Data on the effect of red bioactive pigments were from beta-cryptoxanthin (n  = 15 systematic literature reviews (SLRs)SLRs reporting n  = 33 meta-analyses (MAs)MAs) and lycopene (n = 25 SLRs of n = 65 MAs). Anthocyanins may also be red in an acidic environment, but were reported with the blue/purple bioactive pigments ^[1][38].

Only two of the included MAs on beta-cryptoxanthin were based on randomized controlled trials (RCTs) data (12-weeks; 6 mg/day; mixed sources of beta-cryptoxanthin), and the remaining 31 MAs were based on cohort data (1–26 years). Most cohort MAs compared an unspecified highest category with the lowest; however, where the highest categories were specified, they provided 56–200 µg/day compared with the lowest at <1.8 to 20 µg/day. Cohort data were derived from diet (n = 21 MAs), mixed sources (n = 3 MAs), or serum levels (n = 7 MAs); and included seven dose–response MAs.

The highest category of beta-cryptoxanthin intake was associated with a 28% decreased risk of hip fracture (OR 0.72; 95% CI 0.60, 0.87) ^[2][39] and up to a 27% decreased risk of all-cause mortality (RR 0.73; 95% CI, 0.58, 0.88) ^[3][40], compared to the lowest category of intake. A small effect was also found for the inflammatory biomarker C-reactive protein (CRP; MD −0.35 mg/L; 95% CI −0.54, −0.15), after individuals consumed 6 mg beta-cryptoxanthin over 12-weeks ^[4][37].

In relation to cancer, the highest category of dietary beta-cryptoxanthin intake was associated with a 69% decreased risk of larynx cancer (OR 0.41; 95% CI 0.33, 0.51) ^[5][41], 64% decreased risk of oral cavity and pharynx cancer (OR 0.46; 95% CI 0.29, 0.74) ^[5][41], 42% decreased risk of bladder cancer (RR 0.58; 95% CI 0.36, 0.94) ^[6][42], and 20% decreased risk of lung cancer (RR 0.80; 95% CI 0.72, 0.89) ^[7][43], compared to the lowest category of intake.

In dose–response MAs of serum levels, each 0.1 mg/day increase in beta-cryptoxanthin, decreased the risk of all-cause mortality by 6% (RR 0.94; 95% CI 0.89, 0.99) ^[3][40], whereas for every daily increase of 0.5 µmol/L, the risk of type 2 diabetes (T2DM) decreased by 15% (RR 0.85; 95% CI 0.76, 0.94) ^[8][44].

No differences were found for beta-cryptoxanthin and risk of: cataracts ^[9][45], early age-related macular degeneration ^[10][46], osteoporosis ^[2][39], Parkinson’s disease ^[11][47], non-Hodgkin lymphoma ^[12][48], breast cancer ^[13][29], colorectal cancer ^[14][49], pancreatic cancer ^[15][50] or lung cancer mortality ^[7][43].

The strongest evidence for the health effect of beta-cryptoxanthin was for a decreased risk of all-cause mortality (dose–response relationship, moderate to large effect size, GRADE: very low), bladder cancer (dose–response relationship, very large effect size, GRADE: medium), oral, laryngeal, or pharyngeal cancer (very large effect size, GRADE: very low to medium), and T2DM (dose–response relationship, large effect size, GRADE: low).

There was n = 14 MAs included based on RCT data (1-day to 6-months duration; 2–50 mg/day); with the remaining n = 51 MAs based on observational cohort data (3-months to 26-years duration; 2035–10,000 µg/day), n = 11 of which were dose–response MAs (per 1000 µg/day or incremental serum levels). Most MAs analyzed dietary intake data (n = 32 MAs), followed by mixed sources (n = 22 MAs), serum values (n = 10 MAs), and one MA measured supplemental intake.

