Purpose: The objective of this meta-analysis was to evaluate the evidence on the effectiveness of glycerin trinitrate (NTG) measured by pain severity.

Materials and methods: CENTRAL, MEDLINE, EMBASE, SCOPUS, and WEB of SCIENCE databases were searched in November 2020. The study selection was performed by two independent reviewers. The risk of systematic bias was assessed according to the Cochrane Collaboration's domain-based evaluation framework.

Results: Of 87 identified records, the meta-analysis was conducted on eight RCTs. When grouping by the type of tendinopathy and combining the estimates obtained from all available time points, the pooled standardized difference in means (SMD) was -0.80 (95% CI -1.34 to -0.26), and the number needed to treat 3.53 (95% upper confidence limit 12.4). When combining all the available data (all types of tendinopathy and all time points), the pooled SMD was -1.57 (95% CI -2.47 to -0.67). Overall heterogeneity was high. The risk of systematic bias was low in most of the selected studies.

Conclusions: There is no evidence that NTG is more effective to reduce pain in tendinopathy than placebo. The effects of NTG were insignificant or borderline significant (probably insignificant clinically) concerning rotator cuff tendinopathy, Achilles tendinopathy, patellar tendinopathy, and lateral epicondylitis.IMPLICATIONS FOR REHABILITATIONA meta-analysis conducted on eight RCTs found no evidence that topical glycerin trinitrate is more effective to reduce pain in tendinitis than placebo.The effects were insignificant or borderline significant concerning rotator cuff tendinitis, Achilles tendinitis, patellar tendinitis, and lateral epicondylitis.The effects were independent of the dosage used.