In this paper, we have proposed and examined a model for drivers of QoL that is grounded both in the ICF framework and in participant’s own perception of QoL. Virtually all aspects of the ICF framework made a material difference to the self-reported QoL of people with SCI. Social and psychological factors had key direct impacts and were also key mediators of burdens related to secondary health conditions, difficulty in completing desired activities and environmental barriers. Activity/participation problem burden and pain intensity also directly affected QoL, even after considering all other factors. While functional independence level was important, other aspects of the ICF framework had a greater overall impact on QoL. Primary analyses and sensitivity analyses both support these messages.

These findings underline opportunities to improve QoL outcomes for people with SCI at every level of the health system and suggest that improvements in each and any of the ICF domains are likely to be worthwhile in improving QoL. At system level, participants clearly struggled if access to services, care, assistive devices or supplies was a barrier, with very few participants reporting good QoL in these circumstances. At community level, the importance of community acceptance and inclusiveness is highlighted, with very few participants reporting good QoL if they perceived that the attitudes of people around them was a barrier. The negative impact of social attitudes and issues with access to care and services is supported by the 2020 international report from the InSCI survey across 22 countries, in which the health system indicator showing the strongest association with overall QoL was social attitudes, followed by access to health care, nursing care and access to public spaces. These issues were not limited to any single participating country, although country-specific issues for improvement varied [14]. At an individual level, the path analyses emphasise the role of social and psychological factors.

Substantial existing literature supports our finding of central roles for social and psychological factors as drivers of QoL ratings in people with SCI [9]. In relation to social factors, Whalley Hammell, for example, concluded in a qualitative review (2004) that dissatisfaction with life after SCI arises substantially due to social disadvantage [3]. Post and van Leeuwen in their review identified that emotion-focussed social support was more strongly related to life satisfaction than problem-focussed social support, that quality rather than quantity of social support was more strongly associated with mental health, and that mental health was most strongly associated with social integration and reassurance of worth compared with other subscales of the Social Provisions Scale [9]. Other recent studies have examined links between life satisfaction and loneliness, relationship quality, social network intimacy, and peer support [15,16,17,18]. A path analysis in an African sample of people with SCI found that motor function, sensory function, self-esteem, functional independence level and social support all contributed to QoL, with the largest contribution from social support [19]. Their finding mirrors ours on the importance of social integration. From the psychological perspective, reviews of empirical evidence on the connections between personal psychological factors and QoL indicate that self-efficacy, perceived control, resilience, motivation/purpose in life, and hope are consistently associated with life satisfaction and wellbeing, and social skills, self-efficacy and personality traits are associated with better mental health and higher participation [2, 20].

Thus, both our findings and other accumulating evidence indicate that social and psychological factors substantially influence QoL and life satisfaction in people with SCI. Where the quality of social connections and personal relationships can be strengthened, the strongest individual effect on QoL seen in our main path model was that of social integration. Where psychological resources, personal outlook and aspects of mental health can be strengthened, these improvements would translate not only to direct benefits but also indirect ones since these factors mediated perceived burdens from other issues and the response to them.

Our findings also support an influence of perceived injustice, alongside self-efficacy, on QoL in people with SCI. Perceived injustice is associated with adverse QoL and mental health outcomes in trauma care settings and in people with pain [21, 22]. It may represent a core appraisal influencing adjustment after SCI, and a potentially modifiable target for intervention to improve QoL [23].

Our finding that effects of overall burden of secondary health conditions on QoL were substantially indirect after accounting for injury characteristics, functional independence level, mobility and general covariables accords with a concentric biopsychosocial model for QoL. The concentric biopsychosocial model specifies that social factors mediate or moderate the effects of biological factors on QoL, and psychological factors mediate or moderate the effects of both biological and social factors. Notably, individual conditions such as pain demonstrated direct effects in addition to indirect effects when modelled separately from the global burden score. Smedema evaluated a concentric biopsychosocial model centred around life satisfaction in 235 people with SCI in the United States and also found that psychosocial factors mediated effects of pain on life satisfaction substantially, but not entirely [24]. Muller et al. found a mediating role of participation in the effect of pain on depressive mood and QoL, which did not apply across the board but depended on time since injury, severity and completeness [25].

