• Please Remember: Members are only permitted to share their own experiences. Members are not qualified to give medical advice. Additionally, everyone manages their health differently. Please be respectful of other people's opinions about their own diabetes management.
  • We seem to be having technical difficulties with new user accounts. If you are trying to register please check your Spam or Junk folder for your confirmation email. If you still haven't received a confirmation email, please reach out to our support inbox: support.forum@diabetes.org.uk

Heart age > QRISK2, QRISK3, NHS Heart Age and other calculators

N

Nimbargent

Member
Relationship to Diabetes
Type 2
I am 69 years old. My GP surgery has just calculated that, using QRISK2, my 10-year risk of heart attack or stroke is 36.25%. I was surprised.

I calculated my score using the more up-to-date QRISK3. It is 28.5%. The risk for a healthy person would be 15.7%.

Then I consulted heart age calculators, inputting all my medical data as required.

My age 69
NHS tool heart age 95
QRISK3 heart age 79
A2Z heart age 75
Australian Heart Foundation 79

This is a significant disparity in results, and the NHS tool differs substantially from the others.

QRISK3 is most skewed by ticking or unticking the diabetes box:
Chance of heart attack or stroke in 10 years
Healthy person 15.7%
My result without diabetes is 18.9% (heart age 72)
My result with diabetes is 28.5% (heart age 79)

So diabetes with QRISK3 adds 7 years to my heart age

Using the NHS tool, my heart age without diabetes would be 79 and with diabetes is 95

So diabetes with the NHS tool adds 16 years to my heart age.

The disparities are so wild that I do not know what to make of the results. The models give little consideration to well-controlled or poorly-controlled diabetes, food habits, exercise or medication. HbA1c is not used.

Has anyone else looked at these calculators?

NHS Heart Age https://www.nhs.uk/health-assessment-tools/calculate-your-heart-age
QRISK3 https://www.qrisk.org/index.php
A2Z Heart Age https://a2zcalculators.com/health-and-medicine-calculators/heart-age-calculator
Australian Heart Foundation https://www.heartfoundation.org.au/heart-age-calculator use post code 3978 to complete the questions.
QRISK2 appears to be no longer available as a link but is now integrated within the GP surgery software.

The QRISK calculation includes:​

  • Age and sex
  • Ethnicity
  • Smoking status
  • Diabetes (type 1 or type 2)
  • Blood pressure readings
  • Cholesterol levels
  • Body mass index (BMI)
  • Family history of heart disease
  • Medical conditions such as chronic kidney disease, rheumatoid arthritis, and atrial fibrillation
  • Medications that may influence risk, including corticosteroids or antipsychotics
  • Socioeconomic factors, including postcode-linked deprivation scores
 
Last edited:
No, do you have the links?

Would be good to take a look.
 
Just looked at the NHS calculator.... My age is 79 but with normal cholesterol, ideal blood pressure and healthy BMI I have the heart of a 95 year old. Still a 95 year old if I remove the pills I take for mild elevated blood pressure. Can easily get back to my real age if I increase the HDL figure from what is considered normal.

Figure that one out.

PS... Have not looked at QRISK. I'd be a bit sceptical of anything that quoted probabilities to four decimal places ( 36.25% is the same as a probability of 0.3626 of an event happening in the next 10 years). Looks to be another case of unjustified implied precision.
 
Docb said:
Just looked at the NHS calculator.... My age is 79 but with normal cholesterol, ideal blood pressure and healthy BMI I have the heart of a 95 year old. Still a 95 year old if I remove the pills I take for mild elevated blood pressure. Can easily get back to my real age if I increase the HDL figure from what is considered normal.
Click to expand...
Thank you, Docb, for your interesting observation pertaining to HDL. You are correct, increasing the HDL alone aggressively within the model brings the heart age down rapidly. My HDL is 0.9 on a weak statin. An HDL of 1 would be considered normal for a man. I can increase HDL within the model to 3.5.
 
Docb said:
Have not looked at QRISK. I'd be a bit sceptical of anything that quoted probabilities to four decimal places ( 36.25% is the same as a probability of 0.3626 of an event happening in the next 10 years). Looks to be another case of unjustified implied precision.
Click to expand...
I think that medical results are often presented in this manner. My last leukaemia result, determined by reverse transcription polymerase chain reaction, is reported as a 0.00157% ratio between leukaemia cells and normal cells. I often see forum posters say something like: "My result six months ago was 0.00157%, but today it is 0.00256%. I am devastated". Many will not understand the concept of statistical significance.
 
