A systematic review of pharmacological and non-pharmacological interventions for improving diabetes outcomes in people with severe mental illness

Yorkshire & HumberGeneric Health Relevance
Start Date: 1 Mar 2014

Project summary

Compared to the general population, people with severe mental illness (SMI) are more likely to experience poor physical health, with increased prevalence of cardiovascular disease, diabetes, stroke, asthma and some cancers. Mortality rates are also significantly higher in the SMI population, with life expectancy estimated to be around 15 years less than for the general population. Although cardiovascular disease accounts for the largest proportion of deaths caused by physical illness, increased mortality can in part be explained by higher prevalence of type 2 diabetes in people with SMI, a co-morbid relationship that carries a three- to four-fold increased risk of death than for the general population.

Diabetes is more than twice as prevalent among people with SMI, and compared to the general diabetic population, is associated with poorer outcomes. The reasons for this are not well understood, but are related to a combination of features of the mental illness, metabolic side effects of psychotropic medication, lifestyle factors, and the organisation of health services. Wider socio-economic inequalities facing people with SMI may also increase the risk of developing diabetes, and the multi-factorial nature of risk can make it difficult to prevent and manage diabetes in this population. Despite the increased risk and poor outcomes associated with diabetes, the physical health needs of people with SMI have long been overlooked, in part due to poor assessment, monitoring and recording practices, but also because of diagnostic overshadowing, poor co-ordination between primary and secondary care, a lack of evidence and clarity about who should manage physical health needs, and barriers to accessing and receiving appropriate care and interventions.

Primary outcomes

In people without diabetes, the study must measure one of the following outcomes:

Incidence of diabetes (diagnosis should have been established using the standard criteria valid at the time of the trial, for example ADA 1999; ADA 2008; WHO 1998)

Glycaemic control measured via HbA1c or fasting blood glucose

In people with diabetes, the study must measure at least one of the following outcomes:

Glycaemic control measured via HbA1c or fasting blood glucose


Body Mass Index

Diabetic complications (which include: cardiovascular disease (myocardial infarction and angina), renal failure, micro albuminuria, amputations, diabetic eye disease, diabetic neuropathy, and stroke)

Secondary outcomes

Particular outcomes of interest include:- blood pressure; lipid profile; hypoglycaemia; medication adherence; physical activity; diet; smoking; A&E attendance; hospital admissions (non-mental health); mental health admissions; healthcare costs; mortality; self-management; self-efficacy; quality of life; psychological symptoms (for example depression and anxiety; positive and negative symptoms); adverse events of the intervention.

Dr Najma Siddiqi