Diabetes Drug Class Tied to Lower Dementia Risk in People With Psych Disorders

Use of SGLT2 inhibitors was linked to lower dementia incidence and fewer related psychiatric outcomes in older adults with mood and psychotic disorders, according to a target trial emulation study.

In an intention-to-treat analysis, patients with major depressive disorder, bipolar disorder, or schizophrenia spectrum disorder who used SGLT2 inhibitors were less likely to develop all-cause dementia (OR 0.61, 95% CI 0.52-0.73) and had fewer psychiatric emergency department visits (OR 0.80, 95% CI 0.66-0.97) compared with similar patients who did not use the drugs, reported Jaime Ramos-Cejudo, PhD, of New York University Grossman School of Medicine in New York City, and co-authors.

In a per-protocol analysis, SGLT2 inhibitor use was associated with lower odds of all-cause dementia (OR 0.54, 95% CI 0.40-0.73) and psychiatric hospitalizations (OR 0.56, 95% CI 0.31-1.00), but not psychiatric emergency department visits (OR 0.74, 95% CI 0.53-1.05), the researchers wrote in JAMA Network Open.

The results support a growing body of work suggesting that metabolic and brain health are interconnected, Ramos-Cejudo observed. “While SGLT2 inhibitors were originally developed to treat diabetes, accumulating evidence indicates they may also influence biological pathways involved in neurodegeneration, including brain energy metabolism, mitochondrial function, and inflammation,” he told MedPage Today.

Previous studies have evaluated relationships between dementia risk and diabetes drugs including GLP-1 receptor agonists and SGLT2 inhibitors. GLP-1 agents have been investigated extensively in neurologic and psychiatric conditions, but SGLT2 inhibitors are comparatively understudied, Ramos-Cejudo noted.

“What makes this study particularly novel is the population we examined,” he pointed out. “People living with conditions such as major depression, bipolar disorder, and schizophrenia are known to be at substantially increased risk of dementia, yet they are often underrepresented in dementia prevention research.”

Ramos-Cejudo and colleagues used a target trial emulation framework based on data from the Department of Veterans Affairs healthcare system from 2016 to 2024 to evaluate people ages 65 and older with major depressive disorder, bipolar disorder, or schizophrenia spectrum disorder. The analysis excluded people who had a dementia or a personality disorder diagnosis, and those who had previously used an SGLT2 inhibitor.

The researchers identified 112,725 people who met inclusion criteria for the all-cause dementia trial. These individuals had a median age of 74.1 years; 92.8% were men, 49.3% had obesity, and 6.8% had SGLT2 inhibitor exposure. Most (86.9%) had major depressive disorder. Over a median follow-up of 3.3 years, 4.1% of participants developed dementia.

Initiation and non-initiation of SGLT2 inhibitors were assessed using an intention-to-treat analysis. Sustained and non-sustained use of SGLT2 inhibitors for 3 months or longer were assessed using a per-protocol analysis. SGLT2 inhibitor use was analyzed as a single category, with switching between individual SGLT2 inhibitors treated as continuous therapy.

Incident all-cause dementia was defined by ICD codes. Covariates included demographic data, lifestyle factors, body mass index, HbA1c levels, and medical history. Models were also adjusted for psychiatric diagnosis and medication history.

“SGLT2 inhibitors are unique among antidiabetic medications in that they induce ketogenesis,” Ramos-Cejudo noted. “Ketogenic diets have recently been associated with symptomatic improvement among those with psychiatric disorders in some preliminary clinical trials, and this study lends support to the possibility that they offer a pharmacological means of achieving similar benefits.”

Despite adjustments and the trial emulation design, residual confounding from variables like severity of psychiatric illness, baseline cognitive function, and diabetes indications may persist, the researchers acknowledged. Prescription and ICD codes have limited precision, and the predominantly male population in this study limits generalizability.