Nimbargent said:
I think that medical results are often presented in this manner. My last leukaemia result, determined by reverse transcription polymerase chain reaction, is reported as a 0.00157% ratio between leukaemia cells and normal cells. I often see forum posters say something like: "My result six months ago was 0.00157%, but today it is 0.00256%. I am devastated". Many will not understand the concept of statistical significance.
Click to expand...

Agree with you there @Nimbargent.

If I were in charge I would ban the use of percentages (especially by salesmen and politicians) and would not allow anybody to quote a measurement without its associated error.
 
Heart check - still beating. - good to go.
Only check I do on my heart.
Life is too short for evaluations of risk. Do it later when you can laugh about it in the pub, not before anything happens.
 
I agree that the decimal places are ridiculous and meaningless.

Personally I don't accept the QRisk calculator because it is too general in terms of diabetes in particular. and as you say, there is a big mark up for ticking that box. It lumps me in with many other people with diabetes, some of whom make little or no effort to manage their diabetes, whereas I put a lot of effort into it and I have an active lifestyle.

I have refused statins because the nurse bases her decision on QRisk and because I am a 61 year old woman with diabetes I score higher than they would like, however I can't change any of those parameters but I am fitter and healthier than most of my friends who are the same age and female but without diabetes and they don't apparently need statins. My cholesterol has been mid 4s for the past 6 years so not high, but higher than they like us people with diabetes to be ie under 4. I am happy to make my own decisions on this based on the fact that I am the only person who really knows my lifestyle and results and can therefore best assess my risk.

I haven't calculated my heart age and I have no intention of doing so because I simply don't feel these equations give a realistic reflection of it for me. There is enough to worry about in life without this sort of thing and I prefer to focus my energy on being as healthy as I can, rather than calculating how unhealthy I might be IF these formulae are in some way relevant to me, which I doubt.
 
My QRISK always shoots through the roof the second I tick the 'close family history of heart disease' box.

Interesting to see migraines as an option on there. Was that an option to select on the QRISK2 as well?
 
I think QRISK models were extracted from dozens of research papers, observational studies, and long-term published population outcome data. Then the model was tweaked and revised to improve it based on its predictive nature.

There will always be exceptions, and it’s not very good at precise individualisation, but I think they can be helpful to give a general idea balancing some of the better know and understood risk-factors
 
I use these calculators but also tick the box saying not diabetic I like to see the comparison and as I am a well managed diabetic I put my risk somewhere between the two percentages. Not an exact science but nevertheless it gives me some idea. However, the surgery is very keen on QRISK 2 I put my details in QRISK 3.
 
rebrascora said:
I haven't calculated my heart age and I have no intention of doing so because I simply don't feel these equations give a realistic reflection of it for me. There is enough to worry about in life without this sort of thing and I prefer to focus my energy on being as healthy as I can, rather than calculating how unhealthy I might be IF these formulae are in some way relevant to me, which I doubt.
Click to expand...

Totally agree, could have wrote above myself @rebrascora
 
Thankfully, my surgery has never mentioned QRISK to me (I wouldn't have heard of it if it wasn't for this forum). I understand the logic of it and that it can be a useful tool for HCPs.
Obviously, the criteria that concerns me most is diabetes as it is the only box I can tick. And that is the point for me - it is just a box with no value associated with it. They do not ask you to tick a box for high cholesterol, they ask what your last readings were to quantify it. I am guessing there is not enough data to determine the impact of different current HBA1c. I appreciate the damage may have been done with a high HBA1c in the past but I would expect the same argument to be applied that a low cholesterol reading now does not reduce the impact of high cholesterol in the past.
 
Nimbargent said:
I am 69 years old. My GP surgery has just calculated that, using QRISK2, my 10-year risk of heart attack or stroke is 36.25%. I was surprised.

I calculated my score using the more up-to-date QRISK3. It is 28.5%. The risk for a healthy person would be 15.7%.

Then I consulted heart age calculators, inputting all my medical data as required.

My age 69
NHS tool heart age 95
QRISK3 heart age 79
A2Z heart age 75
Australian Heart Foundation 79

This is a significant disparity in results, and the NHS tool differs substantially from the others.

