into whether there is a link between diabetes levels and the availability of high fructose corn syrup (HFCS). Availability is a measurement of how much of a substance is produced or imported into a country – it does not automatically relate to consumption.

HFCS is used as a sweetener in a wide range of processed food and drinks, but its use and consumption varies widely between countries.

The study found that countries that produced and sold the most HFCS also had higher levels of diabetes when compared with countries with the lowest levels of HFCS availability.

Prevalence of diabetes was 8.0% in the countries with high HCFS availability, compared with 6.7% in countries with lower availability – a difference of approximately 20%.

However, this informative study has some limitations and did not set out to prove that high levels of HFCS consumption caused an increased prevalence of diabetes. Importantly, it did not show that the people with diabetes were consuming more HFCS.

Ecological studies such as these are useful but should be interpreted alongside other studies looking into associations between dietary intake (including HFCS), weight and diabetes at an individual level, so that a complete picture of the potential relationships involved can emerge.

Biscuit-loving UK readers of the Mail's alarming headline will be pleased to hear that consumption of fructose syrup in this country is negligible – a measly 0.38kg per person per year. In the US a whopping 24.78kg per person per year is consumed – more than 65 times that consumed in the UK.

Where did the story come from?

The study was carried out by researchers from the University of Oxford (UK) and the University of Southern California (US). No funding source was reported.

The study was published in the peer-reviewed journal Global Public Health.

Despite a typically arresting headline, the Daily Mail's reporting of this research is well balanced. Especially useful is the reporting of absolute differences between diabetes rates in the countries: "Rates of diabetes were 8% in high-consuming nations and 6.7% among low consumers – a difference of 20%."

This is useful for readers to get a feel for the magnitude of the difference being talked about.

The usual temptation for media outlets is to only report the headline-grabbing "20% higher" figure without any further explanation, which can leave readers thinking the news is more startling than it actually is.

The Mail should also be praised for including a useful graph that shows readers the sharp differences between HFCS availability in different countries, which is a good visual aid.

What kind of research was this?

This was an ecological study looking at the relationship between the availability of high fructose corn syrup (HFCS) and the prevalence of type 2 diabetes across different countries.

An ecological study is an epidemiological study that analyses data at a population level, rather than at an individual level.

HFCS is a corn syrup modified to increase the level of fructose and is used a lot in some processed foods and beverages as a sweetener to replace sugar, as well as prolong shelf life and appearance.

It is found in a host of items, from soft drinks and breakfast cereals to breads, fast food and yoghurt.

Due to historical and economic reasons – namely a series of US trade tariffs – the use of HFCS is particularly widespread in the US, as it serves as a cheaper substitute for more expensive imported sugar.

The researchers report that a growing body of evidence supports the hypothesis that in addition to overall sugar intake, fructose is especially detrimental to health and increases the risk of type 2 diabetes.

It states that the epidemics of obesity and type 2 diabetes we're currently seeing constitute an "alarming public health concern", and that global increases in the use of HFCS in food and beverage production may be contributing to this.

What did the research involve?

Using published resources, the researchers estimated country level estimates of:

  • total sugar availability
  • HFCS availability
  • total calorie availability
  • obesity
  • diabetes prevalence 

The information sources used by the researchers included:

  • diabetes prevalence – International Diabetes Federation (IDF), Diabetes Atlas (fourth edition) and global estimates reported by the Global Burden of Metabolic Risk Factors of Chronic Diseases Collaborating Group (GBMRF)
  • food availability – the Food and Agriculture Organization of the United Nations (FAOSTAT) database of 200 countries
  • HFCS production – an international sugar and sweetener report and data on HFCS quotas for EU countries by F.O. Licht, a commercial organisation that provides information and analysis on some aspects of the global commodity market

Information from 43 different countries was analysed, some of which did not use HFCS at all. The researchers then looked for correlations between the dietary elements (total sugar, HFCS and total calories availability) and the rates of obesity and diabetes.

Some of the analysis adjusted for the effects of body mass index (BMI), as well as population and gross domestic product (GDP) obtained from International Monetary Fund (IMF) tables.

What were the basic results?

