C-reactive protein and erythrocyte sedimentation rate discordance: frequency and causes in adults

C-reactive protein and erythrocyte sedimentation rate discordance: frequency and causes in adults MARK FELDMAN, BILAL AZIZ, GHA NA KANG, MILDRED A. OPONDO, RANDALL K. BELZ, and CONNIE SELLERS DALLAS, TX

C-reactive protein (CRP) levels and the erythrocyte sedimentation rate (ESR) are widely used tests of inflammation that sometimes show opposite results. We performed a retrospective cohort study to clarify the frequency and causes of CRP/ ESR discordance in adults. Between January and December of 2011, the laboratories of Texas Health Presbyterian Hospital performed 2150 paired CRP/ESR measurements in 1753 patients, 1731 of whom were nonpregnant adults aged $ 18 years. Initial CRP and ESR results for each patient were divided into quartiles. CRP/ESR discordance, predefined as results differing by 2 or 3 quartiles, occurred in 212 patients (12%), 105 of whom had high CRP/low ESR discordance (6%) and 107 of whom had high ESR/low CRP discordance (6%). The 212 patients in the CRP/ESR–discordant group (128 women and 84 men) were subdivided into 1 of 6 diagnostic categories, and the causes of discordances were compared. The high CRP/low ESR–discordant group had more patients with infections than the high ESR/low CRP–discordant group (P 5 0.001), particularly infections in the urinary tract (P 5 0.03), gastrointestinal tract (P 5 0.001), lungs (P 5 0.005), and bloodstream (P 5 0.03). However, they had fewer bone and joint infections than the high ESR/low CRP–discordant group (P 5 0.001). Connective tissue diseases, such as systemic lupus erythematosus, were less common in the high CRP/low ESR–discordant group than in the high ESR/low CRP–discordant group (P 5 0.001). Ischemic strokes or transient ischemic attacks almost invariably occurred in the high ESR/low CRP–discordant group (P 5 0.001), whereas myocardial infarction or venous thromboembolism was limited to the high CRP/low ESR–discordant group (P 5 0.001). Our findings provide information to physicians who order these 2 tests together and receive discordant results, which occurs in approximately 1 in 8 patients. (Translational Research 2013;161:37–43) Abbreviations: BMI ¼ body mass index; CRP ¼ C-reactive protein; ESR ¼ erythrocyte sedimentation rate; MI ¼ myocardial infarction; TIA ¼ transient ischemic attack; VTE ¼ venous thromboembolism

he serum C-reactive protein (CRP) level and the erythrocyte sedimentation rate (ESR) are widely used tests that are useful markers of inflammation in many conditions.1-4 However, CRP and ESR may

T

not correlate highly in some conditions,5 and the extent to which CRP and ESR results agree with each other in an unselected patient population is not clear. In clinical practice, the results of the 2 tests are sometimes different.

From the Texas Health Presbyterian Hospital Dallas, Dallas, Texas.

Reprint requests: Mark Feldman, Department of Internal Medicine, Texas Health Presbyterian Hospital Dallas, 8200 Walnut Hill Lane, Dallas, TX; e-mail: [email protected].

None of the authors have a financial conflict to report. They have read the Journal’s policy on conflicts of interest. Submitted for publication April 30, 2012; revision submitted July 5, 2012; accepted for publication July 28, 2012.

1931-5244/$ - see front matter Ó 2013 Mosby, Inc. All rights reserved. http://dx.doi.org/10.1016/j.trsl.2012.07.006

37

38

Translational Research January 2013

Feldman et al

AT A GLANCE COMMENTARY Feldman M, et al. Background

CRP and ESR are widely used tests of inflammation that sometimes give opposite results. The incidence and causes of CRP/ESR discordances have not been carefully characterized. Translational Significance

