Posted by on Wednesday 19 December 2012

Unusual suspects : age at diagnosis

A short while ago I was musing about one of the questions which form part of the scope of my little side project. It got me thinking, and in order to have a slim chance of having something to say when the subject eventually comes up for discussion I began to wonder just how old people generally are when they are diagnosed with various types of diabetes.

In the old days, of course, things were much more straightforward - even the conditions themselves were referred in terms of 'juvenile' and 'maturity' onset. But I'm forever coming across folks in forums who were diagnosed with 'type 1' (which is typically thought of as a condition that develops in childhood) well into their autumn years.

There was an obvious answer to satisfy my curiosity - just ask you lot. So I decided to conduct a quick straw poll and asked around the good people of the DOC for their age at diagnosis. Now of course this was not carried out with much, if any scientific rigour - but I find the results fascinating nonetheless.

First off I was absolutely humbled and amazed as ever by people's generosity of spirit in the DOC. Within hours of tweeting, "I'm just curious..." I'd been inundated by responses from people who had never heard of me before and my little tweet was bouncing all over the twitosphere with RT after RT. Facebookers and forum peeps waded in and before you could say, "well I wasn't expecting that" I had over 220 responses. Two hundred. And twenty. Many of the 'proper' research projects I see and read have sample populations in the 10s. People even continued to respond for a short while after I'd gathered the results in from the various places.

While I recognise that I was asking a particularly skewed demographic, I found it interesting that after logging 70 or 80 anonymised responses into a spreadsheet the proportions of people in the different age groups remained almost entirely unchanged right up until that 221st person chipped in. So while I completely accept the limitations of the survey I do wonder whether it isn't that far off the real picture.

So what did I discover?
The first and perhaps most surprising thing I discovered was that members of the DOC are much more likely to be type 1. I'm not absolutely sure why this is - perhaps because the management of type 2 is often downplayed by healthcare professionals and people are (mistakenly) led to believe that type 2 is somehow 'less serious'. People with type 1 it seems are more likely to feel like they are struggling and to seek help and support online. So while in the UK population type 1's make up only 10% of diabetics, in my survey they represent 72% of responses!

Oldest and youngest
Of the type 1's the age at diagnosis ranged from just a few months old right up to 66. Type 2's were similarly stereotype-defying ranging from 17 up to 66 again. The average age for diagnosis with type 1 was 18, while the average age for type 2 diagnosis was 45. I didn't specifically ask the question, but at least a dozen people volunteered the information that they had first been diagnosed as type 2 and later reclassified as type 1 or LADA. The average age for the first diagnosis in this group was 35. It would seem that if you are diagnosed with diabetes in your 30s then you stand the highest chance of your Dr putting a tick in the wrong box, and if the treatment you are receiving isn't working despite your best efforts it might be worth asking for some more tests!

Just the facts ma'am
For those who fancy poring over the numbers/proportions for themselves here are the basic analysis tables that Excel squirted out:

Type 1 Type 2 Total
Number 158 63 221
Average Dx Age 18 45  
Dx uncertainty/change 12  
Avg Age at initial Dx if changed 35  
By age group Type 1 Type 2 T1 % T2 %
<= 17 90 0 57% 0%
18-34 47 7 30% 15%
35-44 14 24 9% 38%
45-64 5 29 3% 38%
65+ 2 3 1% 5%

So while, as you might expect, over 50% of T1s are diagnosed by the age of 18, almost half of them aren't. Coupled with which the number of people diagnosed with type 2 in their 20s - early 30s is a pretty meagre 15%, with the several of those being later reclassified). One T1 respondent mentioned that at 18 they were told they were 'too old' to be type 1. For another this happened in their mid-20s.

It would seem that getting diagnosed at an unusual age is, well... er... not that unusual.


  1. I like this post. Thank you Mike for writing it up. If this sort of information was widely available to the medical profession perhaps necessary tests would be more commonplace and correct diagnoses would be a matter of course. As you know, it took me 8 years to get a proper diagnosis after being misdiagnosed due to my age.

  2. Hear Hear,

    But when I was diagnosed in 1972 at 22, from a pee dipstick with all the classic symptoms I appeared to be the only person who was surprised it could strike in adulthood. None of the HCPs were at all phased by it. But then, it was immediate hospital admission for about 10 days and involved a whole armful of blood twice a day ..... and we did have a very good consultant. Plus no fingertip tests or HbA1cs, still had glycosulated haemoglobin then! LOL

    Often wonder if they did GAD or C-pep.

    And how they could be so 'wise' then and now aren't?

  3. I was diagnosed at 31 as a Type 1 in 2001 by my GP, and was told it was unusual but I obviously was T1 because of my C-Peptide Blood Test results. I'd already self diagnosed mind as it was clear that I was T1. Was given insulin and told to go teach myself how to inject before the practice nurse could see me in a month's time.

    Ever since I seem to have had to defend myself to healthcare professionals who upon learning I was diagnosed at 31 that I am not a T2 diabetic, and "Are You Sure you're T1?" is frequently asked of me by professionals not savvy with my history, or seemingly unaware that you don;t have to be under 18 to be diagnosed as T1