A total of 964 articles were identified through searching databases and other sources. Following removal of duplicates, 952 articles remained. After screening by title and abstract, 45 articles remained. Following application of specific inclusion and exclusion criteria, critical appraisal, and quality checking of articles, 16 were deemed acceptable for inclusion. See Fig. 1 PRISMA Chart.
Description of included studies
Sixteen studies were included in the review. One included sport performance, oxidative stress measures and side effects , eight included sport performance data only, [14, 18,19,20,21,22,23,24], three had oxidative stress measures [1, 25, 26], one had oxidative and cognitive measures , two included cognitive measures only [28, 29], and one included oxidative measures and side effects . Among the included studies, six were conducted in New Zealand, the remaining in the United Kingdom. Of the nine performance studies with included performance data, all used a cross-over design, and report the effect of New Zealand blackcurrant on either cycling, running or climbing hang time.
Blackcurrant and its effect on performance
The performance effect of NZ BC vs. placebo during exercise was investigated in nine studies. The meta-analysis using the fixed-effects model calculated the standardised mean percent effect of NZ BC on performance to be 0.45 (95% CI 0.09–0.81, p = 0.01), which is significant. The heterogeneity among studies included in the meta-analysis did not significantly affect the outcome (see Fig. 2).
Risk of bias in included studies
Whilst there are some gaps in the reported methods of included studies, there is an overall low to unclear risk that the true performance effect of NZ BC has been influenced by bias from included studies. The classification of bias is presented alongside Fig. 2. In summary, the nine studies were all classified as low risk bias for ‘random sequence generation’ as the cross over design was appropriate and randomisation was conducted. ‘Allocation concealment’ was unclear as the authors did not state their respective methods. Similarly the nine studies were assigned unclear risk bias of ‘incomplete outcome data’ as they were crossover and only included results of participants who had completed both arms of the trial. One study reported drop-out rates with reason, and rated low risk for incomplete data. All studies were classified as low risk of bias for ‘blinding’ as they were stated double-blind and mentioned intervention and control capsules were identical. A search for the associate protocol of each study was completed, however no protocol was found for all included studies. The studies were therefore all classified as unclear ‘selective reporting’ bias.
The distinction in ‘other’ bias classification resulted from inadequate conflict of interest statements. Five studies [14, 16, 19, 20, 23] were classified as low ‘other’ bias as the respective authors specified the study funder had no input in study design, results and/or choice to publish (See Fig. 3). The reviewers classified the remaining studies [18, 21, 22, 24] as unclear or high risk as they provided limited conflict of interest statement regarding trial design and choice to publish by the funder. Industry funded projects were deemed high risk in other sources of bias. Missing data was deemed unclear as data on drop-outs was only available in one study. Other sources of bias include inappropriate wash-out periods which was not detected.
Oxidative stress biomarkers
Outcomes from MDA and PC conducted during exercise trials were graphed and generally indicate a reduction in oxidative stress under the conditions of exercise. One study did report a greater increase in MDA with the intervention than the control; the other two studies indicating reductions. Both MDA and PC were raised after exercise with x/÷ factor SDs of 2.58 and 3.36 respectively. The qualitative statement indicating the magnitude of change in the geometric mean were varied between moderate and large; however the overall change in the intervention group was similar to the control group. There were insufficient papers to graphically represent any other oxidative stress or inflammatory markers.
One additional study measured MDA and PC without an exercise trial and reported trivial differences in PC when taking NZ BC (PC, BC 0.2 vs PL 0.18 nmol/mg protein, at baseline; BC 0.13 vs 0.11 nmol/mg protein, after 12 weeks supplementation), and lower MDA when taking NZ BC (MDA, BC 17.3 vs PL 23.5 ng/mL, at baseline; BC 7 vs PL 10 ng/mL after 12 weeks supplementation) .
Outcome measures have been reported for TNFα, IL-6, IL-10 and FRAP pre and post-exercise in those taking NZ BC or placebo [25, 26]. The pre-exercise inflammatory markers were higher with the intervention treatment than placebo (TNFα, BC 3 vs PL 2.1 pg/mL; IL-6, BC 12 vs 10 pg/mL and IL-10, BC 80 vs 60 pg/mL) and oxidative stress marker higher on the intervention than placebo (FRAP, BC 0.15 vs 0.14%), although none of these reached significance. The post-exercise inflammatory markers were the same or lower with the intervention treatment than placebo (TNFα, BC 9 vs PL 10 pg/mL; IL-6, BC 20 vs 21 pg/mL and IL-10, BC 60 vs 60 pg/mL) and oxidative stress lower on the intervention than placebo post exercise (FRAP, BC 0.12 vs PL 0.13%).
Cognitive biomarkers and effects of blackcurrant
Three studies reported platelet MAO-B concentrations after consuming NZ BC or placebo. All three show clear reduction with NZ BC (Platelet MAO-B, BC 1.6 vs PL 22.1 nM H202 production/μg protein/min ; Platelet MAO-B, BC − 280 vs PL 10 nmol H202 change from baseline ; Platelet MAO-B, BC 1417.86 vs PL 1739.99 nmol H202 ). Data were reported using heterogeneous methods thus making graphing or meta-analysis problematic.
Cognitive questionnaire data were reported by one study  with a Stroop cognitive test accuracy (F = 0.014, p > 0.01) and reaction time (F = 0.33, p > 0.1) no better with NZ BC. However, Rapid Visual Information Processing accuracy improved (F = 5.88, p = 0.005) with NZ BC.
Reported side effects
Three of sixteen studies made specific mention of reported side effects from taking dietary supplements or drinks containing NZ BC [2, 16, 27]. One reported minor gastrointestinal upset in two of the 24 participants consuming NZ BC , which provided the blackcurrant intervention as a drink from a combination of powder extract and concentrated syrup. In two of the studies there were no adverse/ side effects, stomach issues or other consequences from the intervention product [2, 27]. It is worth noting that data on side effects is not systematically collected and reported in all research, we also can’t be sure the side effects reported are the direct result of NZ BC or other ingredients or products.