Investigating the effect on post-prandial glycemic and venous serum insulin response of an apple drink following the conversion of its glucose to gluconate. In a double-blind randomized placebo-controlled clinical trial with cross-over design, 30 male adults with impaired fasting glucose (IFG) received a drink of 500 ml: 1. Verum: Apple juice treated with invertase, glucose oxidase/catalase (glucose 0.05 g; gluconate 18.2 g); 2. Untreated apple juice (free glucose 8.5 g; bound glucose 6.7 g; gluconate below detection limit). Postprandial fingerprick capillary blood glucose and venous serum insulin were measured twice at baseline and at times 0 (start of drink), 15, 30, 45, 60, 90 and 120 min. Gastrointestinal symptoms, stool consistency and satiety were also assessed. The incremental area under the curve (iAUC120) of glucose levels (primary parameter) was significantly lower after verum (mean ± SD: 63.6 ± 46.7 min × mmol/l) compared to control (mean ± SD: 198 ± 80.9 min × mmol/l) (ANOVA F = 137.4, p < 0.001; α = 0.05). Also, iAUC120 of venous serum insulin levels (secondary parameter) was significantly lower after verum (mean ± SD: 2045 ± 991 min × mmol/l) compared to control (3864.3 ± 1941 min × mmol/l), (ANOVA F = 52.94, p < 0.001; α = 0.025). Further parameters of glucose metabolism and ISI = 2/[AUC venous serum insulin × AUC glucose +1] were also improved after verum compared to control. Verum increased stool frequency and decreased stool consistency, as assessed by Bristol stool form scale. By enzymatic treatment of apple juice its sugar content could be reduced by 21% and postprandial glycemic and venous serum insulin response by 68 and 47%, respectively resulting in a reduction of glycemic load by 74.6% without any adverse gastrointestinal side-effects.