The Health and Care of People with Learning Disabilities dataset: Diabetes

Note: if you want to skip the graphs, go straight to the conclusions for a quick summary...

We've known for a long time about the health inequities experienced by people with learning disabilities. We know these health inequities are big, resulting in much earlier typical ages of death compared to other people, and we know that these health inequities cut across a wide range of health conditions.

This blogpost will summarise some recent information about people with learning disabilities with a diagnosis of diabetes. 

This blogpost is the second of a series going into a lot of detail using information from the 'Health and Care of People with Learning Disabilities' dataset, compiled by NHS England. Every year (going April to March), this takes information from GP information systems in England to produce statistics about the health and health treatment of people registered with a GP as a person with learning disabilities, often comparing them to the health and healthcare of people without learning disabilities. It's a brilliant source of information that isn't used as much as it should be. The interactive information tool is quite hard to get your head around but it does make it quite flexible - for many things you can look at how things are by age, sex, region and even more locally than that, and for most things you can look back over at least the last five years. The most recent information available covers April 2024 to March 2025, with updates usually released every year in December.

Before we start, there are a few cautions to bear in mind about this dataset. The biggest one is that the information only covers about 6 in 10 people registered with a GP (54.6%) - not all the companies that own the electronic information systems used by GPs allow their data to be used (!). So coverage across different parts of the country is really patchy. The second caution is that there are likely to be many people with learning disabilities who are not registered as such with their GP. A third caution is that the extent to which this dataset includes people with learning disabilities in inpatient units really isn't clear. Finally, it's really important to remember that all this information is based on what health systems and GPs record, which is likely to have its own biases, particularly when it comes to people with learning disabilities.

Type 1 diabetes

Type 1 diabetes is an autoimmune condition where the pancreas stops producing insulin. The body attacks the healthy pancreas cells by mistake which causes it to stop producing insulin which the body needs to regulate blood sugar. Type 1 diabetes often runs in families and you are more likely to develop the condition if a close relative has the disease. Unlike Type 2 diabetes, there is no link between being overweight or inactive and developing the condition.

In 2024/25, 0.8% of people with learning disabilities had a diagnosis of Type 1 diabetes mellitus, compared to 0.4% of people without learning disabilities. These percentages haven't changed much from 2017/18 (0.7% of people with learning disabilities; 0.4% of people without learning disabilities). [I know the graph looks odd in terms of changes over time for different groups, but I've checked it multiple times and it's a quirk of rounding numbers up or down].

In terms of sex (the term used in the dataset), slightly more girls/women (0.8% in 2024/25) than boys/men (0.7% in 2024/25) with learning disabilities had a diagnosis of Type 1 diabetes, a similar difference (0.5% vs 0.4%) to people without learning disabilities. 

In terms of age bands, there are higher rates of Type 1 diabetes amongst people with learning disabilities at all age bands up to 64 compared to people without learning disabilities, peaking at 1.0% of people with learning disabilities aged 25-34. From the age of 65 onwards people with learning disabilities are equally or less likely than other people to have Type 1 diabetes, with no-one with learning disabilities aged 75+ reported as having Type 1 diabetes (compared to 0.3% of people without learning disabilities).

Type 2 diabetes

The Health and Care dataset has very similar information about diagnosis of what it describes as non-type 1 diabetes mellitus, which I am taking to mean Type 2 diabetes. Type 2 diabetes is a condition affecting the pancreas which leads to disruption of the hormone insulin and high blood sugars. Risk factors include being overweight, inactivity, having a family history of diabetes, or being from certain ethnic backgrounds. If left untreated, diabetes can lead to uncontrolled blood sugar, poor circulation and consequent damage to feet (including amputations) and eye problems.

The graph below shows rates of non-Type 1 diabetes recorded for people with and without learning disabilities over time, from 2017/18 to 2024/25. Overall, the percentage of people with learning disabilities diagnosed with non-Type 1 diabetes has increased over time, from 6.7% of people in 2017/18, to 8.0% of people in 2024/25. There has also been a less steep increase for people without learning disabilities (from 4.7% of people in 2017/18 to 5.5% of people in 2024/25). Rates of diagnosed non-Type 1 diabetes overall are higher for people with learning disabilities compared to other people, with the gap widening over time.