The highest category of dietary lycopene intake was associated with reductions in the risk of cervical (OR 0.54; 95% CI, 0.39, 0.75) ^[16][51], larynx (OR 0.50; 95% CI, 0.28, 0.89) ^[5][41], lung (RR 0.71; 95% CI, 0.51, 0.98) ^[7][43], oral cavity and pharynx (OR 0.74; 95% CI, 0.56, 0.98) ^[5][41] and prostate (RR 0.88; 95% CI, 0.79, 0.99) ^[17][34] cancers, compared to the lowest category of dietary lycopene intake. Reductions in the risk of breast cancer were only reported in case control studies, where greater reductions in risk up to 29% (OR 0.71; 95% CI, 0.56, 0.92) ^[13][29] were found with greater dietary lycopene intake.

Higher lycopene intake was also associated with cardiovascular improvements with small to moderate clinical significance, including a lower risk of CHD (RR 0.87; 95% CI, 0.76, 0.98) ^[18][52], cardiovascular disease (CVD) (HR 0.86; 95% CI 0.77, 0.95) ^[19][53], stroke (HR 0.74; 95% CI 0.62, 0.89) ^[19][53], T2DM (RR 0.85; 95% CI 0.76, 0.96) ^[8][44], mortality (HR 0.63; 95% CI 0.49, 0.81) ^[19][53] and all-cause mortality (RR 0.72; 95% CI 0.49, 0.95) ^[3][40]. In dose–response MAs of serum levels, each 0.5 µmol/L increase in serum lycopene decreased the risk of T2DM by 17% (RR 0.83; 95% CI 0.74, 0.92) ^[8][44].

Lycopene status had no effect on preeclampsia ^[20][54], early age-related macular degeneration ^[10][46], risk of hip fracture ^[21][55], advanced prostate cancer ^[17][22][34,56], colon/colorectal/rectal cancer ^[14][49], bladder cancer ^[6][42], gastric cancer ^[23][57], non-Hodgkin lymphoma ^[12][48], ovarian cancer ^[24][58], Parkinson’s disease ^[11][47], inflammatory biomarkers (except a small effect in interleukin-6 (MD −1.08 pg/mL; 95% CI −2.03, −0.12) ^[4][37], lipid profiles ^[25][26][59,60], blood pressure ^[26][60] and prostate specific antigen (PSA) levels ^[27][61].

The strongest evidence for the health effects of lycopene were decreased risk of breast cancer (dose–response relationship, large to very large effect size, GRADE: very low) and T2DM (dose–response relationship, moderate to large effect size, GRADE: very low to low).

The health effects of consuming orange bioactive pigments from FV were reported by MAs of beta-carotene (bioactive pigment subclass; n = 34 SLRs reporting n = 75 MAs including n = 16 dose–response MAs).

Evidence for the effects of beta-carotene was largely represented by MAs of cohort studies (n = 59 MAs) measured over 1–26 years via dietary intake (n = 32 MAs), mixed sources (n = 16 MAs), serum levels (n = 10 MAs) or supplementation (n = 1 MA). Doses in the highest categories of intake were not usually reported, but when reported ranged from 2473–7000 µg/day intake or 16 to >120 µg/dL serum level. The one supplemental study provided a dose of 600 to 1991 µg/day. Sixteen of the observational MAs were dose–response, examining effects per 500–5000 µg/day intake or per 0.1 µmol/L serum level.

The highest category of beta-carotene intake was associated with a decreased risk of several types of cancers, including cervical (OR 0.68; 95% CI 0.55, 0.84) ^[16][51], gastric (OR 0.74; 95% CI 0.61, 0.91) ^[28][62], larynx (OR 0.43; 95% CI 0.24, 0.77) ^[5][41], non-Hodgkin lymphoma (RR 0.80; 95% CI 0.68, 0.94) ^[12][48], oral cavity (OR 0.54; 95% CI 0.37, 0.80) ^[5][41], ovarian (RR 0.84; 95% CI 0.75, 0.94) ^[29][63] and pancreatic (OR 0.78; 95% CI 0.66, 0.92) ^[15][50] cancers, compared to the lowest category of beta-carotene intake. Reductions in breast cancer risk were supported by dose–response MAs that found dietary beta-carotene intakes of 2000 µg/day, 3000 µg/day or 5000 µg/day reduced the risk of breast cancer by 3%, 4% and 7%, respectively ^[13][29]. For each 1000 µg/1000 kcal increase in dietary beta-carotene the risk of endometrial cancer decreased by 26% (RR 0.74; 95% CI 0.61, 0.91) ^[30][64].