CRP/ESR discordance occurs in approximately 1 in 8 adult patients. High CRP/ESR discordances are most often associated with infections and MIs or VTE, whereas high ESR/low CRP discordances are most often linked to connective tissue diseases and ischemic cerebrovascular disease. Thus, judicious use of these laboratory tests will assist the clinical diagnosis and management of patients with suspected inflammation. To better understand the frequency and causes of CRP/ ESR discordance (both high CRP/low ESR discordance and high ESR/low CRP discordance), we performed a retrospective study of unselected patients in whom CRP and ESR had been measured simultaneously in the laboratories of Texas Health Presbyterian Hospital in 2011. SUBJECTS AND METHODS

This research conformed to the relevant ethical guidelines for human research and was approved by the institutional review board of Texas Health Resources, Arlington, Texas. Our intention was to include approximately 200 adult (aged $ 18 years), nonpregnant patients with CRP/ESR discordances, which had been predefined as CRP and ESR results differing by 2 or 3 quartiles. The CRP assay (CPT 86140), performed in our core chemistry laboratory, is based on a reaction between CRP in a serum or plasma sample and polyclonal anti-CRP antibodies that have been adsorbed to latex particles (Multigent CRP Vario Reagent, Milan, Italy). Agglutination is detected as an absorbance change at 572 nm (Architect c8000, Abbott Laboratories, Abbott Park, Ill). The rate of change in absorbance is proportional to the quantity of CRP in the sample. The CRP concentration in the sample (milligrams/deciliter) is determined by interpolation from a calibration curve prepared from calibrators of known concentration. This latex immunoassay has a sensitivity of 0.2 mg/dL (19.05 nmol/L). The ESR assay (CPT 85652), performed in our core hematology laboratory, is initiated by placing well-mixed venous blood in a vertical tube. The erythro-

cytes suspended in venous plasma are then allowed to fall toward the bottom of the tube for 30 minutes. Samples are read with an ESR-Auto Plus instrument in QuickMode (Streck, Omaha, Neb) and converted to Westergren units (millimeters/hour) by a mathematic algorithm. Results can range from 0 to .120 mm/h. We estimated a requirement for at least 2000 pairs of CRP/ESR measurements to achieve our goal of 200 patients. Between January 1 and December 16, 2011 (the study period), our laboratories performed 2150 paired CRP/ESR measurements in 1757 inpatients and outpatients; 746 of them were male, and 1011 were female. Their mean (6 standard deviation) age was 58 6 20 years. CRP and ESR results were separated into quartiles using the first paired measurements from each patient (Table I). Adult, nonpregnant patients were then classified into those with CRP/ESR concordance (n 5 1521, or 88%), those with high CRP/low ESR discordance (n 5 105, or 6%), and those with high ESR/low CRP discordance (n 5 107, or 6%). Our study hypothesis was that the causes of high CRP/low ESR discordance and high ESR/low CRP would differ in their frequencies. Electronic medical records of the patients in the CRP/ESR–discordant group were reviewed for the clinical diagnosis, age, gender, body mass index (BMI), smoking history, and current anti-inflammatory medication use (systemic glucocorticoids, aspirin or nonsteroidal anti-inflammatory drugs, other anti-inflammatory drugs including biologic agents, and statins). On the basis of their clinical diagnosis, the groups of CRP/ESR–discordant patients were subdivided into 1 of 6 categories: infection; inflammatory disorder not due to infection; ischemic stroke or transient ischemic attack (TIA); myocardial infarction (MI) or venous thromboembolism (VTE); neoplasm; or another condition. Some patients were in more than 1 category (see ‘‘Results’’). Differences between the high CRP/low ESR– and the high ESR/low CRP–discordant groups were tested for significance with t tests or Fisher exact tests, as appropriate.6,7 Correlations were expressed as Pearson correlation coefficients. P values less than 0.05 were considered statistically significant. RESULTS

CRP and ESR results during the study period correlated significantly (r 5 0.56, P , 0.001) (Fig 1). However, there were many examples of discordant results. The ages of the patients in the high CRP/low ESR– discordant group (n 5 105) and the high ESR/low CRP–discordant group (n 5 107) were similar (Table II). There was a slight male preponderance in the former

Translational Research Volume 161, Number 1

Feldman et al

Table I. C-reactive protein and erythrocyte sedimentation rate quartiles in 1753 patients Quartile

CRP (mg/dL)

ESR (mm/h)

1 2 Median 3 4

,0.2–0.30 0.31–1.60 1.61 1.62–6.02 .6.02

0–17 18–36 37 38–65 .65

Abbreviations: CRP, C-reactive protein; ESR, erythrocyte sedimentation rate. Note: To convert CRP concentration from mg/dL to nmol/L, multiply by 95.24.

