The prevalence of adverse mental health symptoms increased during the initial phase of the COVID-19 pandemic.1, 2 Establishing whether the rise in symptoms has persisted is crucial. The finding of Daisy Fancourt and colleagues3 was encouraging, in that anxiety and depression symptoms decreased among participants with at least two longitudinal follow-up measures in the UCL COVID-19 Social Study. However, our optimism was tempered by our own data revealing non-response bias in a retrospective analysis4 of participants in The COVID-19 Outbreak Public Evaluation (COPE) Initiative. Adjusting for demographic differences in longitudinal survey participation, respondents who completed more than two out of four invited surveys between April and September, 2020, had significantly lower prevalence of adverse mental health symptoms at the first timepoint (April, 2020) than those who did not. Furthermore, respondents who had anxiety or depression symptoms in May, 2020, after not having had these symptoms in April, 2020, had higher odds of completing fewer follow-up surveys compared with respondents without these symptoms (anxiety symptoms, adjusted odds ratio [aOR] 1·7 [95% CI 1·2–2·3], p=0·0015; depression symptoms, aOR 1·6 [95% CI 1·2–2·1], p=0·0046).4 Together, these data suggest that respondents who consistently completed surveys had better mental health initially and had more favourable trajectories than those who did not do the follow-up surveys. Fancourt and colleagues analysed data from the subset of the UCL COVID-19 Social Study participants who completed at least three surveys.3 Although no response rate was reported, we do know that their criteria excluded 22 828 (38·5%) of 59 348 participants who completed at least one survey. This raises the question: were the 22 828 individuals who they excluded from analysis different from the participants analysed? If so, the reported mental health improvement might partially reflect survivorship bias.
The answer to this question is consequential, as a premature conclusion that adverse mental health symptoms have decreased could affect mental health service planning. Assessment and management of mental health during the pandemic remain urgently needed. However, sampling and design choices should inform interpretation of findings. Although Fancourt and colleagues implemented measures to address biases, including population weighting and latent growth models to reduce sociodemographic and confounder biases, neither of these measures address potential survivorship bias. Our data suggest that individuals who completed more than two surveys in 2020 might be more resilient than those who did not. Future research should further characterise survivorship bias in longitudinal mental health surveys. Repeated cross-sectional surveillance can be used to estimate population-level mental health with time. Indeed, findings obtained by this design in the USA indicate that adverse mental health symptoms documented early in the pandemic have not abated.5
For more on The COPE Initiative see www.thecopeinitiative.org
Acknowledgments
All authors report institutional grants to Monash University from the CDC Foundation, with funding from BNY Mellon, WHOOP, and Hopelab. MÉC reports grants from the Australian-American Fulbright Commission sponsored by The Kinghorn Foundation, and personal fees from Vanda Pharmaceuticals. CAC reports grants and personal fees from Teva Pharmaceuticals Australia, grants from the National Institute of Occupational Safety and Health of the US Centers for Disease Control and Prevention (grant number R01-OH-011773), educational and research support from Philips Respironics, an endowed professorship provided to Harvard Medical School from Cephalon, an institutional gift to Brigham and Women's Hospital from Alexandra Drane, personal fees and equity interest in Vanda Pharmaceuticals, and a patent on Actiwatch-2 and Actiwatch Spectrum devices with royalties paid from Philips Respironics. The interests of CAC were reviewed and managed by Brigham and Women's Hospital and Partners HealthCare in accordance with their conflict of interest policies. CAC also served as a voluntary board member for the Institute for Experimental Psychiatry Research Foundation. SMWR reports grants and personal fees from the Australian Government-funded Cooperative Research Centre for Alertness, Safety and Productivity, grants and institutional consultancy fees from Teva Pharmaceuticals Australia, and institutional consultancy fees from Vanda Pharmaceuticals, Circadian Therapeutics, BHP Billiton, and Herbert Smith Freehills. All other authors declare no competing interests. SMWR and MEH were co-senior authors.
References
- 1.Pierce M, Hope H, Ford T, et al. Mental health before and during the COVID-19 pandemic: a longitudinal probability sample survey of the UK population. Lancet Psychiatry. 2020;7:883–892. doi: 10.1016/S2215-0366(20)30308-4. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Czeisler MÉ, Lane RI, Petrosky E, et al. Mental health, substance use, and suicidal ideation during the COVID-19 pandemic—United States, June 24–30, 2020. MMWR Morb Mortal Wkly Rep. 2020;69:1049–1057. doi: 10.15585/mmwr.mm6932a1. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Fancourt D, Steptoe A, Bu F. Trajectories of anxiety and depressive symptoms during enforced isolation due to COVID-19 in England: a longitudinal observational study. Lancet Psychiatry. 2020;8:141–149. doi: 10.1016/S2215-0366(20)30482-X. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Czeisler MÉ, Wiley JF, Czeisler CA, Rajaratnam SMW, Howard ME. Uncovering survivorship bias in longitudinal mental health surveys. medRxiv. 2021 doi: 10.1101/2021.01.28.21250694. published online Feb 1. (preprint). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.US Centers for Disease Control and Prevention. National Center for Health Statistics Indicators of anxiety or depression based on reported frequency of symptoms during the last 7 days. Household Pulse Survey. 2020. https://www.cdc.gov/nchs/covid19/pulse/mental-health.htm