Non-alcoholic fatty liver disease, Mediterranean diet and 10-year cardiovascular disease incidence: the mediating role of adiponectin. Highlights from the ATTICA prospective (2002–2012) study

Abstract

Abstract Background/Introduction The most updated set of guidelines (EASL-EASD-EASO 2016) suggests Mediterranean diet as a promising dietary pattern for non-alcoholic fatty liver disease (NAFLD) management. Purpose The association between NAFLD and adherence to Mediterranean diet as well as their interaction on 10-year fatal/non fatal cardiovascular disease incidence were evaluated. Methods Study population included 1,514 men and 1,528 women (>18 years old) in our city in Greece, free of CVD recruited during the 2001–2002 period. Liver steatosis and fibrosis assessment indices were calculated at baseline and specific cut-offs were used to capture NAFLD. Level of adherence to Mediterranean diet was assessed through MedDietScore. National and international food databases were used to estimate total daily carbohydrate intake. The cut-off point of 35% of total daily energy intake was set to define low vs. high carbohydrate intake. Circulating adiponectin level was measured at baseline (4.0 (2.0) μg/mL). Results MedDietScore was inversely associated with NAFLD indices of both steatosis and fibrosis yet only in Triglycerides-Glucose index [Odds Ratio (OR)3rd vs. 1st MedDietScore tertile = 0.53, 95% Confidence Interval (95% CI) (0.29, 0.95)] persisted in multi-adjusted models including adjustments for metabolic syndrome components, energy intake and overall and central obesity. This was more evident in men (p for sex interaction = 0.03). The protective role of Mediterranean diet was retained irrespective to the total carbohydrate content; however, sex-based stratified analysis revealed that in woman subsample only Mediterranean diet with low carbohydrate content (i.e. <35%) had an independent protective role against NAFLD presence [OR3rd vs. 1st MedDietScore tertile = 0.79, 95% CI (0.40, 0.96)]. NAFLD predicted CVD [Hazard Ratio (HR) = 3.01, 95% CI (2.28, 3.95)] in crude model as well as in multi-adjusted model [HR=1.37, 95% CI (1.10, 2.10)]. NAFLD effect on CVD remained significant only in MedDietScore below median [HR=1.38, 95% CI (1.00, 1.93)] whereas it was essentially null [HR=1.00, 95% CI (0.38, 2.63)] among subjects with lower score. Mediation analysis revealed that adiponectin was the strongest mediator of the examined associations. In particular, per 1 mg/dL serum adiponectin about 21% lower likelihood to have NAFLD was observed. Stratified analysis using the median value of adiponectin level (4.1 mg/dL) revealed that the inverse association between Mediterranean diet and NAFLD reached the level of significance only in case of adiponectin levels over the median values (OR (3rd vs. 1st MedDietScore tertile) = 0.69 95% CI (0.35, 0.98)). Conclusion Beyond the inverse association between Mediterranean diet and NAFLD several novelties were revealed; low-carbohydrate Mediterranean diet was more protective against NAFLD in women while an adiponectin-related mechanism was revealed. Funding Acknowledgement Type of funding source: Other. Main funding source(s): This work was supported by a research grant from Hellenic Atherosclerosis Society. The ATTICA study is supported by research grants from the Hellenic Cardiology Society [HCS2002] and the Hellenic Atherosclerosis Society [HAS2003]