Scientific Opinion on the substantiation of health claims related to
walnuts and maintenance of normal blood LDL-cholesterol concentrations
(ID 1156, 1158) and improvement of endothelium-dependent vasodilation
(ID 1155, 1157) pursuant to Article 13(1) of Regulation (EC) No 1924/2006[sup]1[/sup]
EFSA Panel on Dietetic Products, Nutrition and Allergies (NDA)2, 3
European Food Safety Authority (EFSA), Parma, Italy
Słowa kluczowe:
Walnuts
cholesterol
fatty acids
health claims
vasodilation
1. Charakterystyka żywności / składnika
The food that is the subject of the health claims is walnuts.
Walnuts are seeds from the walnut tree (genus Juglans) of which about 20 different species are known in different parts of the world. Walnuts are drupes, rather than nuts. One of the most popular varieties of walnut is the Persian or English walnut, which has a large seed and a thinner shell, yielding more edible walnut meat by weight than other species. Black walnuts are another commonly sold walnut species, as are white walnuts, also called butternuts. The nut kernels contain about 57-65 % fat, of which about 3-6 % are saturated fatty acids (SFAs), 9-15 % are monounsaturated fatty acids (MUFAs), and 35-47 % are polyunsaturated fatty acids (PUFAs) (33-38 % linoleic acid (LA), and 2-9 % alpha-linolenic acid (ALA)). The nut kernels also contain about 5-7 % fibre, of which about 25 % is soluble fibre, 15-29 % protein, and small amounts of plant sterols and other phytochemicals. The Panel notes the variation in terms of macronutrient composition between different types of walnuts.
The Panel considers that the food, walnuts, which is the subject of the health claims, is sufficiently characterised in relation to the claimed effects.
3. Naukowe uzasadnienia wpływu na zdrowie człowieka -
3.1. Utrzymanie prawidłowego stężenia cholesterolu LDL we krwi (ID 1156, 1158)
Among the references provided for the scientific substantiation of the claim were consensus opinions or textbook chapters on the effects of different dietary fatty acids, as well as narrative reviews and human studies on the effect of diets with different fatty acid composition on cardiovascular risk factors. Some human studies addressed outcomes unrelated to the claimed effect, such as postprandial endothelial function or cardiovascular risk. The Panel considers that no conclusions can be drawn from these references for the scientific substantiation of the claimed effect.
All of the human intervention studies provided in the consolidated list on the effects of the consumption of walnuts on blood lipids were considered in the systematic review by Mukuddem- Petersen et al. (2005), which addressed the effects of nuts in general, including walnuts, on blood lipid profile. Nine of the studies selected used walnuts as the intervention (Abbey et al., 1994; Almario et al., 2001; Chisholm et al., 1998; Iwamoto et al., 2002; Morgan et al., 2002; Munoz et al., 2001; Ros et al., 2004; Sabate et al., 1993; Zambon et al., 2000). In two of these studies the intervention and control periods were in the same order for all subjects, with no wash-out period in between (Abbey et al., 1994; Almario et al., 2001); the Panel notes that these studies were not controlled for a temporal effect, and considers that no conclusions can be drawn from these studies for the scientific substantiation of the claimed effect.
Among the seven studies with randomised, controlled cross-over designs, three used a Mediterranean diet as control (Munoz et al., 2001; Ros et al., 2004; Zambon et al., 2000), two used a National Cholesterol Education Program (NCEP) step I diet (Morgan et al., 2002; Sabate et al., 1993), one used a low fat diet (Chisholm et al., 1998), and one used a Japanese diet (Iwamoto et al., 2002). The duration of the intervention was 4-6 weeks. Four of the studies, which included a total of 127 normo- and hypercholesterolaemic subjects with a daily walnut intake of 40-84 g, observed a significant decrease in total (from -4.3 to -12.4 %) and LDL-cholesterol (from -6.7 to -16.3 %) concentrations during the walnut diet compared to the “control” diet (Iwamoto et al., 2002; Ros et al., 2004; Sabate et al., 1993; Zambon et al., 2000), whereas in the remaining three studies differences between intervention and control groups were not significant (n=73; walnut intake = 41-78 g/day). No significant differences between groups were observed in any of the studies with respect to HDL- cholesterol and triglyceride concentrations, with the exception of a significant decrease in HDL-
cholesterol (-4.9 %) in the study by Sabate et al. (1993), where walnuts primarily replaced MUFAs. The Panel notes that the data from these studies could not be pooled in a meta-analysis because of the heterogeneity of the designs, particularly with respect to the control diet and type of subjects included (normocholesterolaemic vs. hypercholesterolaemic). The Panel also notes that in these studies walnuts replaced various types of dietary fats in different amounts (SFAs, MUFAs), which had variable effects on the lipid profile (Mensink et al., 2003). None of these studies controlled for the fatty acid composition of the intervention diet, and therefore it could not be assessed whether the consumption of walnuts had an effect on blood lipids beyond what could be expected from their fatty acid profile (e.g. LA and ALA content). In the systematic review by Mukuddem-Petersen et al. (2005), the predictive equation by Mensink and Katan (1992) was used to compare the actual observed changes in total and LDL-cholesterol with the predicted changes in total and LDL-cholesterol concentrations, based on differences in the dietary fatty acid composition in the individual nut and control diets. However, all studies on nuts included in the review were pooled for this calculation, and no particular figure for walnuts only was provided.
