{"slug":"hs-crp","name":"hs-CRP","aliases":["High-sensitivity C-reactive protein"],"kind":"biomarker","system":"gut-health","unit":"mg/L","unitNote":"Some laboratories report mg/dL; multiply by 10 for mg/L.","answer":"Across 54 cohorts of 160,309 adults, the risk of heart disease, stroke and death rose continuously with hs-CRP, and the banding most laboratories print (under 1 mg/L low, 1 to 3 average, above 3 high) comes from a 2003 AHA/CDC statement; genetic studies show CRP itself does not cause heart disease, so the number is a gauge of inflammation rather than the problem, and a single reading above 10 usually reflects an acute illness and should be repeated once well.","summary":"hs-CRP is the liver's response to inflammation anywhere in the body, measured by an assay sensitive enough to read the low values that track cardiovascular risk. Laboratories band it at 1 and 3 mg/L. The cohorts show risk rising with the number; the trials show that treating inflammation prevents events; the genetics show CRP is the messenger, not the cause. Body fat, sleep, gums, smoking and hormones all move it, and one reading is rarely enough.","whatItIs":[{"text":"C-reactive protein is made only by liver cells, mainly under the control of the cytokine interleukin-6. After a single inflammatory stimulus its blood level rises above 5 mg/L by about 6 hours, peaks at about 48 hours and can climb 10,000-fold, from under 0.05 mg/L to over 500; its half-life is about 19 hours whatever the state of health, so the level is set almost entirely by how much is being made.","cites":["pepys2003"],"grade":"C"},{"text":"The 'high-sensitivity' in the name describes the assay, not a different protein. Ordinary CRP tests were built to track infections and lack the sensitivity to sort cardiovascular risk; the high-sensitivity method resolves the span below about 3 mg/L where that risk sits. In healthy young blood donors the median is 0.8 mg/L, the 90th percentile 3.0 and the 99th percentile 10; across the 54 cohorts of the Emerging Risk Factors Collaboration, with a mean age of 60, the median was 1.72 (5th percentile 0.25, 95th 12.4).","cites":["ledue2003","pepys2003","erfc2010"],"grade":"B"},{"text":"The marker earned its place in cardiovascular medicine because it finds risk that cholesterol misses. Among 27,939 apparently healthy US women followed for a mean of eight years, the highest fifth of CRP carried 2.3 times the cardiovascular risk of the lowest fifth, against 1.5 for the highest fifth of LDL cholesterol, and the two were barely correlated (r = 0.08); 77 percent of events occurred in women with LDL cholesterol below 160 mg/dL. The 2018 American cholesterol guideline lists an hs-CRP of 2.0 mg/L or higher, if measured, as a risk-enhancing factor.","cites":["ridker2002","grundy2019"],"grade":"A"}],"labRange":{"typical":"Under 1.0 mg/L low, 1.0 to 3.0 average, above 3.0 high cardiovascular risk; most US laboratories print these three bands","claims":[{"text":"The three bands come from the 2003 scientific statement of the American Heart Association and the Centers for Disease Control and Prevention, which set a uniform threshold above 3 mg/L for high relative risk and under 1 for low. They are thirds of the population distribution, written for adults at intermediate heart risk, not a definition of health; in the Copenhagen cohort that used them, people above 3 had 1.6 times the ischemic heart disease risk of people below 1.","cites":["pearson2003","khera2005","zacho2008"],"grade":"A"},{"text":"The bands are not adjusted for sex or ancestry, and the distributions differ. In the Dallas Heart Study, 2,749 Black and White adults aged 30 to 65, the median was 3.3 mg/L in women against 1.8 in men and 3.0 in Black against 2.3 in White participants; the share above 3 was 31 percent of White men, 40 percent of Black men, 51 percent of White women and 58 percent of Black women. After adjusting for risk factors, estrogen and statin use and body mass index, women still sat above 3 about 1.6 to 1.7 times as often as White men.","cites":["khera2005"],"grade":"C"},{"text":"A value under 1 mg/L on a first draw is usually enough; a higher value should be repeated. In healthy people the