A Dietitian-Led Vegan Program May Improve GlycA, and Other Novel and Traditional Cardiometabolic Risk Factors in Patients with Dyslipidemia: A Pilot Study.

BackgroundSystematic inflammation and lipid profiles are two major therapeutic targets for cardiovascular diseases. The effect of a nutritionally balanced vegan diet on systematic inflammation and lipoprotein subclass awaits further examination.ObjectiveTo investigate the change in novel and traditional cardiometabolic risk factors before and after a dietitian-led vegan program, and to test the bioavailability of vitamin B12 in Taiwanese purple laver as part of a vegan diet.DesignA one-arm pilot intervention study.Participants/SettingNine patients with dyslipidemia participated in this 12-week vegan program.Main Outcome MeasuresNuclear Magnetic Resonance (NMR) detected GlycA signals (systematic inflammation) and lipoprotein subclass (atherogenicity); trimethylamine N-oxide (TMAO); and other cardiometabolic risk factors.Statistical Analyses PerformedWilcoxon signed-rank test.ResultsIn this 12-week vegan intervention emphasizing whole foods, systematic inflammation improved as indicated by a reduction in GlycA (median: −23 μmol/L, p = 0.01). LDL-c (low-density lipoprotein cholesterol) (median −24 mg/dl, p = 0.04) and LDL-p (low-density lipoprotein particles) (median −75 nmol/L, p = 0.02) both decreased significantly. VLDL (very-low-density lipoprotein) and chylomicron particles showed a decreasing trend (−23.6 nmol/L, p = 0.05). Without caloric restriction, body mass index (BMI) (−0.7 kg/m2, p = 0.03), waist circumferences (−2.0 cm, p < 0.001), HbA1c (−0.2%, p = 0.02), and (HOMA-IR) homeostatic model assessment for insulin resistance (−0.7, p = 0.04) have all improved. The change in the TMAO and vitamin B12 status as measured by holo-transcobalamin appeared to depend on baseline diets, TMAO, and vitamin B12 status.ConclusionsA dietitian-led vegan program may improve systematic inflammation and other novel and traditional cardiometabolic risk factors in high-risk individuals.