King's College London

Research portal

Extent of disease at first cancer presentation and previous anxiety and depressive symptoms: the HUNT study

Research output: Contribution to journalArticle

Robert Stewart, Sophie Dorothea Fosså, Matthew Hotopf, Arnstein Mykletun

Original languageEnglish
Pages (from-to)427-433
Number of pages7
JournalThe British journal of psychiatry : the journal of mental science
Issue number2
Publication statusPublished - 1 Aug 2020

King's Authors


BACKGROUND: Depressive symptoms are associated with higher cancer mortality, whereas anxiety symptoms are associated with lower than expected risk. AIMS: This study aimed to investigate the prospective association between depressive/anxiety symptoms and the extent of disease (EOD) of first cancer at diagnosis. METHOD: Prospective population-based study conducted from the second wave of the Nord-Trøndelag Health (HUNT) study. Of 65 000 residents comprehensively interviewed and examined for health status, 407 received first lifetime cancer diagnoses 1-3 years later, ascertained from the Cancer Registry of Norway, and had EOD recorded. Patients with localised disease or regional/distant spread at cancer diagnosis were analysed for earlier depressive/anxiety symptoms ascertained by the Hospital Anxiety and Depression Scale in HUNT. RESULTS: Beyond-local EOD was present in 59.8% of those with neither anxiety nor depression, in 76.6% of those with depression alone (odds ratio, 2.20; 1.08-4.49), in 39.3% of those with anxiety alone (odds ratio, 0.44; 0.20-0.96) and in 57.7% of those with both anxiety and depression (odds ratio, 0.92; 0.41-2.06). After adjustment for demographic and health status, and cancer type, these associations were marginally stronger, but no longer statistically significant (odds ratios, 2.26; 0.84-6.11; 0.43; 0.15-1.26; and 1.00; 0.98-1.03, respectively). CONCLUSIONS: In people who develop cancer, beyond-local EOD at diagnosis was more common in people with previous depression and less common in people with previous anxiety; however, independence from confounding factors could not be concluded.

View graph of relations

© 2018 King's College London | Strand | London WC2R 2LS | England | United Kingdom | Tel +44 (0)20 7836 5454