Nutrition, Metabolism & Cardiovascular Diseases (2014) xx, 1e14
Available online at www.sciencedirect.com
Nutrition, Metabolism & Cardiovascular Diseases journal homepage: www.elsevier.com/locate/nmcd
SYSTEMATIC REVIEW
A systematic review of vascular and endothelial function: Effects of fruit, vegetable and potassium intake N. Blanch, P.M. Clifton, J.B. Keogh* School of Pharmacy and Medical Science, University of South Australia, Australia Received 2 June 2014; received in revised form 2 October 2014; accepted 4 October 2014 Available online - - -
KEYWORDS Potassium; Fruit; Vegetable; Endothelial function; Vascular function
Abstract Aim: To review the relationships between: 1) Potassium and endothelial function; 2) Fruits and vegetables and endothelial function; 3) Potassium and other measures of vascular function; 4) Fruits and vegetables and other measures of vascular function. Data synthesis: An electronic search for intervention trials investigating the effect of potassium, fruits and vegetables on vascular function was performed in MEDLINE, EMBASE and the Cochrane Library. Potassium appears to improve endothelial function with a dose of >40 mmol/ d, however the mechanisms for this effect remain unclear. Potassium may improve measures of vascular function however this effect may be dependent on the effect of potassium on blood pressure. The effect of fruit and vegetables on endothelial function independent of confounding variables is less clear. Increased fruit and vegetable intake may improve vascular function only in high risk populations. Conclusion: Increasing dietary potassium appears to improve vascular function but the effect of increasing fruit and vegetable intake per se on vascular function is less clear. ª 2014 Elsevier B.V. All rights reserved.
Introduction Vascular function measures are utilised as surrogate, noninvasive measures of the risk of heart disease and stroke. Multiple methodologies used to assess vascular function and can be categorised as measures of arterial stiffness and measures of endothelial function. Arterial stiffness measures include pulse wave velocity (PWV), augmentation index (AI) and pulse pressure (PP). Endothelial function measures include flow mediated dilatation (FMD), peripheral artery tonometry (PAT) and laser Doppler imaging (LDI); and are outlined below.
* Corresponding author. School of Pharmacy and Medical Science, University of South Australia, GPO Box 2471, Adelaide, SA 5000, Australia. Tel.: þ61 8 8302 2579; fax: þ61 8 8302 2389. E-mail address: [email protected] (J.B. Keogh).
PWV is considered as the gold standard method for assessing aortic stiffness, and is a strong predictor of future cardiovascular (CV) events and all-cause mortality [1]. It is assessed by measuring the difference in arrival time of the upstroke of the pressure wave at the femoral artery and at the common carotid artery combined with the distance between these two points and fully automated measurement is now available. Faster velocities indicate stiffer arteries. AI is also an index of arterial stiffness [2] strongly correlated with PWV and is an independent predictor of CV events [3]. It is measured by examining the radial (or femoral or carotid) artery profile and examining the time of arrival of the reflected wave and its contribution to systolic pressures. Higher values indicate stiffer arteries. Ambulatory Arterial Stiffness Indexes (AASI) is the linear regression slope of SBP vs DBP obtained through 24-h noninvasive ambulatory BP monitoring and s-AASI
http://dx.doi.org/10.1016/j.numecd.2014.10.001 0939-4753/ª 2014 Elsevier B.V. All rights reserved.
Please cite this article in press as: Blanch N, et al., A systematic review of vascular and endothelial function: Effects of fruit, vegetable and potassium intake, Nutrition, Metabolism & Cardiovascular Diseases (2014), http://dx.doi.org/10.1016/j.numecd.2014.10.001
2
(symmetric ambulatory arterial stiffness index) has been shown to provide a clinically relevant evaluation of arterial stiffness [4]. Systemic arterial compliance (SAC) is a measure of large artery stiffness determined from carotid pressure and aortic volume flow measures and has been shown to contribute to higher brachial and carotid systolic and pulse pressure [5]. Endothelial dysfunction is strongly associated with traditional cardiovascular risk factors and precedes the development of clinical disease [6]. One of the major functions of the endothelium is to produce vasoactive molecules including nitric oxide (NO), a potent vasodilator. The assessment of endothelial function by measurement of ischaemia-induced FMD, an NO mediated response in the conduit arteries of the peripheral circulation allows the non-invasive study of vascular physiology in human subjects using ultrasound or MRI [7]. FMD is correlated with the extent and severity of coronary artery disease [8] and predicts events, although whether this is independent of traditional risk factors is not clear. PAT measures endothelial function by assessing finger pulse wave amplitude (PWA) at rest and during shear stress [9]. Reactive Hyperaemic-PAT (RH-PAT) is a ratio calculated using the post-occlusion (shear stress) PWA to the baseline (rest) PWA, expressed as reactive hyperaemic index (RHI) [9]. The lnPAT (log EndoPAT) ratio is the natural logarithm of the ratio of the pulse amplitude recorded 90e120 s after cuff release to the baseline amplitude divided by the hyperaemic to baseline ratio in the contralateral control finger [9]. PAT is an independent predictor of CV events and all-cause mortality in populations with established CVD [10]. Endothelial function is commonly measured using forearm blood flow (FBF) assessed by strain gauge venous plethysmography during intra-arterial infusion of agonist (e.g. acetylcholine which releases NO) and antagonists (e.g. NG-monomethyl L-arginine [L-NMMA] which blocks NO production) or after ischaemia [11]. Reduced FBF measurements in response to intra-arterial infusion of acetylcholine are associated with increased risk of CV events [12]. LDI with iontophoresis of acetylcholine measures the microvascular perfusion across a selected region of the peripheral circulation [13]. Acetylcholine acts on muscarinic receptors of the endothelium producing endothelium-dependent vasodilation [14]. Impaired endothelium dependent vasodilation as assessed by LDI with iontophoresis of acetylcholine is associated with increased CHD [15]. The measurement of arterial compliance (e.g. AI) following administration of salbutamol (a b2adrenergic receptor agonist that reduces arterial stiffness in an NO-dependent manner) can be used as a measure of endothelial function [16]. A reduced response to salbutamol administration has been demonstrated in hypercholesterolemic populations and is related to FBF responses to acetylcholine [17]. The association of high potassium diets with reduced stroke risk is now generally accepted within the literature independent of various confounding variables including sodium intake [18]. The mechanisms for this are unclear but may be via an effect on blood pressure [19].