within-person variation of CRP is about 42 percent, and a change has to be about 118 percent to be beyond measurement and biological noise. A level that stays above 10 mg/L signals a significant acute-phase response and calls for a history and physical examination to find the cause, not a cardiovascular interpretation.","cites":["pepys2003","macy1997"],"grade":"C"}]},"evidence":{"headline":"Risk rose continuously with hs-CRP in 54 cohorts; above 3 mg/L carried 1.6 times the heart disease risk of under 1; treating inflammation cut events in two large trials; genetic studies show CRP itself is a marker, not a cause.","scale":{"min":0,"max":10,"ticks":[0,1,2,3,5,10]},"bands":[{"label":"Common laboratory low band","low":0,"high":1,"tone":"lab","source":"AHA/CDC 2003"},{"label":"Common laboratory average band","low":1,"high":3,"tone":"lab","source":"AHA/CDC 2003"},{"label":"Common laboratory high band","low":3,"high":10,"tone":"lab","source":"AHA/CDC 2003"},{"label":"Reference group with the lowest risk in cohorts","low":0,"high":1,"tone":"evidence","source":"Zacho 2008; Pepys 2003"},{"label":"Risk-enhancing factor and trial entry threshold at 2.0 and above","low":2,"high":10,"tone":"neutral","source":"AHA/ACC 2018; JUPITER; CANTOS"},{"label":"1.6 times the ischemic heart disease risk of under 1","low":3,"high":10,"tone":"caution","source":"Zacho 2008"},{"label":"Likely acute inflammation; find the cause and repeat","low":10,"tone":"caution","source":"Pepys 2003"}],"claims":[{"text":"The Emerging Risk Factors Collaboration pooled individual records of 160,309 people without vascular disease from 54 prospective studies, 1.31 million person-years and 27,769 outcomes. Risk rose nearly log-linearly with CRP. Per standard deviation higher log CRP (a threefold higher concentration), coronary heart disease risk was 1.63 times higher adjusted for age and sex and 1.37 after conventional risk factors; ischemic stroke 1.44 and 1.27; vascular death 1.71 and 1.55; and non-vascular death, including cancers and lung disease, 1.55 and 1.54. Adjusting for fibrinogen cut the coronary figure to 1.23. The authors concluded that most of the association with vascular disease depends on conventional risk factors and other inflammatory markers.","cites":["erfc2010"],"grade":"A"},{"text":"In the Copenhagen cohort of 10,276 people, those with CRP above 3 mg/L had 1.6 times the risk of ischemic heart disease and 1.3 times the risk of ischemic stroke of those below 1. Yet combinations of CRP gene variants that raised CRP by up to 64 percent, which should have raised heart disease risk by up to 32 percent if CRP were causal, carried no excess risk at all. Variants in the gene for apolipoprotein E, by contrast, raised both cholesterol and heart disease in the same people.","cites":["zacho2008"],"grade":"B"},{"text":"The largest Mendelian randomization, 194,418 participants and 46,557 coronary cases from 47 studies, found that CRP gene variants shifting the level by up to 30 percent per allele carried a coronary risk ratio of 1.00 (0.90 to 1.13) per standard deviation of genetically raised CRP, against 1.33 observed for the same difference in measured CRP. The authors' conclusion: CRP itself is unlikely to be even a modest cause of coronary heart disease. An umbrella review of 113 meta-analyses and 196 Mendelian randomization analyses later found convincing observational evidence only for cardiovascular mortality and venous thromboembolism, and causal support for no outcome at all.","cites":["ccgc2011","markozannes2021"],"grade":"B"},{"text":"That does not make the number useless; it makes it a gauge. In the JUPITER trial, 17,802 apparently healthy adults with LDL cholesterol below 130 mg/dL and hs-CRP of 2.0 mg/L or higher were randomized to a statin or placebo. The statin cut LDL by 50 percent and hs-CRP by 37 percent, and the combined endpoint of heart attack, stroke, revascularization, unstable angina or cardiovascular death fell from 1.36 to 0.77 per 100 person-years (hazard ratio 0.56), with all-cause death 0.80, over a median of 1.9 years; physician-reported diabetes was more common on the drug. Because the drug lowered both, the trial cannot say which reduction did the