Antipsychotics, Diabetes and the Monitoring Nobody Does

Antipsychotics, diabetes and the monitoring nobody does: glucose was checked in 12.9 per cent

An Australian study used the MedicineInsight primary care database to follow 3,840 children and adolescents who were newly prescribed an antipsychotic between 2011 and 2017, across 488 general practices. The question was simple: in the twelve months after that first prescription, did anybody check them for the metabolic side effects these drugs are known to cause?

About 56 per cent had at least one parameter measured. In the authors’ own words, rates of cardiometabolic monitoring sat “well below guideline recommendations”.

Read the glucose line again. These are children started on a class of medication whose best-documented adverse effects are weight gain, insulin resistance and dyslipidaemia. Fewer than one in eight had their blood sugar checked in the following year. Fewer than one in sixty had a waist measured.

And this matters disproportionately for the population this site serves, because antipsychotics are prescribed at high rates to people with intellectual disability and to autistic people in Australia — frequently for behaviours of concern rather than for a psychotic illness. Which produces the finding that organises this entire guide:

A meaningful share of the excess diabetes in people with intellectual disability was prescribed to them, and then not monitored.

A 2022 meta-analysis pooling 33 studies and 55,548 people found diabetes prevalence of 8.5 per cent among people with intellectual disabilities and odds 2.46 times higher than age- and gender-matched controls. Medication is not the whole explanation for that gap. It is a substantial and almost entirely preventable part of it.

We are a registered provider working across New South Wales. We do not prescribe, we are not a clinical service, and nothing here is medical advice — and specifically, nothing here is a reason to stop or change anybody’s medication. Abrupt withdrawal of an antipsychotic can be harmful. What this guide is for is knowing what monitoring should be happening, and asking for it.

What these drugs actually do

Second-generation or atypical antipsychotics cause metabolic change through several mechanisms — appetite stimulation, altered satiety signalling, direct effects on insulin sensitivity, and effects on fat distribution.

The consequences cluster:

Weight gain, which is frequently substantial and is most rapid in the early weeks and months of treatment. That timing matters enormously, because it means the period when monitoring would be most useful is the period when it is least likely to happen.

Insulin resistance and raised blood glucose, progressing in some people to type 2 diabetes. The effect is not purely a consequence of weight gain — some of it appears to be a direct drug effect, which means a person who has not gained much weight is still not in the clear.

Dyslipidaemia — raised triglycerides, lowered HDL.

Raised blood pressure.

Together those make up metabolic syndrome, which is the cluster of central obesity, raised glucose, raised triglycerides, low HDL and raised blood pressure, and which substantially raises cardiovascular risk.

Risk varies by drug, and this is clinically useful to know because it means switching is sometimes an option. Olanzapine and clozapine carry the highest metabolic risk. Quetiapine and risperidone sit in the middle. Aripiprazole, ziprasidone and lurasidone are generally considered lower risk. The trade-off is that the higher-risk agents are sometimes the more effective ones for a particular person, and clozapine in particular has a specific role nothing else fills. So this is a genuine clinical balance, not a simple hierarchy — but it is a balance that should be actively weighed rather than left where it landed years ago.

Other effects worth knowing, because they appear in the monitoring list above: raised prolactin, which can cause breast development, lactation, sexual dysfunction and, over time, reduced bone density; QTc prolongation on ECG with some agents; sedation; and movement disorders.

The restrictive practice framing nobody applies

Here is the connection that is almost never made, and it reframes the whole problem.

Under the NDIS Quality and Safeguards Commission’s framework, chemical restraint is the use of medication for the primary purpose of influencing a person’s behaviour, other than medication prescribed to treat a diagnosed mental health, physical or medical condition. It is one of the five regulated restrictive practices, which means its use requires authorisation under the relevant state or territory process, a behaviour support plan prepared by an NDIS behaviour support practitioner, and reporting to the Commission — with an interim plan required within one month and a comprehensive plan within six.

