Generalised anxiety disorder and excess cancer deaths: findings from a large, longitudinal population study

Abstract Background Generalised anxiety disorder is a common and persistent disorder, and is associated with increased risk for disability and suicide. Previous studies have examined whether anxiety contributes to excess mortality from principal causes of death, with mixed findings. Possible mechanisms linking anxiety, pathogenesis, and premature mortality include dysregulation of the hypothalamus–pituitary–adrenal axis, impaired immunity, and poor coping. We aimed to assess whether generalised anxiety disorder contributes to excess cancer deaths using a large, longitudinal population study. Methods We used data from over 15 000 British participants over the age of 40 years from the EPIC-Norfolk study. We used health and lifestyle questionnaires to capture sociodemographics and medical history, and death certificates from the UK Office of National Statistics to identify cancer deaths. Generalised anxiety disorder (Diagnostic and Statistical Manual of Mental Disorders, fourth edition) was measured in 1996–2000, and deaths from all cancers in participants were recorded between 2000 and 2015. To determine the association between anxiety and cancer mortality, Cox proportional hazards regression was conducted while adjusting for psychiatric and physical comorbidity, disability, medication use, body-mass index, behavioural risk factors, and sociodemographics. This study received ethics approval. Findings 126 of 7139 men and 215 of 8799 women had generalised anxiety disorder; 796 of 7139 men and 648 of 8799 women died of cancer. Men with anxiety were over twice as likely as men without anxiety to die from cancer over 15 years of follow-up, but this association between anxiety and excess cancer deaths did not exist in women (hazard ratio 2·14, 95% CI 1·32–3·46 vs 1·03, 0·60–1·76). When we excluded those with a history of cancer, findings remained unchanged. Also, the association in men persisted in the long term, after the first 5 years were discounted. Interpretation In men, anxiety has been linked to overexpression of tumour necrosis factor α and interleukin 6, implicated in inflammation-associated carcinogenesis. Compared with women, men are more likely to develop substance misuse to cope with anxiety, but when we accounted for behavioural risk factors, the association remained unchanged. Anxiety can lead to more frequent medical consultations and earlier detection of disease: when we discounted the first 5 years, the association remained strong in the long ter