Background/Purpose

Inflammation is hypothesized to have direct effects on arterial endothelial and vasomotor function, functions which regulate blood pressure (BP).  While inflammation has been implicated in the development of elevated BP, few studies have examined BP longitudinally after diagnosis of an inflammatory disease.  The objective of this study is to examine the longitudinal association between changes in levels of inflammation and changes in BP in patients with rheumatoid arthritis (RA).

Methods

We studied RA subjects classified using a validated electronic medical record (EMR) algorithm (positive predictive value 94%) linked with Medicare Claims Data (patients age>65 years) between 2006-2010.  We extracted EMR data on age, gender, systolic BP (SBP), diastolic BP (DBP), erythrocyte sedimentation rates (ESR), smoking status (ever/never), and calendar year of measurements.  We extracted RA treatment data from Medicare prescription data including non-steroidal anti-inflammatory drugs (NSAIDs), anti-hypertensive drugs and dates of claims.  We studied all subjects with ≥2 concurrent BP and ESR measurements at separate visits at least one week apart.  The baseline was defined as the 1^st concurrent ESR and BP measurement 6 months after the 1^st Medicare claims date.  We examined the association between the change in ESR and change in SBP between baseline and follow-up using multiple linear regression models adjusted by age, gender and smoking status.  We performed sensitivity analyses by including potential treatments received +/- 15 days of ESR measurements that may affect blood pressure individually into the models; the treatments included anti-TNF, NSAIDs, anti-hypertensive medications, and steroids.

Results

We identified 313 subjects with ≥2 instances with ESR and BP measured on the same date.  The mean age was 69.2 years (SD 10.3), 83.7% female, and 57.6% ACPA positive.  The mean SBP was 134.4(SD 18.8) mm/Hg, mean DBP was 75.7(SD 12) mm/Hg, and the mean ESR was 32 (SD 27.1) mm/hour.  We observed an inverse association between change in ESR and SBP, where every 10mm/hour increase in ESR was associated with a 1.2mm/Hg lower in SBP, adjusted by age, gender and smoking status (Table).  A similar relationship was found with DBP: each 10mm/hour increase in ESR was associated with a 0.43 mm/Hg lower DBP (95% CI -0.75, -0.01, p=0.009).  The effect size of the association between change in ESR and change in BP was similar with the addition of indicator variables for anti-hypertensive treatments, anti-TNF, and NSAIDs into the model.   

Conclusion

In a linked dataset containing clinical data from the EMR and detailed prescription data from Medicare, we observed that increases in ESR were associated with a modest reduction in blood pressure.  These findings have potential implications for CV risk assessment in RA patients who commonly experience large fluctuations in inflammation.