3636 clustered Guangzhou HIV-1 pol gene sequences (2008-2020)

Age mixing patterns are critical for understanding HIV-1 transmission and informing targeted public health interventions, yet they remain understudied in China and other concentrated-prevalence settings. HIV-1 pol gene sequences and sociodemographic data were collected from 8,395 people with HIV (PWH) newly diagnosed in Guangzhou, China (2008-2020). Sequences were grouped by birth year of PWH, and HIV-1 molecular transmission network were reconstructed. A total of 11,766 links were detected among 3636 PWH, primarily connecting PWH born in the 1970s and 1980s (16.7%), 1980s and 1990s (16.5%), 1980s and 1980s (15.0%), and 1990s and 1990s (11.5%). Significant age-assortative mixing patterns were observed overall (assortativity coefficient = 0.407, P < 0.001) and within networks of CRF07_BC (0.338), CRF01_AE (0.497), CRF55_01B (0.146) and subtype B (0.756) by transmission intensity matrices and assortativity coefficients. Multivariable exponential random graph models were further applied to assess associations of age mixing patterns with HIV-1 transmission links in the network, indicating that, compared with male-male pairs born in the 1980s , those born in the same decade—whether earlier (＜1960: aOR = 7.018, 95% CI: 5.733-8.590; 1960s: aOR = 2.297, 95% CI: 1.973-2.674; 1970s: aOR = 1.430, 95% CI: 1.309-1.563) or later (≥1990: aOR = 1.539, 95% CI: 1.430-1.656) were more likely to form a transmission link. These patterns were consistent across major HIV-1 genotypes. This molecular transmission network analysis revealed age-assortative mixing as a key driver of HIV-1 transmission in Guangzhou, highlighting the need for age-specific strategies to enhance HIV prevention efforts.