Outcome, time point (number of included studies)	Assessment of publication bias
Resilience, post‐intervention (9 studies)a	We drew a funnel plot (see Effects of interventions and Appendix 16), which is rather symmetrical in shape and shows no clear visual evidence of asymmetry No statistical evidence of asymmetry (see also Effects of interventions; Egger's test: t = −1.42, df = 7, P = 0.20) Results of grey literature (Stephens 2012; no evidence of effect of positive direction) does not differ from other published studies (Barry 2019; Houston 2017), which also found no evidence of effect Difficult to assess small‐study effects due to lack of larger studies; but overestimation of effects in smaller studies unlikely, as the meta‐analysis also included small studies with non‐significant results (Erogul 2014; Mathad 2017; Peng 2014) No relevant conflicts of interest for included studies during the study period
Anxiety, post‐intervention (7 studies)a	We drew a funnel plot (see Effects of interventions and Appendix 16), which is rather symmetrical in shape and provides no clear visual evidence of asymmetry No statistical evidence of asymmetry (see also Effects of interventions; Egger's test: t = −1.61, df = 5, P = 0.17) No grey literature that could have differed from published studies Difficult to assess small‐study effects due to lack of larger studies; but they are unlikely to be relevant here, since included studies (which had small sample sizes) found significant (Wang 2012) as well as non‐significant results (Houston 2017) One potential conflict of interest for one of the included studies disclosed. One of the researchers of Kötter 2016 is a certified coach of a coaching programme (wingwave®), which was tested in the study. There were no further relevant conflicts of interest for the other included studiesb during the study period
Depression, post‐intervention (6 studies)a	We drew a funnel plot (see Effects of interventions and Appendix 16), which is rather symmetrical in shape and shows no clear visual evidence of asymmetry No statistical evidence of asymmetry (see also Effects of interventions; Egger's test: t = −0.85, df = 4, P = 0.44) No grey literature that could have differed from published studies Difficult to assess small‐study effects due to lack of larger studies; but overestimation of effects in smaller studies seems unlikely, as the meta‐analysis included small studies with significant (Wang 2012) as well as non‐significant results (Barry 2019; Recabarren 2019) One potential conflict of interest for one of the included studies disclosed. One of the researchers of Kötter 2016 is a certified coach of a coaching programme (wingwave®), which was tested in the study. There were no further, relevant conflicts of interest for the other included studiesb during the study period
Stress or stress perception, post‐intervention (7 studies)a	We drew a funnel plot (see Effects of interventions and Appendix 16), which is rather symmetrical in shape and provides no clear visual evidence of asymmetry No statistical evidence of asymmetry (see also Effects of interventions; Egger's test: t = −1.55, df = 5, P = 0.18) Results of grey literature (Stephens 2012; no evidence of effect of negative direction) does not differ from other published studies (Barry 2019; Houston 2017; Warnecke 2011), which also found no evidence of effect Difficult to assess small‐study effects due to lack of larger studies; but overestimation of effects in smaller studies seems unlikely, as the meta‐analysis included small studies with significant (Erogul 2014) as well as non‐significant results (Stephens 2012; Warnecke 2011) One potential conflict of interest for one of the included studies disclosed. One of the researchers of Kötter 2016 is a certified coach of a coaching programme (wingwave®), which was tested in the study. There were no further, relevant conflicts of interest for the other included studiesb during the study period
Well‐being or quality of life, post‐intervention (4 studies)a	We drew a funnel plot (see Effects of interventions and Appendix 16), which is rather symmetrical in shape and shows no visual evidence of asymmetry No statistical evidence of asymmetry (see also Effects of interventions; Egger's test: t = 0.12, df = 2, P = 0.91) No grey literature that could have differed from published studies Difficult to assess small‐study effects due to lack of larger studies; but overestimation of effects in smaller studies seems unlikely, as the meta‐analysis included small studies with significant (Wang 2012) as well as non‐significant results (Mathad 2017; Recabarren 2019; Smeets 2014) No relevant conflicts of interest for included studies during the study period
df: Degrees of freedom; P: P value of Egger's test; t: T value of Egger's test