Fig. 2. Fasting effects are distinct from those of a modified DASH diet only, and connected to vascular health benefits.

a Fasting followed by a modified DASH diet, but not a DASH diet alone, significantly improves 24 h ambulatory SBP and MAP 3 months post-intervention (two-sided MWU, FDR-corrected P-values are shown). Lines show individual participant trajectories. b MetS subjects beginning a modified DASH diet post-fasting significantly reduce their intake of antihypertensive medication by 3 months post-intervention, compared to subjects beginning a DASH diet only. Two-sided χ² test, P = 0.035. c Changes in 24 h ambulatory SBP in responders and non-responders including change in antihypertensive medication (two-sided MWU). d, e One week of fasting followed by modified DASH diet, but not DASH diet alone, caused significant (two-sided MWU, FDR-corrected P values are shown) BMI and body weight reduction in MetS patients, persisting 3 months later. f Comparison of changes in 24 h ambulatory SBP and body weight, respectively between baseline and follow-up in both study arms. Each dot represents an individual. g Body weight change is not significantly different between responders and non-responders in the fasting arm between baseline and follow-up (two-sided MWU). h Selected cardiometabolic risk parameters (vertical axis) altered in the fasting arm compared to the DASH arm. Heatmap hues show Cliff’s delta signed effect sizes, with asterisk indicating post-hoc univariate significance after compensating for drug dosage changes (see Methods). Horizontal axis shows each time point comparison: change during fasting/week three of DASH, change during refeeding/3 months of DASH, and change during the study period as a whole. Boxplot hinges denote 25th–75th percentile. Line within the boxplot indicates median. Whiskers on (c, g) are drawn from minimum to maximum values. Whiskers on (d, e) are drawn to minimum and maximum values, but not further than 1.5 × IQR.