Effects of nuts consumption on blood lipids: a meta-analysis

It is well known that diet may influence cardiovascular disease (CVD) occurrence. In the search for dietary bioactive components able to decrease CVD risk, nuts have begun to attract large attention. In particular,most studies in literature have mainly focused the attention on almonds and walnuts cardio-protective effects. Nuts are rich sources of mono and poly-unsaturated fatty acids which may improve blood lipids and other CVD risk factors. Moreover, several studies indicates that other compounds (i.e. fiber, sterols, copper, magnesium, phenolic compounds) in nuts can reduce total cholesterol and LDL-cholesterol beyond the effect on fatty acids. The aim of our work was to evaluatethe effects of nuts consumption on serum lipid profile in humans by means of a meta-analysis including those nuts which have been less investigated, i.e. pecan nuts and pistachios. The meta-analysis included international peer rewieved articles in the Medline database (www.pubmed.com; National Library of Medicine, Bethesda, MD) with no time limits. Published controlled trials that compared nuts-enriched diet (almonds, walnuts, pecan nuts, macadamia nuts, pistachios, pea nuts, Brazil nuts, cashews, hazelnuts) with a control diet in humans were included. Outcomes of interest were: mean serum total cholesterol (TC), low-density lipoprotein (LDL) cholesterol, high-density lipoprotein (HDL) and triglycerides (TG) and the LDL/HDL ratio. These values were expressed as the mean change from baseline to follow-up. Study quality was assessed by the Jadad score level of evidence rating for randomized controlled trials (Jadad et al., 1996).The mean change from baseline to follow-up for each intervention and control group was calculated, if not reported by using Student’s ttest. A DerSimonian and Laird random-effects model was used in calculating the weighted mean difference (WMD) and its 95% confidence interval. Statistical heterogeneity was addressed using the I2statistic. Visual inspection of funnel plots was used to assess the presence of publication bias. Statistics were performed using Review Manager (RevMan, Version 5.0, Copenhagen, 2008). Thirty studies were included in the analysis. Nuts consumption range from 2 and 24 weeks. Initiallyan analysis with all studies accepted was conducted. The weighted mean difference (WMD) show a reduction in serum TC, LDL levels and LDL/HDL ratio in patients consuming nuts than those fed a control diet:TC [-0.28 (-0.35; -0.21) ]; LDL [-0.25 (-0.32; 0.04) ]; and LDL/HDL [-0.17 (-0.29; 0.11)]. TheWMD for TG and HDL was negligible, even though its trend in favor of diets enriched with nuts: TG [-0.04 (-0.08; 0.00)]; HDL [0.01 (-0.03; 0.04)]. Then we carried out a meta-analysis of subgroups, each formed by the studies concerning only one type of nut consumption. Results show that pistachios [-0.13 mmol/L (I.C. 95% -0.18, -0.07)] and pecan nuts [-0.12 mmol/L (I.C. 95% -0.19, -0.05)] have a higher capacity to reduce serum TG than walnuts [-0.04 mmol/L (I.C. 95% -0.12, 0.03)] and almonds [-0.07 mmol/L (I.C. 95% -0.16, 0.01)]. Reduction of serum TC and LDL levels is high especially for pistachios and pecan nuts [-0.45 mmol/L (I.C. 95% -0.51, -0.40) and -0.39 mmol/L (I.C. 95% -0.47,-0.31), respectively]. Furthermore, pecan nuts and pistachios act favorably on LDL/HDL ratio. Our study shows that nuts can positively influence blood lipids. Best results may be obtained for pecan nuts and pistachios; these positive effects are very likely due to their lipid composition, in particular the low content of saturated fatty acids.