Amongst people with learning disabilities, girls/women are more likely to be diagnosed with non-Type 1 diabetes (9.0% in 2024/25) than boys/men (7.4% in 2024/25). This is the opposite pattern to people without learning disabilities, where girls/women are less likely to be diagnosed with non-Type 1 diabetes (4.9% in 2024/25) than boys/men (6.1% in 2024/25).

In terms of age bands, there are higher rates of non-Type 1 diabetes for people with learning disabilities overall at all ages from 10 onwards, with 8.5% of people with learning disabilities aged 35-44 having a diagnosis of non-Type 1 diabetes compared to 2.2% of other people, and 15% or more of people with learning disabilities having a diagnosis of non-Type 1 diabetes from the age of 45 onwards. While rates of non-Type 1 diabetes are higher for girls/women with learning disabilities than boys/men with learning disabilities up to the age of 44, this flips from the age of 45 onwards.

Measuring blood glucose

The Health and Care of People with Learning Disabilities dataset also records how many people with a diagnosis of diabetes mellitus (Type 1 and Type 2) have had their blood glucose measured in the last 12 months (HbA1c), as a way of checking how well people are generally doing in managing their diabetes. This is a basic measure of how well health services are regularly monitoring people with diabetes - it doesn't say anything about how well people are managing their diabetes or what help people are getting beyond regular monitoring.

The graph below shows what percentage of people have a diagnosis of diabetes and have had their blood glucose checked in the last 12 months, from 2017/18 to 2024/25. For people with learning disabilities in 2024/25, 8.1% of people had had their blood glucose checked - this is compared to the 8.8% of people with learning disabilities in 2024/25 with a diagnosis of either Type 1 or Type 2 diabetes. So most people with learning disabilities and diabetes, but by no means everyone, had had their blood glucose checked in the last 12 months. For people without learning disabilities in 2024/25, 5.4% of people had had a blood glucose test in the last 12 months, out of the 5.9% of people with a diagnosis of Type 1 or Type 2 diagnosis. 

For people with and without learning disabilities, the percentage of people getting a blood glucose test has increased over time. For people with learning disabilities, this is an increase from 6.8% of people in 2017/18 to 8.1% of people in 2024/25; for other people this is an increase from 4.7% of people in 2017/18 to 5.4% of people in 2024/25. This has been a steady year-on-year increase, except for a dip during the Covid-19 pandemic in 2020/21.

But as we've seen earlier on, the percentage of people with learning disabilities with a diagnosis of diabetes has also increased over time, and at a faster rate than for people without learning disabilities. This means that the gap between people with a diagnosis of diabetes and those with a blood glucose test has increased slightly over time for people with learning disabilities (the gap was 0.6% of people in 2017/18 compared to 0.7% of people in 2024/25). 

As we saw earlier, in relation to diabetes diagnoses, more girls/women with learning disabilities than boys/men in 2024/25 had had their blood glucose checked in the last 12 months (9.1% vs 7.5%). This is the opposite way round for girls/women vs boys/men without learning disabilities (4.8% vs 5.9% in 2024/25).

In terms of age bands, the age pattern of blood glucose checks closely mirrors the age pattern of diabetes diagnoses (particularly for the more common non-Type 1 diabetes) that we saw earlier, being much more common for people with learning disabilities at virtually all ages.

Taking into account differences in the populations of people with and without learning disabilities in things like age and sex, in 2024/25 a diagnosis of Type 1 diabetes was 1.6 times as common and a diagnosis of non-Type 1 diabetes 1.9 times as common among people with learning disabilities than you would expect. 

Conclusions

Both Type 1 and Type 2 diabetes are more commonly diagnosed amongst people with learning disabilities compared to other people, and are increasing faster over time. In 2024/25, around 1 in 11 people with learning disabilities had some form of diabetes diagnosis, compared to 1 in 17 people without learning disabilities.

These differences are biggest in almost all adult age groups.

Girls/women with learning disabilities are more likely than boys/men with learning disabilities to have a diagnosis of Type 2 diabetes and to have blood glucose checks - this is the opposite way round to the population of people without learning disabilities.

This dataset doesn't tell us anything about ethnicity and diabetes amongst people with learning disabilities, which we know is an important issue among people without learning disabilities.

Most (but not all) people with learning disabilities with a diagnosis of diabetes are getting a regular blood glucose check - this is increasing over time, but is not quite keeping pace with the increasing numbers of people with learning disabilities with a diagnosis of diabetes.

