The authors of the current study wanted to examine more of the potential psychiatric effects of GLP-1RAs, particularly how they relate to depression.

This target trial emulation study compared GLP-1RAs with two other medications used in diabetes treatment: sodium-glucose cotransporter-2 inhibitors (SGLT2is) and dipeptidyl peptidase-4 inhibitors (DPP4is).

While researchers were not able to randomize what participants received which treatment, they did emulate randomization by doing one-to-one propensity score matching.

This allowed them to match participants based on certain covariates like the use of medications, comorbid conditions, and age. Participants were aware of the treatment they were receiving.

Researchers had one group of 14,665 participants taking GLP-1RAs matched to 14,665 participants taking SGLT2is. The other group matched 13,711 participants taking GLP-1RAs to 13,711 participants taking DPP4is.

Participants in this study were at least 66 years old and had type 2 diabetes. Further, they had no previous history of depression, use of antidepressants in the previous year, or type 1 diabetes.

They also had no use of the medications of interest within the past year. In both comparison groups, the average age of participants was around 73 years old, and just under a third of participants used insulin at baseline.

Additionally, researchers looked at data in subgroups based on components like age, obesity at baseline, and treatment duration.

Researchers used data from United States’ national Medicare administrative claims. Participants were enrolled in Medicare Parts A, B, and D for at least 1 year before starting any of the medications examined in this study.

Researchers utilized “an intention-to-treat analysis,” which looks at the treatment participants were originally assigned whether or not participants received or properly followed the treatment protocol.

The results of the study showed a similar number of cases of depression among GLP-1RAs and SGLT2is. Throughout the study timeframe, 961 participants using GLP-1RAs developed depression, and 902 participants using SGLT2is developed depression.

This research has several components that limit the findings. For example, the researchers acknowledge several potential unmeasured confounders that could have affected the results.

They did not have data on components like body mass index (BMI) and hemoglobin A1C, a measure of blood sugar levels, which could have affected factors like participant selection and risk for depression.

They included a broad range of participants, which could have introduced bias. Researchers could not address “potential clustering effects” that could happen at the clinical site level because of the nature of the study.

The study’s results also cannot necessarily be generalized to all people who use GLP-1RAs. For example, the results cannot necessarily be generalized to younger individuals since this research focused on adults aged 66 and older. It also may not apply to people who use GLP-1RAs for weight loss who do not have type 2 diabetes.

Since researchers used data from Medicare administrative claims, it is possible that they missed some health information or that there was misclassification, such as misclassification of depression cases.

The authors note that they determined depression based on a claims data algorithm. Most participants were white, so more diversity could be important in future trials.

Researchers also noted several limitations related to their finding about how increased time taking GLP-1RAs was associated with a decreased depression risk. First, people who adhere to GLP-1RA treatment may choose healthier behaviors that also decrease depression risk.

Second, people with greater risk for mood disorders or depression could be less likely to take GLP-1RAs in the long term, potentially leading to reverse causality. Third, obesity has a strong correlation with depression, and people with obesity may be more often prescribed GLP-1RAs.

Weight loss for these individuals could indirectly lower the risk for depression. Finally, this study had a fairly short follow-up time, so it does not necessarily show how GLP-1RAs affect long-term depression risk.

It is also critical to look at this research in light of other studies completed in this area and the potential psychiatric risks associated with GLP-1RAs. Researchers acknowledge that it is possible that SGLT2is could also help reduce depression risk.

This study only followed participants for up to 2 years, so more research could examine the long-term effects. Future research can include randomized controlled trials and see if there are similar effects in other populations.

This study adds another factor to consider when it comes to improving depression outcomes in people with diabetes in the future. It also draws attention to a mental illness that affects many people with diabetes.

Andres Splenser, MD, an endocrinologist affiliated with Memorial Hermann, also not involved in the study, told MNT that, “if a patient with [type 2 diabetes] also has depression, this can affect their motivation to eat healthy, exercise, and impact compliance with taking their medications.“

“Today, however, there are many treatment options for diabetes that allow patients to remain healthy and not have to deal with any [diabetes] complications in their lifetime as long as their blood glucose values are well controlled,” he added.

Regarding this study, Splenser noted that it offered “reassurance that in patients 66 years and older, GLP1-RA therapy may not increase risk of depression.” However, it is likely that there are also several other factors at play.

He explained:

“More importantly — and not related to this study — is that improving glycemic control and allowing obese patients to lose weight tends to improve patients’ mood and outlook on their disease. Mood and appetite are closely related (think of stressful eating or eating when we are sad, ‘comfort food’), and a benefit of GLP1-Ra is that they help control appetite and satiety, which allows patients to make better meal choices and become healthier. So, the modest improvement in depression seen in the study may be related to weight loss, improved blood sugars, [fewer diabetes] symptoms, and likely an overall sense of patients getting healthier.”