Observational study suggested SGLT2 inhibitors might promote healthy aging. However, whether brain-related phenotypes mediate this association. We applied Mendelian randomization (MR) to investigate the effect of SGLT2 inhibition on chronological, biological age and cognition and explore the mediation effects of brain imaging-derived phenotypes (IDPs). We selected genetic variants associated with both expression levels of SLC5A2 (GTEx and eQTLGen data; N=129 to 31,684) and HbA1c levels (UK Biobank; N=344,182) and used them to proxy the effect of SGLT2 inhibition. Aging related outcomes, including parental longevity (N=389,166) and epigenetic clocks (N=34,710), and cognitive phenotypes, including cognitive function (N=300,486) and intelligence (N= 269,867) were derived from genome-wide association studies. Two-step MR were conducted to explore the associations between SGLT2 inhibition, IDPs, and aging outcomes, cognition. SGLT2 inhibition was associated with longer father's attained age (years of life increase per SD (6.75 mmol/mol) reduction in HbA1c levels = 6.21, 95%CI 1.95 to 11.15), better cognitive function (beta = 0.17, 95%CI 0.03 to 0.31) and higher intelligence (beta = 0.47, 95%CI 0.19 to 0.75). Two-step MR identified two IDPs as mediators linking SGLT2 inhibition with chronological age (total proportion of mediation = 22.6%), where four and five IDPs were mediators for SGLT2 inhibition on cognitive function and intelligence respectively (total proportion of mediation = 61.6% and 68.6% respectively). Our study supported that SGLT2 inhibition increases father's attained age, cognitive function and intelligence, which was mediated through brain images of different brain regions. Future studies are needed to investigate whether similar effect could be observed for users of SGLT2 inhibitors.