There are also a few things I want to say about this that will come up time and time again (if I ever get round to other blogposts about this dataset)...

1) I think this info shows the value of the huge effort involved in setting up regular health checks for people with learning disabilities, as long as they are done well. I suspect that annual health checks have been crucial in spotting diabetes (especially Type 2 diabetes) and in regularly monitoring people's diabetes, particularly among younger adults with learning disabilities.

2) However, while this monitoring clearly has important benefits for aspects of people's health, monitoring on its own is not enough. The evidence for weight management or physical activity programmes working well for people with learning disabilities is not brilliant, and we are not paying any attention to the 'causes of the causes' (in Sir Michael Marmot's phrase) like poverty and where people are living. 

3) I think the information about age is really important and is often overlooked. For diabetes (and for a whole host of other health issues) we see a clear pattern of health problems starting much earlier for more people with learning disabilities. Which is unlikely to be where GPs and health services are generally expecting to see, or be screening for, these health problems. It's also the case that among people without learning disabilities, many health conditions are most common at much older ages. Bluntly, many people with learning disabilities don't live that long, so overall figures across all age groups tend to minimise health inequities between people with and without learning disabilities.

4) Relating to the point just above, it's important to remember that although percentages of health issues like diabetes are more common amongst people with learning disabilities, in terms of absolute numbers of people they can be swamped by the numbers of people without learning disabilities. The Health and Care of People with Learning Disabilities dataset covers 55% of people in GP practices in England - scaling up figures from the dataset would mean there are just over 28,000 people with learning disabilities diagnosed with Type 2 diabetes in England, compared to 3.5 million people without learning disabilities. Easy to ignore in mainstream health services, or an easily manageable number of people for health services to pay attention to and make a big difference?

The Health and Care of People with Learning Disabilities dataset - how many people are taking mental health medications?

Note: if you want to skip the graphs, go straight to the conclusions for a quick summary...

For over 30 years, it's been clear that huge numbers of people with learning disabilities in England (and in other places too) are prescribed (told by their doctor to take) what are sometimes called 'psychotropic' drugs - drugs that are prescribed to change something about a person's mental state. Typically, these are drugs prescribed for reasons of 'mental health', although they might also be prescribed to change a person's behaviour. Recent research led by Sue Caton at Manchester Metropolitan University with co-researchers with learning disabilities has shown that people with learning disabilities are often not involved in decisions about what mental health medications they're taking (this link goes to a toolkit produced by the project that might help people get more involved in these decisions).

In England, NHS England and other organisations have been running the STOMP programme since 2016. STOMP stands for 'Stopping over medication of people with a learning disability, autism or both'. More recently STAMP ('Supporting Treatment and Appropriate Medication in Paediatrics) has been added by NHS England. Part of the aim of both these programmes is to reduce the number of people with learning disabilities and autistic people who are prescribed these drugs.

This blogpost will summarise some recent information about the percentages of people with learning disabilities being prescribed a range of mental health drugs. 

This blogpost will be the first of a series going into a lot of detail using information from the 'Health and Care of People with Learning Disabilities' dataset, compiled by NHS England. Every year (going April to March), this takes information from GP information systems in England to produce statistics about the health and health treatment of people registered with a GP as a person with learning disabilities, often comparing them to the health and healthcare of people without learning disabilities. It's a brilliant source of information that isn't used as much as it should be. The interactive information tool is quite hard to get your head around but it does make it quite flexible - for many things you can look at how things are by age, sex, region and even more locally than that, and for most things you can look back over at least the last five years. The most recent information available covers April 2024 to March 2025, with updates usually released every year in December.

There are a few reasons that I want to go into mind-numbing detail this time around (it's not ONLY that I like graphs). First, I'm not going to be doing these blogposts for ever, so I want to make the ones I do as full of information as possible for completeness. Second, I want to show what these kinds of datasets can do, in the hope that people will keep using them. Third, a recent 'consultation' has said that this is the last year that this dataset in this format will be produced. It will be replaced by something that might have more information and will cover all GP practices (a very welcome development), but will not have any way of directly comparing people with learning disabilities to people without learning disabilities (which is such a useful feature of the dataset as it stands).

