A PILOT STUDY OF SODIUM-GLUCOSE COTRANSPORTER-2 (SGLT2) INHIBITOR EMPAGLIFLOZIN IN MAJOR DEPRESSIVE DISORDER

Background

Sodium-glucose cotransporter-2 (SGLT2) inhibitors are unique among antidiabetic medications for the ketogenic properties. In population-based studies, they have associated with the lowest risk of incident depression of any antidiabetic medications, reduced dementia risk, and reduced risk of suicidal behaviors. Evidence from imaging, transcriptomic, genetic and cerebrospinal fluid studies points to mitochondrial impairments as associated with mood disorders, and SGLT2 inhibitors have demonstrated energetic-enhancing effects in preclinical studies. Given these data and encouraging reports of the antidepressant properties of ketogenic diets, this pilot study sought to test the feasibility, safety, and efficacy, in a preliminary way, of SGLT2 inhibitor empagliflozin in major depressive disorder.

Methods

Nine (of a planned twelve) participants with MDD confirmed by DSM-V criteria have completed a six-week, single-arm, open-label study of empagliflozin. Participants are administered 10mg of empagliflozin for two weeks and then 25mg of empagliflozin for four weeks for a total of six weeks on the medication. Baseline measures included weight, BMI, age, C-reactive protein (CRP), Montgomery-Åsberg Depression Rating Scale (MADRS), Snaith-Hamilton Pleasure Scale (SHAPS), Clinical Global Impression-Severity (CGI-S), and serum β-hydroxybutyrate (BHB) levels. Here, we report the comparisons between baseline and after six weeks of treatment.

Results

Among the first nine patients, mean BMI was 30.8 kg/m² (SD 7.8), age was 33.5 years (SD 12.0), hsCRP 5.2 mg/L (SD 7.3), and HbA1c 5.2% (SD 0.35). There were six females and three males. MADRS total scores were found to be significantly lower than baseline starting from week 4 of treatment (median change = −9.5, V = 1, p-value = 0.02071) and continuing through week 6 (median change = −13, V = 0, p = 0.007813). The effect sizes for the changes in MADRS total scores from baseline at weeks 4 (r=0.843, Cohen’s d = 3.134) and 6 (r=0.891, Cohen’s d = 3.925) were large. BHB levels trended towards increase with a baseline mean of 0.14mM and endpoint mean of 0.36mM, but the change was not significantly different from baseline at week 2 (median change = 0.05, V = 22, p-value = 0.2049) or week 6 (median change = 0.1, V = 29, p-value = 0.1484) of treatment. Spearman correlation between change in BHB and MADRS was non-significant at 6 weeks (rho 0.71, p=0.06). There was no subject dropout, however, one participant was only partially adherent.

Conclusion

In this small pilot study of individuals with MDD on empagliflozin, the medication was well tolerated, in line with studies in diabetes, heart failure, and chronic kidney disease. Empagliflozin treatment was associated with decreased depression severity scores and a trend towards increased peripheral fasting ketones. We did not observe a significant association between change in ketones and change in depression scores. Further studies of SGLT2 inhibitors in psychiatric disorders may complement studies of other ketogenic treatments.