work.","cites":["ridker2008"],"grade":"B"},{"text":"CANTOS separated the two. In 10,061 people with a previous heart attack and hs-CRP of 2 mg/L or more, an antibody against interleukin-1β lowered hs-CRP by 26 to 41 percentage points more than placebo without changing lipids, and at the 150 mg dose cut heart attack, stroke or cardiovascular death by 15 percent (hazard ratio 0.85, 0.74 to 0.98) over a median of 3.7 years, at the cost of more fatal infections and with no difference in all-cause death. Inflammation upstream of CRP is a treatable cause; CRP is how the trial chose its patients and watched the treatment work.","cites":["ridker2017"],"grade":"B"},{"text":"The predictive gain over standard risk scores is real but small. In the Reykjavik study of 18,569 people, the top third of CRP carried 1.45 times the coronary risk of the bottom third after adjustment, against 2.35 for high total cholesterol and 1.87 for smoking, and the authors called CRP a relatively moderate predictor. The 2018 systematic review for the US Preventive Services Task Force, 43 studies and 267,244 people, found inconsistent gains from hs-CRP, with the largest analysis of 166,596 people showing a C-statistic change of 0.0039, and no adequately powered trial of whether testing changes outcomes.","cites":["danesh2004","lin2018"],"grade":"A"}]},"levers":[{"name":"Weight loss","effect":"lowers it, roughly in proportion to the weight lost","claim":{"text":"Across 33 weight-loss intervention studies, lifestyle and surgical, CRP fell by a mean of 0.13 mg/L for each kilogram lost (weighted correlation 0.85), with a shallower slope of 0.06 in the lifestyle studies alone. Body fat is the commonest reason a healthy person's CRP sits above the low band: in 16,616 US adults, obesity carried 2.1 times the odds of an elevated CRP in men and 6.2 times in women, and waist-to-hip ratio added risk independent of body mass index.","cites":["selvin2007","visser1999"],"grade":"B"}},{"name":"Exercise training","effect":"lowers it, with or without weight loss","claim":{"text":"Pooling 83 controlled trials with 3,769 participants, exercise training lowered CRP with a mean effect size of 0.26; the effect was larger when body mass index fell (0.38) but still present when weight did not change (0.19), regardless of age or sex.","cites":["fedewa2017"],"grade":"A"}},{"name":"Mediterranean diet","effect":"lowers it modestly","claim":{"text":"In the PREDIMED pilot, 772 adults aged 55 to 80 at high cardiovascular risk were randomized to a low-fat diet or a Mediterranean diet with olive oil or with nuts; at three months the olive-oil group's CRP was 0.54 mg/L lower than the low-fat group's (confidence interval 0.03 to 1.04), alongside lower glucose and blood pressure.","cites":["estruch2006"],"grade":"B"}},{"name":"Sleep","effect":"disturbed and very long sleep track with higher values","claim":{"text":"Across 72 studies of more than 50,000 adults, sleep disturbance was associated with higher CRP (effect size 0.12) and so were shorter sleep (0.09) and very long sleep (0.17); experimental sleep deprivation in a laboratory did not raise CRP, so the cohort signal reflects chronic sleep problems rather than one bad night.","cites":["irwin2016"],"grade":"B"}},{"name":"Stopping smoking","effect":"lowers it over years","claim":{"text":"In 15,489 US adults, CRP and other inflammatory markers rose with the intensity of smoking and fell with time since quitting, more slowly than blood pressure and lipids, returning to never-smoker levels about 5 years after cessation.","cites":["bakhru2005"],"grade":"C"}},{"name":"Treating gum disease","effect":"lowers it by about half a milligram per liter","claim":{"text":"Periodontitis is a common hidden source. In 25 controlled trials of 1,748 patients, periodontal treatment lowered hs-CRP by 0.50 mg/L (0.22 to 0.78), and by 0.71 in patients who also had heart disease or diabetes; a 2026 meta-analysis restricted to 12 randomized trials found 0.58 mg/L (0.33 to 0.83) together with a 4.6 mm Hg fall in systolic blood pressure, with moderate to low certainty.","cites":["teeuw2014","gandhi2026"],"grade":"A"}},{"name":"Medication that lowers inflammation","effect":"lowers it