The Royal Commission into Violence, Abuse, Neglect and Exploitation of People with Disability examined psychotropic prescribing to people with disability at length, and the practice of prescribing antipsychotics for behaviour in the absence of a psychiatric diagnosis is well documented in Australia.

So for a substantial number of people, the antipsychotic they are taking is a regulated restrictive practice.

And here is the gap: the restrictive practice framework concerns itself with authorisation, consent, documentation and reduction. It says very little about the fact that the practice causes diabetes.

A behaviour support plan for chemical restraint will set out the behaviour, the strategies, the authorisation and the reduction plan. It will very rarely say: this medication causes weight gain, insulin resistance and dyslipidaemia, and here is the monitoring schedule that will detect them.

That seems to us like an obvious omission in a safeguarding framework. If a practice is restrictive enough to require authorisation and a reduction plan, the physical harm it causes should be monitored as part of the safeguard — not left to a GP who may not know the prescription was for behaviour, in a system where glucose gets checked in 12.9 per cent of cases.

The practical action: if somebody you support is on an antipsychotic under a behaviour support plan, look at the plan and see whether it mentions metabolic monitoring. If it does not, raise it with the behaviour support practitioner and the prescriber. It is a reasonable thing to ask to be added, and it creates a documented accountability that currently does not exist.

What monitoring should look like

The precise schedule comes from the relevant clinical guideline and the prescriber, and guidelines are updated, so confirm the current version rather than relying on a website. The general shape across guidelines is consistent:

At baseline, before starting — personal and family history of diabetes and cardiovascular disease, weight, height, BMI, waist circumference, blood pressure, fasting glucose or HbA1c, a full lipid profile, and for some agents an ECG and prolactin.

In the early weeks and months — weight in particular, because this is when gain is most rapid. Guidelines commonly specify checks at around four weeks and again at around three months, with glucose and lipids repeated at the three-month mark.

Then at regular intervals ongoing — commonly at least annually, more often where there are abnormalities, a dose increase, a drug change, or rapid weight gain.

Weight more frequently than the rest. Weight is free, non-invasive, takes thirty seconds, and is the earliest signal. Monthly weighing is not a clinical imposition; it is the cheapest safeguard available and it is something a support service can simply do.

Waist circumference matters more than BMI for metabolic risk, and it was measured in 1.6 per cent of that Australian cohort. It requires a tape measure.

And for anybody with reduced muscle mass — after spinal cord injury, in advanced neurological conditions, with limited mobility — BMI read against general population thresholds understates adiposity, so a normal BMI is not reassurance. Waist circumference and clinical judgement matter more.

What to do with an abnormal result

This is the part that makes monitoring worth doing, and it is a prescriber’s decision in every case.

Ask whether the medication is still needed. This is the first and most important question and it frequently has not been asked for years. Antipsychotics started for a crisis, or during a period of distress, or in a previous placement, or in childhood, are often simply continued. A documented review of whether the original indication still applies is a legitimate thing to request.

Consider dose reduction. Metabolic effects are generally dose-related.

Consider switching to a lower-risk agent, where that is clinically appropriate.

Consider metformin. There is reasonable evidence supporting metformin as an adjunct to limit antipsychotic-induced weight gain, and it is used for this purpose. Whether it suits an individual is a prescribing decision, but it is worth knowing the option exists, because many families have never been told.

Address the environment. Not as a substitute for the above, and not as a way of making this the person’s responsibility — but funded exercise support, a food environment that is not purely the product of a shift roster and a budget, and exercise physiology in the NDIS plan all genuinely help. This is one of the better arguments for exercise physiology funding in a plan, because the clinical rationale is specific rather than general.

Treat the diabetes if it has developed, through the health system, while also reviewing the medication that contributed to it.

What not to do: stop the medication abruptly, or on a provider’s initiative. Abrupt withdrawal of an antipsychotic can cause rebound symptoms, withdrawal effects and a genuine deterioration, and the person may become unsafe. Reduction is a planned, slow, clinically supervised process, and it needs behaviour support in place alongside it — because the behaviour the medication was suppressing may re-emerge and will need something other than a tablet to respond to. A reduction attempted without that support will fail and will usually end in a higher dose than before.

