Diabetic retinopathy screening & prevention
in children and young people

David Kinshuck



When does retinopathy develop in children?

Incidence in UK Eye 16 .  Natural history Eye 16

In developed countries it is unusual for children to develop retinopathy, but it is not at all uncommon in poorer countries. Thus Porta's group reported it was unusual in Italy, and linked to the duration of diabetes, but Betts's group found that 12% of Russian diabetic children had retinopathy. I have heard of children with severe proliferative retinopathy in Moscow, age 10y. See

Retinopathy has even been reported after 1 months duration of diagnosed diabetes, in a 16y old patient, by Hamilton. But this is unusual. Whilst there are genetic factors, it is well recognised that an improvement of diabetic control can precipitate retinopathy (discussed here, identified in DCCT papers and here, very well recognised in pregnancy). This may be the mechanism in Hamilton's patient...although this patient may have also had their pubertal growth spurt which will also stimulate retinopathy.

Similarly accelerated retinopathy development is not uncommon after 12 months diabetes, if control has improved a lot, as here.  

But as Porta reported retinopathy is most unusual in children whose control has been reasonable without much glucose fluctuation. Generally screening is not needed <13y.


How should retinpathy screening be carried out?

As a rule at 12y-16y many children can be refered to adult optometrist screening program, and this should be the case if parents and child are happy with this and are reliable.  Younger children or those not attending the screening program, should be examined by their diabetologist (or expert ophthalmologist) yearly if possible. This is not only to pick up the occasional case with retinopathy, but to get the child and family into the pattern of having yearly eye examinations.

Until recently nearly all type 1 insulin dependant patients developed retinopathy, so yearly examinations are critical after 16y. Strictly speaking examinations every 2 years may suffice if there is no retinopathy whatsoever, but this makes it harder to create a pattern and riskier if a mistake is made.

Examine carefully if there has been >8y of very poor diabetic control. Consider other factors that influence retinopathy ...such as wildly fluctuating control, high blood pressure, or sickle trait or equivalent...examinations are more critical.

Consider genetic factors...there are genes that influence retinopathy, and some of these will work by influencing blood pressure. For example, if there is a family history of hypertension that child will be at much greater risk of developing complications. Similarly older sibs may develop retinopathy.

Expect retinopathy with 14y of diabetes with mediocre or poor control. Logically, retinopathy is uncommon in children not simply because they are children, but probably more because they have not been diabetic long enough. Take a brief history ...are there black spots or other problems may be due to bleeding?


Are you an expert in detecting retinopathy?

Studies have shown that retinal photography through a dilated pupil combined with slit lamp stereo-microscopy is the most accurate screening method, and this will soon be available in Birmingham. At present in the city, selected optometrists who are accredited in diabetic screening use a slit lamp, generally with 85% accuracy, and such screening is ideal, and more accurate than many ophthalmologists.

If your patient cannot be sent to the optometry program, then they should be examined in clinic. As there may be some compliance issues in such patients, this should be on the same day they attend for their regular check up.

What should you look for in terms of retinopathy?

Early new vessels are difficult to photograph and see, so look for clues. There should be no haemorrhages or dilated or tortuous vessels. Any abnormal squiggly vessels, even if very narrow and smaller than regular vessels, are usually IRMAs...that is intra-retinal microvascular abnormalities....and hence part of pre-proliferative retinopathy or proliferative. Any microaneurysms in a child are sign of a serious problem, and such a patient should be examined very carefully, usually by the optometry program or ophthalmologist expert in retinopathy.

Care is needed...even expert ophthalmologists and cameras miss early new vessels at times. Once again, any retinopathy in a child would be very worrying, even one microaneurysm or retinal haemorrhage. This would be a sign of that proliferation may follow, and examinations should then be carried out every 6 months. As mentioned retinopathy can develop over a month...although usually it is much more gradual.


Preventing retinopathy

This author is not happy with a twice daily insulin mixture regime. This is covered here. This author, because they cause much less retinopathy, prefers insulin pumps (CSII) or if that is not possible or appropriate an MDI (multiple dose insulin or basal bolus regime) using glargine or detemir as the long acting insulin, with rapid acting insulin before meals. I accept the published evidence for this preference is limited, but I have also been advised by experts. Pumps can be used with children, even with learning difficulties, but they may be problems, as below (some discussed here).