substantially","claim":{"text":"A statin lowered hs-CRP by 37 percent in JUPITER, and an interleukin-1β antibody by 26 to 41 percentage points more than placebo in CANTOS, with fewer cardiovascular events in both; the antibody carried more fatal infections. Whether any drug belongs in the picture is a decision for a physician who knows the person's whole risk, not something a number settles.","cites":["ridker2008","ridker2017"],"grade":"B"}}],"caveats":[{"text":"One reading is a weak foundation. In 113 healthy adults measured five times over a year, only 63 percent of first and second hs-CRP readings landed in the same quartile, though 90 percent were within one quartile, a stability about the same as total cholesterol; over 12 years in the Reykjavik study the within-person correlation was 0.59. Read a first value above 1 mg/L as a prompt to measure again a week or more later, when well, rather than as a verdict.","cites":["ockene2001","danesh2004","pepys2003"],"grade":"C"},{"text":"Any infection, injury or flare of an inflammatory disease can push CRP far above the cardiovascular range within hours, and the level falls back only as fast as the stimulus resolves; a value above 10 mg/L is an acute-phase response until proven otherwise and says nothing about long-term risk until repeated.","cites":["pepys2003"],"grade":"C"},{"text":"Oral contraceptives and menopausal hormone therapy taken by mouth roughly double baseline CRP without any other sign of inflammation. In 30 healthy non-smoking young women, users of a low-dose pill ran 2.0 mg/L against 0.9 in non-users across two menstrual cycles; in 493 healthy postmenopausal women, median CRP on hormone therapy was 2.7 mg/L against 1.4 off it (converted from 0.27 and 0.14 mg/dL), independent of weight, age, blood pressure, lipids and smoking.","cites":["dreon2003","ridker1999"],"grade":"C"},{"text":"CRP is a marker of inflammation, not its cause, so lowering the number by itself has never been shown to help: the genetics say people born with higher CRP are at no higher coronary risk, and no trial has lowered CRP as its target in healthy adults. The guideline role is narrow, a risk-enhancing factor that may tip a borderline statin decision, and the Preventive Services evidence review found no trial showing that testing improves outcomes.","cites":["ccgc2011","grundy2019","lin2018"],"grade":"A"}],"seeAPhysician":["An hs-CRP above 10 mg/L, or one that stays above 10 when repeated, which points to an infection, an injury or an inflammatory disease rather than to cardiovascular risk.","A value above 3 mg/L on two readings a week or more apart with no obvious explanation, particularly alongside high blood pressure, diabetes, smoking, a high ApoB or a family history of early heart disease.","Fever, night sweats, unexplained weight loss, swollen or painful joints, persistent diarrhea, mouth ulcers or a new rash with a raised hs-CRP, which are the questions an autoimmune or infectious workup answers.","An hs-CRP of 2.0 mg/L or higher with a borderline or intermediate 10-year heart risk, where the 2018 cholesterol guideline treats it as a reason to revisit treatment."],"openQuestions":["No trial has treated healthy adults to an hs-CRP target or tested whether lowering CRP itself changes outcomes; CANTOS treated interleukin-1β in people who had already had a heart attack, and JUPITER lowered cholesterol at the same time.","The incremental value of hs-CRP over standard risk scores was minimal in the largest analysis and inconsistent elsewhere, and no trial has shown that testing changes what happens to people.","Women and Black adults run higher CRP distributions than White men, and whether the 1 and 3 mg/L bands should differ by sex or ancestry is unresolved.","CRP tracks death from cancers and lung disease about as strongly as it tracks vascular death, and the authors of the largest pooled analysis call the relevance of CRP to such a range of disorders unclear."],"questions":[{"q":"Is an hs-CRP of 2.5 bad?","a":"It sits in the 1 to 3 mg/L 'average' band of the AHA/CDC banding and above the 2.0 line the 2018 cholesterol guideline uses as a risk-enhancing factor, so it is not low. Risk rose continuously with CRP in the cohorts, but within-person variation is about 42 percent, so one reading does not settle a baseline; a second draw a week or more later, when well, does. Body fat, poor sleep, gum disease, smoking, a recent cold, the pill and oral hormone therapy all push it up."