N. Blanch et al.
Diets high in fruit and vegetables are associated with reduced cardiovascular disease. Fruit and vegetables are a source of nutrients including potassium and non-nutrients such as polyphenols, and the high content of potassium in fruits and vegetables is a plausible mechanism to explain the apparent CV protection of fruits in vegetables in epidemiological reports [20]. The purpose of this review is to summarise the effects of fruits, vegetables and potassium on endothelial and vascular function in intervention studies. Method A systematic search was performed in MEDLINE, EMBASE and the Cochrane Library for original research articles investigating the relationship between potassium, fruits and vegetables and vascular function published before 01 June 2014. Reference lists of retrieved articles were also searched for relevant articles. No restrictions were placed on publication date. Studies were limited to those published in English, studies in humans and clinical trials. Key search terms were “potassium” OR “fruit” OR “vegetable” AND “endothelial function” OR “vascular function”. Individual fruits or vegetables were not included in the search strategy but were identified if “fruit” or “vegetable” was listed as a keyword, or from reference lists of retrieved articles. Eligibility criteria To be included in this review a published study had to meet the following criteria: 1) original research; 2) controlled intervention studies aiming to determine the effect of potassium, fruits or vegetables on endothelial and vascular function. Studies including vitamin, mineral supplementation (other than potassium), dietary changes (other than fruits and vegetables), mixed interventions (e.g. Low fat combined with increased fruit/vegetables), lifestyle modification (e.g. exercise or stress management) or pharmacological treatment were excluded. Dietary studies specifically increasing nitrate intake were excluded as this area has been reviewed recently [21].Vascular function was defined as measures of arterial compliance (PWV, AI, PP) and endothelial function. Blood pressure was excluded as this has been recently reviewed [19,22]. Biochemical markers of endothelial dysfunction included in this review were nitrate/nitrite as indices of NO production, adhesion molecules, Von Willebrand’s factor (vWF) which can increase the risk of clotting, and endothelin-1 (ET-1) which is a pro-inflammatory, proproliferative vasoconstrictor peptide [11]. Inflammatory molecules and cytokines were not included in this review and have been published elsewhere [23,24]. Crosssectional, cohort, animal and cellular studies were not the focus of this review but were included to help explain the rationale and mechanisms behind the observed effect. Searches identified 4063 potential articles of which all but 44 were excluded on the basis of title and abstract (Fig. 1).
Please cite this article in press as: Blanch N, et al., A systematic review of vascular and endothelial function: Effects of fruit, vegetable and potassium intake, Nutrition, Metabolism & Cardiovascular Diseases (2014), http://dx.doi.org/10.1016/j.numecd.2014.10.001
A systematic review of vascular and endothelial function
3
Review
sensitivity may play a role in explaining the inconsistency of findings, with one study showing potassium supplementation (4.5 g/d potassium chloride) increased plasma and urine NO levels in normotensive and mildly hypertensive salt sensitive subjects during salt loading [30]. Serum potassium may be transiently increased within the physiological range after supplementation, which may directly affect the small and intermediate calcium-activated potassium channels of the endothelial cells which are then electrically coupled to the smooth cells e i.e. an endothelium derived hyperpolarizing factor effect [31]. Evidence suggests the vascular response to increases in extracellular potassium from potassium supplementation is mediated by NO, as infusion of L NMMA (an NO synthase inhibitor) abolishes the enhanced vasodilatory response to potassium with acetylcholine [32]. Inwardly rectifying K channels (KIR) and the Naþ/Kþ ATPase are involved in the control of basal tone of human forearm resistance vessels [33] and these are involved in Kþ induced vasorelaxation within these vascular beds [33,34]. Furthermore, intra-arterial administration of glibenclamide, a KIR channel closer, has been shown decrease reactive hyperaemic flow [32]. Combining administration of glibenclamide and L-NMMA did not further affect the reduction in reactive hyperaemic flow, suggesting the KIR channels and NO response are not independent pathways [32].
Potassium and endothelial function
Fruit and vegetables and endothelial vasodilator function
Table 1 summarises studies investigating the effects of increasing potassium intake with diet or supplementation on endothelial function. We have recently shown in a group of healthy adults, that a 36 mmol potassium load can attenuate the postprandial reduction in FMD within 2 h [25]. Furthermore an increase of 48 mmol/d in potassium intake achieved by including bananas and potatoes in place of apples and rice significantly improved fasting FMD after only one week [26]. In addition, a study by He et al. demonstrated that fasting FMD was significantly improved following 4 weeks of 64 mmol/d potassium chloride and 64 mmol/d potassium bicarbonate compared to placebo [27]. There was no difference between the potassium chloride and potassium bicarbonate treatments. Conversely, Berry et al. [28] found no effect on fasting FMD by increasing potassium intake by 20 mmol/d and 40 mmol/d through fruit and vegetable intake or with a 40 mmol/d potassium citrate supplement when compared to placebo. In another study, six weeks of 64 mmol/d potassium chloride supplementation did not evoke changes in endothelial function, as assessed by changes in pulse wave analysis following GTN and salbutamol administration, when compared to placebo [29]. Given the inconsistency of these findings, it remains unclear whether increasing dietary potassium can improve endothelial function. There is no evidence suggesting increased potassium intake is harmful to endothelial function. The mechanisms by which increased dietary potassium may improve endothelial function are unclear. Salt-
Increases in fruit and vegetable intake have been shown to improve endothelial function in a dose dependent manner [35]. McCall et al. [35] demonstrated that an additional portion (80 g) of fruit and vegetable consumed daily led to a 6.2% improvement in endothelium-dependent FBF responses after 8 weeks of intervention. In addition, it has recently been demonstrated that tomato paste supplementation (70 g/d) significantly improved FMD after two weeks [36] compared with a no supplement control period but it is not practical to expect individuals to consume this amount of tomato paste on a regular basis. This effect was not seen in the 24 h following a single dose (70 g). Conversely, another study showed no effect on FMD after 7 days of supplementation with tomato puree (70 g/d) in postmenopausal, healthy women compared with placebo [37]. Fruit and vegetable supplement capsules have been shown to attenuate the FMD reduction in response to a high fat meal compared with placebo capsules [38] while a recent study demonstrated polyphenols-rich juice from Concord grapes significantly improved FMD compared with polyphenols-poor juice and prevented a postcigarette reduction in FMD after only 7 days, suggesting polyphenols are important in determining endothelial function [39]. Encapsulated grape solids (6 500 mg/d for two weeks) had no effect on FMD response to a high fat meal [40]. In a study investigating the effects fruit and vegetable extracts on FMD in an overweight population with metabolic syndrome, there were no differences following 8 weeks of supplementation when compared to
Figure 1
Flow diagram of article selection.
