Partner Bereavement and Risk of Herpes Zoster: Results from Two Population-Based Case-Control Studies in Denmark and the United Kingdom

• Published in: Clinical infectious diseases 2017-03, Vol.64 (5), p.572-579
• Main Authors: Schmidt, Sigrun A. J., Vestergaard, Mogens, Pedersen, Henrik S., Schønheyder, Henrik C., Thomas, Sara L., Smeeth, Liam, Mansfield, Kathryn E., Sørensen, Henrik T., Forbes, Harriet J., Langan, Sinéad M.
• Format: Article
• Language: English
• Subjects: Adult; Aged; Aged, 80 and over; ARTICLES AND COMMENTARIES; Bereavement; Case-Control Studies; Denmark - epidemiology; Epidemiology; Female; Grief; Herpes viruses; Herpes Zoster - epidemiology; Herpes Zoster - etiology; Humans; Major; Male; Middle Aged; Odds Ratio; Population Surveillance; Risk; Risk factors; Sampling; Socioeconomic Factors; Stress, Psychological; Time Factors; United Kingdom - epidemiology
• ISSN: 1058-4838; 1537-6591; 1537-6591
• DOI: 10.1093/cid/ciw840

Background. Psychological stress is commonly thought to increase the risk of herpes zoster by causing immunosuppression. However, epidemiological studies on the topic are sparse and inconsistent. We conducted 2 parallel case-control studies of the association between partner bereavement and risk of zoster using electronic healthcare data covering the entire Danish population and general practices in the UK Clinical Practice Research Datalink. Methods. We included patients with a zoster diagnosis from the primary care or hospital-based setting in 1997–2013 in Denmark (n = 190 671) and 2000–2013 in the United Kingdom (n = 150 207). We matched up to 4 controls to each case patient by age, sex, and general practice (United Kingdom only) using risk-set sampling. The date of diagnosis was the index date for case patients and their controls. We computed adjusted odds ratios with 99% confidence intervals for previous bereavement among case patients versus controls using conditional logistic regression with results from the 2 settings pooled using random-effects meta-analysis. Results. Overall, the adjusted odds ratios for the association between partner bereavement and zoster were 1.05 (99% confidence interval, 1.03–1.07) in Denmark and 1.01 (.98–1.05) in the United Kingdom. The pooled estimates were 0.72, 0.90, 1.10, 1.08, 1.02, 1.04, and 1.03 for bereavement within 0–7, 8–14, 15–30, 31–90, 91–365, 366–1095, and >1095 days before the index date, respectively. Conclusions. We found no consistent evidence of an increased risk of zoster after partner death. Initial fluctuations in estimates may be explained by delayed healthcare contact due to the loss.