},{"q":"What is an optimal hs-CRP level?","a":"No trial defines one. Under 1 mg/L is the lowest band in the laboratory banding and the reference group in the cohorts, and healthy young blood donors run a median of 0.8. The genetics show that CRP itself does not cause heart disease, so the number is a gauge of inflammation rather than a target; what the evidence supports is finding and treating the source of a persistently raised value, not chasing the number."},{"q":"How do I lower my CRP?","a":"Lose excess weight (about 0.13 mg/L per kilogram in pooled studies), train (exercise lowered CRP even without weight loss across 83 trials), eat a Mediterranean pattern (0.54 mg/L lower at three months in a randomized trial), treat gum disease (about 0.5 mg/L), sleep well and stop smoking (inflammatory markers normalize over about 5 years). Medications that lower inflammation lower it more and belong with a physician who knows the whole picture."}],"citations":[{"id":"pepys2003","authors":"Pepys MB, Hirschfield GM","title":"C-reactive protein: a critical update","journal":"Journal of Clinical Investigation","year":2003,"pmid":"12813013","doi":"10.1172/JCI18921","type":"review","finding":"CRP is made by hepatocytes under IL-6, rises above 5 mg/L by 6 hours and up to 10,000-fold, has a 19-hour half-life; healthy donors median 0.8 mg/L, 90th percentile 3.0, 99th 10; under 1 one draw suffices, higher values need serial samples a week or more apart; persistently above 10 needs a history and examination.","url":"https://pubmed.ncbi.nlm.nih.gov/12813013/"},{"id":"ledue2003","authors":"Ledue TB, Rifai N","title":"Preanalytic and analytic sources of variations in C-reactive protein measurement: implications for cardiovascular disease risk assessment","journal":"Clinical Chemistry","year":2003,"pmid":"12881440","doi":"10.1373/49.8.1258","type":"review","finding":"Traditional CRP tests lack the sensitivity for coronary risk stratification; high-sensitivity assays were developed because results are read against fixed cutpoints.","url":"https://pubmed.ncbi.nlm.nih.gov/12881440/"},{"id":"erfc2010","authors":"Emerging Risk Factors Collaboration; Kaptoge S, Di Angelantonio E, et al.","title":"C-reactive protein concentration and risk of coronary heart disease, stroke, and mortality: an individual participant meta-analysis","journal":"Lancet","year":2010,"pmid":"20031199","doi":"10.1016/S0140-6736(09)61717-7","type":"meta-analysis","finding":"160,309 people in 54 cohorts: per SD higher log CRP, coronary risk 1.63 (age and sex) and 1.37 (fully adjusted), stroke 1.44 and 1.27, vascular death 1.71 and 1.55, non-vascular death 1.55 and 1.54; median CRP 1.72 mg/L.","url":"https://pubmed.ncbi.nlm.nih.gov/20031199/"},{"id":"ridker2002","authors":"Ridker PM, Rifai N, Rose L, Buring JE, Cook NR","title":"Comparison of C-reactive protein and low-density lipoprotein cholesterol levels in the prediction of first cardiovascular events","journal":"New England Journal of Medicine","year":2002,"pmid":"12432042","doi":"10.1056/NEJMoa021993","type":"cohort","finding":"In 27,939 women over eight years, top versus bottom quintile relative risk 2.3 for CRP against 1.5 for LDL cholesterol; 77 percent of events at LDL below 160 mg/dL.","url":"https://pubmed.ncbi.nlm.nih.gov/12432042/"},{"id":"grundy2019","authors":"Grundy SM, Stone NJ, Bailey AL, et al.","title":"2018 AHA/ACC/AACVPR/AAPA/ABC/ACPM/ADA/AGS/APhA/ASPC/NLA/PCNA guideline on the management of blood cholesterol","journal":"Circulation","year":2019,"pmid":"30586774","doi":"10.1161/CIR.0000000000000625","type":"guideline","finding":"hs-CRP of 2.0 mg/L or higher, if measured, is a risk-enhancing factor that may favor statin therapy at borderline 10-year risk of 5 to 7.5 percent.","url":"https://pubmed.ncbi.nlm.nih.gov/30586774/"},{"id":"pearson2003","authors":"Pearson TA, Mensah GA, Alexander RW, et al.; Centers for Disease Control and Prevention; American Heart Association","title":"Markers