Back

Youth co-creation in health interventions: a systematic review and meta-analysis

Tao, Y.; Adedeji, O. J.; Sakuma, Y.; Conklin, J.; Magista, M.; Aung, M. T.; Septiawa, K. G. A. P.; Prado, E. O.; Siddique, M. K. B.; Azugo, N. O.; Koledowo, A. A.; Walker, A.; Chamouni, S.; Condoy, J. S. I.; Gonzales, J. E. V.; Prieto, G. J.; Sangucho, I. A. S.; Adeboye, A.; Mutisya, L.; Nono, J.; Mawuli, K.; Aribodor, O. B.; Mier, J. D.; Babatunde, Y.; Li, C.; Nwaozuru, U. C.; Day, S.; Kpokiri, E. E.; Tang, W.; Walker, D.; Tucker, J. D.

2025-10-22 public and global health
10.1101/2025.10.20.25338251 medRxiv
Show abstract

IntroductionCo-creation is an iterative, bidirectional collaboration between researchers and laypeople to create knowledge. Co-creation has increasingly been recognized as an effective strategy for developing youth (10-35 years old) health interventions. This systematic review and meta-analysis examines the effectiveness of youth co-created interventions using quantitative approaches. MethodWe followed the Cochrane Handbook and searched PubMed, CINAHL, and Global Health on December 4, 2024, for studies reporting on youth co-creation in health. Studies were included if they involved youth co-creators in planning, designing, implementing, or evaluating interventions and reported quantitative health-related outcomes. We extracted data on youth engagement, implementation, and health outcomes, required resources, and implementation factors. Random-effects meta-analysis was used to pool the effects of the co-creation interventions on health outcomes. The study was registered in PROSPERO, CRD42024615528. ResultsFrom 17,869 citations, 112 studies were included (75,906 participants). Studies were in high-income countries (97 studies, 86.6%), middle-income countries (12 studies, 10.7%), and low-income countries (two studies, 1.8%). Interventions focused on mental health (36 studies, 32.1%), physical activity (19, 17.0%), and sexual health (11 studies, 9.8%). Most studies focused on adolescents aged 10-19 years (71 [63.4%]). Co-creation outputs included digital platforms (30.4%), toolkits or curricula (28.6%), and creative media (25.9%). The co-creation phase ranged from one-time sessions to over a year, with 43.8% of studies lasting longer than three months. Youth were universally involved as co-designers (100%) and frequently served as evaluators (40.2%), facilitators (20.5%), and presenters (13.4%), with additional roles including peer educators, outreach supporters, and co-researchers. While 63.4% of studies reported some level of youth decision-making power, only 15.2% granted lead roles and 4.5% offered final authority. Training for youth was provided in 53.6% of studies, and training for adults was provided in 38.4%, most commonly related to intervention delivery (10.7%) and facilitation or power-sharing (9.8%). However, only 49.1% of studies reported offering compensation to youth participants, and just 58.0% provided public or academic credit. Facilitators of successful co-creation included youth leadership, supportive partnerships, stakeholder engagement, flexible and relevant content, and youth training. Barriers included limited stakeholder involvement, structural barriers, insufficient resources, digital access issues, and implementation challenges. Meta-analysis showed that youth co-creation interventions improved mental health outcomes, with significant reductions in depressive (four studies, WMD: -8.63, 95% CI: -13.52 to -3.75; evidence level: low) and anxiety symptoms (four studies, WMD: -8.47, 95% CI: -12.55 to -4.38; evidence level: moderate). Co-creation interventions may increase psychological well-being (four studies; WMD: 2.31, 95% CI: -1.22 to 5.84; evidence level: very low). ConclusionYouth co-creation interventions are associated with improved health outcomes and have been implemented across diverse settings. However, youth involvement remains limited beyond design. More research is needed to optimize co-creation processes and evaluate their effectiveness.

Matching journals

The top 4 journals account for 50% of the predicted probability mass.

50% of probability mass above

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.