39

Table II. Demographic features in the 212 patients with C-reactive protein/erythrocyte sedimentation rate discordances

Feature

High CRP group (n 5 105)

High ESR group (n 5 107)

P value

Age, y Female sex, No. (%) BMI Never smoker, No. (%) Current smoker, No. (%) Former smoker, No. (%) Glucocorticoid use, No. (%) Aspirin or NSAID use, No. (%) Statin use, No. (%) Other anti-inflammatory, No. (%)

58 6 21 50 (48) 27.0 6 6.3 55 (52%) 19 (18%) 24 (23%) 18 (17%) 41 (39%) 30 (29%) 11 (10%)

59 6 17 78 (73) 27.9 6 8.4 69 (64%) 18 (17%) 17 (16%) 15 (14%) 35 (33%) 34 (32%) 12 (11%)

0.70 0.005 0.95 0.09 0.86 0.23 0.57 0.39 0.66 0.99

Abbreviations: BMI, body mass index (calculated as weight in kilograms divided by height in meters squared); CRP, C-reactive protein; ESR, erythrocyte sedimentation rate; NSAID, nonsteroidal anti-inflammatory drug. Note: Plus-minus values are means 6 standard deviation.

Fig 1. Correlation between CRP and ESR, measured simultaneously, in 1753 patients. Although the correlation coefficient of 0.56 was statistically significant (P , 0.001), many patients had discordant results. CRP, C-reactive protein; ESR, erythrocyte sedimentation rate.

group, whereas there was a strong female preponderance in the latter (P 5 0.005). BMI, smoking history, and use of anti-inflammatory medication were similar in the 2 discordant groups. The diagnostic categories in the 2 CRP/ESR–discordant groups differed considerably and are compared in Fig 2. The high CRP/low ESR–discordant group had 62 patients with infections versus 38 patients in the high ESR/low CRP–discordant group (P 5 0.001; Table III). The types of infections encountered also differed. The high CRP–discordant group had significantly more patients with urinary tract, gastrointestinal tract, lungs, and bloodstream infections, but they also had significantly fewer patients with bone and joint infections (Table IV). In most cases, the cause of the infection was confirmed by culture of the organism or by another method (Table V); in the other cases, the infection was diagnosed clinically. The high CRP/low ESR–discordant group had numerically fewer patients with inflammatory conditions than the high ESR–discordant group (Table III). The types of inflammatory conditions encountered also differed. Connective tissue diseases were significantly less com-

Fig 2. Diagnostic categories in 105 patients with high CRP/low ESR discordance (blue) and 107 patients with high ESR/low CRP discordance (gold). *P , 0.05. CRP, C-reactive protein; ESR, erythrocyte sedimentation rate; MI, myocardial infarction; PE, pulmonary embolism; TIA, transient cerebral ischemic attack.

mon in the high CRP/low ESR–discordant group than in the high ESR/low CRP–discordant group (Table VI). Systemic lupus erythematosus was almost exclusively seen in the high ESR–discordant group (Table VII). Ischemic strokes and TIAs were rare in the high CRP/ low ESR–discordant group (1 with stroke) compared with the high ESR–discordant group (9 with stroke and 4 with TIA; P 5 0.001). In contrast, MI and VTE occurred in 10 patients with CRP/ESR discordance, all in the high CRP/low ESR–discordant group (Table III, P 5 0.001). Of the 5 patients with MI, 3 had a non– ST-elevation MI, 1 had an ST-elevation MI, and 1 had an MI associated with a left bundle branch block. Of the 5 patients with VTE, 3 had deep vein thrombophlebitis and 3 had pulmonary embolism (1 had both).