QRISK3 is most skewed by ticking or unticking the diabetes box:
Chance of heart attack or stroke in 10 years
Healthy person 15.7%
My result without diabetes is 18.9% (heart age 72)
My result with diabetes is 28.5% (heart age 79)

So diabetes with QRISK3 adds 7 years to my heart age

Using the NHS tool, my heart age without diabetes would be 79 and with diabetes is 95

So diabetes with the NHS tool adds 16 years to my heart age.

The disparities are so wild that I do not know what to make of the results. The models give little consideration to well-controlled or poorly-controlled diabetes, food habits, exercise or medication. HbA1c is not used.

Has anyone else looked at these calculators?

NHS Heart Age https://www.nhs.uk/health-assessment-tools/calculate-your-heart-age
QRISK3 https://www.qrisk.org/index.php
A2Z Heart Age https://a2zcalculators.com/health-and-medicine-calculators/heart-age-calculator
Australian Heart Foundation https://www.heartfoundation.org.au/heart-age-calculator use post code 3978 to complete the questions.
QRISK2 appears to be no longer available as a link but is now integrated within the GP surgery software.

The QRISK calculation includes:​

  • Age and sex
  • Ethnicity
  • Smoking status
  • Diabetes (type 1 or type 2)
  • Blood pressure readings
  • Cholesterol levels
  • Body mass index (BMI)
  • Family history of heart disease
  • Medical conditions such as chronic kidney disease, rheumatoid arthritis, and atrial fibrillation
  • Medications that may influence risk, including corticosteroids or antipsychotics
  • Socioeconomic factors, including postcode-linked deprivation scores
Click to expand...
It is said that a person who has had a heart attack has a 20% risk of another one within 7 years. The Actuaries have established that a person diagnosed with Type 2 Diabetes also has a 20% chance of a heart attack/cvd within 7 years of dx. That's why a leading diabetologist ( whose name I have forgotten) said, at the 2009 DiabetesUK Professional Conference, that being dxed with Type 2 was the equivalent of having had your first heart attack. Sobering news irrespective of how risk calculations are done.
 
Sharron1 said:
I use these calculators but also tick the box saying not diabetic I like to see the comparison and as I am a well managed diabetic I put my risk somewhere between the two percentages. Not an exact science but nevertheless it gives me some idea. However, the surgery is very keen on QRISK 2 I put my details in QRISK 3.
Click to expand...
My thinking is much the same as yours, Sharron1. The model suggests that, objectively, diabetes adds many years to the heart age, making it appear like a dreadful disease. However, it gives no credit for the diabetic who takes all sorts of measures to manage his/her diabetes: perfect drug compliance, good diet, adequate exercise.

I have discovered that QRISK2 is built into the software used by some GP surgeries. My GP uses "systmonline". It appears to draw the data directly from the database, eliminating the need for the doctor to fill in the form.
 
everydayupsanddowns said:
I think QRISK models were extracted from dozens of research papers, observational studies, and long-term published population outcome data. Then the model was tweaked and revised to improve it based on its predictive nature.

There will always be exceptions, and it’s not very good at precise individualisation, but I think they can be helpful to give a general idea balancing some of the better know and understood risk-factors
Click to expand...
I am sure that you are correct; objectively, the model is well-supported by wide-scale research. The ticking of the diabetes box impacts very severely on heart risk, heart age and consequently life expectancy. The cholesterol ratio adjustment significantly affects the model's predictions. Most surprisingly, at least in my case, changing the BMI has barely any impact on the risk.
 
I looked quite a long time ago via the NICE references.

My general feeling is here: https://forum.diabetes.org.uk/board...al-diabetes-checks.118283/page-2#post-1450499

iirc I was unimpressed by the number of (T1) diabetics in the cohorts used to generate the statistics (which makes one wonder how representative they were - iirc the data are reported by some subset of doctors' surgeries, and the raw data are unavailable, with only the summary statistics provided due to data protection of the patients), linked to this I'm not keen on the massive effect this tickbox has based on a very small volume of data.

While one assumes that all T1s must be registered with a doctor, which probably removes selective reporting as a problem, the question remains of what a typical T1 looks like (in terms of exercise, weight, and importantly HbA1C - the stats for the latter are available for the UK as a whole up to, iirc, 2021 on a national level, and the numbers weren't good)

For a T1 at least, who sits towards the end of the Gaussian tail of the HbA1C stats, I don't think there's enough data to be useful for me.

I must admit I didn't pay all that much attention to T2 diabetics, so this commentary is perhaps not all that useful for the OP.
 
You must log in or register to reply here.
Back
Top