Screening for and preventing type 2 diabetes 

NICE published its public health guidance on screening for diabetes and prediabetes in 2017; it was broadly unchanged from the previous 2012 iteration (NICE 2017, PH38). In 2025, SIGN stepped in with fresh guidance which broadly agrees with NICE PH38, and has additional information on remission, ultra-low-calorie diets, GLP-1 mimetics and tirzepatide for prediabetes (SIGN 2025, 172). Here, we summarise the (rather complicated!) guidance. But, first – is screening for diabetes worth it?

This article was last updated in June 2026.

Is screening for and treating prediabetes worth it?

Yes! (JAMA 2023;329:1206)

  • Clinical trials show that progression to diabetes can be prevented with intensive lifestyle modification.
  • Interventions improve macrovascular and possibly microvascular outcomes.

All this is discussed in more detail in the article Prediabetes.

‘Know your diabetes risk’

Diabetes UK encourages people to use an online score to assess their diabetes risk. Those who use the risk score and are at increased risk (a moderate or high score) are encouraged to seek a screening blood test from their GP.

The Know Your Risk score asks about:

  • Sex.
  • Age.
  • Ethnicity.
  • Height and weight.
  • Waist circumference.
  • Personal history of hypertension.
  • Family history of diabetes in a first-degree relative.

Could you add a link on your practice website or display in the waiting room?

NICE guidance on screening for diabetes and prediabetes

In 2019, the UK National Screening Committee declared that population-level screening was not supported by evidence.

Instead, NICE asks us to perform targeted screening.

The guidance asks us to screen a LOT of people, but the screening work isn’t just for us – community pharmacists, dentists, opticians, community leaders, occupational health teams and those working in prisons, as well as people providing services for those with learning disabilities, are all encouraged to identify high-risk patients and risk stratify them (and then send them to us for the blood test!), or get patients to do this for themselves via the Diabetes UK website (see useful resources, below, for the link).

The key message is that, for most, the guidance recommends a two-stage screening process:

  • Stage 1: assess risk of diabetes in high-risk people. This should be done using a validated scoring tool, similar to QRISK for CVD risk.
  • Stage 2: in those deemed high risk on the scoring tool, do a blood test (HbA1c or fasting plasma glucose) and offer the appropriate intervention according to the result of the blood test.

Which risk scoring tool should I use?

NICE does not recommend a particular risk scoring tool; it gives examples of the Leicester and Cambridge scores (which can interrogate clinical record systems) or the Diabetes Risk Score (which the patient completes with data such as height, weight, waist circumference and family history). SIGN suggests QDIABETES or the Diabetes UK Know Your Risk score. Below, we summarise the NICE guidance, which broadly aligns with SIGN. 

The ‘eyeball’ test and BMI in minority ethnic groups

Are we objective when offering screening? Or do we subjectively administer the ‘eyeball’ test (i.e. ‘guess the BMI’) then decide if they warrant screening? It turns out that the ‘eyeball’ test is not particularly accurate!

Most GPs tend to underestimate people’s BMI. We know this from a study that showed GPs photos of people and asked them to classify them as underweight, healthy weight or overweight (BJGP 2014;64:e703). Of course, remote consulting is making this even worse because we don’t even get the visual cue that weight may be an issue!

Another problem with remote consulting is that we may forget about the differential risk of BMI in different ethnicities.

A massive study from the UK CPRD compared the diabetes risk for different ethnicities (Lancet Diabetes Endocrinol 2021;9:419-26). Taking the risk of developing diabetes in a White person with a BMI of 30, the equivalent risk occurred at much lower BMIs in other ethnicities:

Ethnic groups (self-classified)  Compared with a White person with a BMI of 30, the equivalent risk of developing diabetes over the next 6.5 years occurred at these BMIs
South Asian 23.9
Arab 26.6
Chinese 26.9
Black 28.1
BMI cut-offs for overweight and obesity in the UK (SIGN 2025, 172)
Ethnicity  BMI lower cut-off for ‘overweight’ BMI lower cut-off for obesity
White 25.0 30.0
South Asian 19.2 23.3
Black or African-Caribbean 23.4 25.9
Arab 22.1 26.6
Chinese 22.2 24.6

What does this mean in practice?

  • ‘Eyeballing’ is not a good enough way to work out whether people are overweight (and it REALLY doesn’t work when remote consulting!).
  • Ethnicity affects risk, and some ethnicities have a much higher risk of diabetes at normal, or just above normal, BMIs.
  • This is one of the reasons that the NICE guidance says:
    • Those of certain ethnicities should be offered a diabetes risk assessment from 25y, even in the absence of obvious risk factors.
    • Consider a screening HbA1c in those of South Asian or Chinese ethnicity without doing a risk assessment first if age ≥25y and BMI >23.

Screening for and preventing type 2 diabetes 

  • A systematic approach to identify those at risk of type 2 diabetes is advocated by NICE and SIGN. This involves identification and risk stratification to identify highest-risk individuals, who are then offered blood test screening.
  • For those at increased risk of diabetes, lifestyle modification with weight loss is key (for those with raised BMI).
  • The evidence for this approach shows a reduction in progression to diabetes, a reduction in macrovascular outcomes and possibly a reduction in microvascular outcomes.
What do you and your practice currently offer to those at high risk of diabetes? How might this change in light of the NICE guidance and the drive to detect more diabetes?  

Useful resources:

Websites (all resources are hyperlinked for ease of use in Red Whale Knowledge)