40

Translational Research January 2013

Feldman et al

Table III. Diagnostic categories in the 212 patients with C-reactive protein/erythrocyte sedimentation rate discordances Category

High CRP group (n 5 105)

High ESR group (n 5 107)

Difference (95% CI)

P value

Infection, No. (%) Inflammation, No. (%) Stroke/TIA, No. (%) MI/VTE, No. (%) Neoplasia, No. (%) Other, No. (%)

62 (59%) 23 (22%) 1 (1%) 10 (10%) 4 (4%) 17 (16%)

38 (36%) 33 (31%) 13 (12%) 0 (0%) 8 (7%) 20 (18%)

24% (10%–36%) 29% (220% to 3%) 211% (219% to 25%) 10% (4%–17%) 24% (211% to 3%) 23% (213% to 8%)

0.001 0.16 0.001 0.001 0.37 0.72

Abbreviations: CI, confidence interval; CRP, C-reactive protein; ESR, erythrocyte sedimentation rate; MI, myocardial infarction; TIA, transient cerebral ischemic attack; VTE, venous thromboembolism. Notes: Twelve patients in the high CRP group and 5 patients in the high ESR group fit into 2 categories.

Table IV. Types of infections in the 212 patients with C-reactive protein/erythrocyte sedimentation rate discordances Infection type

High CRP group (n 5 105)

High ESR group (n 5 107)

Difference (95% CI)

P value

Skin/soft tissue, No. (%) Bone/joints, No. (%) Urinary, No. (%) Gastrointestinal, No. (%) Pulmonary, No. (%) Bloodstream, No. (%) Viral illness, No. (%) Brain, No. (%)

18 (17%) 6 (6%) 11 (10%) 13 (12%) 10 (10%) 5 (5%) 2 (2%) 0 (0%)

13 (12%) 19 (18%) 3 (3%) 1 (1%) 1 (1%) 0 (0%) 0 (0%) 1 (1%)

5% (25 to 15) 212% (221 to 24) 8% (1–15) 11% (5–19) 9% (3–16) 5% (0–11) 2% (22 to 7) 1% (23 to 5)

0.34 0.001 0.03 0.001 0.005 0.03 0.24 0.99

Abbreviations: CRP, C-reactive protein; ESR, erythrocyte sedimentation rate. Notes: In the high CRP group, 1 patient with a skin/soft tissue infection also had a gastrointestinal infection, 1 patient with a urinary infection also had a bloodstream infection with a different organism, and 1 patient with a urinary infection also had a gastrointestinal infection.

There were numerically fewer patients with neoplasia in the high CRP/low ESR–discordant group than in the high ESR/low CRP–discordant group. A variety of other conditions were diagnosed, none of which occurred at a significantly different frequency in the high CRP/ low ESR and the high ESR/low CRP–discordant groups. These included patients with pain syndromes (7 vs 11), encephalopathy (2 vs 3), symptomatic anemia (0 vs 4), hypertensive urgency or emergency (2 vs 1), renal failure (1 vs 1), atrial fibrillation or flutter (2 vs 0), and fever of unknown origin or small bowel obstruction (1 vs 0 for each). When women and men were analyzed separately, all 3 statistically significant differences (Fig 2, Table III) between the high CRP/low ESR and the high ESR/low CRP–discordant groups persisted for the women, and the significant difference in the stroke/TIA category persisted for the men. Table VIII compares the frequencies of CRP/ESR discordance by diagnostic category among the 1731 adults. Frequencies ranged from 38% to 6% among the 6 categories. In 33 of the original 1753 patients, CRP and ESR had been reordered (and remeasured) simultaneously within

48 hours of the initial paired measurements. There were minimal, statistically insignificant changes in mean CRP and ESR values, with good correlations between test 1 and test 2 (r 5 0.81 for CRP and r 5 0.96 for ESR; P , 0.001 for each). In these 33 ‘‘repeaters,’’ 29 had CRP/ESR concordance in both sets, 2 had high CRP/low ESR discordance in both sets, and 2 had concordance in set 1 and high CRP/low ESR discordance in set 2. Thus, our initial classification of concordance or discordance remained consistent in 31 of 33 (94%) of these early ‘‘repeaters’’ (95% confidence interval, 79–99). DISCUSSION

