Background Selective Serotonin Reuptake Inhibitors (SSRIs) represent the standard treatment for Major Depressive Disorder (MDD). However, their efficacy is variable and incomplete [1]. In order to explain, at least in part, such variable efficacy, it has been proposed that SSRI administration does not affect mood per se but, by enhancing neural plasticity, amplifies the influence of the living conditions on mood. Consequently, in a favorable environment, SSRI treatment leads to a reduction of symptoms while, in stressful conditions, it could lead to a worse prognosis [2,3]. Therefore, the pharmacological modulation of factors underlying the link between living conditions and SSRI efficacy represents a novel and desirable strategy to improve treatment outcome. Since metabolic profile is markedly affected by the quality of the living environment and, in turn, affects antidepressant efficacy, the regulation of metabolism represents a potential target to improve treatment outcome. Aim The aim of our preclinical study was to test whether, in mice living in adverse conditions, the combined administration of the SSRI fluoxetine and metformin, a drug able to improve metabolic profile, results in an increased antidepressant efficacy compared to fluoxetine alone. Methods Adult C57BL6 male mice were exposed to 5 weeks of stress. The first two were aimed at inducing a depression-like phenotype. For remaining 3 weeks, mice received one of the following treatments: vehicle (VEH), fluoxetine (FLX), metformin (MET) or a combination of FLX and MET. We assessed molecular and cellular endpoints in both the dorsal and ventral area of the hippocampus. We measured the expression levels of genes involved in stress response, metabolism and MDD such as glucocorticoid receptor, insulin growth factor 1 and 2, and BDNF and the long term potentiation (LTP) at Schaffer collateral-CA1 synapses. In addition, we assessed liking- (saccharin preference) and wanting-type anhedonia (progressive ratio) as behavioral phenotypes. Data were analyzed with ANOVA with Fisher-PLSD post-hocs. Results The efficacy of combined treatment is higher compared to fluoxetine alone. Liking-type anhedonia was significantly affected at week 1 [F(3,34)=3.406, p=0.0285], FLX-MET showing a higher saccharin preference compared to VEH and FLX mice (respectively, p=0.0276 and p=0.0251). Wanting-type anhedonia was affected at both week 1 [F(3,35)=3.047, p=0.414] and 3 [F(3,35)=4.329,p=0.0107], FLX-MET displaying an improvement compared to FLX (p=0.0141) at week 1. With regard to electrophysiology, metformin did not affect hippocampal plasticity neither in combination with fluoxetine nor alone. Fluoxetine led to an enhanced brain plasticity [mean fEPSP slope potentiation: 1.28±0.008; F(1,40)=4.152,p=0.0482]. Gene expression analysis showed that IGF2 gene was significantly modulated according to treatment and hippocampal area [F(1,25)=17.188,p=0.00034]. In FLX-MET, IGF2 was higher in the dorsal area (p Conclusions Behavioral results show that the combined treatment is more effective than fluoxetine alone in ameliorating the depressive phenotype. Electrophysiological results suggest that changes in plasticity are not responsible for the behavioral effects. IGF2 expression appears critically associated to the effects of the different treatments.