of inflammation and cardiovascular disease: application to clinical and public health practice: a statement for healthcare professionals from the Centers for Disease Control and Prevention and the American Heart Association","journal":"Circulation","year":2003,"pmid":"12551878","doi":"10.1161/01.cir.0000052939.59093.45","type":"consensus","finding":"The source of the hs-CRP bands laboratories print: under 1 mg/L low, 1 to 3 average, above 3 high relative risk.","url":"https://pubmed.ncbi.nlm.nih.gov/12551878/"},{"id":"khera2005","authors":"Khera A, McGuire DK, Murphy SA, et al.","title":"Race and gender differences in C-reactive protein levels","journal":"Journal of the American College of Cardiology","year":2005,"pmid":"16053959","doi":"10.1016/j.jacc.2005.04.051","type":"cross-sectional","finding":"Dallas Heart Study, 2,749 adults: median CRP 3.3 mg/L in women versus 1.8 in men, 3.0 in Black versus 2.3 in White participants; above 3 in 31, 40, 51 and 58 percent of White men, Black men, White women and Black women.","url":"https://pubmed.ncbi.nlm.nih.gov/16053959/"},{"id":"zacho2008","authors":"Zacho J, Tybjaerg-Hansen A, Jensen JS, et al.","title":"Genetically elevated C-reactive protein and ischemic vascular disease","journal":"New England Journal of Medicine","year":2008,"pmid":"18971492","doi":"10.1056/NEJMoa0707402","type":"mendelian-randomization","finding":"CRP above 3 versus below 1 mg/L carried 1.6 times the ischemic heart disease and 1.3 times the ischemic stroke risk in 10,276 people; CRP genotypes raising the level up to 64 percent carried no excess risk.","url":"https://pubmed.ncbi.nlm.nih.gov/18971492/"},{"id":"ccgc2011","authors":"C Reactive Protein Coronary Heart Disease Genetics Collaboration (CCGC); Wensley F, Gao P, et al.","title":"Association between C reactive protein and coronary heart disease: mendelian randomisation analysis based on individual participant data","journal":"BMJ","year":2011,"pmid":"21325005","doi":"10.1136/bmj.d548","type":"mendelian-randomization","finding":"194,418 participants: coronary risk ratio 1.00 per SD of genetically raised CRP against 1.33 observed; CRP itself unlikely to be even a modest causal factor.","url":"https://pubmed.ncbi.nlm.nih.gov/21325005/"},{"id":"markozannes2021","authors":"Markozannes G, Koutsioumpa C, Cividini S, et al.","title":"Global assessment of C-reactive protein and health-related outcomes: an umbrella review of evidence from observational studies and Mendelian randomization studies","journal":"European Journal of Epidemiology","year":2021,"pmid":"32978716","doi":"10.1007/s10654-020-00681-w","type":"meta-analysis","finding":"Of 113 meta-analyses and 196 Mendelian randomization analyses, only cardiovascular mortality and venous thromboembolism had convincing observational evidence, and no outcome had substantial causal support.","url":"https://pubmed.ncbi.nlm.nih.gov/32978716/"},{"id":"ridker2008","authors":"Ridker PM, Danielson E, Fonseca FA, et al.; JUPITER Study Group","title":"Rosuvastatin to prevent vascular events in men and women with elevated C-reactive protein","journal":"New England Journal of Medicine","year":2008,"pmid":"18997196","doi":"10.1056/NEJMoa0807646","type":"rct","finding":"17,802 healthy adults with LDL under 130 mg/dL and hs-CRP 2.0 mg/L or higher: statin cut LDL 50 percent and hs-CRP 37 percent; primary endpoint hazard ratio 0.56, all-cause death 0.80, over a median 1.9 years.","url":"https://pubmed.ncbi.nlm.nih.gov/18997196/"},{"id":"ridker2017","authors":"Ridker PM, Everett BM, Thuren T, et al.; CANTOS Trial Group","title":"Antiinflammatory therapy with canakinumab for atherosclerotic disease","journal":"New England Journal of Medicine","year":2017,"pmid":"28845751","doi":"10.1056/NEJMoa1707914","type":"rct","finding":"10,061 people after heart attack with hs-CRP 2 mg/L or more: interleukin-1β blockade lowered hs-CRP 26 to 41 percentage points more than placebo without changing lipids and cut events 15 percent at 150 mg (hazard ratio 0.85), with more fatal infections.","url":"https://pubmed.ncbi.nlm.nih.gov/28845751/"},{"id":"danesh2004","authors":"Danesh J, Wheeler JG, Hirschfield GM, et al.","title":"C-reactive protein and other