Highest categories of beta-carotene intake were also associated with a lower risk of all-cause mortality (RR 0.82; 95% CI 0.77, 0.86) ^[3][40], coronary heart disease (CHD) (RR 0.73; 95% CI 0.65, 0.82) ^[31][65], CVD mortality (RR 0.68; 95% CI 0.52, 0.83) ^[32][66], total fracture (RR 0.63; 95% CI 0.52, 0.77) ^[33][67], hip fracture (OR 0.72; 95% CI 0.54, 0.95) ^[21][55] and the incidence of cataract (RR 0.90; 95% CI 0.83, 0.99) ^[9][45] and preeclampsia (SMD −0.40; 95% CI −0.72, −0.08) ^[20][54], when compared to the lowest intakes. In dose–response MAs, each 1 mg/day increase in beta-carotene intake decreased the risk of all-cause mortality by 5% (OR 0.95; 95% CI 0.92, 0.99) ^[3][40], whereas for every 0.5 µmol/L serum increase the risk of T2DM decreased by 35% (OR 0.65; 95% CI 0.48, 0.89) ^[8][44].

No differences were found for dietary beta-carotene and risk of bladder cancer ^[6][42], colon cancer ^[14][49], colorectal cancer ^[14][49], lung cancer ^[7][43], lung cancer or lung cancer mortality ^[7][43], melanoma ^[34][68], prostate cancer ^[22][28][56,62], rectal cancer ^[14][49], COPD ^[35][69], total fracture ^[36][70] and Alzheimer’s disease ^[37][71].

The strongest evidence for the health effects of beta-carotene were decreased risk of all-cause and CVD mortality (dose–response relationship, very large effect size, GRADE: very low to low), T2DM (dose–response relationship, large to very large effect size, GRADE: low to medium), bladder cancer (dose–response relationship, very large effect size, GRADE: very low), breast cancer (dose–response relationship, large effect size, GRADE: very low to low) and endometrial cancer (dose–response relationship, large effect size, GRADE: very low).

The evidence for the health effects of yellow bioactive pigments were from MAs reporting on alpha-carotene (n = 16 SLRs reporting n = 41 MAs), lutein (n = 7 SLRs reporting n = 10 MAs), zeaxanthin (n = 2 SLRs reporting n = 3 MAs) or lutein and zeaxanthin as a combined group (n = 13 SLRs reporting n = 31 MAs).

All n = 41 MAs reporting on the health effects of alpha-carotene were based on cohort and/or case–control research, measured via dietary intake (n = 27 MAs with 1–25 years follow-up); serum levels (n = 10 MAs with 2–26 years follow-up); or mixed diet, serum levels, and/or supplements (n = 4 MAs with 9-months to 26-years follow-up). Most MAs compared an unspecified highest category against the unspecified lowest category; however, some category groups were defined as >881–2000 µg/day compared against <180 to 300 µg/day dietary intake or serum levels of >1 to >5 µg/dL compared against <1 to <2 µg/dL.

The highest category of alpha-carotene intake was associated with a reduced risk of gastric (OR 0.58; 95% CI 0.44, 0.76) ^[23][57], non-Hodgkin lymphoma (RR 0.87; 95% CI 0.78, 0.97) ^[12][48], oral cavity and pharynx (OR 0.57; 95% CI 0.41, 0.79) ^[5][41] and prostate (RR 0.87; 95% CI 0.76, 0.99) ^[22][56] cancers, and a reduced risk of T2DM (RR 0.91; 95% CI 0.85, 0.96) ^[8][44] and all-cause mortality (RR 0.79; 95% CI, 0.63, 0.94) ^[3][40]. In dose–response MAs, for each 1000 µg/day increase in alpha-carotene the risk of breast cancer decreased by up to 18% (RR 0.82; 95% CI 0.73, 0.93) ^[13][29] and the risk of non-Hodgkin lymphoma decreased by 13% (RR 0.87; 95% CI 0.78, 0.97) ^[12][48].

