Background Psychosocial factors are associated with various manifestations of severe asthma. Wider research and theory highlight complex, bi-directional pathways by which interactions may occur. However, few studies have explored relationships between multiple psychosocial factors and outcomes in severe asthma with these pathways in mind.

Objectives This study investigated combined and independent cross-sectional and longitudinal relationships of self-management behaviours, an index of self-management (ISM), anxiety, depression, perceived control (PC) and socio-demographic/economic characteristics with asthma control, quality of life (QoL) and severe attacks amongst asthma patients on high levels of treatment and/or with a history of admissions.

Methods Cross-sectional data from 132 adults recruited to a previously reported trial of an intervention and accompanying comparative study were subjected to multiple regression analyses. These systematically examined relationships between psychosocial factors and asthma outcomes, and were used to build final hierarchical regression models in which key clinical variables were controlled for. More limited data from a maximum of 112 patients were used to explore longitudinal relationships, primarily with asthma control.

Results Final hierarchical regression models accounted for up to 69% and 73% of the variability in asthma control and QoL respectively (p<0.001) and significantly predicted experience of a severe attack (p<0.001). Variables showing significant independent relationships to outcomes in these models are highlighted in the Abstract P175 Table 1. Some individual behaviours and the ISM showed independent and differing cross-sectional relationships to each outcome. Other psychological factors were related to subjective outcomes but not severe attacks. Relationships of some psychosocial factors (eg, depression, unemployment) to outcomes were not fully mediated by other variables, including self-management behaviours. In longitudinal analyses, there was some evidence for depression directly contributing to poorer subjective outcomes, whilst relationships of PC and anxiety with outcomes were more variable and inconsistent.

Conclusions Emotional and cognitive factors appear at least as important as self-management behaviours in relation to subjective outcomes in severe asthma. Along with a growing body of other research, findings suggest a particular need to identify and address depression amongst patients with severe asthma in practice, as in other chronic diseases.