In lupus, review glucocorticoid exposure over the last one to three years, not just the current dose

Design

Nationwide register cohort with weighted cumulative exposure models

Population

3,749 patients with newly diagnosed lupus in Sweden, 2006 to 2022

Primary outcome

First organ damage by oral glucocorticoid dosing window

Effect

Past 2 years: adjusted HR 1.42 (95% CI 1.31 to 1.53) per 5 mg/day; past 1 year: 1.34 (1.23 to 1.42)

A Swedish nationwide register study followed 3,749 patients with newly diagnosed lupus (82.1% female, median age 48) from 2006 to 2022, using pharmacy dispensing to model which time window of oral glucocorticoid dosing best explained organ damage.

For first organ damage overall, the last two years of dosing captured the risk well (adjusted HR 1.42, 95% CI 1.31 to 1.53 for 5 mg per day), and the last year fairly well (1.34, 1.23 to 1.42). For cardiovascular damage, dosing over the past three years contributed substantially.

This is an observational analysis, and patients on higher doses may have more active disease, so confounding by indication remains possible. It does support looking at the dose trajectory over at least a year, and three for cardiovascular risk, when judging steroid burden. Assess steroid exposure alongside disease activity, and avoid the idea that damage reflects only the present dose.

• Review steroid dose over at least the previous year when assessing organ damage risk.
• Look back three years when assessing cardiovascular risk.
• Each 5 mg per day was associated with about a 40% higher hazard of first organ damage.
• Disease activity can confound the link; assess both together.

Why it matters

It shifts attention from the current dose to the recent trajectory, which is what the damage tracked.

Don't overread it

Observational register data open to confounding by disease severity; no dose threshold is proven safe.

The statistics, in plain English

An adjusted hazard ratio of 1.42 per 5 mg/day means a 42% higher rate of first organ damage, with weights placed on dose in recent years. It shows association over time windows, and cannot separate drug harm from sicker patients needing more drug.

Read the rest in the app

You have read your two free briefings this month. The app carries all 27 specialties, every morning, free — and this finding is waiting in it.

Get Daily Doctor — free

Scan to keep reading on your phone. No account needed to start.