Alpha-carotene intake was reported to have no effect on risk of pre-eclampsia ^[20][54], cataract ^[9][45], early aged-related macular degeneration ^[10][46], cancer of the larynx ^[5][41], risk of colon, rectal, or colorectal cancer ^[14][49], hip fracture ^[21][55], lung cancer ^[7][43], pancreatic cancer ^[15][50] or Parkinson’s disease ^[11][47].

The strongest evidence for the health effects of alpha-carotene were decreased risk of all-cause mortality (dose–response relationship, large effect size, GRADE: very low to low), bladder cancer (dose–response relationship, very large effect size, GRADE: high), non-Hodgkin lymphoma (dose–response relationship, moderate to large effect size, GRADE: very low) and T2DM (dose–response relationship, large to very large effect size, GRADE: medium).

All n = 10 MAs reporting on the health effects of lutein drew upon cohort or case–control data, measured via dietary intake (n = 4 MAs with 10–12 years follow-up), serum levels (n = 2 MAs with 8–10 years follow-up) or mixed sources (n = 4 MAs with 9-months to 26-years follow-up). Most of the MAs compared the highest unspecified category of lutein to the lowest unspecified category; however, one highest category was defined as 3701 to 4041 µg/day compared to 1413 to 1736 µg/day dietary intake.

When compared to the lowest category, the highest category of lutein intake improved the risk of stroke by 18% (RR 0.82; 95% CI 0.58, 0.78) ^[38][72] and the risk of T2DM by 35% (RR 0.65; 95% CI 0.55, 0.77). In dose–response MAs of serum levels, each 0.2 ugmol/L increase in lutein decreased the risk of T2DM by 21% (RR 0.79; 95% CI 0.72, 0.86). Lutein was reported to have no effect on risk of lung cancer ^[7][43], gastric cancer ^[23][57], Parkinson’s disease ^[11][47], pre-eclampsia ^[20][54] or all-cause mortality ^[3][40]

The strongest evidence for the health effects of lutein was for decreased risk of T2DM (dose–response relationship, large to very large effect size, GRADE: medium).

All three MAs which measured the effect of zeaxanthin on human health were based on cohort studies, with n = 2 MAs based on serum zeaxanthin levels (8–10 years follow-up) and n = 1 MA based on mixed sources (2–26 years follow-up). When comparing the highest unspecified category against the lowest unspecified category, zeaxanthin had no effect on all-cause mortality ^[3][40] or risk of T2DM ^[8][44].

Thirty-one MAs investigated the effect of combined lutein and zeaxanthin on human health, n = 29 of which were based on cohort of case–control studies, and n = 2 were based on RCTs. The observational research primarily measured lutein and zeaxanthin via dietary intake (n = 22 MAs with 1–25 years follow-up) or serum levels (n = 5 MAs with 2–26 years follow-up), with only one MA considering mixed sources (5–18 years follow-up). Only n = 3 MAs defined the intake of lutein and zeaxanthin in the highest category (>1815 to 5000 µg/day) as compared to the lowest category (<775 to 1000 µg/day). The two RCT MAs both considered dietary or supplemental intake of lutein and zeaxanthin with interventions ranging from 8–32 weeks and doses of 8 mg/day to 27 mg/day.

When comparing the highest category versus lowest category of lutein and zeaxanthin status, higher dietary intakes reduced the risk of non-Hodgkin lymphoma by 18% (RR 0.82; 95% CI 0.69, 0.97) ^[12][48], while higher serum intakes reduced the risk of bladder cancer (RR 0.53; 95% CI 0.33, 0.84) ^[6][42] and all-cause mortality (RR 0.85; 95% CI 0.74, 0.97) ^[3][40]. Lutein and zeaxanthin intakes were associated with decreased CRP levels (SMD −0.3 mg/L; 95% CI −0.45, −0.15) ^[4][37] and a dose–response relationship found a favorable 17% decreased risk of breast cancer for every 3000 µg/day increase in lutein and zeaxanthin intake (RR 0.83; 95% CI 0.77, 0.89) ^[13][29]. Dose–response relationships were not found for any other level of lutein or zeaxanthin intake.

