Transmission Dominance Under Random-Contact Intensification in Epidemic Networks: Multilayer Contact Network Simulation Study.

Background: In the context of COVID-19, infection spread through human contact networks remains a major public health challenge. Beyond cumulative infections and deaths, it is necessary to understand which contacts matter most, and which population segments contribute most to transmission under different social conditions. In multilayer urban networks with community structure, routine contacts coexist with incidental encounters, and it remains unclear whether incidental encounters can alter epidemic burden and the main contributors to transmission when per-layer contact caps and routine-contact minima are unchanged (for the nonrandom layers). Objective: Under explicit daily-contact constraints, we examined (1) how changing overall contact opportunities affects epidemic speed and burden when incidental encounters are held fixed, and (2) whether increasing incidental encounters alone, per-layer contact caps, and routine-contact minima fixed (for the nonrandom layers), shifts the main contributors to transmission from a high-contact group to a medium-contact group, and the underlying network mechanism. Methods: We constructed a multilayer potential contact network for a synthetic urban population of 10,038 individuals, representing household, school, workplace, distance-driven activities, and incidental encounters as separate layers. Daily contact networks were sampled from the potential network each day, and transmission was simulated for 120 days using a Susceptible-Exposed-Infectious-Removed model with vaccination. Individuals were classified into high-contact and medium-contact groups based on baseline contact intensity, and group contribution combined each group's share of infectious individuals and its per-infectious effective transmission yield. Contact-constraint parameters were calibrated using an online survey in Tokyo and Kanagawa (n=1089), and scenario comparisons and parameter sweeps were used to locate the transition point. Results: With incidental encounters held fixed, higher overall contact opportunities produced earlier and higher epidemic peaks and larger cumulative infections and deaths, whereas reduced opportunities slowed and prolonged spread. Holding overall contact opportunities and routine contacts fixed, increasing incidental encounters shifted the main contributors to transmission: higher-contact individuals accounted for more effective transmissions at low incidental contact, whereas medium-contact individuals accounted for more beyond a clear transition point. Network visualization and schematics suggest a bridge-allocation mechanism, where stronger incidental contact adds cross-community bridges that more often terminate at medium-contact individuals and carry infection into less-affected communities. Across R=30 replicate runs under fixed settings, the dominance flip was consistently observed, and the estimated threshold W∗ showed a narrow but nonzero distribution (reported as median and IQR). Conclusions: In multilayer urban contact networks with community structure, our results indicate that intensifying incidental encounters can change the main contributors to transmission even when overall contact opportunities and routine contacts are unchanged. We present an analysis framework under explicit daily-contact constraints to identify this contributor shift and its transition point, supporting comparisons of intervention priorities across social contact conditions.