Climate change impacts humans: A cross-sector narrative review of impact pathways, cascading risks, and accountable quantification

Climate change affects human well-being through interacting physical, ecological, infrastructural, and socioeconomic processes rather than through a single exposure or disease pathway. Major assessments identify observed harms to health, food and water security, livelihoods, settlements, and ecosystems, alongside increasing risks with additional warming (IPCC, 2022a). This narrative review develops a cross-sector framework from a source inventory of 154 human-impact pathways in 14 domains, supported by major assessments, selected peer-reviewed studies, and institutional datasets. Literature selection used targeted web-index searches and source verification through 2 October 2026; the review does not estimate pooled effects or claim exhaustive systematic coverage. The synthesis distinguishes direct injury and illness, ecosystem-mediated harms, failures of essential services, financial transmission, and losses of social and cultural continuity. Representative quantitative findings illustrate substantial existing burdens, but attribution studies, ecological associations, exposure models, displacement records, and conditional forecasts measure different quantities and cannot be combined into a universal five-year trajectory (Vicedo-Cabrera et al., 2021; Wang et al., 2026; Clement et al., 2021). Compound hazards and service interdependencies can increase losses beyond the directly affected sector, while repeated shocks can reduce the capacity to recover or adapt (Zscheischler et al., 2018; IPCC, 2023). A reporting framework is proposed that records the outcome, population, baseline, geography, time horizon, emissions and development assumptions, adaptation, and uncertainty for each indicator. The central contribution is an integrated interpretation and measurement architecture, not a new forecast of global mortality, migration, or societal collapse. Effective assessment should preserve evidence of serious and interacting harms without converting uncertain thresholds or heterogeneous studies into unsupported calendar-specific predictions.