Please cite this article in press as: Blanch N, et al., A systematic review of vascular and endothelial function: Effects of fruit, vegetable and potassium intake, Nutrition, Metabolism & Cardiovascular Diseases (2014), http://dx.doi.org/10.1016/j.numecd.2014.10.001
4
Reference
Study design
Study participants
Intervention
Outcome measures
Effects of intervention
Effect on EF
Effect on VF
Berry [28]
Randomised cross-over study of four 6 week interventions
48 individuals aged 22 e65 y, BMI 20e35 kg/ m2
Fruit and vegetable (20 mmol/d), Fruit and vegetable (40 mmol/d), and potassium citrate (40 mmol/d) compared to placebo capsules
FMD, PWV, peripheral AI, ICAM
4
4
Blanch [25]
Randomised cross-over acute postprandial test meal (2 h)
32 individuals aged 18 e70 y, BMI 18e30 kg/ m2
High potassium meal (36 mmol) compared to a low potassium meal (6 mmol)
FMD
Blanch [26]
Randomised cross-over study of two 6 day interventions
35 individuals aged 18 e70 y, BMI 18e30 kg/ m2
High potassium diet (150 mmol/d) compared to low potassium diet (80 mmol/d)
FMD, PWVcef, AI, Eselectin, endothelin-1, ADMA
Graham [29]
Randomised cross-over study of two 6 week interventions
40 individuals aged 40 e70 y, with increased CVD risk
64 mmol potassium chloride compared to placebo
PWV, AI, ICAM, VCAM, E-selectin
He [27]
Randomised cross-over study of three 4 week interventions
42 individuals aged 18 e75 y
Potassium bicarbonate (64 mmol/d) and potassium chloride (64 mmol/d) capsules compared to placebo
FMD, PWV
Liu Z [56]
4-phase stepwise study design of 3 day run in and three 7 day interventions
155 normotensive and mild hypertensive individuals aged 36 e61 y
51.3 mmol/d (low) sodium chloride then 307.7 mmol/d (high) sodium chloride compared to 307.7 mmol/d sodium chloride and 60 mmol/ d potassium chloride
AASI, s-AASI, endothelin-1
No difference between treatment and placebo for FMD, PWV, peripheral AI. Y ICAM in female participants only following 40 mmol/ d fruit and vegetable treatment compared to placebo (p < 0.05). [ FMD over 2 h following the high potassium meal compared to the low potassium meal. [ fasting FMD and Eselectin following high potassium diet compared to low potassium diet. No effect on PWVcef, AI, ADMA, endothelin-1. Y PWV following potassium chloride compared to placebo (p Z 0.033), reduction lost after adjustment for SBP. No effect on AI, ICAM, VCAM or Eselectin. [ fasting FMD following potassium bicarbonate (p < 0.05) and potassium chloride (p < 0.001). Y PWV following potassium bicarbonate (p < 0.05) and potassium chloride (p < 0.001). Y AASI (p < 0.05), sAASI (p < 0.05), endothelin-1 (p < 0.05) following sodium chloride and potassium chloride intervention when compared to the high sodium intervention.
[
[
4
4
[
[
[
[
[
N. Blanch et al.
Please cite this article in press as: Blanch N, et al., A systematic review of vascular and endothelial function: Effects of fruit, vegetable and potassium intake, Nutrition, Metabolism & Cardiovascular Diseases (2014), http://dx.doi.org/10.1016/j.numecd.2014.10.001
Table 1 Effect of potassium interventions on endothelial and vascular function.
Abbreviations: AASI e ambulatory arterial stiffness indexes; ADMA e asymmetric dimethylarginine; AI e augmentation index; BMI e body mass index; EF e endothelial function; FMD e flow mediated dilatation; ICAM e intercellular adhesion molecule; NO e nitric oxide; PWV e pulse wave velocity; PWVcef e pulse wave velocity (carotidefemoral); s-AASI e symmetric ambulatory arterial stiffness index; VF e vascular function; VCAM e vascular adhesion molecule; [ e improvement; Y e deterioration; 4 e no effect.