Who is actually responsible, and why nothing happens

The honest answer is that responsibility is distributed in a way that means nobody holds it.

The prescriber may be a psychiatrist who sees the person twice a year, or a GP who inherited a prescription from a psychiatrist years ago, or a paediatrician whose patient is now 23.

The GP may not know that the antipsychotic was prescribed for behaviour rather than for a psychiatric illness, and may not be the one reviewing it.

The provider holds the daily observations and the weight scales and is usually the only party in a position to notice gradual change — and has no prescribing role and frequently no clear instruction to monitor anything.

The family may not have been told about metabolic risk at all.

The behaviour support practitioner writes the plan governing the restrictive practice and typically does not address physical health monitoring.

The authors of the Australian study pointed at exactly this: alongside education and training, they identified that “communication between primary and secondary care providers needs to be addressed.”

Which means, practically, that it happens when somebody decides to own it. Usually a family member, a support coordinator or a house manager.

What a provider should simply do

None of this requires clinical authority.

• Weigh the person monthly, on the same scales, and record it with the date.
• Measure waist circumference, with a tape measure, at the same interval.
• Flag a trend, not a single reading. Three kilograms over three months is a trend and it is worth a letter.
• Keep a current medication list that identifies which medications are psychotropic and which are subject to a behaviour support plan.
• Know whether the antipsychotic is authorised as chemical restraint, and if so, whether the plan addresses metabolic monitoring.
• Report symptoms that might indicate diabetes — increased thirst, increased urination, fatigue, recurrent thrush or skin infections, slow-healing wounds, blurred vision — in writing, with dates, rather than noting them in a shift report nobody reads longitudinally.
• Book the annual health assessment. For people with intellectual disability, MBS item 707 funds a prolonged assessment of more than 60 minutes annually, booked directly with a GP, no referral needed, using the free CHAP tool. A companion item, MBS 699, covers an annual Heart Health Assessment for that same group. This is the appointment where glucose, lipids, blood pressure, weight and a medication review all happen in one place. Given the 12.9 per cent figure, it is the most realistic mechanism available.
• Ask for the review. In writing, with the weight data attached, addressed to the prescriber.

Children and young people specifically

The Australian study was about children, and children have additional considerations.

Growth and puberty. Weight gain in a growing child needs interpreting against growth charts, and rapid weight gain during puberty has consequences that persist into adult life.

Prolactin. Raised prolactin in adolescence can delay or disrupt puberty, cause breast development in boys, and affect bone mineral density accrual at exactly the age when peak bone mass is being established. It was measured in 1.1 per cent of that cohort.

Duration. A medication started at nine and continued without review is a forty-year exposure. The question of whether it is still needed should be asked at every developmental transition, not just when something goes wrong.

Diagnostic clarity. An antipsychotic prescribed for behaviour in a distressed nine-year-old may follow that child into adulthood as though it were treatment for a diagnosed illness. The original indication should be documented and revisited.

The questions to ask, and who to ask

To the prescriber:

1. What is this medication for, and is that indication still current?
2. What metabolic monitoring should be happening, and when was it last done?
3. Can we see the baseline measurements from when it was started?
4. Is this a lower-risk agent, and has switching been considered?
5. Is dose reduction possible?
6. Is metformin worth considering for the weight gain?
7. What is the plan for review, and when?

To the GP:

1. Was a baseline glucose and lipid profile done, and can we repeat it now?
2. Can we book the annual health assessment, item 707, and use the CHAP?

To the behaviour support practitioner:

1. Is this medication authorised as chemical restraint?
2. Does the behaviour support plan address metabolic monitoring, and can it?
3. What is the reduction plan, and what supports need to be in place for it to succeed?

To the provider:

1. Are you weighing the person monthly and recording it?
2. Who reviews that record, and how often?