CRP and ESR, alone or in combination, are helpful in diagnosing and excluding many diseases caused by infection and inflammation, and following their levels over time can be valuable in assessing the response to treatment.1-4 The major purpose of our study was to provide some clarity as to how often and why CRP and ESR yield discordant results when ordered and measured together. The study demonstrated that CRP and ESR test results are concordant in approximately

Translational Research Volume 161, Number 1

Feldman et al

Table V. Organisms identified in the 100 patients with C-reactive protein/erythrocyte sedimentation rate discordances and infections

Site

Skin, soft tissues MRSA MSSA CNS GBS Pseudomonas aeruginosa Bone MRSA MSSA CNS GAS Corynebacterium (not jeikeium) Polymicrobial infection Joint MSSA CNS GBS Urinary Aerococcus sp. Escherichia coli Klebsiella pneumoniae Mycobacterium tuberculosis Polymicrobial infection Gastrointestinal/perianal/perirectal GAS E. coli Salmonella thompson Clostridium difficile Polymicrobial infection Hepatitis C virus Pulmonary Streptococcus pneumoniae Propionibacterium acnes Atypical mycobacterial infection Histoplasma capsulatum Bloodstream MSSA E. coli P. aeruginosa Histoplasma capsulatum Plasmodium falciparum Central nervous system Toxoplasma gondii Viral illness

High CRP group (n 5 62)

High ESR group (n 5 38)

5 2 0 1 1

0 1 1 1 2

0 1 0 1 0 2

1 1 1 0 1 5

1 0 0

1 1 1

1 2 2 0 2

0 2 0 1 0

1 1 1 3 1 1

0 0 0 1 0 0

1 1 2 0

0 0 0 1

1 1 1 1 1

0 0 0 0 0

0 2

1 0

Abbreviations: CNS, coagulase-negative staphylococcus; CRP, C-reactive protein; ESR, erythrocyte sedimentation rate; GAS, group A streptococcus; GBS, group B streptococcus; MRSA, methicillin-resistant Staphylococcus aureus; MSSA, methicillinsensitive Staphylococcus aureus. Note: Infection was diagnosed clinically in some cases (see ‘‘Results’’).

7 of 8 adult patients and are discordant in approximately 1 in 8 (95% confidence interval, 1 in 7 to 1 in 9). The frequency of discordance ranged from 38% in patients

41

with MI or VTE to 6% in patients with other noninfectious, noninflammatory conditions. Moreover, CRP/ESR concordance or discordance was consistent in more than 90% of patients in whom the 2 tests were rerun together within 48 hours. High ESR/low CRP discordance was seen predominately in women, probably because of their predilection to connective tissue diseases, such as systemic lupus erythematosus, in which despite inflammation, the serum CRP is often suppressed in the absence of infection.8,9 Although CRP levels increase with adiposity,10 the high CRP/low ESR group had BMIs that were similar to those in the high ESR/low CRP group. Compared with patients with high ESR/low CRP discordance, patients with high CRP/low ESR discordance had significantly more infections, particularly those involving the urinary tract, gastrointestinal tract, lungs, and bloodstream. Despite the lower overall prevalence of infection in patients with high ESR/low CRP discordance, infection (when present) was most commonly in the bones or joints. Why bone and joint infections, unlike most other infections, were so strongly associated with high ESR/low CRP discordance is unclear, but the often insidious onset of these orthopedic infections and their chronicity may contribute. In addition to infection and inflammation, CRP and ESR may also increase in response to tissue injury, including burn injury3 and myocardial, pulmonary, or cerebral infarctions,11-14 although the CRP and ESR responses to these injuries may not always be parallel. In our CRP/ ESR–discordant cohort, TIA and stroke were seen almost invariably in the high ESR/low CRP–discordant group, whereas MI and VTE were seen only in the high CRP/low ESR–discordant group. Why patients with TIA without cerebral infarction would have a high ESR with a low or normal CRP is uncertain. Some of them may have had a small cerebral infarction that could not be detected by brain imaging. Another intriguing possibility is that some of these patients may have had some form of chronic inflammation that predisposed them to atherosclerosis and TIAs, as was observed in the large Reykjavik study.15 One can only speculate as to the mechanism(s) for the strong trend toward high CRP/low ESR discordance in the discordant subset of patients with infection or with MI/VTE and the strong trend for low CRP/high ESR discordance in the discordant subset of patients with connective tissue diseases and TIA/ stroke. CRP concentrations fluctuate rapidly, whereas serum fibrinogen concentrations that affect the ESR change relatively slowly.16 This disparity may partially explain why in various types of acute infection there is a disproportionate elevation of CRP in many patients,

