In this study, we aimed to determine the relative effectiveness of common dietary polyphenols or the isoquinoline alkaloid berberine in protecting against molecular mechanisms underlying non-alcoholic fatty liver disease (NAFLD) involving changes to cellular lipid metabolism and bioenergetics. In a model of steatosis using HepG2 hepatocytes, exposure of the cells to 1.5 mM oleic acid (OA) for 24 h caused steatosis and distorted cell morphology, induced the expression of mRNA for enzymes that are involved in lipogenesis and fatty acid oxidation (FAS and CPT1A), and impaired indices of aerobic energy metabolism (PPARγ mRNA expression, mitochondrial membrane potential (MMP), and galactose-supported ATP production). Co-treatment with 10 µM of selected polyphenols all strongly protected against the steatosis and changes in cell morphology. All polyphenols, except cyanidin, inhibited the effects on FAS and PPARγ and further increased CPT1A1 expression, suggesting a shift toward increased β-oxidation. Resveratrol, quercetin, catechin, and cyanidin, however not kuromanin or berberine, ameliorated the decreases in MMP and galactose-derived ATP. Berberine was unique in worsening the decrease in galactose-derived ATP. In further investigations of the mechanisms involved, resveratrol, catechin, and berberine increased SIRT1 enzyme activity and p-AMPKα^Thr172 protein, which are involved in mitochondrial biogenesis. In conclusion, selected polyphenols all protected against steatosis with similar effectiveness, however through different mechanisms that increased aerobic lipid metabolism and mitochondrial function.

Non-invasive monitoring of insulin could hold promise to identify those with, or at risk for developing, insulin resistance. We aimed to examine saliva insulin responses across the day following high- and low-carbohydrate meals and evaluate whether changes in saliva insulin might accurately reflect changes in plasma insulin. In two randomized crossover studies, young normal weight men (NW; n = 8; Study 1) and adults with overweight/obesity (OO; n = 8; Study 2) completed two 9-h experimental trials in which the participants consumed isocaloric mixed high-carbohydrate (HC) or low-carbohydrate (LC) meals at 0, 3, and 6 h. Plasma and saliva samples were collected at fasted baseline and every 30 min for a total of 19 samples across 9 h. Overall, findings revealed a similar trend for postprandial saliva and plasma insulin responses regardless of the time of the day with a ∼30–45 min lag between saliva and plasma insulin responses. In both NW and OO groups, saliva and plasma insulin area under the curve (AUC) and incremental AUC were significantly higher in HC condition as compared to LC condition (all P ≤ 0.002). Nine-hour plasma and saliva insulin total AUCs were strongly and very strongly correlated in both HC ( r = 0.68; P = 0.007) and LC ( r = 0.84, P < 0.001) conditions, respectively. Saliva insulin is proportionate to and appears to reasonably track plasma insulin across the day with a ∼30–45 min delay. Saliva insulin shows promise as a non-invasive method to discern between low and high plasma insulin and may have utility in predicting the degree of insulin resistance (NCT03374436).