Children/parents should not be offered a free choice.....they should be offered the best possible/best in practice regime, and only be offered twice daily regime if there are really major problems with a more effective regime.


Insulin pumps & good control

Pumps (CSII) do lower HbA1c. Linkeschova reported an improved quality of life with fewer severe hypos and an HbA1c 0.5% lower (17.5 % less retinopathy). Hoogma's team 0.3% lower (12% less retinopathy) with no worse a quality of life. Doyle's short term study (age 8-21y) noticed a 0.9% improvement (34% less retinopathy).....and this was compared to an MDI glargine/rapid acting insulin regime. There was even a 1.8% improvement (67% less retinopathy) in adult patients (I believe) in Brazil. Low HbA1c help to prevent complications, even in children, see editorial and Amin.


Contact between health professionals and patients

As an ophthalmologist I will see patients who the system has failed or whose diabetes was aggresssive, who develop retinopathy. Looking though the literature it is clearly that regular contact between the diabetic team and professional and patient is critical. In the UK until recently there were were long gaps between consultations.

Even now, (with adult patients) problems are identified, and patients are not followed up but told to 'contact me' if problems persist...in practice even many adult patients do not follow up on such issues, which then continue for years on end. This is partly a resource issue...but it is partly because team members are spending so long with patients with complications that patients without are not followed up as frequently.

Targets may not be considered logically...the able child, with a wonderful family, no problems at school, no emotional problems, using glargine/radid acting basal-bolus regime, could be expected to maintain an HbA1c of ~6.5. Such a child may be 7.0 without a pump, 6.5% with a pump, This difference is equivalent to an 18% lower risk of complications. (As soon as pumps monitor glucose levels all children will need one...do we need to get our children ready?). But if that same child was using a twice daily insulin mixture, 7.5% might be the best achievable.

In practice many children canot achieve such low levels, and average levels in clinics are often 8.9%, but levels are dropping. See. There is a seasonal variation, and this can lead to more hypos  in April. Similarly, there were some clinics in Eastern European clinics with very little retinopathy in their patients, even using older insulins. Such clinics had plenty of staff and organised regular contact between patients and professional. Similarly in the UK I have attended meetings when patients were presented 'this bad control was for 6 months..patients should really be seen each month, and problems addressed right away.

Many problems can be prevented by having regular contact. A problem develops one month, and is reviewed the next, and if worse further action is needed, until improved. In this way the problem is usually sorted after 3 months rather that persisting for 9 months. Even now many adult patients will have a period or poor control for 6 months-2 years when DKA or hypos are likely to occur, before their problem is sorted. Generally the quality of care is improving year by year, as it must if retinopathy and renal failure is to be prevented.

At a recent dicussion at Good Hope we heard that in clinics with children and young people with well controlled diabetes, patients see their doctor and their nurse and their dietician each month, each for one hour. This contact simply does not happen in most UK clinics.

We know the high risk groups  ..80% of complications occur among the 20% of children who had recurrent events.

Some girls (and boys) will have an eating disorder, and use lack of insulin to slim...they might need psychotherapy, a healthy high fibre diet, and a lot more exercise...the whole family might need support which will include dietary advice and home visits by an expert dietician (see recent TV programs)...is this done? A low HbA1c may not be possible at such a stage....the priority is to avoid DKA and hypos. Any obesity will be due to insufficient exercise or the wrong/too much food, rather than too much insulin. Girls need proper family planning advice.....pregnancy is a really big problem, not least because it can precipitate retinopathy.

Similarly smoking, binge drinking are issues....are there any techniques that you can use to increase the child's self esteem...the higher the self-esteem, the better the diabetic control and so on?Young people or children with poor control have behaviour problems and lower social competence, higher levels of family conflict, and their parents reported lower levels of family cohesion, expressiveness and organization (see also).  We desperately need to improve the care for this group...is there anyway we can do this effectively? The current literature suggests we have a long way to go.

I have been advised that Primary Care Trusts are aiming to provide good quality care to at least 80% of their patients. But as we all know preventing diabetic complications is a 'total quality' issue, and any break in the chain will mean that the other 20% will develop complications. The diabetic care has to identify and treat the missing 20%, or at least all the patients who are able to comply.


This author

This author is not an expert in this field specifically, but has reviewed published online literature to prepare this webpage.