42

Translational Research January 2013

Feldman et al

Table VI. Types of inflammation in the 212 patients with CRP/ESR discordances Inflammation type

High CRP group (n 5 105)

High ESR group (n 5 107)

Difference (95% CI)

P value

Connective tissue disease, No. (%) Cholecystitis/hepatitis, No. (%) Colitis/enteritis, No. (%) Dermatitis, No. (%) Pericarditis/pleuritis, No. (%) Acute gouty arthritis, No. (%) Other, No. (%)

9 (9%) 1 (1%) 3 (3%) 3 (3%) 3 (3%) 2 (2%) 4 (5%)

24 (22%) 2 (2%) 0 (0%) 0 (0%) 0 (0%) 1 (1%) 6 (6%)

214% (24 to 224) 21% (26 to 4) 3% (21 to 8) 3% (21 to 8) 3% (21 to 8) 1% (23 to 6) 21% (28 to 6)

0.001 0.99 0.12 0.12 0.12 0.62 0.99

Abbreviations: CRP, C-reactive protein; ESR, erythrocyte sedimentation rate. Note: One patient in the high CRP group had rheumatoid arthritis and acute gouty arthritis, and 1 patient had rheumatoid arthritis and a ruptured ovarian cyst (‘‘Other’’).

Table VII. Types of connective tissue diseases in the 212 patients with C-reactive protein/erythrocyte sedimentation rate discordances

Type

Ankylosing spondylitis Dermatomyositis Polymyalgia rheumatica/giant cell arteritis Progressive systemic sclerosis Rheumatoid arthritis €gren’s syndrome Sjo Systemic lupus erythematosus

High CRP group (n 5 105)

High ESR group (n 5 107)

Table VIII. Frequencies of C-reactive protein/ erythrocyte sedimentation rate discordances by diagnostic categories in 1731 adult, nonpregnant patients

Category

CRP/ESR Concordant (n 5 1519)

CRP/ESR Discordant (n 5 212)

Percent discordant (95% CI)

Infection, No. Inflammation, No. Stroke/TIA, No. MI/VTE, No. Neoplasia, No. Other, No.

619 299 34 16 43 544

100 56 14 10 12 37

14% (12–17) 16% (12–20) 29% (18–43) 38% (22–58) 22% (13–35) 6% (5–9)

P value

0 0 0

1 1 4

0.99 0.99 0.12

0 8 0 1

1 6 1 13

0.99 0.59 0.99 0.001

Abbreviations: CRP, C-reactive protein; ESR, erythrocyte sedimentation rate. Note: One patient with rheumatoid arthritis also had progressive € gren’s syndrome, systemic sclerosis, 1 patient with lupus also had Sjo and 1 patient with lupus also had ankylosing spondylitis.

whereas in more chronic inflammation associated with connective tissue disease and bone/joint infection, the ESR may remain elevated long after the CRP has returned toward normal. CONCLUSIONS

This study does not address whether a clinician facing a clinical question about the presence or absence of systemic or local inflammation should order a CRP alone, an ESR alone, both tests simultaneously, or the tests in sequence (CRP, followed by ESR if the CRP is normal; ESR, followed by CRP if the ESR is normal). Such a cost-effectiveness study has yet to be performed. In clinical practice, the 2 tests are often ordered together. We hope the findings reported in this article will provide useful information to physicians who order these 2 tests together, because discordant results will be encountered frequently in a wide variety of clinical conditions.