circulating markers of inflammation in the prediction of coronary heart disease","journal":"New England Journal of Medicine","year":2004,"pmid":"15070788","doi":"10.1056/NEJMoa032804","type":"cohort","finding":"Reykjavik study of 18,569 people: top versus bottom third of CRP, odds ratio 1.45 for coronary disease after adjustment; within-person correlation 0.59 over 12 years; a relatively moderate predictor.","url":"https://pubmed.ncbi.nlm.nih.gov/15070788/"},{"id":"lin2018","authors":"Lin JS, Evans CV, Johnson E, et al.","title":"Nontraditional risk factors in cardiovascular disease risk assessment: updated evidence report and systematic review for the US Preventive Services Task Force","journal":"JAMA","year":2018,"pmid":"29998301","doi":"10.1001/jama.2018.4242","type":"meta-analysis","finding":"43 studies, 267,244 people: hs-CRP gains inconsistent, the largest analysis of 166,596 showing a C-statistic change of 0.0039; no adequately powered trial of testing on outcomes.","url":"https://pubmed.ncbi.nlm.nih.gov/29998301/"},{"id":"macy1997","authors":"Macy EM, Hayes TE, Tracy RP","title":"Variability in the measurement of C-reactive protein in healthy subjects: implications for reference intervals and epidemiological applications","journal":"Clinical Chemistry","year":1997,"pmid":"8990222","type":"laboratory","finding":"In 143 healthy donors the median CRP was 0.64 mg/L; within-subject variation 42.2 percent and a critical difference of 118 percent between sequential values.","url":"https://pubmed.ncbi.nlm.nih.gov/8990222/"},{"id":"ockene2001","authors":"Ockene IS, Matthews CE, Rifai N, Ridker PM, Reed G, Stanek E","title":"Variability and classification accuracy of serial high-sensitivity C-reactive protein measurements in healthy adults","journal":"Clinical Chemistry","year":2001,"pmid":"11238295","type":"laboratory","finding":"113 adults measured quarterly: 63 percent of first and second hs-CRP values in the same quartile, 90 percent within one quartile, stability similar to total cholesterol.","url":"https://pubmed.ncbi.nlm.nih.gov/11238295/"},{"id":"selvin2007","authors":"Selvin E, Paynter NP, Erlinger TP","title":"The effect of weight loss on C-reactive protein: a systematic review","journal":"Archives of Internal Medicine","year":2007,"pmid":"17210875","doi":"10.1001/archinte.167.1.31","type":"meta-analysis","finding":"Across 33 weight-loss studies, CRP fell 0.13 mg/L per kilogram lost (weighted r 0.85); slope 0.06 in lifestyle interventions alone.","url":"https://pubmed.ncbi.nlm.nih.gov/17210875/"},{"id":"visser1999","authors":"Visser M, Bouter LM, McQuillan GM, Wener MH, Harris TB","title":"Elevated C-reactive protein levels in overweight and obese adults","journal":"JAMA","year":1999,"pmid":"10591334","doi":"10.1001/jama.282.22.2131","type":"cross-sectional","finding":"NHANES III, 16,616 adults: obesity carried an odds ratio of 2.13 for elevated CRP in men and 6.21 in women; waist-to-hip ratio added risk independent of BMI.","url":"https://pubmed.ncbi.nlm.nih.gov/10591334/"},{"id":"fedewa2017","authors":"Fedewa MV, Hathaway ED, Ward-Ritacco CL","title":"Effect of exercise training on C reactive protein: a systematic review and meta-analysis of randomised and non-randomised controlled trials","journal":"British Journal of Sports Medicine","year":2017,"pmid":"27445361","doi":"10.1136/bjsports-2016-095999","type":"meta-analysis","finding":"83 controlled trials, 3,769 participants: exercise training lowered CRP (effect size 0.26), 0.38 with a fall in BMI and 0.19 without weight loss.","url":"https://pubmed.ncbi.nlm.nih.gov/27445361/"},{"id":"estruch2006","authors":"Estruch R, Martínez-González MA, Corella D, et al.; PREDIMED Study Investigators","title":"Effects of a Mediterranean-style diet on cardiovascular risk factors: a randomized trial","journal":"Annals of Internal Medicine","year":2006,"pmid":"16818923","doi":"10.7326/0003-4819-145-1-200607040-00004","type":"rct","finding":"772 adults at high cardiovascular risk: at three months the Mediterranean diet with olive oil lowered CRP by 0.54 mg/L compared with a low-fat diet.","url":"https://pubmed.ncbi.nlm.nih.gov/16818923/"},{"id":"irwin2016","authors":"Irwin MR, Olmstead R, Carroll