[ Plasma and urine NO Low-salt loading (3 g/ d), high-salt loading (18 g/d) and high-salt loading plus potassium chloride (4.5 g/d) 39 normotensive and mild hypertensive individuals ( 28 kg/m2
20 g Marie Menard cider apple powdered (high polyphenols) compared to 19.3 golden delicious apple powdered (low polyphenols) 500 mL red orange juice/d compared to 500 mL placebo/d
FMD, plasma NO
[
Del Bo [43]
Randomised cross-over acute postprandial test meal (1 h)
10 healthy male subjects, aged 20.8 1.6 y, BMI 22.5 2.1 kg/m2
RHI, NO
Dohadwala MM [44]
Randomised cross-over study of two 4 week interventions
44 patients with stable coronary artery disease
Homogenised blueberries (300 g) compared to control jelly (sugar added equivalent to blueberry intervention) Double-strength cranberry juice (480 ml/ d) compared to calorie, taste, and appearancematched placebo beverage containing no polyphenols (480 ml/d)
[ FMD following red orange juice supplementation in the compared to placebo No effect of intervention on plasma NO No differences in RHI, NO compared to control
4
Eccleston C [66]
Randomised parallel design of two 8 week interventions Randomised cross-over study of two 6 week interventions
20 men, aged 18e55 y
sICAM-1
12 postmenopausal women, aged 57 1 y, BMI 38.1 2.1 kg/m2
300 ml/d Sea buckthorn juice compared to 300 ml/d placebo drink Watermelon extract (6 g/d) compared to placebo
Randomised cross-over acute postprandial test meal (2 h)
20 individuals aged 44.55 13.34 y, BMI 23.81 2.46 kg/m2
Y PWVcef following juice supplementation compared to placebo (p Z 0.003) No differences in FMD, lnPAT ratio, PWVcer, sICAM-1 between interventions sICAM-1 unaffected by either Sea buckthorn or placebo juice treatment Y PWVbea following watermelon supplementation compared to baseline (p < 0.001) and placebo (p < 0.01) No effect on AI following both interventions No effect of intervention or placebo on LDI, plasma NO, sICAM-1, sVCAM-1
Fruit Auclair [52]
Figueroa A [58]
Jin Y [68]
Lynn A [60]
250 ml 20% blackcurrant juice drink compared to 250 ml placebo drink
FMD, lnPAT ratio, PWVcer, PWVcef, sICAM-1
PWVbea, AI
LDI with iontophoresis of acetylcholine chloride, plasma NO, sICAM-1, sVCAM-1 PWV
Effect on VF
4
[
4
[
4
4
N. Blanch et al.
Please cite this article in press as: Blanch N, et al., A systematic review of vascular and endothelial function: Effects of fruit, vegetable and potassium intake, Nutrition, Metabolism & Cardiovascular Diseases (2014), http://dx.doi.org/10.1016/j.numecd.2014.10.001
Table 2 Effect of fruit or vegetable interventions on endothelial and vascular function.
48 healthy individuals, aged 30e50 y
Morand [47]
Randomised cross-over study of three 4 week interventions
Morand [47]
Randomised cross-over acute postprandial test (6 h)
24 healthy men aged 51 e63 y, BMI 25.2 e30.5 kg/m2
Poreba R [45]
3-phase stepwise study design of 6 week intervention, washout, repeated intervention
35 men with mild hypercholesterolaemia, mean age 53.9, BMI 27.4 kg/m2
Riso P [42]
Randomised cross-over study of two 6 week interventions
18 male volunteers mean age 47.8 9.7 y, BMI 24.8 2.6 kg/m2
Siasos G [39]
Randomised cross-over study of two 2 week interventions
26 smokers, aged 26 5 y
Pomegranate juice (330 ml/d) compared to control drink of lemonade (330 ml/d; 7 UP) Orange juice (500 ml) compared to a control drink (500 ml) plus 2 placebo capsules (2 146 mg starch) compared to control drink (500 ml) plus 2 capsules of pure hesperidin (2 146 mg) Orange juice (500 ml) compared to a control drink (500 ml) plus 2 placebo capsules (2 146 mg starch) compared to control drink (500 ml) plus 2 capsules of pure hesperidin (2 146 mg) 250 ml/d 100% chokeberry juice compared to abstinence from chokeberry juice
Wild blueberry drink (prepared by suspending 25 g of wild blueberry freeze-dried powder in 250 mL of water) compared to placebo drink (250 mL water, 7.5 g fructose, 7 g glucose, 0.5 g citric acid and 0.03 g blueberry flavour) 100% Concord grape juice (high polyphenol) compared to low polyphenol grapefruit juice placebo (7 cc/kg/ d)
A systematic review of vascular and endothelial function
No effect of intervention or placebo on PWV
LDI with iontophoresis of acetylcholine chloride, ICAM, VCAM, NO, vWF
No significant effect on LDI, ICAM, VCAM, NO, vWF between treatments compared to placebo drink
4
LDI with iontophoresis of acetylcholine chloride
[ LDI with acetylcholine following orange juice compared to control drink plus hesperidin (p < 0.05) and placebo (p < 0.01)
[
FMD, NO
[ FMD after juice and abstinence compared to baseline (p < 0.001). [ FMD after juice compared to abstinence (p < 0.01) [ NO throughout the study compared to baseline (p < 0.01) No differences in RHI, FRHI, AI, AI@75 bpm NO, sVCAM-1 following each intervention compared to baseline
[
RHI, FRHI (Framingham reactive hyperaemia index), AI, AI@75 bpm, NO, sVCAM-1
FMD, PWVcef measured before and after smoking a single cigarette (0 and 20 min postcigarette)
[ FMD following Concord grape juice supplementation prior to cigarette (p Z 0.02) and attenuated post cigarette decrease
4
4
[
[
(continued on next page)
7
Please cite this article in press as: Blanch N, et al., A systematic review of vascular and endothelial function: Effects of fruit, vegetable and potassium intake, Nutrition, Metabolism & Cardiovascular Diseases (2014), http://dx.doi.org/10.1016/j.numecd.2014.10.001
Randomised cross-over study of two 4 week interventions
8
Reference
Van Mierlo [40]
Study design
Study participants
Intervention
Outcome measures
Effects of intervention when compared to placebo (p < 0.001). Y PWVcef following Concord grape juice supplementation prior to cigarette (p Z 0.04) and attenuated post cigarette increase when compared to placebo (p < 0.001). No differences in FMD, large artery elasticity index small artery elasticity index after a high fat meal compared to placebo
Effect on EF
Effect on VF
4
4
Randomised cross-over study of three 2 week interventions
35 healthy men aged 18 e45 y, BMI 18e32 kg/ m2
Grape solids (wine grape and grape seed) 6 500 mg/d compared to placebo capsules
FMD, large artery elasticity index small artery elasticity index after a high fat meal
Randomised cross-over study of two 24 h and 7 day interventions
19 postmenopausal women
FMD
No effect of intervention or control meal on FMD
4
Upritchard JE [67]
Randomised parallel design of four 4 week interventions
52 Type 2 diabetics aged
Available online at www.sciencedirect.com
Nutrition, Metabolism & Cardiovascular Diseases journal homepage: www.elsevier.com/locate/nmcd
SYSTEMATIC REVIEW
A systematic review of vascular and endothelial function: Effects of fruit, vegetable and potassium intake N. Blanch, P.M. Clifton, J.B. Keogh* School of Pharmacy and Medical Science, University of South Australia, Australia Received 2 June 2014; received in revised form 2 October 2014; accepted 4 October 2014 Available online - - -
KEYWORDS Potassium; Fruit; Vegetable; Endothelial function; Vascular function
Abstract Aim: To review the relationships between: 1) Potassium and endothelial function; 2) Fruits and vegetables and endothelial function; 3) Potassium and other measures of vascular function; 4) Fruits and vegetables and other measures of vascular function. Data synthesis: An electronic search for intervention trials investigating the effect of potassium, fruits and vegetables on vascular function was performed in MEDLINE, EMBASE and the Cochrane Library. Potassium appears to improve endothelial function with a dose of >40 mmol/ d, however the mechanisms for this effect remain unclear. Potassium may improve measures of vascular function however this effect may be dependent on the effect of potassium on blood pressure. The effect of fruit and vegetables on endothelial function independent of confounding variables is less clear. Increased fruit and vegetable intake may improve vascular function only in high risk populations. Conclusion: Increasing dietary potassium appears to improve vascular function but the effect of increasing fruit and vegetable intake per se on vascular function is less clear. ª 2014 Elsevier B.V. All rights reserved.
