Three RCTs were included in the review (306 participants); one parallel group design and two cross-over designs. Two studies received a Jadad score of four and one study received a Jadad score of five.
Pain reduction (three RCTs): A statistically significant effect in favour of rose-hip was found (ES 0.37, 95% CI 0.13 to 0.60, p=0.0019). No evidence of statistical heterogeneity was found.
Use of rescue medication (three RCTs): A small but statistically significant difference was found in favour of rose-hip (ES 0.28, 95% CI 0.05 to 0.51, p=0.018). No evidence of statistical heterogeneity was found.
Number of patients responding to therapy (three RCTs): Patients receiving rose-hip were significantly more likely to respond to treatment than patients receiving placebo (OR 2.19, 95% CI 1.38 to 3.48, p=0.00089). No evidence of statistical heterogeneity was found. The combined osteoarthritis corresponded to a number needed to treat of six patients (95% CI 4 to 13).
Adverse events: Mild cases of gastrointestinal discomfort and acid regurgitation were reported in both study arms. In one trial, mild unwanted effects that did not cause withdrawal were reported for frequent voiding, diarrhoea, constipation and short episode of mild urticaria. No statistically significant between group differences were found.