JE","title":"Sleep disturbance, sleep duration, and inflammation: a systematic review and meta-analysis of cohort studies and experimental sleep deprivation","journal":"Biological Psychiatry","year":2016,"pmid":"26140821","doi":"10.1016/j.biopsych.2015.05.014","type":"meta-analysis","finding":"72 studies, more than 50,000 people: sleep disturbance (effect size 0.12), shorter sleep (0.09) and long sleep (0.17) were associated with higher CRP; experimental deprivation was not.","url":"https://pubmed.ncbi.nlm.nih.gov/26140821/"},{"id":"bakhru2005","authors":"Bakhru A, Erlinger TP","title":"Smoking cessation and cardiovascular disease risk factors: results from the Third National Health and Nutrition Examination Survey","journal":"PLoS Medicine","year":2005,"pmid":"15974805","doi":"10.1371/journal.pmed.0020160","type":"cross-sectional","finding":"15,489 adults: inflammatory markers rose with smoking intensity and returned to baseline about 5 years after cessation, more slowly than traditional risk factors.","url":"https://pubmed.ncbi.nlm.nih.gov/15974805/"},{"id":"teeuw2014","authors":"Teeuw WJ, Slot DE, Susanto H, et al.","title":"Treatment of periodontitis improves the atherosclerotic profile: a systematic review and meta-analysis","journal":"Journal of Clinical Periodontology","year":2014,"pmid":"24111886","doi":"10.1111/jcpe.12171","type":"meta-analysis","finding":"25 controlled trials, 1,748 patients: periodontal treatment lowered hs-CRP by 0.50 mg/L, and by 0.71 in patients with cardiovascular disease or diabetes.","url":"https://pubmed.ncbi.nlm.nih.gov/24111886/"},{"id":"gandhi2026","authors":"Gandhi KK, Batra C, Affendi H","title":"Cardiovascular and anti-inflammatory effects of periodontal therapy: a systematic review and meta-analysis of randomized trials","journal":"Journal of Periodontology","year":2026,"pmid":"42400461","doi":"10.1002/jper.70162","type":"meta-analysis","finding":"12 randomized trials: periodontal therapy lowered CRP by 0.58 mg/L and systolic blood pressure by 4.64 mm Hg, with moderate to low certainty.","url":"https://pubmed.ncbi.nlm.nih.gov/42400461/"},{"id":"dreon2003","authors":"Dreon DM, Slavin JL, Phinney SD","title":"Oral contraceptive use and increased plasma concentration of C-reactive protein","journal":"Life Sciences","year":2003,"pmid":"12850240","doi":"10.1016/s0024-3205(03)00425-9","type":"cross-sectional","finding":"In 30 healthy young women, CRP was 2.0 mg/L in low-dose oral contraceptive users against 0.9 in non-users, independent of cycle phase and diet.","url":"https://pubmed.ncbi.nlm.nih.gov/12850240/"},{"id":"ridker1999","authors":"Ridker PM, Hennekens CH, Rifai N, Buring JE, Manson JE","title":"Hormone replacement therapy and increased plasma concentration of C-reactive protein","journal":"Circulation","year":1999,"pmid":"10449692","doi":"10.1161/01.cir.100.7.713","type":"cross-sectional","finding":"In 493 healthy postmenopausal women, median CRP was twice as high on hormone replacement therapy (0.27 versus 0.14 mg/dL), independent of BMI, age, diabetes, hypertension, lipids, alcohol and smoking.","url":"https://pubmed.ncbi.nlm.nih.gov/10449692/"}],"written":"2026-10-07","updated":"2026-10-07","reviewed":true,"reviewedOn":"2026-10-10","related":{"essays":["ferritin-of-20-what-the-lab-calls-normal","gut-bacteria-and-immune-function"],"glossary":["hsCRP","Ferritin","ApoB","Reference range","Optimal range"],"symptoms":["joint-pain","fatigue","weight-loss-resistance","high-cholesterol"],"entries":["ferritin","apob","fasting-insulin"]},"url":"https://taggehealth.com/labs/hs-crp","author":"Daniel Tagge, MD","publisher":"Tagge Health","license":"Free to read and cite with attribution to the URL above.","inPractice":"I like to see hs-CRP under 1.0, measured when a person is well and repeated when it is not. That is the lowest band in the AHA/CDC banding and the reference group in the cohorts, and I am honest that no trial has treated healthy adults to that number. I read it as a gauge, not a cause: when it sits above 1, I look for what is feeding it, body fat, sleep, gums, a smoldering infection, a hormone pill, and expect it to fall as the cause is addressed, with ApoB and fasting insulin read beside it."}