One observational human study addressed the association between blood cholesterol and walnuts (oil and kernel) consumption. Lavedrine at al. (1999) carried out a cross-sectional survey on 793 subjects (426 males and 367 females) aged 18-65 years, from Dauphine, France, where walnuts and walnut oil are part of the usual diet. To assess the level of walnut consumption, a food frequency questionnaire (1-year recall) was used. Blood samples were taken to measure LDL and total cholesterol. No significant differences in blood LDL or total cholesterol concentrations were observed between non- consumers, intermediate consumers and frequent consumers of walnuts and walnut oil. The Panel notes that this cross-sectional study does not show an association between walnut or walnut oil consumption and blood LDL-cholesterol concentrations.
In weighing the evidence, the Panel took into account that the evidence provided did not establish that consumption of walnuts had an effect on blood LDL-concentrations beyond what could be expected from their fatty acid composition and that the LDL-cholesterol-lowering effect of walnuts could be attributed to their content of MUFAs and PUFAs.
The Panel concludes that a cause and effect relationship has not been established between the consumption of walnuts and maintenance of normal blood LDL-cholesterol concentrations beyond what could be expected from the fatty acid composition of walnuts.
A claim on the replacement of mixtures of SFAs with cis-MUFAs and/or cis-PUFAs in foods or diets and maintenance of normal blood LDL-cholesterol concentrations has already been assessed with a favourable outcome (EFSA Panel on Dietetic Products Nutrition and Allergies (NDA), 2011).
A claim on linoleic acid and maintenance of blood cholesterol concentrations (EFSA Panel on Dietetic Products Nutrition and Allergies (NDA), 2009a) and a claim on alpha-linolenic acid and maintenance of blood cholesterol concentrations (EFSA Panel on Dietetic Products Nutrition and Allergies (NDA), 2009b) have also already been assessed with favourable outcomes.
3.2. Poprawa rozszerzenia naczyń krwionośnych zależnego od śródbłonka (ID 1155, 1157)
The references provided for the scientific substantiation of the claim included narrative reviews, consensus opinions, and epidemiological and intervention studies on the effects of PUFAs, including ALA and LA, on health outcomes other than improvement of endothelium-dependent vasodilation (e.g. blood lipids, blood pressure and cardiovascular risk). In addition, some epidemiological studies which investigated the association between nut consumption, including walnuts, on health outcomes not related to the claimed effect (e.g. blood lipids, blood pressure and cardiovascular risk) were provided. The Panel considers that no conclusions can be drawn from these references for the scientific substantiation of the claimed effect.
In a randomised, controlled cross-over trial by Ros et al. (2004) 21 untreated hypercholesterolaemic men and women consumed a Mediterranean type diet low in saturated fatty acids (about 5 % energy) and a diet of similar energy and fat content (33 E%) in which 40-65 g/day of walnuts contributed to about 18 % total energy, and replaced approximately 32 % of the energy obtained from monounsaturated fat (olive oil, olives and avocados) in the control diet, for 4 weeks each without a wash-out period. There was a 4-week run-in period to allow for diet and weight stabilisation. The order of the diets was randomised. After each intervention, ultrasound measurements of brachial artery vasomotor function were taken. Endothelium-dependent vasodilation (EDV) was the primary outcome of the study. Endothelium-independent vasodilation (EIDV) was evaluated by administering 0.4 mg sublingual glyceryl trinitrate prior to the measurement, and was used as internal control. Reproducibility of measurements over one month was documented in 15 healthy volunteers. The
repeatability coefficient was 5.16, and the mean SD difference between EDV values was
0.52 2.66 %. It was calculated that 19 subjects were needed to detect a mean difference in EDV of
2 %, with a type I error of 5 % and a power 90 %. A total of 20 subjects completed the study and suitable brachial artery ultrasound measurements were available for 18 subjects. Compared with the
Mediterranean diet, the walnut diet significantly improved EDV (from 3.6 3.3 % to 5.9 3.3 %, a relative increase of 64 %), and significantly reduced concentrations of vascular cell adhesion molecule-1 (VCAM-1), which is a marker of endothelial activation. EIVD and concentrations of intercellular adhesion molecule-1 (ICAM-1) were not significantly different between diets. The walnut diet significantly decreased total cholesterol (-4.4±7.4 %) and LDL-cholesterol (-6.4±10.0 %) concentrations. Changes in EDV significantly correlated with changes in the total/HDL-cholesterol ratio. There was no evidence of carry-over effects between treatment periods. The Panel notes that this study provides evidence for a positive and sustainable effect of walnut consumption on endothelium-dependent vasodilation.