Abbreviations: CI, confidence interval; CRP, C-reactive protein; ESR, erythrocyte sedimentation rate; MI, myocardial infarction; TIA, transient cerebral ischemic attack; VTE, venous thromboembolism. Note: A total of 46 patients in the concordant group and 17 patients in the discordant group fit into 2 categories.

The authors thank Drs. Edward Goodman and Beverly Dickson for advice and guidance. REFERENCES

1. Osei-Bimpong A, Meek JH, Lewis SM. ESR or CRP? A comparison of their clinical utility. Hematology 2007;12:353–7. 2. Crowson CS, Rahman MU, Matteson EL. Which measure of inflammation to use? A comparison of erythrocyte sedimentation rate and C-reactive protein measurements from randomized clinical trials of golimumab in rheumatoid arthritis. J Rheumatol 2009;36:1606–10. 3. Barati M, Alinejad F, Bahar MA, et al. Comparison of WBC, ESR, CRP and PCT serum levels in septic and non-septic burn cases. Burns 2008;34:770–4. 4. Buess T, Ludwig C. Diagnostic value of C-reactive protein in comparison with erythrocyte sedimentation as routine admission diagnostic test. Schweiz Med Wochenschr 1995;125:120–4. 5. Keenan RT, Swearingen CJ, Yazici Y. Erythrocyte sedimentation rate and C-reactive protein are poorly correlated with clinical measures of disease activity in rheumatoid arthritis, systemic lupus erythematosus and osteoarthritis patients. Clin Exp Rheumatol 2008;26:814–9. 6. Quick Calcs. Online calculators for scientists. Available at: http:// www.graphpad.com. Accessed April 29, 2012.

Translational Research Volume 161, Number 1

7. Student’s T-test. Available at: http://studentsttest.com. Accessed April 29, 2012. 8. Cengic M, Heljic B, Rasic S, Dilic M. Role of C-reactive protein in systemic lupus erythematosus. Med Arh 2002;56:147–9. 9. Suh CH, Chun HY, Ye YM, Park HS. Unresponsiveness of C-reactive protein in the non-infectious inflammation of systemic lupus erythematous is associated with interleukin 6. Clin Immunol 2006;119:291–9. 10. Visser M, Bouter LM, McQuillan GM, Wener MH, Harris TB. Elevated C-reactive protein levels in overweight and obese adults. JAMA 1999;282:2131–5. 11. Dedobbeleer C, Melot C, Renard M. C-reactive protein increase in acute myocardial infarction. Acta Cardiol 2004;59:291–6. 12. Aujesky D, Hayoz D, Yersin B, et al. Exclusion of pulmonary embolism using C-reactive protein and D-dimer. A prospective comparison. Thromb Haemost 2003;90:1198–203.

Feldman et al

43

13. Scherer R, Mararescu A, Ruhenstroth-Bauer G. The specific effect of plasma proteins in erythrocyte sedimentation. An analysis of the correlation coefficients between ESR and the concentration of twenty individual plasma proteins in healthy and sick persons, with special reference to myocardial infarction. Kloin Wochenschr 1978;53:265–73. 14. Zaremba J, Skrobanski P, Losy J. Acute ischaemic stroke increases the erythrocyte sedimentation rate, which correlates with early brain damage. Folia Morphol (Warsz) 2004;63: 373–6. 15. Andresdottir MB, Sigfusson N, Sigvaldason H, Gudnason V. Erythrocyte sedimentation rate, an independent predictor of coronary heart disease in men and women. The Reykjavik study. Am J Epidemiol 2003;158:844–51. 16. Kusher I. Acute phase reactants. UpToDate 2011. Available at: http://www.uptodate.com. Accessed April 28, 2012.