Steroid tablets: linked to more anxiety, no clear link to suicidal behaviour.

Cohort study. 1,105,964 people. 2006-2020; mean follow-up about 12 years; target-trial cohorts followed 365 days.

People were watched for years. Nobody was told what to do. A clue, not proof.

Who: everyone in Sweden aged 15-54 who collected at least one oral glucocorticoid prescription 2006-2020 (58.1% female); Addison's disease excluded. Outcomes = unplanned SPECIALIST (hospital/outpatient specialist) psychiatric contacts or suicide death, not primary care

• Anxiety care visits+25%on vs off tablets
• Anxiety, first 2 weeks+37%highest early on
• Suicidal behaviournoneno clear link

More to know, from our read of the paper:

• Observational: linked, not proven. Within-person design removes stable confounders (genes, upbringing) but NOT time-varying confounding by indication: people start steroids when ill (e.g. cancer relapse, autoimmune flares), which itself is distressing. No cancer diagnosis data
• Outcomes are only severe, unplanned specialist-care contacts; milder problems treated in primary care were not captured
• Dispensed, not necessarily taken; treatment periods built with a 120-day gap rule that relies on future prescriptions
• In autoimmune-diagnosis subgroups associations were attenuated with wide CIs; target-trial ITT showed higher anxiety risk only in the any-autoimmune cohort, per-protocol CIs all included 1, nothing in the GI-autoimmune cohort
• Between-person comparisons gave higher HRs (anxiety 1.70) and a significant suicidal-behaviour association; authors attribute this to between-person confounding. Do not quote between-person numbers as the finding

Swedish health records covered 1,105,964 people aged 15 to 54 who collected steroid tablets (oral glucocorticoids) between 2006 and 2020. Comparing each person's time on versus off the tablets, the risk of unplanned specialist care for anxiety was 25% higher on treatment, and highest in the first two weeks. There was no clear link with suicidal behaviour. People with past mental health problems had higher rates overall. This is observational: linked, not proven. If you take steroids and feel different, talk to your doctor. If you are struggling, please reach out to a doctor or a local crisis line.

We read the abstract and the full text of this paper and checked every number on this page against it (16 facts, checked on 2026-10-29). The study design is named above. Cohort and other observational studies are described as linked, not proven. Found an error? Email theis@onestudyaday.com.