Introduction Vascular function measures are utilised as surrogate, noninvasive measures of the risk of heart disease and stroke. Multiple methodologies used to assess vascular function and can be categorised as measures of arterial stiffness and measures of endothelial function. Arterial stiffness measures include pulse wave velocity (PWV), augmentation index (AI) and pulse pressure (PP). Endothelial function measures include flow mediated dilatation (FMD), peripheral artery tonometry (PAT) and laser Doppler imaging (LDI); and are outlined below.
* Corresponding author. School of Pharmacy and Medical Science, University of South Australia, GPO Box 2471, Adelaide, SA 5000, Australia. Tel.: þ61 8 8302 2579; fax: þ61 8 8302 2389. E-mail address: [email protected] (J.B. Keogh).
PWV is considered as the gold standard method for assessing aortic stiffness, and is a strong predictor of future cardiovascular (CV) events and all-cause mortality [1]. It is assessed by measuring the difference in arrival time of the upstroke of the pressure wave at the femoral artery and at the common carotid artery combined with the distance between these two points and fully automated measurement is now available. Faster velocities indicate stiffer arteries. AI is also an index of arterial stiffness [2] strongly correlated with PWV and is an independent predictor of CV events [3]. It is measured by examining the radial (or femoral or carotid) artery profile and examining the time of arrival of the reflected wave and its contribution to systolic pressures. Higher values indicate stiffer arteries. Ambulatory Arterial Stiffness Indexes (AASI) is the linear regression slope of SBP vs DBP obtained through 24-h noninvasive ambulatory BP monitoring and s-AASI
http://dx.doi.org/10.1016/j.numecd.2014.10.001 0939-4753/ª 2014 Elsevier B.V. All rights reserved.
Please cite this article in press as: Blanch N, et al., A systematic review of vascular and endothelial function: Effects of fruit, vegetable and potassium intake, Nutrition, Metabolism & Cardiovascular Diseases (2014), http://dx.doi.org/10.1016/j.numecd.2014.10.001
2
(symmetric ambulatory arterial stiffness index) has been shown to provide a clinically relevant evaluation of arterial stiffness [4]. Systemic arterial compliance (SAC) is a measure of large artery stiffness determined from carotid pressure and aortic volume flow measures and has been shown to contribute to higher brachial and carotid systolic and pulse pressure [5]. Endothelial dysfunction is strongly associated with traditional cardiovascular risk factors and precedes the development of clinical disease [6]. One of the major functions of the endothelium is to produce vasoactive molecules including nitric oxide (NO), a potent vasodilator. The assessment of endothelial function by measurement of ischaemia-induced FMD, an NO mediated response in the conduit arteries of the peripheral circulation allows the non-invasive study of vascular physiology in human subjects using ultrasound or MRI [7]. FMD is correlated with the extent and severity of coronary artery disease [8] and predicts events, although whether this is independent of traditional risk factors is not clear. PAT measures endothelial function by assessing finger pulse wave amplitude (PWA) at rest and during shear stress [9]. Reactive Hyperaemic-PAT (RH-PAT) is a ratio calculated using the post-occlusion (shear stress) PWA to the baseline (rest) PWA, expressed as reactive hyperaemic index (RHI) [9]. The lnPAT (log EndoPAT) ratio is the natural logarithm of the ratio of the pulse amplitude recorded 90e120 s after cuff release to the baseline amplitude divided by the hyperaemic to baseline ratio in the contralateral control finger [9]. PAT is an independent predictor of CV events and all-cause mortality in populations with established CVD [10]. Endothelial function is commonly measured using forearm blood flow (FBF) assessed by strain gauge venous plethysmography during intra-arterial infusion of agonist (e.g. acetylcholine which releases NO) and antagonists (e.g. NG-monomethyl L-arginine [L-NMMA] which blocks NO production) or after ischaemia [11]. Reduced FBF measurements in response to intra-arterial infusion of acetylcholine are associated with increased risk of CV events [12]. LDI with iontophoresis of acetylcholine measures the microvascular perfusion across a selected region of the peripheral circulation [13]. Acetylcholine acts on muscarinic receptors of the endothelium producing endothelium-dependent vasodilation [14]. Impaired endothelium dependent vasodilation as assessed by LDI with iontophoresis of acetylcholine is associated with increased CHD [15]. The measurement of arterial compliance (e.g. AI) following administration of salbutamol (a b2adrenergic receptor agonist that reduces arterial stiffness in an NO-dependent manner) can be used as a measure of endothelial function [16]. A reduced response to salbutamol administration has been demonstrated in hypercholesterolemic populations and is related to FBF responses to acetylcholine [17]. The association of high potassium diets with reduced stroke risk is now generally accepted within the literature independent of various confounding variables including sodium intake [18]. The mechanisms for this are unclear but may be via an effect on blood pressure [19].
