Sciencey bit: Rising Incidence

The incidence of disease in the UK in under 15s is something like 20 per 100,000 per year. Prevalence is often quoted at 0.4% (four in every thousand) but I challenge you to identify the paper where this is actually calculated. Incidence is one of those words you kind of have a feel for without actually knowing quite what it is. It is the measure of NEW cases of a disease and is often measured per 100,000 (people) per year. This is distinct from prevalence which is about the total number of people with a disease. As you can imagine one problem with measuring incidence is knowing how many people have the disease, the other is having an accurate idea of how many people make up the general population. You could say 1 out of 100 children were admitted to A&E  because of new onset type 1 diabetes (hence forth abbreviated to T1D for the sake of my sanity) but it wouldn’t really tell you anything about anything except that A&E department (the numbers used here are as an example and are based on no factual information). You might think we have a good record of who has T1D, after all it is fatal if untreated, but when you start including new adult cases of T1D you start to muddy the water with cases of type 2 diabetes. ‘Adult Cases?’ I hear you cry! ‘but T1D is a disease of childhood, isn’t it?’. Actually not so much; as I understand it the majority of individuals actually develop disease after the age of 15 but per age group the incidence is lower in adults (at least with our current definitions of T1D which are slightly circular). For the reasons mentioned above incidence of diabetes is easier to calculate in some places (generally those with better developed health care) and in children. Even in western countries however national registers of diabetes are few and far between therefore another important part of calculating incidence is to try and figure out who you might have missed. This is done by cross checking the cases you found with cases identified by an independent source and is known as secondary ascertainment. If you see incidence without ascertainment be suspicious, very suspicious! The other challenging thing about incidence is that you get a lot of variation year on year. This means you need data from a lot of years to get a real sense of the variation and any change in the average incidence.

 

So now we are caught up on the tricksiness of incidence calculating what is all this about incidence rising? Over the last 30-40 years we have started to get some good data collected over a number of years on the incidence of disease in children aged under 15 years. Studies of the incidence of disease have been spearheaded by a European study (EURODIAB) and a worldwide study (DIAMOND) which have analysed data from a number of countries simultaneously. These studies have shown almost universal increased incidence of disease in childhood at an average rate of 3-4%. Of course that is a increase in a small starting incidence but it is an increase that adds up; the most dramatic data are from Finland (which is also the country with the highest incidence of disease) where incidence increased from 31 per 100,000 per year to 64 per 100,000 per year in just 25 years (1980-2005). The increased rate of disease is most obvious in children under the age of 5 years old. There are two possible explanations 1) that disease is becoming more common in all age groups, 2) that individuals are developing disease at an earlier age. We really need good data on disease in adults to figure out which of these possibilities is true and as yet no consensus has been established.  So though I say ‘rising incidence’ this is actually short hand for ‘rising incidence of type 1 diabetes in under 15s in ‘more developed’ countries because I have no idea about incidence in adults or most of the world’s population’. There have been a few reports of a plateau in incidence but as yet none have been sustained.

 

Why is incidence rising? One might suppose that more people with diabetes are now surviving long enough to reproduce and so maybe we are passing more disease on to new generations genetically (for more info see the genetics section). In a small way this is true but the increase in incidence is too dramatic to be caused by proliferation of ‘at risk’ genes. Some researchers have looked at genetic risk in populations of individuals who have been diagnosed with disease at different time points. These studies show is that people who are diagnosed more recently have (on average) less genetic risk for diabetes. This means those with high genetic risk are still developing disease but more people with less genetic risk are now developing disease. Why might this be happening? Pretty much all biology boils down to the product of genes and environment. So if the genetic contribution is reducing the environmental contribution must be increasing. What might this environmental factor be? Short answer, we don’t have a clue. For some more ideas see the environmental factors post.