We eliminated the secondary outcome 'mean duration of vaccine protection'. We added 'dropouts' as a secondary outcome because this relates to the safety of the intervention.

We considered blinding of outcome assessment to be at low risk of bias when participants in double‐blind trials filled out cards received from the investigator themselves.

We considered the adverse event 'death' separately from serious adverse events as a secondary outcome for the review. We based this decision on the importance of death as a concept in both studies and clinical practice.

In Methods > Data collection and analysis > Measures of treatment effect > Continuous data, we added: "we inserted these data into an additional table".

In Methods > Unit of analysis issues, we used data from cross‐over studies (separated or grouped) when available.

In Methods > Sensitivity analysis, we added cross‐over studies.

In Data collection and analysis > Data synthesis, we changed the text to: "we conducted meta‐analyses using a random‐effects model".

Brenda NG Andriolo was previously known as Brenda NG Silva.

The first author Anna Maria Zaragoza Gagliardi 2019 (Gagliardi 2019) was changed to Juliana de Oliveira Gomes in 2022.

In this update (2022), the 'incidence of herpes zoster' was changed to 'cumulative incidence of herpes zoster' since the primary studies defined the incidence of outcome over a period of time.

The outcomes 'Interference of herpes zoster in activities' and 'ZBPI pain score during the first 14 days after HZ onset', which reflect how much herpes zoster pain limits people from performing their daily tasks, were removed from the analyses, since they were not pre‐specified in the protocol.