Influence of resistant starch from potatoes on glycemic response in healthy, overweight females



Journal Title

Journal ISSN

Volume Title



Resistant starch (RS) improves postprandial glycemia and insulin sensitivity in healthy adults and those with insulin resistance. RS is found naturally in potatoes, where the amount varies based on cooking method and refrigeration storage time. The purpose of this study was to determine if the cooking method and storage duration of potatoes influenced the postprandial glucose response in healthy, overweight women. The study also attempted to develop a precision nutrition model that could be used to predict glucose incremental area under the curve (iAUC) following potato intake.

Thirty overweight or obese females were enrolled in a randomized, crossover study to consume 250 g of Russet potatoes either boiled and served hot or baked then chilled five days and served cold. Fasting and postprandial blood was collected to analyze glucose, insulin, glucagon-like peptide-1, peptide YY (PYY), and glucose-dependent insulinotropic peptide (GIP) over 120 minutes and compared between potato types. Baseline body composition, usual dietary patterns, and stool were measured. Lastly, a penalized regression model was using an elastic net algorithm was used to predict postprandial glucose iAUC following potato consumption.

Significant reductions were observed for glucose, insulin, and GIP iAUCs (49.9%, p=0.021; 40.5%, p<0.0001; and 71.1%, p<0.0001, respectively) between the two potato types. Twenty-one of 30 participants (70%) exhibited a lower glucose iAUC following consumption of the cold potato compared to the hot potato. Only seven baseline characteristics were selected via elastic net regression (using 5-fold cross validation) as important features that accounted for 32.15% of variation in iAUC glucose following potato intake. The potato type (hot or cold) and Faecalibacterium were significantly associated with iAUC glucose (β coefficient 547, 95% CI 131.61, 963.68, p = 0.01; β coefficient -73.50, 95% CI -128.52, -18.47, p = 0.01, respectively).

A statistically significant difference in postprandial glycemic response was observed between hot and cold potatoes, with more favorable responses occurring after the consumption of cold potatoes. The increased amount of RS likely contributed to this response. Baseline features predicted postprandial glycemic response following potato intake, and a precision nutrition model was developed that can be used to personalize recommendations for promoting glycemic control.



Resistant starch, Potato, Postprandial glucose response, Microbiome, Machine learning