The increased mortality of people with a severe mental illness
The findings of the audit








In March 2018, the National Diabetes Audit1 published a sub-audit of the diabetic care received by people with a severe mental illness (SMI) living in the community. The National Diabetes Audit (NDA) is a long-standing, high quality, audit of diabetes care carried out annually by NHS Digital.  The audit incorporates both primary and secondary care data and covers more than 95% of the known population with diabetes. The sub-audit of people with a SMI were identified by their inclusion in the Mental Illness register—one of the Quality and Outcomes Framework (QOF) domains. The diagnoses for inclusion in the QOF register include schizophrenia, bi-polar disorder, and other psychoses. The sub-audit described the care that this group received and compared it to care received by people without a SMI, using NICE guidelines for diabetes as the gold standard.


The increased mortality of people with a severe mental illness

People with a SMI die 15–20 years earlier than they would have done, had they not had a SMI.2 The principle cause of this increased mortality is from long-term conditions such as diabetes, cardiovascular disease, respiratory disease and cancer.  Over the last 30 years, health outcomes across the general population for people with long-term conditions have improved, but for those with a SMI, it has remained unchanged; the gap between health outcomes for those with, and those without, a SMI is increasing.3

The underlying causes of this increased mortality is complex. Anti-psychotic medication is without doubt a factor, as it is obesogenic, diabetogenic4 and thrombophilic.5 However, anti-psychotic medication is not the only cause of the increased mortality and morbidity. Sir Henry Maudsley noted in the 1890s that diabetes was more common amongst people with dementia praecox—as schizophrenia was then known. This was some 30 years before anti-psychotic medication was invented. Evidence has now shown that there is an apparent genetic link between diabetes and schizophrenia.6

It is a feature of SMI that sufferers are more likely to be unemployed, need state benefits, and have poor accommodation.7 All these social circumstances make managing a long-term condition more difficult. In the same way, the chaotic nature of these disorders makes adhering to complex drug regimens and attending healthcare appointments (primary or secondary care) more difficult.

Around 45% of people with a SMI living in the community smoke, and this figure rises to 70% for those who are long-term residents in mental health trusts.8 In the general population less than 20% smoke, and this figure is falling. Smoking exacerbates all the long-term conditions.

There is one final factor that contributes to the physical health problems experienced by this group—diagnostic overshadowing. This is the term used to describe how clinicians of all types, when faced with a physical health symptom in a person with a SMI, ascribe it to the mental health problem and not to a possible physical health cause, as they would have done in somebody without a mental health problem.

The consequence of these various factors, which are complex and inter-related, is that diabetes (and other long-term conditions) are 2–3 times more common amongst people with a SMI, and contributes to the increased mortality and morbidity experienced by this group.

Which is why the National Diabetes Audit of diabetic care received by this group is so important.  This is the first audit, anywhere in the world, of its type. Other surveys and audits have looked at the diabetic care received by some people with a SMI, but there has been no national audit of diabetic care—so the results are significant.


The findings of the audit

The audit shows that diabetes, type 2 diabetes, is twice as common amongst those with a SMI. There are some 60,000 people with both a SMI and type 2 diabetes. Type 1 diabetes occurs at much the same rate in people with a SMI as those without—nationally there are about 2,500 people with type 1 diabetes and SMI. What the audit also shows, is that the age distribution of these conditions is quite different to the general population. Type 2 diabetes is more likely to be found in younger people with a SMI, than the general population, whereas type 1 diabetes is more likely to be found in an older group with SMI than the general population (tables 1 and 2). Type 2 diabetes is also a little more common amongst women than men with SMI.

Deprivation does have a significant effect on diabetes and SMI. Diabetes and SMI is much more common in those living in the most deprived areas of the country, compared to the least deprived areas (table 3). Diabetes and SMI is also much more common in deprived areas than diabetes alone.




The NICE guidelines describe the care of diabetes in two distinct groups—the process of care, and the treatment targets for diabetes. Processes of care are the nine interventions that all diabetics should be offered: measure HbA1c, blood pressure, serum cholesterol, serum creatinine, urine albumin/creatinine ration, undertake a foot risk surveillance, measure the body mass index (BMI) and record the smoking history. The ninth intervention—diabetic retinopathy screening —is carried out by Public Health England, as one of the national screening programmes, and therefore is not included in this audit. The audit demonstrates that people with type 1 diabetes and SMI receive the same processes of care as those with type 1 diabetes in the general population—in fact maybe even slightly greater care. So far as type 2 diabetes is concerned, people with SMI receive significantly less processes of care than those in the general population. The audit examined some factors that may influence people receiving all eight processes of care.  Deprivation seemed not to play a role receiving processes of care—there was no obvious difference between the most deprived quintile and the least deprived quintile for both type 1 diabetes and type 2 diabetes. Smokers, however, were less likely to receive all eight processes of care.



NICE guidelines describe the achievement of treatment targets; the achievement of blood pressure control, and blood cholesterol and glucose control. The audit reports on the achievement of all three of these outcomes. It shows that there is no significant difference between those with SMI and those in the general population achieving the NICE-specified treatment targets. Deprivation also appears to have little or no association. Smoking (predictably) does demonstrate a slight effect, in that smokers are less likely to achieve all the treatment targets.




The NDA audit of people with a SMI is the first of its type in the world. It is one of the largest audits of its kind, and the data needs to be considered carefully. But, the findings raise more questions than answers. Other evidence from around the world described generally worse control and outcomes of diabetes in this group—but that does not seem to have been demonstrated by this audit. The NDA will be reporting next year on the complications of diabetes in people with SMI—which is not included in this audit. It may be that this will throw more light on the subject.