In a randomised, controlled cross-over trial by Ma et al. (2010), 24 type 2 diabetic subjects aged 30-75 years on stable medication use for at least three months (17 on blood pressure-lowering medication, 13 on cholesterol-lowering medication, 10 on oral antidiabetic medication) consumed a walnut-enriched ad libitum diet, and an ad libitum diet without walnuts, for 8 weeks each in a random order with an 8-week wash-out period in between, and after a 4-week run-in period to allow for diet and weight stabilisation. During the walnut period, subjects consumed 56 g English walnuts daily. EDV was the primary outcome of the study. A stimulus-adjusted response measure was used as internal control. The sample size was determined to allow for about 20 % attrition and
non-compliance, and to provide 80 % power to detect a minimal difference of 2.5 % in EDV between treatments, with a maximum allowable type I error of 5 %. A total of 96 % of subjects
complied with the consumption of walnuts (defined as intake of 80 % of the assigned dose). EDV improved significantly after consumption of the walnut-enriched diet as compared with the ad libitum
diet without walnuts (2.2 1.7 % vs. 1.2 1.6 %, p=0.04), whereas no significant changes between groups were observed in the stimulus-adjusted response measure. No significant differences were observed between diets with respect to changes in blood lipid profile, body weight, blood glucose control or insulin sensitivity, whereas a significant decrease in systolic and diastolic blood pressure was observed in the walnut diet compared to the control diet. The Panel notes that the results of this study are consistent with findings in human intervention studies in untreated subjects on the effect of walnut consumption on endothelium-dependent vasodilation.
In a randomised cross-over study, Cortes et al. (2006) investigated whether the addition of walnuts or olive oil to a fatty meal had different effects on post-prandial vasoactivity, lipoproteins, markers of oxidation and endothelial activation, and plasma asymmetric dimethylarginine (ADMA). Endothelium-dependent vasodilation was the primary outcome of the study. A total of 12 healthy subjects and 12 subjects with hypercholesterolaemia who received no pharmacological treatment were randomised to two high-fat meal sequences (separated by one week) to which 25 g olive oil or 40 g walnuts had been added. Both test meals contained 63 % of the energy as fat (80 g fat) and 35 E% saturated fatty acids. Ultrasound measurements of brachial artery endothelial function were performed
after fasting and 4 h after test meals. In both study groups, flow-mediated dilation (FMD) was lower after the olive oil meal than after the walnut meal (p=0.006). Flow-independent dilation and plasma ADMA concentrations were unchanged, and concentrations of oxidised low-density lipoproteins decreased (p=0.051) after both meals. The plasma concentrations of soluble inflammatory cytokines, and of adhesion molecules, decreased (p<0.01) independently of meal type, except for E-selectin, which decreased significantly more after the walnut meal. E-selectin is an adhesion molecule involved in the early steps of monocyte recruitment to the endothelium. The Panel notes that this study is an acute study, and does not allow conclusions to be drawn on the sustainability of the effects of the consumption of walnuts on EDV.
In weighing the evidence, the Panel took into account that one intervention study in healthy subjects, adequately powered and controlled, showed a sustained effect of the consumption of walnuts on endothelium-dependent vasodilation, that the results of one additional intervention study in type 2 diabetic subjects on blood pressure-lowering, cholesterol-lowering, and/or oral antidiabetic medication are consistent with these findings, and that an acute intervention study also showed a positive effect of the consumption of walnuts on endothelium-mediated vasodilation.
The Panel concludes that a cause and effect relationship has been established between the consumption of walnuts and improvement of endothelium-dependent vasodilation.