Before we start, there are a few cautions to bear in mind about this dataset. The biggest one is that the information only covers about 6 in 10 people registered with a GP (54.6%) - not all the companies that own the electronic information systems used by GPs allow their data to be used (!). So coverage across different parts of the country is really patchy. The second caution is that there are likely to be many people with learning disabilities who are not registered as such with their GP. A third caution is that the extent to which this dataset includes people with learning disabilities in inpatient units really isn't clear. Finally, it's really important to remember that all this information is based on what health systems and GPs record, which is likely to have its own biases, particularly when it comes to people with learning disabilities.

'Severe mental illness' and antipsychotics

Antipsychotics are a set of drugs that are supposed to be prescribed for people with mental health issues with labels such as psychosis, which in GP-record world (along with some other mental health issues) are classed as 'severe mental illness'.

In 2024/25, 7.6% of people with learning disabilities were diagnosed with a 'severe mental illness', compared to 1.0% of people without learning disabilities. A clutch of research projects have shown that mental health issues such as psychosis might be more common amongst people with learning disabilities, but nowhere near to the extent that GPs are diagnosing severe mental illness, with the suspicion that some people with learning disabilities are getting this label to justify them being prescribed antipsychotic drugs for reasons essentially of attempted behavioural control. These percentages haven't changed much from 2017/18 (7.6% of people with learning disabilities; 0.9% of people without learning disabilities).

In terms of sex (the term used in the dataset), more girls/women (8.0% in 2024/25) than boys/men (7.4% in 2024/25) with learning disabilities had a severe mental illness diagnosis, while there was no difference amongst people without learning disabilities. 

In terms of age bands, substantially more people with learning disabilities are diagnosed with severe mental illness from an early age (e.g. 10% or more from the age of 35-44 onwards) than people without learning disabilities, peaking at 16.3% of people with learning disabilities aged 65-74.

Antipsychotics. Overall, in 2024/25 13.8% of people with learning disabilities were being prescribed antipsychotics, compared to 0.9% of people without learning disabilities. For people with learning disabilities, this was a reduction from 15.5% of people being prescribed antipsychotics in 2017/18 (there was no change over time for people without learning disabilities). 

Fewer girls/women with learning disabilities were being prescribed antipsychotics (12.9% in 2024/25) than boys/men with learning disabilities (14.3% in 2024/25). 

In terms of age bands, the percentage of people with learning disabilities prescribed antipsychotics increases sharply from the age of 10 - from the ages of 25-34 years onwards, more than 10% of all people with learning disabilities in 2024/25 were being prescribed antipsychotics, peaking at 23.5% of people aged 65-74.

From 2020/21 onwards, the Health and Care dataset has added information on people being prescribed antipsychotics when they don't have a diagnosis of severe mental illness or palliative care needs. In 2024/25, 8.2% of people with learning disabilities (6.9% of girls/women; 9.1% of boys/men) without a severe mental illness diagnosis were still getting prescribed antipsychotics, compared to 0.4% of people without learning disabilities or a severe mental illness diagnosis. The percentage of people with learning disabilities prescribed antipsychotics in the absence of a diagnosis of severe mental illness has reduced over time, from 9.3% of people in 2020/21 to 8.2% in 2024/25.

These prescriptions in the absence of a severe mental illness diagnosis also rise steeply for people with learning disabilities from an early age - 10% of adults with learning disabilities without a diagnosis of severe mental illness are prescribed antipsychotics by the age of 25-34.

Taking into account differences in the populations of people with and without learning disabilities in things like age and sex, in 2024/25 being prescribed antipsychotics was 14 times more common among people with learning disabilities than you would expect from general population prescribing rates; for those without a diagnosis of severe mental illness this rose to being 18 times more common.

Depression and anti-depressants

The Health and Care dataset has very similar information about GP diagnosis of depression and the prescribing of anti-depressants.

Overall, the percentage of people with and without learning disabilities diagnosed with depression has increased over time, for people with learning disabilities (15.9% in 2024/25, up from 14.6% in 2020/21) and for people without learning disabilities (14.7% in 2024/25, up from 13.6% in 2020/21). Rates of diagnosed depression overall are higher for people with learning disabilities.

Amongst people with learning disabilities, girls/women were more likely to be diagnosed with depression (20.0% in 2024/25) than boys/men (13.2% in 2024/25). In terms of age bands, from the ages of 25-34 years onwards, more than 15% of people with learning disabilities had a diagnosis of depression, peaking at 25.5% of people aged 55-64.