The group that the NDA did not include in their audit, are those who are long-term residents of mental health trusts. Every acute hospital, and every general practice contributes data to the NDA. The only group that does not contribute data are mental health trusts. There are some 10,000 people who are long-term residents of mental health trusts—those receiving rehabilitation or those that are detained under the Mental Health Act. Evidence suggests that up to 30% of these people will have diabetes,9 and that in most cases diabetic care is provided by training psychiatrists, rather than experienced GPs or diabetologists.  

More people with a SMI smoke,8 and they tend to smoke more tobacco, than those without a SMI.  Smoking induces liver enzymes, which metabolises anti-psychotic medication, so that to achieve the same clinical effect a higher dose is needed in a smoker than a non-smoker.10

The NDA shows that those who smoke have less good treatment target achievements, than those who don’t smoke. However, the audit doesn’t provide quantitative information on the amount of tobacco smoked, nor the type of antipsychotic medication used by patients. This information would help to understand the complex interrelationship between lifestyle, medication, long-term conditions and SMI, and the effect that these have on cardiovascular risk.

In June 2016, an update to QRISK11 was published that added new risk factors to the algorithm used for calculating cardiovascular risk and the risk of acquiring diabetes. These additional risk factors included whether or not the patient was taking anti-psychotic medication or had a SMI. QRISK3 and QDIABETES are now the only risk algorithms in the world that include SMI and antipsychotic medication—which linked to the NDA starts to provide some unique and valuable information, as to the causes of the increased mortality and morbidity experienced by this group. Taken together, the QRISK3, QDIABETES, and the NDA results, give individual practices the opportunity to innovate new ways to address the needs of this group of high cardiovascular risk patients.

The NDA has provided a unique set of data; a set of data that challenges some of the accepted theories and ideas about this group of “hard to reach” people.  It certainly raises research questions, about the quality of diabetic care that the SMI receive. The NDA raises a number of questions about service innovation and development—the role of smoking, the inclusion of mental health trusts in the NDA as examples.

But there are broader questions that the NDA raises—are there other national audits that could replicate the work of the NDA by looking at the care of the SMI? There is a national Stroke audit, and a national respiratory disease audit—both of these could provide valuable information as to how to improve care. Diagnostic overshadowing, and the level of knowledge of psychiatrists about the management of long-term conditions are significant training challenges. The same could be said about physicians and GPs about their knowledge of severe mental illness.  

In 2016, the Academy of Medical Royal Colleges produced a report12 detailing the actions needed to be taken by the NHS and its partner Arm’s Length Bodies to improve the physical health outcomes for people with severe mental illness. Recommendations made in the report covered training, data reporting and national audits, access to national screening programmes, registration with GPs, commissioning advice, etc. In September 2018, at the Royal College of Nursing, a national learning network and collaborative was formally launched to implement the recommendations of that report.  More information on this network can be found at

The website also holds a number of different documents and resources that may be of interest including the report from the NDA, and a set of fact sheets and case scenarios from WONCA—the World Association of Family Doctors—providing advice on the physical health care of people with a severe mental illness.


Alan Cohen,  Institute for General, Family and Preventive Medicine; Paracelsus Medical Private University, Salzburg, Austria.

Conflict of interest: none declared



1. National Diabetes Audit.  (

2. Mental Health Taskforce I. THE FIVE YEAR FORWARD VIEW FOR MENTAL HEALTH. 2016 [cited 2017 Mar 13]; Available from:

3. Ashworth M, Schofield P, Das-Munshi J. Physical health in severe mental illness. Br J Gen Pract 2017; 67(663): 436–7

4. Leucht S, Cipriani A, Spineli L, et al. Comparative efficacy and tolerability of 15 antipsychotic drugs in schizophrenia: a multiple-treatments meta-analysis. Lancet 2013; 382(9896): 951–62

5. Parker C, Coupland C, Hippisley-Cox J. Antipsychotic drugs and risk of venous thromboembolism: nested case-control study. BMJ 2010; 341: c4245–c4245.

6. Jurczyk A, Nowosielska A, Przewozniak N, et al. Beyond the brain: disrupted in schizophrenia 1 regulates pancreatic β-cell function via glycogen synthase kinase-3β. FASEB J 2016; 30(2): 983–93

7. Cohen A. Addressing comorbidity between mental disorders and major noncommunicable diseases Addressing comorbidity between mental disorders and major noncommunicable diseases. 2017

8. Royal College of Physicians; Royal College of Psychiatrists. Smoking and mental health A joint report by the Royal College of Physicians and the Royal College of Psychiatrists. Smok Ment Heal 2013; 195–201

9. Puzzo I, Gable D, Cohen A. Using the National Diabetes Audit to improve the care of diabetes in secure hospital in-patient settings in the UK. J Forensic Psychiatry Psychol 2017

10. Desai HD, Seabolt J JM. Smoking in patients receiving psychotropic medications: a pharmacokinetic perspective. CNS Drugs 2001; 15(6): 469–94

11. Hippisley-Cox J, Coupland C, Brindle P. Development and validation of QRISK3 risk prediction algorithms to estimate future risk of cardiovascular disease: prospective cohort study. BMJ 2017; 357:j 2099

12. The Royal College of Psychiatrists. Improving the physical health of adults with severe mental illness: essential actions. 2016;1–81. Available from: