Thread by @jennDSN

A 🧵 on #COVID's impact on #adolescents, summarizing our preprint compiling longitudinal data from 1,339 adolescents in the U.S., Netherlands, and Peru, studied by 12 research groups. 1/ psyarxiv.com/hn7us/

Studies included participants aged 9-18 years old. Sample was 43% White, 16% Hispanic/Latinx, 9% Biracial/Multiracial, 8% Black, 2.5% Asian, 1.5% other, 20.5% missing; the international studies did not collect ethnic/racial identity information. 2/

We converted measures of depression and anxiety symptoms to proportion of maximum scores (POMS). We also measured disease burden (local case and death rates) and severity of government restrictions. 3/ ncbi.nlm.nih.gov/pmc/articles/P…

Linear mixed effects models showed that depressive symptoms significantly increased from before to during the pandemic. The median POMS of depression symptoms, adjusted for repeated measures, increased from 0.20 to 0.25 (a 28% increase). 4/

Age did not moderate the increased depressive symptoms, but did for anxiety symptoms. On the whole anxiety symptoms did not significantly increase; for younger adolescents, anxiety symptoms decreased during the pandemic. 5/

Strongest increases in both depression and anxiety symptoms were seen in biracial/multiracial adolescents. Hispanic/Latino adolescents showed decreases in anxiety symptoms and Black adolescents showed smaller increases in depression symptoms. 6/

Change in depression & anxiety symptoms was moderated by case rate (more symptom increases in areas with lower disease burden). But government restrictions & case rates were negatively correlated. Stricter government restrictions were associated with more symptom increases. 7/

Takeaways (oversimplified): Longitudinal data are invaluable. Anxiety is multifaceted, subtypes may have been affected differently by pandemic (eg, social vs general). Inequities in US disproportionately affect biracial/multiracial adolescents' response to COVID. 8/

Government restrictions seem to have more negative impact on adolescent mental health than actual case or death rates, perhaps due to mismatch between restrictions and disease burden. Leave-one-out analyses showed general stability in all findings except death rate analyses. 9/

Led by @marjoleinbar @j_flan @Cait_Cavanagh & myself; contributions from @spbecker @estelleLberger @church_lang @EACrone @SuzannevdGroep @TiffanyCheingHo @luciamawe @anilak_99 @mitchprinstein @EvaTelzer @AprilGileThomas and many others! (Please tag yourself in...) 10/

Background story: We were 60% through the third wave of a longitudinal study of adolescent girls* when COVID hit. I knew we needed to have a record of the pandemic's impact on our participants. 11/

https://twitter.com/jennDSN/status/1224573259355545603?s=20

I assumed others were in the same position. Our lab built up an OSF repository of measures, IRB materials, etc. to facilitate adolescent researchers who wanted to ask similar questions. Also joined by @CD_Ladouceur in this effort. 12/

https://twitter.com/jennDSN/status/1243771271621435392?s=20

We saw an opportunity to ensure researchers were not exclusively focusing on the bad, but considering sources of #adolescent resilience, including social connection and prosocial behavior, by including such measures in the repo. @develadolescent 13/

The repo linked to a registry for researchers who planned to study adolescents' response to COVID. Nearly 60 groups signed up. I hoped we would be able to harmonize our data and ask bigger questions than any group alone. Now we're on our way! 14/

Contact us if you want to join the next round of inquiries! Lots to be asked including about pubertal/neural development, social support/connection, and more. Thanks to @Hopelab for support as we work to stand up this Consortium on Adolescent Pandemic Experiences (CAPE). /end