Research

Can psychiatric treatment change high-risk diabetes outcomes?

Diabetes psychiatry begins with a physical-health problem.

Recurrent DKA, persistent extreme hyperglycaemia, severe hypoglycaemia and dangerous interactions between eating behaviour and insulin have many possible causes.

In some people, psychological, psychiatric, cognitive, relational or social processes may help maintain these outcomes.

The research question is therefore not simply:

How common are mental-health problems in people with diabetes?

It is:

Can understanding and treating psychological and psychiatric processes change a high-risk diabetes trajectory?

Research principle

The purpose is not simply to demonstrate that mental health matters in diabetes.

The more useful questions are:

Which psychological and psychiatric processes are producing this particular physical-health outcome?

Can we identify them systematically?

Can we intervene?

Does the physical-health trajectory change?

1. Can diabetes outcomes identify people who need psychiatric treatment?

Most mental-health pathways begin with symptoms or a psychiatric diagnosis.

Diabetes psychiatry can also begin with the physical outcome.

Recurrent DKA

Recurrent DKA provides the clearest example.

Published research has associated recurrent DKA with mental-health difficulties, psychological distress and socioeconomic disadvantage.

Our own published case-control work found greater anxiety, diabetes distress, difficulties with emotion regulation and personality dysfunction in adults experiencing recurrent DKA than in people with type 1 diabetes without recurrent DKA.

This supports a clinically useful proposition:

Recurrent DKA can function as a psychiatric case-finding signal.

It does not identify a diagnosis.

It does not imply that every episode of DKA has a psychological cause.

It suggests that repeated DKA provides a reason to look systematically.

Current programme research is examining what becomes visible when larger recurrent-DKA populations undergo detailed psychiatric assessment rather than relying upon diagnoses already recorded in routine care.

Persistent extreme hyperglycaemia

The evidence is less developed.

Depression, diabetes distress, attachment style, psychosocial adversity and difficulties with diabetes self-management are associated with poorer glycaemic outcomes.

Much less is known about the population at the extreme end of the HbA1c distribution.

This raises two questions:

Who are the people whose HbA1c remains ≥100 mmol/mol?

and:

Does persistent extreme hyperglycaemia itself provide a useful psychiatric case-finding signal?

This is an active area of current work.

2. Should specialist intervention be concentrated in the extreme-risk tail?

Risk in diabetes is not evenly distributed.

Long-term observational studies of type 1 diabetes show substantially increasing mortality and complication risk at high HbA1c levels. High DKA frequency indicates increasing mortality risk.

This creates an important service-design question.

Should specialist psychological and psychiatric intervention be offered broadly to people above conventional glycaemic targets?

Or can scarce specialist resource produce greater health benefit by concentrating on a smaller population experiencing extreme and persistent physical-health risk?

A fall in HbA1c from 105 to 90 mmol/mol and a fall from 75 to 60 mmol/mol are numerically similar.

They should not automatically be assumed to have the same absolute health or economic consequences because the starting risks differ.

This leads to a developing hypothesis:

Concentrating specialist intervention towards the extreme diabetes-risk tail may produce greater absolute health benefit per unit of specialist resource than applying the same intervention across a broader population with moderately elevated risk.

This is plausible rather than established and we have already tested its feasibility.

Understanding its utility requires measurement of absolute physical-health risk, acute healthcare use, complications, mortality, intervention costs and treatment effects.

3. What kind of integration works?

There is established evidence that integrated approaches addressing diabetes and mental health can improve psychological outcomes and glycaemic control.

Programmes including 3 Dimensions of Care for Diabetes (3DFD) have demonstrated the feasibility of integrating medical, psychological and social approaches for people with persistently poor glycaemic control and psychosocial complexity.

Much of this literature conceptualises integration through a multidisciplinary team:

DIABETES EXPERTISE

PSYCHIATRY

SOCIAL INTERVENTION

↓

COORDINATION BETWEEN PROFESSIONALS

A related possibility is integration within the clinician.

An integrated clinician may hold sufficient expertise to understand the diabetes, assess and treat psychiatric illness, formulate psychotherapeutically, assess risk and cognition, intervene practically and remain involved longitudinally — while adding specialist expertise when necessary.

The principle is:

Add expertise without unnecessarily transferring ownership.

This leads to another testable hypothesis:

For a small population with very high complexity and physical-health risk, concentrating broad clinical capability within one longitudinal specialist clinician may be more efficient than distributing the same functions across multiple services and professionals.

This is not yet established.

Relevant outcomes would include clinical effectiveness alongside referrals, non-attendance, duplication, professional contacts, handoffs, coordination time, hospital use and total cost.

4. Can psychiatric treatment change physical-health trajectories?

Psychological outcomes matter in their own right.

But where psychiatric intervention is triggered by a high-risk diabetes outcome, the physical-health trajectory should also remain visible.

Relevant questions include:

Did DKA recur?

Did HbA1c change?

Did severe hypoglycaemia decrease?

Did emergency attendance, hospital admission or ICU use change?

Over longer periods:

Did complications, morbidity or premature mortality change?

This creates an important methodological principle for diabetes psychiatry:

The physical-health outcome provides one test of the psychiatric formulation.

If an intervention directed at the proposed mechanism does not produce the expected change, the formulation itself may need reconsideration.

5. DEVELOPING HYPOTHESES

Ideas requiring formal evaluation.

These currently include:

  • extreme diabetes phenotypes as psychiatric case-finding signals

  • the importance of trajectory rather than a single HbA1c

  • HEAT as a psychotherapeutic lens on affective experience in diabetes

  • making diabetes psychologically tolerable to think about

  • formulation-derived levers

  • the integrated-clinician model

  • concentrating specialist intervention within the extreme-risk tail.

These ideas are presented as questions for clinical and research development rather than established mechanisms.

Current research questions

Current work asks:

  • Can recurrent DKA and persistent extreme hyperglycaemia provide useful psychiatric case-finding signals?

  • What distinguishes transient from persistent extreme hyperglycaemia?

  • Does moving further into the extreme-risk tail identify increasing psychological, psychiatric or social complexity?

  • Where on the diabetes-risk distribution does specialist intervention produce the greatest absolute health benefit?

  • Which interventions function as meaningful levers?

  • Does longitudinal continuity improve outcomes?

  • Can reducing professional handoffs improve engagement and treatment fidelity?

  • Can psychiatric intervention reduce DKA, HbA1c, severe hypoglycaemia and hospital use?

  • Are complications and premature mortality increasingly concentrated within a relatively small population experiencing persistent high risk?

Wider evidence and national context

In 2026, the Health Services Safety Investigations Body (HSSIB) published a national investigation into insulin self-administration among people with diabetes and mental health problems. The report identified continuing gaps between specialist diabetes and mental health services and called for improved integration at national, regional and local levels.

Its local learning also highlighted the importance of identifying people with recurrent DKA or hypoglycaemia who may require mental health input.