“Obesity and diabetes cause around 800,000 cancers a year,” reports the Mail Online. Researchers found more than 5% of cancers worldwide were caused by being overweight (having a body mass index – BMI – above 25) or having diabetes.
The link between being overweight and some cancers has been known for years. More recently, researchers have linked diabetes to several cancers, including liver, pancreatic and breast cancer.
This is the first time anyone has calculated the combined effect of being overweight and having diabetes on cancer worldwide. Researchers collected data from 175 countries, using information about BMI and diabetes in 2002 and cancers recorded in 2012.
Researchers say their findings are “particularly alarming” because the number of people who are overweight and / or have diabetes is rising worldwide. On current trends, they say the number of cancers attributable to BMI and diabetes could rise by 20-30% by 2035.
Type 2 diabetes and being overweight are closely linked. Both cause a variety of health problems, apart from cancer.
Where did the story come from?
The study was done by researchers at Imperial College London, the University of Kent and the World Health Organization. It was published in the peer-reviewed medical journal The Lancet Diabetes and Endocrinology. It was funded by the National Institute for Health Research and the Wellcome Trust.
The Sun’s headline says: “Diabetes and obesity to blame for 21,000 cancer cases in UK every year”. The study does not include a figure for UK cancers (it pools the 175 countries into larger regional groups), but this seems to be a “back of an envelope” calculation by The Sun that applied the global figure of 5.6% to the 356,860 cases of cancer that Cancer Research UK says were recorded in the UK in 2014.
The global figure of 5.6% is unlikely to be appropriate for the UK.
The Sun and the Mail Online also mistakenly say that global cases attributable to being overweight and having diabetes are set to rise to 30% by 2035. The study says cases could rise by 30%, not to 30%. In other words, cases attributable to being overweight and having diabetes may rise by 30% of the current level of 5.6%, taking the total to 7.28% in 2035.
What kind of research was this?
The study was a comparative risk assessment, in which researchers used the best estimate from research of how much a risk factor affects an outcome, to compare expected outcome if the level of risk factor was zero, with actual rates. This type of research is useful to model how risk factors might have an effect at a population level, and to make projections about the future. But the models are only as robust as the data on which they are based. There’s lots of room for error if any assumption is incorrect.
What did the research involve?
Researchers gathered data about rates of people being overweight (with a BMI over 25) and having diabetes in 2002 from 175 countries worldwide, broken down by age and sex. They then collected data about numbers of 12 cancers already thought to be linked to being overweight or having diabetes, diagnosed in 2012. They used estimates of how much excess BMI and diabetes increase the risk of these cancers, to calculate how many of the cases from 2012 could be attributed to people being overweight or having diabetes in 2002.
The researchers collected data from the NCD Risk Factor Collaboration, which only uses sources where people have their weight, height and any measure of diabetes such as fasting blood glucose recorded (as opposed to self-reporting their BMI or diabetes status). This should increase the reliability of the data.
Calculations of the contribution of BMI and diabetes to cancer risk came from the World Cancer Research Fund, International Agency for Research on Cancer, and published estimates of cancer risk from previous research. They calculated the “population attributable fraction” or proportion of each of 12 cancers attributable to a risk factor in each individual group of people (age group, sex and country) using each risk factor separately and combined. They then used the GLOBOCAN database to find the numbers of cancers for each country.
Finally, they calculated the global proportion of cancers attributable to excess BMI and diabetes, and estimated how much of this was down to the rise in weight and diabetes from 1980 to 2002. They used this to estimate how this might change in the future.
What were the basic results?
Researchers calculated that 5.6% of cancers diagnosed in 2012 worldwide could be attributed to being overweight and having diabetes (792,600 additional cases). However, the overall figure masks big differences between groups of people, regions and cancers.
- Cancers attributed to diabetes and being overweight were twice as common in women (496,700 cases) as men (295,900 cases).
- Breast and endometrial cancer accounted for the highest proportion of cancers attributed to weight or BMI in women, while liver and colorectal cancers accounted for the highest proportion for men.
- The biggest proportion of the excess cases of cancer attributed to being overweight or diabetes came from high income western countries, such as the UK (38.2% of the 792,600 cases), followed by east and southeast Asia (24.1%).
- 16.4% of cases of cancer in men and 15% in women in high income western countries were attributable to being overweight, compared to 2.7% and 3% respectively in south Asia.
- 31.9% of cases attributed to being overweight in the study might not have happened had the rates of overweight in 2002 been the same as the rates in 1980.
How did the researchers interpret the results?
The researchers said: “Our results suggest that the increases in diabetes and BMI worldwide could lead to a substantial increase in the cancer burden in future decades.” They say this is “particularly alarming” because of the high cost of cancer care.
“Population-based strategies to prevent diabetes and high BMI have great potential impact … but have so far often failed,” say the researchers.
The research puts some useful figures around the impact of growing rates of excess body mass index and diabetes on cancer.
While the 5.6% global rate makes for a good headline, it doesn't tell us much because of the wide variations between countries. You might also think that 5.6% is quite low, as it means 94.4% of cancers are caused by something other than weight and diabetes.
However, the much higher figures of 15% to 16% of cancers in high income western countries, and the increase in the numbers of people who are overweight or have diabetes, point to a potentially worrying trend. As known cancer risk factors such as smoking tobacco decline, so other factors such as BMI become more important.
The study has limitations. The estimates of the effect of BMI and diabetes on cancer are all based on observational studies, some of which are more robust than others. Modelling studies such as this are only as good as the data that goes into them. The researchers assumed that risk factors recorded in 2002 would affect cancers diagnosed 10 years later – but we don’t know whether the 10-year lag represents a good picture of people's overall exposure to risk factors such as diabetes and overweight. A further limitation is grouping both diabetes type 1 and type 2 together in the analysis when their risk profiles for cancers are different.
The overall message of the study remains clear, however. Rising rates of excess BMI and diabetes may lead to increasing levels of certain cancers, which will place a big strain on health services worldwide. Tackling the problem at source, by reducing or at least slowing the rise in unhealthy weigh gain and diabetes, should be a priority.
Analysis by Bazian
Edited by NHS Website
Links to the headlines
The Sun, 29 November 2017
Mail Online, 28 November 2017
Links to the science
The Lancet - Diabetes & Endocrinology