C-reactive protein (CRP), the prototypical acute-phase reactant, is one of the most widely known biomarkers of cardiovascular disease. Circulating levels of CRP are clinically used to predict the occurrence of cardiovascular events and to aide in the selection of therapies based on more accurate risk assessment in individuals who are at intermediate risk. This paper reviews the role of CRP in hypertension. In hypertensive individuals, CRP levels associate with vascular stiffness, atherosclerosis and the development of end-organ damage and cardiovascular events. Data suggest that some anti-hypertensive medications may lower CRP levels in a manner independent of their effect on blood pressure. In individuals who are normotensive at baseline, CRP levels have been shown in multiple cohorts to foretell the development of hypertension on follow-up. Whether genetic variability that influences circulating levels of CRP independent of environmental and behavioral factors can also be used in a similar manner to predict the change in blood pressure and development of hypertension is controversial. In addition to its role as a biomarker, experimental studies have unraveled an active direct participation of CRP in the development of endothelial dysfunction, vascular stiffness and elevated blood pressure. CRP has also been implicated as a mediator of vascular remodeling in response to injury and cardiac remodeling in response to pressure overload. Emerging data may reveal novel vascular inflammatory pathways and identify new targets for treatment of vascular pathology.
There is a growing body of evidence that hypertension, and vascular disease in general, is an inflammatory disease.1, 2 Several threads of evidence point to the direct involvement of inflammation in the initiation and progression of vascular disease including endothelial dysfunction, atherosclerosis, vascular remodeling and hypertension. This vascular inflammatory process involves a complex interplay between inflammatory cells, cytokines, chemokines, adhesion molecules, reactive oxygen species, the renin–angiotensin–aldosterone system and the nervous system.
C-reactive protein (CRP) has long been used as a marker of systemic inflammation as circulating levels of CRP increase several hundred folds within hours of an inflammatory insult.3 With the introduction of sensitive (so-called high sensitivity) assays that can measure low CRP serum levels that were previously considered in the normal range, CRP is now also recognized as an indicator of vascular inflammation. Since this association was first reported in 1997, there has been an increasing interest in the scientific community in CRP and its role in the vascular disease process (Figure 1). Epidemiological studies support a strong relationship between CRP levels and cardiovascular risk in individuals free of cardiovascular disease at baseline. This association has been shown in multiple cohorts to be independent of other risk factors. In an individual participant meta-analysis of 160 309 individuals without a history of vascular disease from 54 long-term prospective studies (1.31 million person-years at risk, 27 769 fatal or non-fatal disease outcomes), CRP showed a strong log-linear association with risk of coronary heart disease (CHD), ischemic stroke and vascular mortality with no obvious risk thresholds. The risk ratios were attenuated upon adjustment for an extensive list of risk factors but remained statistically significant with an estimated adjusted risk ratio per s.d. higher log(e) CRP concentration of 1.37 (95% CI 1.27–1.48) for CHD, 1.27 (95% CI 1.15–1.40) for ischemic stroke and 1.55 (95% CI 1.37–1.76) for vascular mortality.4 In addition, the JUPITER (Justification for the Use of statins in Prevention: an Intervention Trial Evaluating Rosuvastatin) trial demonstrated the benefit of statin therapy in a population of patients with elevated CRP levels in the absence of other indications for treatment.5 It is due to this robust body of data that the 2010 American College of Cardiology Foundation/American Heart Association guideline for assessment of cardiovascular risk in asymptomatic adults endorsed the measurement of CRP levels in men 50 years of age or older or women 60 years of age or older (Class IIa, level of evidence B) and in younger individuals at intermediate risk (Class IIb, level of evidence B) for risk assessment and for selection of patients for statin therapy (Table 1).6 The Canadian Cardiovascular Society guidelines for the diagnosis and treatment of dyslipidemia for the prevention of cardiovascular disease in the adult (updated in 2012) have also supported the use of CRP to direct therapy in individuals that meet the entry criteria for JUPITER.7
The value of CRP in hypertensive patients
Many of the epidemiologic studies that associated CRP levels with future cardiovascular events included patients with hypertension. For example, in the meta-analysis of the 54 prospective studies discussed above, the mean±s.d. for blood pressure was 137/81±18/10 mm Hg, which indicates that a substantial proportion of the population was hypertensive.4 In this study, CRP positively correlated with systolic blood pressure (SBP) (r=0.13 (0.11–0.15)), and the adjusted risk ratio for CHD per s.d. higher levels of log(e) CRP concentration (1.37, 95% CI 1.27–1.48) was similar to that of SBP (1.35 95% CI 1.25–1.45) and higher than that of total cholesterol (1.16 95% CI 1.06–1.28) (Figure 2). Furthermore, the year-to-year consistency in CRP (serial measurements performed in 22 124 participants) was similar to that for SBP and total cholesterol. Thus, serum CRP concentrations are as consistent within individuals during several years as are total cholesterol concentration and SBP, and it provides an additional prognostic information of similar or higher magnitude.
