NSAIDs Lower ESR and Should Be Considered When Interpreting Results
NSAIDs significantly reduce ESR values by approximately 20-25%, which can lead to false-negative or falsely reassuring results in patients being evaluated for inflammatory conditions. While there are no formal guidelines addressing whether to discontinue NSAIDs before ESR testing, the clinical evidence demonstrates a clear suppressive effect that clinicians must account for when interpreting results.
The Evidence for NSAID Effects on ESR
The most relevant study examined patients with giant cell arteritis (GCA) and found that NSAID users had a median ESR of 75 mm/hour compared to 98 mm/hour in non-users (p=0.004) 1. This represents a clinically meaningful 23% reduction. Importantly, NSAIDs did not significantly affect CRP levels in the same patients 1.
In patients with ankylosing spondylitis, NSAID treatment resulted in marked reductions in both ESR and CRP levels as part of their therapeutic anti-inflammatory effect 2. This confirms that NSAIDs actively suppress inflammatory markers through their mechanism of action.
Clinical Implications and Recommendations
When evaluating patients for inflammatory conditions (particularly GCA, polymyalgia rheumatica, or other vasculitides), consider ordering CRP instead of or in addition to ESR if the patient is taking NSAIDs 1. The evidence shows:
- ESR is suppressed by NSAIDs (statistically significant reduction)
- CRP is NOT significantly affected by NSAIDs (p=0.522 in the GCA study) 1
- This makes CRP a more reliable marker in NSAID users
Practical Algorithm for ESR Testing in NSAID Users
If suspecting inflammatory disease in an NSAID user:
- Order CRP as the primary inflammatory marker - it remains unaffected by NSAID use
- If ESR is ordered and comes back normal or mildly elevated, interpret with caution knowing it may be falsely lowered by 20-25%
- Consider the clinical context: a "normal" ESR of 40 mm/hour in an NSAID user might actually represent a true ESR of 50-60 mm/hour
- Do NOT routinely discontinue NSAIDs before testing - there is no guideline support for this, and the delay in diagnosis may be more harmful than the interpretive challenge
Important Caveats
The suppressive effect of NSAIDs on ESR appears to be independent of their effect on the underlying disease. Even in patients where NSAIDs are treating inflammation (like ankylosing spondylitis), the ESR reduction occurs 2. This means you cannot assume an NSAID user with a normal ESR doesn't have inflammatory disease.
Statins also lower ESR (median 57.5 vs 85 mm/hour, p=0.005) 1, so patients on both medications may have even more significant suppression of ESR values.
The mechanism appears related to changes in serum protein fractions and inflammatory mediators rather than a direct effect on red blood cell aggregation 3. This explains why CRP remains a more reliable marker in these patients.