COPD ICS Dosing Linked to No Added CV Risk

Key Takeaways

• In a very large Danish cohort with specialist-verified COPD and no prior cardiovascular event history, 8,948 patients experienced a cardiovascular event during 1 year of follow-up, an event burden of 10.9%.
• Low- and moderate-dose ICS were associated with significantly lower cardiovascular event rates than no ICS use in the primary adjusted analysis.
• No ICS dose category was associated with increased cardiovascular event risk in the primary analysis.
• The lower-risk signal was most consistent for moderate-dose ICS, while the exploratory 3-year analysis was neutral across dose groups.

Cardiovascular disease is a major comorbidity in chronic obstructive pulmonary disease (COPD), and it has remained unclear whether inhaled corticosteroid (ICS) exposure changes cardiovascular event risk when dose is considered rather than treatment status alone. In Danish routine secondary-care practice, that question is complicated by the fact that ICS use often tracks with more severe respiratory disease and greater comorbidity. A nationwide cohort in specialist-treated COPD examined dose-stratified ICS exposure and subsequent cardiovascular outcomes in routine care.

In the nationwide retrospective register-based Danish COPD cohort study of inhaled corticosteroid dose and cardiovascular events, investigators included 82,327 patients aged 40 years or older with specialist-verified chronic obstructive pulmonary disease between 2008 and 2022 after excluding those with prior cardiovascular event (CVE) history. Inhaled corticosteroids (ICS) were defined from redeemed prescriptions in the year before the index date and converted to mean daily budesonide-equivalent dose as no ICS, low-dose more than 0 to less than 400 µg/day, moderate-dose 400 to 800 µg/day, or high-dose more than 800 µg/day. The primary outcome was first cardiovascular event, including CVE-related death; nonfatal CVEs comprised ischemic heart disease, arrhythmias, heart failure or cardiomyopathy, cerebrovascular disease or transient ischemic attack, and coronary revascularization procedures, while CVE-related death was limited to death attributed to ischemic heart disease or cerebrovascular disease. Follow-up was prespecified at 1 year, with adjusted cause-specific Cox models, death treated as a competing risk, multiple imputation, and sensitivity analyses addressing demographic, smoking, disease-severity, comorbidity, and treatment factors.

During 1 year of follow-up, 8,948 patients (10.9%) experienced a CVE. In the adjusted primary analysis, low-dose ICS was associated with a lower CVE rate than no ICS use, with a hazard ratio of 0.93 (95% CI 0.87 to 0.99; p=0.022), and moderate-dose ICS showed a similar association at HR 0.91 (95% CI 0.85 to 0.97; p=0.005). No dose category, including high-dose ICS, was associated with increased cardiovascular risk in the primary analysis.

Cumulative incidence was highest among patients not receiving ICS, while ICS-treated groups had lower and largely overlapping risk curves. The moderate-dose signal remained the most consistent directionally across sensitivity analyses, whereas the low-dose association was less stable across model choices: low-dose lost significance in the competing-risk analysis and shifted direction in the time-updated analysis; after inverse probability of treatment weighting, no ICS dose remained statistically significant, and exploratory follow-up over 3 years was also neutral across dose groups. Subgroup analyses suggested effect modification by age and COPD severity, but not by blood eosinophil count strata, so any apparent benefit was neither uniform across models nor durable over longer follow-up.

Because the study was observational, residual confounding and confounding by indication may have persisted despite extensive adjustment and weighting, especially because patients receiving higher ICS doses entered the cohort with more severe disease and greater comorbidity. Exposure was inferred from redeemed prescriptions rather than confirmed use, leaving adherence, inhaler technique, and changes in clinical status over time outside direct measurement. The authors interpreted the conflicting low-dose pattern as more consistent with confounding by indication than with a harmful low-dose ICS effect, and they noted possible overadjustment if exacerbations lay on the causal pathway.

In this Danish secondary-care COPD population without previous CVEs, ICS use at any dose was not associated with increased 1-year cardiovascular event risk. Any reduction signal was modest and appeared most reproducible for moderate-dose exposure. Neutral findings over 3 years and attenuation in alternative models limit claims of a durable protective association.

Clinician Questions

Do these findings apply to COPD patients with prior cardiovascular disease?

No. The cohort included Danish patients aged 40 years or older with specialist-verified COPD, but it excluded anyone with a prior cardiovascular event before study entry, so the reported associations apply to a secondary-care COPD population without previous CVE history.

How were cardiovascular events defined in this COPD ICS cohort?

The primary outcome was first cardiovascular event including cardiovascular event-related death, defined from ischemic heart disease, cardiac arrhythmias, heart failure or cardiomyopathy, cerebrovascular disease or transient ischemic attack, and coronary revascularization procedures.

Which COPD subgroups showed the clearest moderate-dose ICS signal for cardiovascular events?

The researchers reported effect modification by age and disease severity: the moderate-dose association was seen in patients younger than 65 years and in those with more severe airflow limitation, while no significant associations were reported across blood eosinophil count strata.