Several recent Canadian studies have examined the role of secondary health conditions alongside other factors in people with SCI. Rivers et al. used path modelling to examine the effects of secondary health conditions on function, health-related QoL and life satisfaction. At each stage of their path model, the number of secondary health conditions had negative effects, not only on function, but also on QoL and life satisfaction [7]. Sweet et al. used structural equation modelling to examine jointly the effect on QoL of latent constructs representing unmet vital needs (health care, financial care, transport, housing and equipment), unmet developmental needs (peer support, job training, recreation, counselling), neurophysiological symptoms (sleep, spasticity, autonomic dysreflexia, fatigue, dizziness), bowel and bladder dysfunction, sexual dysfunction and pressure ulcers. Significant effects were found for both types of unmet need and for neurophysiological symptoms [26]. Engel et al. reported negative total effects of SHC on subjective wellbeing, largely or entirely mediated via capability wellbeing (measured via five domains for feeling settled and secure, having love friendship and support, being independent, achievement/progress, and enjoyment/pleasure) and health-related QoL (measured via three physical domains including pain and five psychosocial domains for mental health, happiness, self-worth, coping and relationships) [27].

While these path models cannot be directly compared with ours, a consistent message is that secondary health conditions are important alongside other factors and that mediating factors are involved in this effect. Our finding that overall burden of secondary health conditions and NEFI-S scores showed mainly indirect effects on overall QoL suggests that their impacts can be mitigated if social, psychological and mental health factors and ability to participate in desired activities can be maintained.

Our study has several limitations. Inter-relationships between factors under the ICF framework are complex and bidirectional. It is not possible to make causal inferences about these relationships from cross-sectional data. Therefore, in this paper, the interpretation of different aspects of lived experience as potential determinants of QoL, and as potential mediating factors, is a conceptual interpretation based on a theoretical framework for QoL, rather than being an evidence based conclusion drawn from the analysis of the cross-sectional data. While our suggested path models are not definitive, nonetheless the evaluation of a reasonable hypothesised model in conjunction with key sensitivity analyses is informative. The hypothesised path model draws on key aspects of the ICF framework, biopsychosocial models for wellbeing and health-related QoL and the Spinal Cord Injury Adjustment Model to reflect a conceptual framework for potential determinants of QoL [28, 29]. In future work, we will be able to examine the temporal nature of relationships under the model using data from the longitudinal phase of Aus-InSCI. This will allow further confirmation of effect direction.

Our path model places current work participation at the most proximate location to QoL, and this aspect of the modelling is likely oversimplified, in so far as work participation potentially fosters positive settings in the rest of the framework, such as positive subjective social position and financial resources, social connections, positive outlook, sense of purpose and sense of self-worth and self-efficacy. Indeed, in a systematic review of qualitative studies, Hilton et al. concluded that there is an intrinsic need for occupational engagement which generally outweighs difficulties involved in regaining work after SCI, and that timely reassurance, support and assistance about the possibilities and value of work provide injured people with a positive message about their future [30]. Therefore, although our analysis did not show a further positive effect of current work participation on overall QoL ratings when all other aspects of the ICF framework were already accounted for, we would not infer any contradiction of other literature supporting positive benefits from work participation.

Our results arise in a relatively high resource setting, where investments in disability care and support have been in focus politically and there is an ongoing open public discourse about supporting the participation and integration of people with disability in society. It is unknown how well the findings would generalise to other contexts where the relative weight of issues may differ. However, the key message that biological, environmental, social and psychological dimensions of lived experience were all impactful on perceived QoL of people with SCI when taken together is unlikely to be particular to only one resource setting or societal context. Patterns of contributing factors to QoL may be different for different subgroups of the population with SCI, which will require future investigation. Data from other countries participating in InSCI will enable us to explore these relationships further.

For our QoL outcome measure in our path models we used the single question of how participants rated their overall QoL on five response levels, as a continuous measure. The key advantage and strength of this approach is that each person’s response is based on their own perception of what overall QoL means [3, 28]. A potential disadvantage is that this measure would have relatively low precision, however, the large cohort size supports this choice for analysis.