Identifying Persuasive Visual Features within Tobacco Pictorial Warnings: Effects on Anticipated Loss of Face, Gifting, and Refrain Intentions Among Chinese Men Who Smoke

= 2,306), we examined which visual features predict these three CSOs. Each participant viewed six randomly presented PWs. The experimental data were merged with manually coded content features across 24 dimensions after recounting all the features and esthetic visual characteristics of each PW extracted through computer vision analysis. Multilevel regression models revealed that thematic categories, including oral health, fertility health, and cardiovascular health, as well as specific visual elements like depictions of diseased organs and medical equipment, consistently predicted all three outcomes. The effects of other features varied. Findings provide insights for designing tobacco control campaigns and implementing cigarette package warning labels in China. Our results show the importance of tailoring PWs to resonate with culturally specific concerns in tobacco control messaging, offering important implications for pictorial warnings targeting Chinese male smokers.