From Anxiety to Work Productivity and Activity Impairment: The Mediating Role of Fatigue in Hereditary Angioedema

Hereditary angioedema (HAE) is a rare genetic disorder characterised by recurrent episodes of swelling [1]. In recent years, the humanistic burden of HAE, including impaired health-related quality of life (HRQoL), work productivity loss and mental health issues, has been increasingly recognised [2, 3]. At a societal level, absenteeism and productivity impairments also indirectly contribute to disease burden. Yet, causal relationships between these psychological and social outcomes are not well-understood in HAE. Thus, we carried out this study utilising causal mediation analysis, a method frequently used in health psychology research.

We analysed data from adult HAE patients in Hong Kong's CaSE-HAE program, a prospective family screening initiative offered in Hong Kong [4]. Their responses to patient-reported outcome measures (PROMs), namely the Hospital Anxiety and Depression Scale (HADS), Angioedema Quality of Life Questionnaire (AE-QoL) and Work Productivity and Activity Impairment Questionnaire (WPAI) at time of diagnosis were retrieved [3]. Their clinicodemographic characteristics (sex, age of diagnosis, HAE type, education level, comorbidities, past availability of HAE mediations, angioedema attacks and days of hospitalisation per year) were extracted. Patients with incomplete data were excluded. This study was approved by the Institutional Review Board of the University of Hong Kong/Hospital Authority Hong Kong West Cluster. All patients provided informed consent.

Causal mediation analysis was performed on R version 4.3.1 (R Foundation, Vienna, Austria) via the mediation package [5]. HADS Anxiety and AE-QoL Fatigue were designated as model exposure and mediator, respectively. To increase the robustness of our results, we constructed three models, respectively with WPAI Work productivity loss, WPAI Activity impairment, and AE-QoL Functioning as model outcomes. Sex, age of diagnosis, education level, angioedema attacks per year and days of hospitalisation per year were included in the three models as covariates. Besides, a bootstrapping analysis was conducted with 5000 simulations to yield a 95% confidence interval for our estimates. The Spearman correlations between PROM domains were also computed using IBM SPSS Statistics version 28 (IBM, Armonk, NY). Two-sided p < 0.05 denotes statistical significance.

This study included 25 HAE patients (52.0% female, median diagnosis age = 44 years, 72.0% Type 1 and 28.0% Type 2). Their clinicodemographic characteristics and PROM results can be found in Online Repository (https://osf.io/3nafm/). Overall, the median (interquartile range) scores were 11.0 (6.0–19.5) for HADS Anxiety, 35.0 (2.5–52.5) for AE-QoL Fatigue, 30.0 (10.0–60.0) for AE-QoL Functioning, 50.0 (10.0–80.0) for WPAI Work productivity loss and 50.0 (10.0–85.0) for WPAI Activity impairment. All five PROM domains demonstrated significant correlations (all p < 0.05; detailed correlation matrix available in Online Repository).

The results of causal mediation analysis are shown in Figure 1. In all three models, anxiety (indicated by HADS Anxiety) showed association with the model outcome (p = 0.022 for WPAI Work productivity loss; p = 0.015 for WPAI Activity impairment; and p = 0.002 for AE-QoL Functioning). All associations were significantly and largely mediated by fatigue (represented by AE-QoL Fatigue; 54.6%–79.3%, all p < 0.05), while all direct effects were insignificant (all p > 0.05). In the bootstrapping analysis, all total and indirect effects remained significant whereas all direct effects remained insignificant (see the Online Repository).

To the best of our knowledge, we are first to characterise the inter-relationships among anxiety, fatigue and work productivity and activity impairment among HAE patients. In addition, we demonstrated the mediating role of fatigue between mental health (anxiety or depression) and impaired HRQoL or work productivity, a pattern also seen among other conditions like systemic lupus erythematosus [6].

Fatigue, the most common prodromal symptom in HAE, is experienced by up to 75% of patients [7, 8]. While previously, physicians primarily focused on its potential to predict angioedema attacks, our findings suggest that fatigue may also be a distressing symptom with significant psychological and social impact. Fatigue can be complex and multifactorial, occurring before, after, or even between attacks [7]. It may be related to frequent emergency department visits and hospitalizations, or secondary to mental health issues such as anxiety, as demonstrated in this study. After adjusting for these factors, we still found significant association between anxiety and fatigue, and between fatigue and activity impairments, suggesting a significant psychological component in the pathogenesis of fatigue in HAE. Given that fatigue and anxiety, which may also trigger HAE, often remain unaddressed in HAE patient, future studies should explore potential dedicated strategies to significantly improve patients' HRQoL and work productivity [3, 7]. Indeed, this could be particularly beneficial for patients who still lack access to HAE-specific medications [9].

The sample size is relatively small, which may limit the generalisability and statistical power of our findings, which require further validation by larger multi-centre, multi-ethnic studies.

A standalone PROM or other objective markers were not used for fatigue assessment, nor did we distinctly differentiate between mental and physical fatigue. The temporal relationship between fatigue and HAE attacks warrants further investigations as longitudinal data was not available in this study. Despite the adjusted covariates, there may still be unconsidered confounders in this study.

In conclusion, among HAE patients, a significant association exists between anxiety and work productivity and activity impairment, and this association is significantly and largely mediated by fatigue. The importance of addressing fatigue among HAE patients is often overlooked and undertreated. In lieu of the lack of HAE-specific studies, it may be of great interest to investigate the effect of interventions extrapolated from other medical disorders such as physical activity/therapy, cognitive behavioural therapy, mindfulness-based stress reduction, health education management or complementary and alternative medicine.

Hugo W. F. Mak researched the data and drafted the manuscript. Jane C. Y. Wong, Valerie Chiang and Dorothy L.Y. Lam researched the data. Philip H. Li conceptualised and supervised the study. All authors contributed to and approved the final version of the manuscript before submission.

The authors declare no conflicts of interest.