Background: Vaccine cold-chain performance in mountainous regions is typically audited using binary threshold breaches, overlooking cumulative thermal variance and transport latency that may biologically attenuate humoral response. We evaluated whether continuous thermal instability gradients and last-mile delay duration are causally associated with spatial clustering of vaccine breakthrough, quantified through neutralizing antibody geometric mean titers (GMTs) derived from publicly available immunogenicity datasets.
Methods: We integrated three data layers such as high-frequency temperature logger datasets (30-minute intervals) from publicly documented mountainous vaccine distribution evaluations in South Asia and East Africa, terrain-derived altitude and slope gradients from NASA SRTM 30 m digital elevation models, and neutralizing antibody GMT trajectories from ImmPort Phase II/III vaccine immunogenicity repositories with individual-level serologic follow-up. Travel delay duration was modeled using OpenStreetMap-derived friction surfaces to compute realistic last-mile transport times. Thermal instability exposure was operationalized as cumulative mean kinetic temperature deviation, freeze-event minutes, and intra-route variance. Breakthrough clustering was detected using spatial scan statistics. Heterogeneous causal effects were estimated using generalized random forests, and population-level intervention contrasts were derived via targeted maximum likelihood estimation under counterfactual reduction of excursion burden.
Results: Across 1,842 monitored transport routes and 6,314 immunogenicity-linked recipients, increasing thermal variance above 1.8°C² during transport was associated with a 14.6% reduction in post-vaccination neutralizing GMT at day 28 (standardized mean difference −0.41). Routes with cumulative freeze exposure exceeding 45 minutes demonstrated a 2.3-fold higher probability of residing within significant breakthrough clusters (spatial likelihood ratio p<0.001). Travel delay duration exceeding 6 hours in high-altitude (>1,500 m) districts amplified GMT attenuation, with an interaction effect coefficient of −0.18 per additional hour of delay. Causal forest analysis identified pronounced heterogeneity among pediatric recipients and mRNA-based platforms. TMLE-estimated intervention modeling showed that reducing excursion variance to <0.8°C² would increase expected GMT by 11.2% and reduce predicted breakthrough clustering risk by 9.7 cases per 10,000 vaccinations. Sensitivity analysis using E-values indicated robustness against moderate unmeasured confounding.
Conclusions: Continuous thermal instability gradients and transport delay duration exert quantifiable, biologically coherent effects on neutralizing antibody response and geographic breakthrough clustering in mountainous districts. Modeling cold-chain performance as a causal exposure gradient rather than a binary failure reveals actionable targets for precision vaccine logistics and immunization program optimization.