GLP-1 RA use and neuropsychiatric outcomes in Medicare adults with type 2 diabetes

The study

This population-based cohort study used claims data from Medicare enrolees with type 2 diabetes. It was published in Diabetes, Obesity & Metabolism. The researchers compared GLP-1 receptor agonists (GLP-1 RAs) with other second-line glucose-lowering medicines. They looked at four treatment patterns: monotherapy, switch, escalation and augmentation. They also looked at differences by obesity status.

Eight neuropsychiatric outcomes were studied. The analysis used 1:1 propensity score-matched, covariate-adjusted stratified Cox regression with time-varying covariates.

What they found

For all psychiatric disorders combined, GLP-1 RA use was associated with a small rise in risk. The hazard ratio was 1.04 (95% confidence interval 1.03-1.06). This was 1.7 excess events/1000 person-years. The authors describe this as of uncertain clinical significance, given the large size of the cohort.

Results differed by outcome:

• Depression: risk rose (hazard ratio 1.11), with 3.0 excess events/1000 person-years.
• Substance use disorders: risk fell (hazard ratio 0.84), with 4.1 fewer events/1000 person-years.
• Dementia: risk fell (hazard ratio 0.87), with 2.8 fewer events/1000 person-years.
• Suicidal or self-harm events: risk fell (hazard ratio 0.81), with 0.4 fewer events/1000 person-years.

Treatment pattern mattered. With monotherapy, psychiatric risk rose (hazard ratio 1.43) and emotional symptom risk rose (hazard ratio 1.32). Escalation also showed higher psychiatric risk. Switch and augmentation showed reduced or null risks.

The abstract says results also varied by obesity status. It does not give the figures for this.

How much weight to give it

The authors conclude that the neuropsychiatric profile of GLP-1 RAs appears context-dependent and is not a uniform class effect. They suggest the higher risks with monotherapy or escalation may reflect baseline clinical differences rather than a pharmacological effect. They call the lower risks of substance use disorders, dementia and suicidal or self-harm events reassuring.

This is an observational cohort study. It can show association only. It cannot show that GLP-1 RAs caused any of these changes.

The absolute difference for the aggregate psychiatric outcome was small (1.7 excess events/1000 person-years). The absolute difference for suicidal or self-harm events was also small (0.4 fewer events/1000 person-years).

The abstract does not report the number of patients, the length of follow-up, which GLP-1 RAs were studied, the comparator drugs in detail, funding, or confidence intervals for the individual outcomes. These should be checked in the full paper.

For your clinic

The authors describe the neuropsychiatric profile of GLP-1 RAs as context-dependent. The study is observational, so it does not show that GLP-1 RAs cause depression, or that they protect against dementia, substance use disorders or self-harm.

The picture is mixed: higher depression risk, but lower risks of dementia, substance use disorders and suicidal or self-harm events. Read the full paper before drawing conclusions. In particular, check how the groups were matched, how outcomes were defined, which drugs were compared and how the obesity analysis was done.

The data come from Medicare enrolees. The abstract does not say whether the results apply to patients in India. Decisions about drug choice should rest on the full evidence and current guidelines, not on this abstract alone.