For lutein and zeaxanthin, no differences were found for the risk of gastric cancer ^[23][57], lung cancer ^[7][43], lung cancer or lung cancer mortality ^[7][43], pancreatic cancer ^[15][50], oral cavity and pharynx cancer ^[5][41], colon ^[14][49], rectal ^[14][49] or colorectal cancer ^[14][49], early aged macular degeneration ^[10][46], hip fracture ^[21][55] or IL-6 ^[4][37].

The strongest evidence for combined lutein and zeaxanthin was for decreased risk of bladder cancer (dose–response relationship, large to very large effect size, GRADE: low to high) and breast cancer (dose–response relationship, large effect size, GRADE: very low to low).

The health effects of consuming pale yellow bioactive pigments from FV were reported by MA of flavonols (bioactive pigment subclass; n = 17 SLRs reporting n = 33 MAs), kaempferol (n = 1 SLR reporting n = 1 MA), myricetin (n = 1 SLR reporting n = 2 MA), and quercetin (n = 10 SLRs reporting n = 25 MAs).

As a group, MAs of flavonols subclass were primarily based on cohort and/or case–control data (n = 25 MAs with 2–28 years follow-up); although there was substantial cause-and-effect investigation via n = 8 MAs of RCTs (14–84 days intervention duration). Over half (n = 14 of 25) of the observational MAs measured flavonols from dietary sources alone (1–27 years follow-up), with the remaining n = 11 MAs measuring flavonols from mixed sources (4–28 years follow-up). The doses of intake in the highest or lowest categories were not reported. Four of the observational MAs were dose–response, examining effects per 10 mg or 20 mg, and were based on cohort data. All the MAs based on RCTs tested the effect of flavonols delivered via supplementation from 14- to 90-days at doses of 6–1000 mg.

When comparing the highest intake or levels with the lowest, flavonols were found to improve the risk of stroke (RR 0.86; 95% CI 0.75, 0.96), CVD (RR 0.85; 95% CI 0.79, 0.91), and CHD (RR 0.88; 95% CI, 0.79, 0.98), as well as CVD- (RR 0.79; 95% CI 0.63, 0.99) and CHD-related death (RR 0.80; 95% CI 0.69, 0.93) ^{[39][40][41][42][43]}[73,74,75,76,77]. High categories of flavonols were also associated with a reduced risk of T2DM (RR 0.92; 95% CI 0.85, 0.98) ^[44][78] and risk of breast (RR 0.88; 95% CI 0.80, 0.96), colorectal (RR 0.71; 95% CI 0.63, 0.81), gastric (OR 0.80; 95% CI 0.70, 0.91), ovarian (RR 0.68; 95% CI 0.58, 0.80) and smoking related cancer (OR 0.77; 95% CI 0.63, 0.95) ^{[45][46][47][48][49]}[79,80,81,82,83]. However, two other MAs found no association with breast cancer ^[50][84], one found no association with CHD ^[51][85], and no differences were found for effect on all-cause mortality ^[39][73], hypertension ^[52][86] or other types of cancer including liver, lung, pancreatic, esophageal or prostate ^[50][53][54][84,87,88].

In dose–response MAs, for each 20 mg/day increase in flavonols the risk of stroke decreased by 14% (RR 0.86; 95% CI 0.77, 0.96) ^[43][77], and for each 10 mg/day increase in flavonols the risk of CVD mortality decreased by 13% (RR 0.87; 95% CI 0.76, 0.99) ^[39][73]. MAs of RCTs examined chronic disease indicators, finding supplementation with flavonols improved systolic (MD −3.05 mmHg; 95% CI −4.83, −1.27) and diastolic (MD −2.63 mmHg; 95% CI −3.83, −1.42) blood pressure, HDL cholesterol (MD 0.05 mmol/L; 95% CI 0.02, 0.07), low density lipoprotein (LDL) cholesterol (MD −0.14 mmol/L; 95% CI −0.21, −0.07), and total cholesterol (MD −0.11 mmol/L; 95% CI −0.20, −0.02), blood glucose (MD −0.18 mmol/L; 95% CI −0.29, −0.08) and triglycerides (MD −0.11 mmol/L; 95% CI −0.18, −0.03) ^[55][89]; however, there was no effect on waist circumference ^[56][90].