N. Blanch et al.
Diets high in fruit and vegetables are associated with reduced cardiovascular disease. Fruit and vegetables are a source of nutrients including potassium and non-nutrients such as polyphenols, and the high content of potassium in fruits and vegetables is a plausible mechanism to explain the apparent CV protection of fruits in vegetables in epidemiological reports [20]. The purpose of this review is to summarise the effects of fruits, vegetables and potassium on endothelial and vascular function in intervention studies. Method A systematic search was performed in MEDLINE, EMBASE and the Cochrane Library for original research articles investigating the relationship between potassium, fruits and vegetables and vascular function published before 01 June 2014. Reference lists of retrieved articles were also searched for relevant articles. No restrictions were placed on publication date. Studies were limited to those published in English, studies in humans and clinical trials. Key search terms were “potassium” OR “fruit” OR “vegetable” AND “endothelial function” OR “vascular function”. Individual fruits or vegetables were not included in the search strategy but were identified if “fruit” or “vegetable” was listed as a keyword, or from reference lists of retrieved articles. Eligibility criteria To be included in this review a published study had to meet the following criteria: 1) original research; 2) controlled intervention studies aiming to determine the effect of potassium, fruits or vegetables on endothelial and vascular function. Studies including vitamin, mineral supplementation (other than potassium), dietary changes (other than fruits and vegetables), mixed interventions (e.g. Low fat combined with increased fruit/vegetables), lifestyle modification (e.g. exercise or stress management) or pharmacological treatment were excluded. Dietary studies specifically increasing nitrate intake were excluded as this area has been reviewed recently [21].Vascular function was defined as measures of arterial compliance (PWV, AI, PP) and endothelial function. Blood pressure was excluded as this has been recently reviewed [19,22]. Biochemical markers of endothelial dysfunction included in this review were nitrate/nitrite as indices of NO production, adhesion molecules, Von Willebrand’s factor (vWF) which can increase the risk of clotting, and endothelin-1 (ET-1) which is a pro-inflammatory, proproliferative vasoconstrictor peptide [11]. Inflammatory molecules and cytokines were not included in this review and have been published elsewhere [23,24]. Crosssectional, cohort, animal and cellular studies were not the focus of this review but were included to help explain the rationale and mechanisms behind the observed effect. Searches identified 4063 potential articles of which all but 44 were excluded on the basis of title and abstract (Fig. 1).
Please cite this article in press as: Blanch N, et al., A systematic review of vascular and endothelial function: Effects of fruit, vegetable and potassium intake, Nutrition, Metabolism & Cardiovascular Diseases (2014), http://dx.doi.org/10.1016/j.numecd.2014.10.001
A systematic review of vascular and endothelial function
3
Review
sensitivity may play a role in explaining the inconsistency of findings, with one study showing potassium supplementation (4.5 g/d potassium chloride) increased plasma and urine NO levels in normotensive and mildly hypertensive salt sensitive subjects during salt loading [30]. Serum potassium may be transiently increased within the physiological range after supplementation, which may directly affect the small and intermediate calcium-activated potassium channels of the endothelial cells which are then electrically coupled to the smooth cells e i.e. an endothelium derived hyperpolarizing factor effect [31]. Evidence suggests the vascular response to increases in extracellular potassium from potassium supplementation is mediated by NO, as infusion of L NMMA (an NO synthase inhibitor) abolishes the enhanced vasodilatory response to potassium with acetylcholine [32]. Inwardly rectifying K channels (KIR) and the Naþ/Kþ ATPase are involved in the control of basal tone of human forearm resistance vessels [33] and these are involved in Kþ induced vasorelaxation within these vascular beds [33,34]. Furthermore, intra-arterial administration of glibenclamide, a KIR channel closer, has been shown decrease reactive hyperaemic flow [32]. Combining administration of glibenclamide and L-NMMA did not further affect the reduction in reactive hyperaemic flow, suggesting the KIR channels and NO response are not independent pathways [32].
Potassium and endothelial function
Fruit and vegetables and endothelial vasodilator function
Table 1 summarises studies investigating the effects of increasing potassium intake with diet or supplementation on endothelial function. We have recently shown in a group of healthy adults, that a 36 mmol potassium load can attenuate the postprandial reduction in FMD within 2 h [25]. Furthermore an increase of 48 mmol/d in potassium intake achieved by including bananas and potatoes in place of apples and rice significantly improved fasting FMD after only one week [26]. In addition, a study by He et al. demonstrated that fasting FMD was significantly improved following 4 weeks of 64 mmol/d potassium chloride and 64 mmol/d potassium bicarbonate compared to placebo [27]. There was no difference between the potassium chloride and potassium bicarbonate treatments. Conversely, Berry et al. [28] found no effect on fasting FMD by increasing potassium intake by 20 mmol/d and 40 mmol/d through fruit and vegetable intake or with a 40 mmol/d potassium citrate supplement when compared to placebo. In another study, six weeks of 64 mmol/d potassium chloride supplementation did not evoke changes in endothelial function, as assessed by changes in pulse wave analysis following GTN and salbutamol administration, when compared to placebo [29]. Given the inconsistency of these findings, it remains unclear whether increasing dietary potassium can improve endothelial function. There is no evidence suggesting increased potassium intake is harmful to endothelial function. The mechanisms by which increased dietary potassium may improve endothelial function are unclear. Salt-
Increases in fruit and vegetable intake have been shown to improve endothelial function in a dose dependent manner [35]. McCall et al. [35] demonstrated that an additional portion (80 g) of fruit and vegetable consumed daily led to a 6.2% improvement in endothelium-dependent FBF responses after 8 weeks of intervention. In addition, it has recently been demonstrated that tomato paste supplementation (70 g/d) significantly improved FMD after two weeks [36] compared with a no supplement control period but it is not practical to expect individuals to consume this amount of tomato paste on a regular basis. This effect was not seen in the 24 h following a single dose (70 g). Conversely, another study showed no effect on FMD after 7 days of supplementation with tomato puree (70 g/d) in postmenopausal, healthy women compared with placebo [37]. Fruit and vegetable supplement capsules have been shown to attenuate the FMD reduction in response to a high fat meal compared with placebo capsules [38] while a recent study demonstrated polyphenols-rich juice from Concord grapes significantly improved FMD compared with polyphenols-poor juice and prevented a postcigarette reduction in FMD after only 7 days, suggesting polyphenols are important in determining endothelial function [39]. Encapsulated grape solids (6 500 mg/d for two weeks) had no effect on FMD response to a high fat meal [40]. In a study investigating the effects fruit and vegetable extracts on FMD in an overweight population with metabolic syndrome, there were no differences following 8 weeks of supplementation when compared to
Figure 1
Flow diagram of article selection.
