Abstract Heat stress remains one of the most consistent environmental challenges limiting swine productivity, welfare, and sustainability, with economic losses that are likely to exceed current estimates. Modern genetic selection has increased productivity but has also elevated metabolic heat production, thereby increasing heat stress sensitivity, particularly in gestating and lactating sows. This work integrates physiological, behavioral, immunological, and genomic evidence to define heat stress resilience in pigs across the life course, with particular emphasis on in utero heat stress (IUHS) and its long-term consequences. Data demonstrate that IUHS, driven primarily by elevated maternal body temperature during early gestation, induces persistent alterations in offspring bioenergetics, thermoregulation, immune response, behavior, growth performance, and reproductive capacity that persist regardless of postnatal environment. These phenotypic outcomes are linked to placental insufficiency and epigenetic modifications, including differential DNA methylation of genes involved in metabolism, immune function, neurobiology, and vascular development. Heritability of thermoregulatory traits is further examined, and limitations of selecting solely on absolute body temperature are highlighted, as this trait is often negatively correlated with performance. Instead, rate of body temperature increase is proposed as a biologically relevant alternative trait that better captures heat loss efficiency without reducing productivity. Results from divergent selection lines demonstrate that improved thermotolerance can be achieved alongside maintained or enhanced reproductive performance, growth, and immune function across generations. Taken together, these observations underscore the need to integrate physiology-based phenotypes with genomic selection strategies, environmental management, and decision-support tools to improve heat stress resilience in pigs while safeguarding productivity and animal welfare.
Jay S Johnson (Wed,) studied this question.