Anti-depressants. Overall, in 2024/25, 22.4% of people with learning disabilities were being prescribed anti-depressants, compared to 11% of people without learning disabilities. Girls/women with learning disabilities were more likely to be prescribed anti-depressants (27.2% in 2024/25, compared to 14.6% of girls/women without learning disabilities) than boys/men with learning disabilities (19.4% in 2024/25, compared to 7.5% of boys/men without learning disabilities). 

In terms of age bands, from the ages of 18-24 years onwards, more than 15% of people with learning disabilities were being prescribed anti-depressants, peaking at 34% of people aged 55-64.

The Health and Care dataset has also been reporting on people being prescribed anti-depressants when they don't have a diagnosis of depression. In 2024/25, 12.4% of people with learning disabilities without a diagnosis of depression were being prescribed anti-depressants, compared to 4.4% of people without learning disabilities or a diagnosis of depression. For both groups, this had increased slightly since 2020/21 (11.6% of people with learning disabilities without a diagnosis of depression; 4.3% of people without learning disabilities or a diagnosis of depression). Again, this was more common for women with learning disabilities without a diagnosis of depression (13.7% in 2024/25) than men (11.5%). 

In terms of age bands, from the ages of 18-24 years onwards, more than 10% of people with learning disabilities but with no diagnosis of depression were being prescribed anti-depressants, peaking at 17.3% of people aged 65-74.

Taking into account differences in the populations of people with and without learning disabilities in things like age and sex, in 2024/25 being prescribed antidepressants was twice as common among people with learning disabilities than you would expect from general population prescribing rates; for those without a diagnosis of depression this rose to being 3 times more common.

Benzodiazepenes

Benzodiazepenes are a type of 'sedative' - they slow down the body and mind and are prescribed for acute anxiety or sleep problems, and it is recommended that they are not prescribed for more than four weeks at a time. The percentage of people with and without learning disabilities being prescribed benzodiazepenes is decreasing over time. In 2024/15, 6.1% of people with learning disabilities were being prescribed benzodiazepenes (down from 7.6% in 2017/18) compared to 1.4% of people without learning disabilities in 2024/25 (down from 2.4% in 2017/18). 

Amongst people with learning disabilities, in 2024/25 similar percentages of women with learning disabilities (6.4%) and men with learning disabilities (6%) were being prescribed benzodiazepenes.

In terms of age bands, more than 5% of people with learning disabilities from the age of 25 onwards were being prescribed benzodiazepenes.

Taking into account differences in the populations of people with and without learning disabilities in things like age and sex, in 2024/25 being prescribed benzodiazepenes was 5 times as common among people with learning disabilities than you would expect from general population prescribing rates. 

Conclusions

Many more people with learning disabilities than other people continue to be prescribed mental health medications, often without a relevant mental health diagnosis. Nearly 1 in 4 of all people with learning disabilities registered with GPs are on antidepressants, and 1 in 7 people with learning disabilities are on antipsychotics. 

While STOMP and STAMP initiatives might be having some impact on reducing the number of people being prescribed antipsychotics and benzodiazepenes, the increasing numbers of people being prescribed antidepressants is threatening to cancel out any of these gains. 

People with learning disabilities, compared to people without learning disabilities, are particularly likely to be prescribed mental health medications in the absence of a matching mental health diagnosis. This isn't news, but it doesn't stop it being a scandal.

There are also some big gender differences amongst people with learning disabilities, with little sign of these gaps closing over time. Women with learning disabilities are more likely to be prescribed antidepressants and than men with learning disabilities,  and it's the other way round for antipsychotics.

Significant numbers of people with learning disabilities are being prescribed these drugs at much earlier ages than people without learning disabilities (often in relatively early adulthood) for potentially longer periods of time. This is really worrying, given what we know about the side effects of many of these drugs over time, their addictive properties, and their potential impact on how they make people feel.

From the Health and Care dataset. we don't know how many people with learning disabilities are being prescribed more than one mental health drug, although a research project in Jersey suggested that almost a quarter (23%) of people with learning disabilities were being prescribed two or more mental health drugs. 

We also don't know any more how many people with learning disabilities in inpatient units are being prescribed these types of medications. 

The reductions in antipsychotic prescriptions show what can be achieved with concerted, consistent, targeted effort over many years, but also just how big the gap continues to be between people with and without learning disabilities. The increases in antidepressant prescriptions show what happens in the absence of such effort - for all the government talk about reducing health inequalities, the general way that societies and health services operate is taking existing horrendous inequities and making them worse.