One SLR reported on highest versus lowest dietary intake of kaempferol, quercetin, and myricetin using case–control data (duration not reported) ^[50][84]; whereas the nine other SLRs reported on 30–1000 mg/day of supplemental quercetin via MA of RCTs (5-days to 12-weeks duration) ^{[57][58][59][60][61][62][63][64][65]}[91,92,93,94,95,96,97,98,99]. There were no dose–response MAs.

When comparing the highest dietary intake with the lowest, kaempferol, but not myricetin or quercetin, reduced the risk of lung cancer by 23% (RR 0.77; 95% CI 0.62, 0.97) ^[50][84]. Supplemental quercetin improved a range of CVD risk factors, including systolic (MD −3.09 mmHg; 95% CI −4.83, −1.27) and diastolic blood pressure (MD −2.86 mmHg; 95% CI −5.09, −0.63) ^[59][93], CRP (MD −0.33 mg/L; 95% CI −0.50, −0.16) ^[60][94], VO2 max (MD 1.94%; 95% CI 0.30, 3.59) ^[63][97] and exercise performance (MD 2.82%; 95% CI 2.05, 3.58) ^[65][99]. RCT evidence for quercetin found no effect on blood lipids ^[57][59][91,93], glycemic or insulin metabolism ^[61][95], other measures of inflammation ^[62][64][96,98] or adiposity ^[58][92].

The strongest evidence for the health effect of flavonols and flavonols sub-classes was for improved blood pressure (cause-and-effect relationship established, large effect size, GRADE: low to high), cholesterol (cause-and-effect relationship established, small effect size, GRADE: low to high), blood glucose (cause-and-effect relationship established, small effect size, GRADE: medium) and risk of CVD or CHD mortality (dose–response relationship, very large effect size, GRADE: medium) and stroke (dose–response relationship, moderate to large effect size, GRADE: very low).

The evidence for the health effects of white bioactive pigments were from MAs reporting on flavones. All n = 19 MAs for flavones were based on cohort data (1–24 years duration) as measured via the diet (n = 13 MAs) or mixed sources (n = 6 MAs), and two MAs were dose–response analyses. The highest category of flavones was associated with a decreased risk of all-cause (RR 0.86; 95% CI 0.80, 0.93) and CVD mortality (RR 0.85; 95% CI 0.75, 0.96) ^[50][84], breast cancer (RR 0.81; 95% CI 0.68, 0.96) ^[49][50][83,84], CHD (RR 0.94; 95% CI 0.89, 0.99) ^[40][74], esophageal cancer (OR 0.78; 95% CI 0.64, 0.95) ^[53][87], liver cancer (RR 0.49; 95% CI 0.30, 0.78) ^[50][84] and smoking-related cancer (OR 0.77; 95% CI 0.69, 0.85) ^[46][80]. In dose–response MAs, for each 1 mg/day increase in flavones, the risk of CVD mortality decreased by 7% (RR 0.93; 95% CI 0.90, 0.97) ^[50][84]. No differences were found for hypertension ^[52][86], risk of CVD ^[42][76], risk of T2DM, or risk of colorectal ^[45][79], lung ^[48][50][82,84], ovarian ^[50][84], pancreatic ^[50][84] or prostate cancer ^[54][88].

The strongest evidence for the health effect of flavones was for decreased risk of all-cause and CVD mortality (dose–response relationship, moderate to large effect size, GRADE: very low to low), liver cancer (very large effect size, GRADE: medium) and smoking-related cancers (moderate effect size, GRADE: medium).

Purple/blue bioactive pigments were contributed to by anthocyanidins, anthocyanins, proanthocyanidins and proanthocyanins.

All n = 7 MAs examining anthocyanidins were based on cohort data derived from the diet (n = 2 MAs, 4–20 years duration) or mixed sources (n = 5 MAs, 4–16 years duration). The highest and lowest categories were not defined, but the single dose–response MA analyzed effects per 10 mg/day.

