Udied CRP stability annually over five years in 8901 placebo-treated individuals within the JUPITER trial. Using box plots and correlation coefficients, the authors concluded that CRP in these individuals with high-risk initial values exhibits `strong tracking’ over the long term. However, because serial box plots track a group, the considerable fluctuation in serial measurements in the same individual could be obscured, if not cancelled out, when medians of a large group are examined. It may also be questioned whether the application of correlation coefficients on log-transformed data in this and the 2 preceding studies is the best means to analyze intra-individual stability. Logtransformation (that was applied to CRP but not to cholesterol) considerably attenuates the 23977191 variance of the data. As well, correlations, especially non-parametric ones that mask outlying values, do not inform about the magnitude of the variability, but about how related measurements are, and hence are not a good means of understanding how CRP varies with time. These latter studies may thus considerably underestimate the variability of CRP over time.ConclusionOur study suggests that the use of CRP to assign an atherosclerotic disease risk 
status to individual subjects may be problematic. It cannot be assumed that a single value or even a pair of values will reliably define an individual’s stable or necessarily unchanging inflammation risk status. This does not detract from the importance of inflammation in the pathogenesis of atherosclerotic vascular disease or from its well-established epidemiological associations despite persistent controversy over its added value for risk stratification. In SRIF-14 site contrast to studies that have estimated the ability of CRP to predict future events averaged across tens of thousands of subjects, we have reported the individual level variation in day-to-day absolute CRP measurements, and subsequently the potential effect that this variability may have on predicting individual level future events. Our findings question the use of isolated CRP values to assign definitive 
risk status and to make long-term management decisions in individual patients in routine clinical practice.Supporting InformationAppendix S1.(PDF)AcknowledgmentsWe gratefully acknowledge the support of Serge Simard for statistical assistance, Remy Theriault for creation of the database, and Fernand ??Bertrand for supervising blood sample measurements. Finally, we are veryCRP Variabilitygrateful to the 100 subjects who volunteered for this study and who, over a year for each, came to our research center on 16 occasions, donating generously their time and offering their blood samples to make this work possible.Author ContributionsAcquisition of data: LB AL. Critical revision of the manuscript for important intellectual content: P. Bogaty GRD LJ P. Belisle LB AL JMB. ?Statistical analysis: P. Belisle LJ JMB. Administrative and technical ?support: LB AL. Conceived and designed the experiments: P. Bogaty LB JMB GRD. Analyzed the data: P. Belisle LJ P. Bogaty JMB GRD. Wrote ?the paper: P. Bogaty LJ JMB GRD.
In May and July 2011 Germany experienced an Entero Haemolytic Escherichia coli (EHEC) O104 infection outbreak. The Robert Koch Institut (RKI), a Federal Institute within the portfolio of the Federal Ministry of Health, reported 2987 cases of Shigatoxin mediated gastroenteritis [1]. The outbreak was declared to have been terminated on July 26th 2011. Most cases occurred Dimethylenastron inNort.Udied CRP stability annually over five years in 8901 placebo-treated individuals within the JUPITER trial. Using box plots and correlation coefficients, the authors concluded that CRP in these individuals with high-risk initial values exhibits `strong tracking’ over the long term. However, because serial box plots track a group, the considerable fluctuation in serial measurements in the same individual could be obscured, if not cancelled out, when medians of a large group are examined. It may also be questioned whether the application of correlation coefficients on log-transformed data in this and the 2 preceding studies is the best means to analyze intra-individual stability. Logtransformation (that was applied to CRP but not to cholesterol) considerably attenuates the 23977191 variance of the data. As well, correlations, especially non-parametric ones that mask outlying values, do not inform about the magnitude of the variability, but about how related measurements are, and hence are not a good means of understanding how CRP varies with time. These latter studies may thus considerably underestimate the variability of CRP over time.ConclusionOur study suggests that the use of CRP to assign an atherosclerotic disease risk status to individual subjects may be problematic. It cannot be assumed that a single value or even a pair of values will reliably define an individual’s stable or necessarily unchanging inflammation risk status. This does not detract from the importance of inflammation in the pathogenesis of atherosclerotic vascular disease or from its well-established epidemiological associations despite persistent controversy over its added value for risk stratification. In contrast to studies that have estimated the ability of CRP to predict future events averaged across tens of thousands of subjects, we have reported the individual level variation in day-to-day absolute CRP measurements, and subsequently the potential effect that this variability may have on predicting individual level future events. Our findings question the use of isolated CRP values to assign definitive risk status and to make long-term management decisions in individual patients in routine clinical practice.Supporting InformationAppendix S1.(PDF)AcknowledgmentsWe gratefully acknowledge the support of Serge Simard for statistical assistance, Remy Theriault for creation of the database, and Fernand ??Bertrand for supervising blood sample measurements. Finally, we are veryCRP Variabilitygrateful to the 100 subjects who volunteered for this study and who, over a year for each, came to our research center on 16 occasions, donating generously their time and offering their blood samples to make this work possible.Author ContributionsAcquisition of data: LB AL. Critical revision of the manuscript for important intellectual content: P. Bogaty GRD LJ P. Belisle LB AL JMB. ?Statistical analysis: P. Belisle LJ JMB. Administrative and technical ?support: LB AL. Conceived and designed the experiments: P. Bogaty LB JMB GRD. Analyzed the data: P. Belisle LJ P. Bogaty JMB GRD. Wrote ?the paper: P. Bogaty LJ JMB GRD.
In May and July 2011 Germany experienced an Entero Haemolytic Escherichia coli (EHEC) O104 infection outbreak. The Robert Koch Institut (RKI), a Federal Institute within the portfolio of the Federal Ministry of Health, reported 2987 cases of Shigatoxin mediated gastroenteritis [1]. The outbreak was declared to have been terminated on July 26th 2011. Most cases occurred inNort.