Please cite this article in press as: Blanch N, et al., A systematic review of vascular and endothelial function: Effects of fruit, vegetable and potassium intake, Nutrition, Metabolism & Cardiovascular Diseases (2014), http://dx.doi.org/10.1016/j.numecd.2014.10.001
4
Reference
Study design
Study participants
Intervention
Outcome measures
Effects of intervention
Effect on EF
Effect on VF
Berry [28]
Randomised cross-over study of four 6 week interventions
48 individuals aged 22 e65 y, BMI 20e35 kg/ m2
Fruit and vegetable (20 mmol/d), Fruit and vegetable (40 mmol/d), and potassium citrate (40 mmol/d) compared to placebo capsules
FMD, PWV, peripheral AI, ICAM
4
4
Blanch [25]
Randomised cross-over acute postprandial test meal (2 h)
32 individuals aged 18 e70 y, BMI 18e30 kg/ m2
High potassium meal (36 mmol) compared to a low potassium meal (6 mmol)
FMD
Blanch [26]
Randomised cross-over study of two 6 day interventions
35 individuals aged 18 e70 y, BMI 18e30 kg/ m2
High potassium diet (150 mmol/d) compared to low potassium diet (80 mmol/d)
FMD, PWVcef, AI, Eselectin, endothelin-1, ADMA
Graham [29]
Randomised cross-over study of two 6 week interventions
40 individuals aged 40 e70 y, with increased CVD risk
64 mmol potassium chloride compared to placebo
PWV, AI, ICAM, VCAM, E-selectin
He [27]
Randomised cross-over study of three 4 week interventions
42 individuals aged 18 e75 y
Potassium bicarbonate (64 mmol/d) and potassium chloride (64 mmol/d) capsules compared to placebo
FMD, PWV
Liu Z [56]
4-phase stepwise study design of 3 day run in and three 7 day interventions
155 normotensive and mild hypertensive individuals aged 36 e61 y
51.3 mmol/d (low) sodium chloride then 307.7 mmol/d (high) sodium chloride compared to 307.7 mmol/d sodium chloride and 60 mmol/ d potassium chloride
AASI, s-AASI, endothelin-1
No difference between treatment and placebo for FMD, PWV, peripheral AI. Y ICAM in female participants only following 40 mmol/ d fruit and vegetable treatment compared to placebo (p < 0.05). [ FMD over 2 h following the high potassium meal compared to the low potassium meal. [ fasting FMD and Eselectin following high potassium diet compared to low potassium diet. No effect on PWVcef, AI, ADMA, endothelin-1. Y PWV following potassium chloride compared to placebo (p Z 0.033), reduction lost after adjustment for SBP. No effect on AI, ICAM, VCAM or Eselectin. [ fasting FMD following potassium bicarbonate (p < 0.05) and potassium chloride (p < 0.001). Y PWV following potassium bicarbonate (p < 0.05) and potassium chloride (p < 0.001). Y AASI (p < 0.05), sAASI (p < 0.05), endothelin-1 (p < 0.05) following sodium chloride and potassium chloride intervention when compared to the high sodium intervention.
[
[
4
4
[
[
[
[
[
N. Blanch et al.
Please cite this article in press as: Blanch N, et al., A systematic review of vascular and endothelial function: Effects of fruit, vegetable and potassium intake, Nutrition, Metabolism & Cardiovascular Diseases (2014), http://dx.doi.org/10.1016/j.numecd.2014.10.001
Table 1 Effect of potassium interventions on endothelial and vascular function.
Abbreviations: AASI e ambulatory arterial stiffness indexes; ADMA e asymmetric dimethylarginine; AI e augmentation index; BMI e body mass index; EF e endothelial function; FMD e flow mediated dilatation; ICAM e intercellular adhesion molecule; NO e nitric oxide; PWV e pulse wave velocity; PWVcef e pulse wave velocity (carotidefemoral); s-AASI e symmetric ambulatory arterial stiffness index; VF e vascular function; VCAM e vascular adhesion molecule; [ e improvement; Y e deterioration; 4 e no effect.