Higher anthocyanidin serum levels were associated with an 11% decrease in both all-cause (RR 0.89; 95% CI 0.85, 0.94) and CVD mortality (RR 0.89; 95% CI 0.83, 0.95). In dose–response MAs, for each 10 mg/day increase in anthocyanidins the risk of CVD mortality improved by 6% (RR 0.94; 95% CI 0.88, 0.99) ^[39][73]. Greater anthocyanidin intake was also associated with a 32% decreased risk of colorectal cancer (RR 0.68; 95% CI 0.56, 0.82) ^[45][79], 14% decreased risk of T2DM (HR 0.86; 95% CI 0.81, 0.91) ^[44][78], but a 12% increased risk of prostate cancer (RR 1.12; 95% CI 1.03, 1.21) ^[54][88]. No association was found for smoking-related cancer ^[46][80].

Most anthocyanin research was based on RCTs (n = 67 MAs) derived from diet (n = 19 MAs of 3-days to 6-weeks duration; dose not reported), mixed sources (n = 32 MAs of 4-h to 6-months duration, dose 1.3–1025 mg/day) or supplementation (n = 16 MAs of 1–96 weeks duration, dose 1.6–1323 mg/day). The n = 14 cohort MAs measured anthocyanins from the diet (n = 11 MAs, 1–24 years duration, dose not reported) or mixed sources (n = 3 MAs, 5–41 years, dose not reported).

The highest category of anthocyanin intake was associated with a decreased risk of CVD (RR 0.82; 95% CI 0.70, 0.96) ^[42][76], CHD (RR 0.90; 95% CI 0.83, 0.98) ^[40][74], CVD mortality (RR 0.92; 95% CI 0.87, 0.97) ^[66][100], hypertension (RR 0.92; 95% CI 0.88, 0.97) ^[52][86] and esophageal cancer (OR 0.60; 95% CI 0.49, 0.74) ^[53][87]. However, no association was found with risk of stroke ^[66][100], or multiple cancers including breast, liver, lung, pancreatic or gastric ^[49][50][67][83,84,101].

Thirty-three of the n = 67 (49%) RCT MAs reported improved inflammatory, oxidative, lipid, or glycemic markers (e.g., adiponectin, apolipoprotein A1/B, CRP, fasting glucose, HbA1c, HOMA-IR, LDL and HDL cholesterol, interleukin-6, tumor necrosis factor alpha (TNF-alpha), triglycerides) ^{[67][68][69][70][71]}[26,101,102,103,104], as well as vascular reactivity (SMD 0.77; 95% CI 0.37, 1.16) ^[72][105] and body mass index (BMI)BMI (SMD −0.36 kg/m²; 95% CI −0.58, −0.13) ^[73][27]. No improvements were found for liver enzymes ^[74][106], uric acid, blood pressure ^[75][107], waist circumference ^[75][107], delayed onset muscle soreness ^[67][101] or vascular stiffness ^[72][105].

The strongest evidence for the health effect of anthocyanins and anthocyanidins was for improved inflammatory and oxidative stress biomarkers (cause-and-effect relationship established, small to large effect size, GRADE: very low to low), glycemic and insulinemic biomarkers (cause-and-effect relationship established, small effect size, GRADE: medium), lipid profiles and vascular function (cause-and-effect relationship established, small to large effect size, GRADE: very low to medium) and adiposity (cause-and-effect relationship established, small effect size, GRADE: low to medium).

There were n = 11 MAs which reported on the effects of proanthocyanidins (n = 4 RCT MAs, n = 7 cohort MAs). Proanthocyanidin RCT MAs were all based on supplemental interventions of 100–400 mg/day delivered over 5 to 16 weeks. Cohort MAs were delivered over 4–16 years with unspecified categories of highest and lowest intakes, measured via diet (n = 3 MAs) or mixed sources (n = 4 MAs). One of the n = 7 cohort MAs was a dose–response analyses examining effects per 100 mg/day.