Data on 43 countries was available covering the use of HFCS (kg per year per person) alongside estimates of total sugar intake (kg per year per person), BMI, and the estimates of diabetes prevalence from two separate sources (IDF versus GBMRF).
Use of high fructose corn syrup per person

The US was by far the highest consumer of HFCS out of the 43 nations assessed at 24.78kg per year per person, far ahead of second place Hungary at 16.85kg per year per person. The UK was far lower, at 0.38kg per year per person. Fourteen countries registered 0kg per year per person – all except India were European.

Countries with high HFCS availability versus countries with low HFCS availability

The researchers compared measures from those countries with low availability of HFCS (21 countries) versus high availability of HFCS (21 countries). Countries with high availability were defined as having an average value of more than 0.5kg HFCS per person per year.

The average HFCS consumption in the low-availability countries was 0.1kg per person per year, compared with 5.8kg per person per year in the countries classed as having high availability.

The report stated that all indicators of diabetes were higher in countries that had high availability of HFCS compared with those that had low availability. This trend was more significant for the IDF measure of diabetes prevalence.

Countries with high HFCS availability had an average diabetes prevalence of 7.8%, compared with 6.3% in those with low availability (p=0.013). So, the high-availability countries had approximately 20% higher diabetes prevalence than those with low availability (23.8%)

Using estimates of fasting glucose levels to estimate diabetes prevalence showed the difference was 5.33mmol/L in high HFCS availability countries, versus 5.23mmol/L in low availability countries.

Other influencing factors

There were no significant differences between countries of different availability of HFCS (high versus low) for BMI, total calorie intake, cereal intake, total sugar intake and "other sweeteners" intake.

The researchers interpreted this as meaning that the differences in diabetes prevalence may have had more to do with the level of HFCS availability, rather than these additional factors.

How did the researchers interpret the results?

The researchers concluded that, "Our analysis revealed that countries electing to use HFCS in their food supply have a diabetes prevalence that is ~20% higher than that in countries that do not use HFCS […] even after adjusting for country-level estimates of BMI, population and gross domestic product."

They linked their own finding to previous research that they reported "showed that increasing consumption of HFCS in the twentieth century was the primary nutritional factor associated with increasing prevalence of type 2 diabetes."

This led them to warn that, "The increasing popularity of HFCS around the world should, therefore, be considered seriously due to its potential contribution to increases in fructose in the global food supply and its association with the global prevalence of type 2 diabetes."

They also make the point that even modest increases in disease prevalence can have a significant economic impact if a disease is both common and its treatment complex. They state that the health costs of treating diabetes in the US during 2007 was $174bn. A 20% reduction in diabetes prevalence would save $34.8bn, or approximately $95m per day.


This ecological study suggests that countries with a high availability of high fructose corn syrup (HFCS) – defined as more than 0.5kg per person per year – may have higher diabetes levels than those defined as having low HFCS availability.

Countries where availability was defined as high had approximately 20% higher rates of diabetes than those defined as having low availability.

While informative, this study does not prove cause and effect. For example, this study does not show that individuals with diabetes consumed higher levels of HFCS or that this consumption contributed to their diabetes.

Ecological studies such as these need to be interpreted alongside other studies investigating the association between calorie intake (including from HFCS), weight and diabetes at an individual level, so that the full picture of the relationships involved can be established.

Neither HFCS nor diabetes was measured at an individual level, so we cannot assume that the link reported at the country level would be found if the study used individual level data – for example, examining individual diet and diabetes diagnosis.

The low versus high availability of HFCS cut-offs were not justified for clinical or other reasons in the study, and this may have been an arbitrary cut-off.

The choice of where to put this cut-off for low versus high availability and the reasons for such a decision are very important, as selecting a different cut-off point could lead to vastly difference results.

The precise country level estimates of HFCS and diabetes levels are also likely to be subject to significant error that could affect the results.

However, without assessing each information source in detail we cannot say how important this limitation may be, but it is important to be aware of it.

This type of study design is a useful starting point to identify country level trends, but further research is needed at an individual level to explore whether HFCS consumption is linked to diabetes in any way.

Finally, the fact that HFCS availability was relatively low in the UK would suggest that this is less of a public health issue here than in the USA.

However, consumption of HFCS may vary considerably person to person so the Great British biscuit lover should be aware that eating high levels of sugar (HFCS or otherwise) – or indeed fat – is known to have detrimental effects to health.

NHS Attribution