[ Plasma and urine NO Low-salt loading (3 g/ d), high-salt loading (18 g/d) and high-salt loading plus potassium chloride (4.5 g/d) 39 normotensive and mild hypertensive individuals ( 28 kg/m2
20 g Marie Menard cider apple powdered (high polyphenols) compared to 19.3 golden delicious apple powdered (low polyphenols) 500 mL red orange juice/d compared to 500 mL placebo/d
FMD, plasma NO
[
Del Bo [43]
Randomised cross-over acute postprandial test meal (1 h)
10 healthy male subjects, aged 20.8 1.6 y, BMI 22.5 2.1 kg/m2
RHI, NO
Dohadwala MM [44]
Randomised cross-over study of two 4 week interventions
44 patients with stable coronary artery disease
Homogenised blueberries (300 g) compared to control jelly (sugar added equivalent to blueberry intervention) Double-strength cranberry juice (480 ml/ d) compared to calorie, taste, and appearancematched placebo beverage containing no polyphenols (480 ml/d)
[ FMD following red orange juice supplementation in the compared to placebo No effect of intervention on plasma NO No differences in RHI, NO compared to control
4
Eccleston C [66]
Randomised parallel design of two 8 week interventions Randomised cross-over study of two 6 week interventions
20 men, aged 18e55 y
sICAM-1
12 postmenopausal women, aged 57 1 y, BMI 38.1 2.1 kg/m2
300 ml/d Sea buckthorn juice compared to 300 ml/d placebo drink Watermelon extract (6 g/d) compared to placebo
Randomised cross-over acute postprandial test meal (2 h)
20 individuals aged 44.55 13.34 y, BMI 23.81 2.46 kg/m2
Y PWVcef following juice supplementation compared to placebo (p Z 0.003) No differences in FMD, lnPAT ratio, PWVcer, sICAM-1 between interventions sICAM-1 unaffected by either Sea buckthorn or placebo juice treatment Y PWVbea following watermelon supplementation compared to baseline (p < 0.001) and placebo (p < 0.01) No effect on AI following both interventions No effect of intervention or placebo on LDI, plasma NO, sICAM-1, sVCAM-1
Fruit Auclair [52]
Figueroa A [58]
Jin Y [68]
Lynn A [60]
250 ml 20% blackcurrant juice drink compared to 250 ml placebo drink
FMD, lnPAT ratio, PWVcer, PWVcef, sICAM-1
PWVbea, AI
LDI with iontophoresis of acetylcholine chloride, plasma NO, sICAM-1, sVCAM-1 PWV
Effect on VF
4
[
4
[
4
4
N. Blanch et al.
Please cite this article in press as: Blanch N, et al., A systematic review of vascular and endothelial function: Effects of fruit, vegetable and potassium intake, Nutrition, Metabolism & Cardiovascular Diseases (2014), http://dx.doi.org/10.1016/j.numecd.2014.10.001
Table 2 Effect of fruit or vegetable interventions on endothelial and vascular function.
48 healthy individuals, aged 30e50 y
Morand [47]
Randomised cross-over study of three 4 week interventions
Morand [47]
Randomised cross-over acute postprandial test (6 h)
24 healthy men aged 51 e63 y, BMI 25.2 e30.5 kg/m2
Poreba R [45]
3-phase stepwise study design of 6 week intervention, washout, repeated intervention
35 men with mild hypercholesterolaemia, mean age 53.9, BMI 27.4 kg/m2
Riso P [42]
Randomised cross-over study of two 6 week interventions
18 male volunteers mean age 47.8 9.7 y, BMI 24.8 2.6 kg/m2
Siasos G [39]
Randomised cross-over study of two 2 week interventions
26 smokers, aged 26 5 y
Pomegranate juice (330 ml/d) compared to control drink of lemonade (330 ml/d; 7 UP) Orange juice (500 ml) compared to a control drink (500 ml) plus 2 placebo capsules (2 146 mg starch) compared to control drink (500 ml) plus 2 capsules of pure hesperidin (2 146 mg) Orange juice (500 ml) compared to a control drink (500 ml) plus 2 placebo capsules (2 146 mg starch) compared to control drink (500 ml) plus 2 capsules of pure hesperidin (2 146 mg) 250 ml/d 100% chokeberry juice compared to abstinence from chokeberry juice
Wild blueberry drink (prepared by suspending 25 g of wild blueberry freeze-dried powder in 250 mL of water) compared to placebo drink (250 mL water, 7.5 g fructose, 7 g glucose, 0.5 g citric acid and 0.03 g blueberry flavour) 100% Concord grape juice (high polyphenol) compared to low polyphenol grapefruit juice placebo (7 cc/kg/ d)
A systematic review of vascular and endothelial function
No effect of intervention or placebo on PWV
LDI with iontophoresis of acetylcholine chloride, ICAM, VCAM, NO, vWF
No significant effect on LDI, ICAM, VCAM, NO, vWF between treatments compared to placebo drink
4
LDI with iontophoresis of acetylcholine chloride
[ LDI with acetylcholine following orange juice compared to control drink plus hesperidin (p < 0.05) and placebo (p < 0.01)
[
FMD, NO
[ FMD after juice and abstinence compared to baseline (p < 0.001). [ FMD after juice compared to abstinence (p < 0.01) [ NO throughout the study compared to baseline (p < 0.01) No differences in RHI, FRHI, AI, AI@75 bpm NO, sVCAM-1 following each intervention compared to baseline
[
RHI, FRHI (Framingham reactive hyperaemia index), AI, AI@75 bpm, NO, sVCAM-1
FMD, PWVcef measured before and after smoking a single cigarette (0 and 20 min postcigarette)
[ FMD following Concord grape juice supplementation prior to cigarette (p Z 0.02) and attenuated post cigarette decrease
4
4
[
[
(continued on next page)
7
Please cite this article in press as: Blanch N, et al., A systematic review of vascular and endothelial function: Effects of fruit, vegetable and potassium intake, Nutrition, Metabolism & Cardiovascular Diseases (2014), http://dx.doi.org/10.1016/j.numecd.2014.10.001
Randomised cross-over study of two 4 week interventions
8
Reference
Van Mierlo [40]
Study design
Study participants
Intervention
Outcome measures
Effects of intervention when compared to placebo (p < 0.001). Y PWVcef following Concord grape juice supplementation prior to cigarette (p Z 0.04) and attenuated post cigarette increase when compared to placebo (p < 0.001). No differences in FMD, large artery elasticity index small artery elasticity index after a high fat meal compared to placebo
Effect on EF
Effect on VF
4
4
Randomised cross-over study of three 2 week interventions
35 healthy men aged 18 e45 y, BMI 18e32 kg/ m2
Grape solids (wine grape and grape seed) 6 500 mg/d compared to placebo capsules
FMD, large artery elasticity index small artery elasticity index after a high fat meal
Randomised cross-over study of two 24 h and 7 day interventions
19 postmenopausal women
FMD
No effect of intervention or control meal on FMD
4
Upritchard JE [67]
Randomised parallel design of four 4 week interventions
52 Type 2 diabetics aged