The highest serum levels of proanthocyanidin compared with the lowest was associated with a 11% improvement in CVD mortality risk (RR 0.89; 95% CI 0.81, 0.97), but this was not significant in a dose–response analysis ^[50][84]. Higher status was also associated with a 28% decreased risk of colorectal cancer (RR 0.72; 95% CI 0.61, 0.85) ^[45][79]. No differences were found with risk of all cause-mortality, T2DM, breast cancer or esophageal cancer ^[39][44][53][73,78,87]. MAs of RCT evidence showed that supplemental proanthocyanidin (100–400 mg for 5–16 weeks) improved systolic (MD −4.60 mmHg; 95% CI −8.04, −1.16) and diastolic (MD −2.75 mmHg; 95% CI −5.09, −0.41) blood pressure and mean arterial pressure (MD −3.37 mmHg; 95% CI −6.72, −0.01), but not pulse pressure ^[76][108].

The two MAs of proanthocyanins were based on cohort data and measured the highest dietary intakes compared with the lowest for up to 16 years, and found an inverse association with risk of CVD (RR 0.83; 95% CI 0.73, 0.95) ^[42][76] and CHD (RR 0.78; 95%CI 0.65, 0.94) ^[40][74].

The strongest evidence for the health effect of proanthocyanidins and proanthocyanins was for decreased blood and arterial pressure (large effect size, GRADE: high).

The health effects of consuming green bioactive pigments from FV were reported by single RCT and cohort evidence for chlorophyll. Ten of the seventeen health outcome measures reported for chlorophyll were based on RCT data (Sweden and Japan, 8–12 weeks of 0.7–3000 mg supplementation/day); the remaining seven were from cohort data (Netherlands, 9-years duration, highest undefined quintile). One RCT reported chlorophyll supplementation improved seasonal allergic rhinitis rescue medication scores (MD −3.09; 95% CI −5.96, −0.22) ^[77][109] and 3000 mg supplementation per day trended towards 1.5 kg weight loss; however, this appeared underpowered (p = 0.06, n = 36 participants) ^[78][110]. RCT evidence reported no effect on other measures of body composition or levels of insulin, glucose, or leptin ^[78][110]. Analysis of cohort data found no association between the highest intakes of chlorophyll and colorectal, colon or rectal cancer ^[79][111].

Many health outcomes were improved by three or more bioactive pigments, such as a decreased risk of all-cause mortality with the highest intakes of lycopene, beta-cryptoxanthin, beta-carotene, alpha-carotene, lutein and zeaxanthin, flavones and anthocyanin/anthocyanidin. Other improved health outcomes which were associated with three or more bioactive pigment colors were body weight; total cholesterol/lipid profiles; inflammatory biomarkers; CVD, CHD, CVD mortality; stroke; T2DM; and multiple cancers including breast, oral, lung, prostate, bladder, colorectal/colon/rectal and gastric.

Some health effects were unique to only one or two bioactive pigments or colors. Every FV bioactive pigment color had a single highly unique health effect which was not associated with any other pigment color, except red and yellow (Table 12). For example, only red bioactive pigments were associated with a decreased risk of pancreatic and laryngeal cancer, and only pale-yellow pigments were associated with improved exercise performance. All bioactive pigment colors also had other unique health effects that were associated with only two bioactive pigment colors. For example, decreased risk of cervical cancer was associated with only red and orange bioactive pigments, and decreased risk of esophageal cancer was only associated with white and blue/purple bioactive pigments (Table 12).

Of the highly unique health effects (i.e., significant effect in a single color of FV), only four outcomes have been confirmed as being truly unique by being tested for association with three or more different bioactive pigments. Waist circumference, unique to carotenoids, was found not to be affected by anthocyanins nor flavonols; risk of hypertension, unique to anthocyanins, was found to have no association with flavones nor flavonols; risk of preeclampsia, unique to beta-carotene, was found to have no association with alpha-carotene, lutein nor lycopene; and risk of liver cancer, unique to flavones, was found to have no association with anthocyanins nor flavanols (Table 12). The remaining highly unique health effects reported in Table 12 were only tested for association with one or two bioactive pigments, and it is therefore unknown if they may be improved by other bioactive pigments also.