Table 1.

Summary of included studies.

Study	Setting	Study Design	Population
Soy protein (contains isoflavones)
Fanti et al. 2006 [40]	USA Unknown number and type of recruitment sites Data collected: dates not specified	Two-arm parallel, placebo-controlled, double-blind randomised trial Allocation method: random allocation 2:1	ESRD on haemodialysis. Primary disease 44% DM, 36% HTN, 20% other. n = 32 participants; n = 19 IG; n = 13 CG. Attrition: 22% (n = 15 IG; n = 10 CG) Mean age 61.0 ± 2.9 years; 40% female.
Chen et al. 2005 [42]	Taiwan 1 hemodialysis centre, from a general hospital. Data collected: dates not specified	Two-arm parallel, placebo-controlled, double-blind randomised trial Allocation method: random allocation 1:1	ESRD on haemodialysis. Primary diseases not described. n = 42 participants; n = 18 IG; n = 19 CG (post attrition numbers). Attrition: 12%. Participants divided into normolipidemic (n = 10 CG; n = 9 IG) and hyperlipidemic (n = 9 CG; n = 10 IG) Mean age 59–66 ± 9–13 years; 92% female.
Chen et al. 2006 [41]	Taiwan 1 hemodialysis centre, from a general hospital. Data collected: dates not specified	Two-arm parallel, placebo-controlled, double-blind randomised trial Allocation method: random allocation 1:1	ESRD on haemodialysis and all with hypercholesterolaemia. Primary diseases not described. n = 29 participants; n = 18 IG; n = 19 CG (post attrition numbers). Attrition: 11%. Participants divided into normolipidemic (n = 13 CG; n = 13 IG) and hyperlipidemic (n = 9 CG; n = 10 IG) Mean age 58–59 ± 11–12 years; 27% female.
Siefker & DiSilvestro 2006 [43]	USA 2 hemodialysis centres, from 2 general hospitals. Data collected: dates not specified	Two-arm parallel, placebo-controlled, double-blind randomised trial Allocation method: random allocation 1:1	ESRD on haemodialysis and all with hypercholesterolaemia. Primary diseases not described. n = 20 participants; n = 8 IG; n = 9 CG (post attrition numbers). Attrition: 15%. Mean age 20 years (range 27–77); 59% female.
Grape (contains various polyphenols)
Janiques et al. 2014 [25]	Brazil 1 nephrology centre Data collected: dates not specified	Two-arm parallel, placebo-controlled, double-blind randomised trial Allocation method: random allocation 1:1	ESRD on haemodialysis. Primary diseases not described. n = 34 participants; n = 17 IG; n = 17 CG. Attrition: 6% (n = 16 IG; n = 16 CG). Mean age 52.7–53.0 ± 9.8–13.7 years; 44% female.
Turmeric (contains curcuminoids)
Pakfetrat et al. 2014 [38]	Iran 1 haemodialysis centre Data collected: 2011–2012	Two-arm parallel, placebo-controlled, double-blind randomised trial Allocation method: random allocation 1:1	ESRD on haemodialysis, all with uremic pruritus. Primary diseases not described. n = 100 participants; n = 50 IG; n = 50 CG. Attrition: 0%. Mean age 51.0–55.6 ± 14.7–16.6 years; 40% female.
Pakfetrat et al. 2015 a [27]	Iran 1 haemodialysis centre Data collected: dates not specified	Two-arm parallel, placebo-controlled, double-blind randomised trial Allocation method: random allocation 1:1	ESRD on haemodialysis, all with uremic pruritus. Primary diseases: 35% DM, 33% HTN. n = 50 participants; n = 50 IG; n = 50 CG. Attrition: 4%. Mean age 47–52 ± 15 years; 45% female.
Cocoa (contains flavanols)
Rassaf et al. 2016 [28]	Germany 1 renal centre Data collected: 2012–2013	Two-arm parallel, placebo-controlled, double-blind randomised trial Allocation method: random allocation 1:1	ESRD on haemodialysis, all with uremic pruritus. Primary diseases: 32% HTN, 23% DM, 19% GN, 26% other. n = 52 participants; n = 26 IG; n = 26 CG. Attrition: 6% (n = 24 IG; n = 25 CG). Mean age 65 ± 13 years; 26% female.
Pomegranate (Phenolic acids & Flavanoids)
Shema-Didi et al. 2012 [37]	Israel 1 dialysis centre in a hospital Data collected: dates not specified.	Two-arm parallel, placebo-controlled, double-blind randomised trial Allocation method: random allocation 2:1	ESRD on haemodialysis. Primary diseases: 64% DM, 60% CVD. n = 101 participants; n = 66 IG; n = 35 CG. Attrition: 34% (n = 41 IG; n = 26 CG). Mean age 66–68 ± 11–13 years; 45% female.
Shema-Didi et al. 2014 [36]	As per Shema-Didi et al. 2012 [37]	Shema-Didi et al. 2012 [37]	Shema-Didi et al. 2012 [37]
Shema-Didi et al. 2013 [26]	Israel 1 dialysis centre in a hospital Data collected: dates not specified.	Two-arm parallel, placebo-controlled, double-blind randomised trial. Allocation method: random allocation 2:1	ESRD on haemodialysis, receiving an iron transfusion during a single haemodialysis session. Primary diseases: 60% DM, 51% CVD. Others not described. n = 27 participants; n = 17 IG; n = 10 CG. Attrition: N/A Mean age 65–67 ± 11–13 years; 40% female.
Wu et al. 2015 [44]	USA Unknown number of dialysis clinics in two cities Data collected: dates not specified.	Two-arm parallel, placebo-controlled, double-blind randomised trial Allocation method: random allocation 1:1	ESRD on haemodialysis. Primary diseases: 49% HTN, 36% DM. Others not described. n = 33 participants; n = 16 IG; n = 17 CG. Attrition: 19% (n = 13 IG; n = 14 CG). Mean age 53–56 ± 3 years; 39% female.

CVD, Cardiovascular Disease; CG, control group; DM, Diabetes Mellitus; HTN, hyptertension; IG, intervention group; N/A, Not Applicable; ^a It is unclear if the sample and study is the same as that described in Pakfetrat et al. 2014 [38]. The samples appear to be the same; where Pakfetrat et al. [27] excluded some participants after recruitment, which could explain slight differences in sample characteristics. However, Pakfetrat et al. 2015 [27] does not refer to the earlier study at all, and therefore it is not clear.