Sodium-glucose cotransporter 2 inhibitors (SGLT2is) are available for individuals with type 1 diabetes, but appropriate use is recommended to prevent ketosis or ketoacidosis. This study aimed to evaluate the risk of ketosis in people with type 1 diabetes, focusing on the relationship between nutritional assessment, glycaemic status, c-peptide immunoreactivity (CPR) index and body composition. In total, 46 Japanese patients with type 1 diabetes were included, and dietary assessment from food photographs and ketone levels were evaluated before and after taking SGLT2is. The effect of diet on morning ketone levels was also investigated. All patients had an increase in mean ketone concentrations after taking SGLT2is (before 0.12 ± 0.06 mmol/L, after 0.23 ± 0.16 mmol/L). A significant negative correlation was found between average morning ketone levels and age (r = -0.514, p < .001) and the CPR index (r = -0.523, p = .038) after taking SGLT2is. Using a mixed-effects model based on the results before starting the inhibitors, it was noted that both patient-to-patient and age, or patient-to-patient and capacity of insulin secretion, influenced the ketone levels. Multiple regression analysis showed that factors associated with the risk of increasing ketone levels after taking SGLT2is were younger age (β = -0.504, p = .003) and a low ratio of basal to bolus insulin (β = -0.420, p = .005). When administering SGLT2is to patients with a low CPR index or younger patients with type 1 diabetes, adequate instructions to prevent ketosis should be given.