China Health and Nutrition Survey (CHNS), a longitudinal cohort using repeated dietary measurement across 7 survey waves (1991-2009), analysed here as egg intake (1991-2009 cumulative average) vs diabetes status ascertained by fasting glucose in 2009 (n=8,545, 11.1% developed diabetes; mean/median follow-up ~11-12y). Compared to the lowest quartile of egg intake (0-9.0 g/d), adjusted ORs for quartiles 2-4 were 1.29, 1.37 and 1.25 respectively (all 95% CIs exclude 1; p-trend=0.029) - i.e. a positive, roughly flat-above-threshold dose-response, not a monotonic one. This is a Chinese cohort finding elevated risk, cutting against the simple “Asia = null/inverse” story and showing heterogeneity within Asia (contrast with the two Japan/Korea sources above, and with Ni 2020’s null Hangzhou cross-sectional study and Shi 2011’s positive Jiangsu cross-sectional study, both excluded here as non-cohort designs - see orientation exclusions). relevance_note: the main China-specific prospective/longitudinal primary and a direct counter-example to “Asian cohorts are null/inverse” - the US/Asia split is a real central tendency, not a clean partition.