Extended Data Figure 3: Subject physiology across diet arms.

a, Gastrointestinal motility, as measured by the initial appearance of a non-absorbable dye added to the first and last lunch of each diet. The median time until dye appearance is indicated with red arrows. Subject motility was significantly lower (P < 0.05, Mann–Whitney U test) on the animal-based diet (median transit time of 1.5 days) than on the plant-based one (1.0 days). b, Range (shaded boxes) and median (solid line) of subjects’ weights over time. Subjects’ weight did not change significantly on the plant-based diet relative to baseline periods, but did decrease significantly on the animal-based diet (asterisks denote q < 0.05, Bonferroni-corrected Mann–Whitney U test). Subjects lost a median of 1.6% and 2.5% of body weight by days 3 and 4, respectively, of the animal-based diet arm. c, Measurements of subjects’ urinary ketone levels. Individual subjects are shown with black dots, and median values are connected with a black solid line. Urinary ketone readings were taken from day 0 of the animal-based diet onwards. Ketone levels were compared to the readings on day 0, and asterisks denote days with significant ketone increases (q < 0.05, Bonferroni-corrected Mann–Whitney U test; significance tests were not carried out for days on which less than four subjects reported their readings.). All subjects on the animal-based diet showed elevated levels of ketones in their urine by day 2 of the diet (≥15 mg dl⁻¹ as compared to 0 mg dl⁻¹ during initial readings), indicating that they experienced ketonuria during the diet arm. This metabolic state is characterized by the restricted availability of glucose and the compensatory extraction of energy from fat tissue⁵⁶.