Studies have associated CRP levels with the development of vascular disease in hypertensive patients. Hashimoto et al.8 evaluated 124 hypertensive patients (age 63±9 years, 53% male patients, 17% diabetic, 82% on anti-hypertensive medications) using serial carotid ultrasonography to assess the development of atherosclerosis. Baseline CRP levels were independently associated with the progression of carotid atherosclerosis even after adjusting for multiple factors including age, sex, pulse pressure, total cholesterol, hemoglobin A1c, cigarette pack years, body mass index, the baseline severity of carotid atherosclerosis and medication use. Importantly, the association of CRP with the progression of disease was equivalent or superior to the association of pulse pressure or SBP with this progression. A recent study confirmed the association of CRP level with the extent of carotid atherosclerosis in hypertensive older adults.9
In addition to its association with carotid atherosclerosis, CRP also associates with the markers of arterial stiffness in patients with hypertension. In a case-series study, pulse-wave velocity, measured using the SphygmoCor system in 258 hypertensive patients without antecedent cardiovascular disease or diabetes, was positively correlated to CRP in both women and men.10 In this study, CRP explained 10.2% and 6.7% of pulse-wave velocity variability in women and men, respectively. In an independent study, CRP levels were also associated with the aortic pulse-wave velocity and the augmentation index in a population of middle-aged and elderly men from the community independent of other cardiovascular risk factors.11
In patients with hypertension, CRP levels associate with end-organ damage such as left ventricular hypertrophy12, 13 and albuminuria,14 suggesting that inflammation may contribute to the development of vascular dysfunction in hypertension leading to cardiac and kidney damage. Elevated levels of CRP are independently associated with a greater rate of kidney function decline15 and the development of chronic kidney disease,16 and CRP continues to be a strong prognostic indicator in patients who have developed chronic kidney disease17 and end-stage renal disease.18, 19
In addition to the strong association between levels of CRP at baseline and future cardiovascular events as outlined above, some studies have suggested that CRP levels in the acute phase of myocardial infarction20, 21, 22 and stroke23, 24, 25, 26 have also been linked to long-term outcomes. The implication of this measure of acute inflammatory response to risk prediction beyond available measures is less clear.
The effects of interventions on CRP
Statin medications have been shown to reduce CRP independent of their effect on cholesterol levels. In fact, the reduction in CRP levels with statin therapy is an important indicator of treatment success and has prognostic implications similar in magnitude and additive to the reduction in LDL cholesterol levels.27 Anti-hypertensive medications also influence CRP levels. Angiotensin II is known to have pro-inflammatory properties and angiotensin II receptor blockers have been shown in randomized studies of hypertensive patients to lower CRP levels independent of their effects on blood pressure.28, 29 Two independent cohorts, the Multi-Ethnic Study of Atherosclerosis (MESA)30 and the Genetic Epidemiology Network of Arteriopathy (GENOA),31 showed that, in hypertensive patients receiving monotherapy, renin–angiotensin–aldosterone system inhibitor use was associated with lower CRP levels than diuretic use. In MESA, beta-blocker use was also associated with lower CRP levels among all participants and in those receiving monotherapy. Whether the effect of these medications on CRP alters the natural history of the vascular disease process or mediates part of the benefit of these anti-hypertensive medications independent of their effect on blood pressure is still unproven.
CRP as a predictor of hypertension
CRP levels have been shown to predict the development of new-onset hypertension. In the Women’s Health Study, out of 20 525 women with normal blood pressure at baseline, 5365 women developed hypertension during a median follow-up of ∼8 years. CRP levels at baseline were independently predictive of the development of incident hypertension even after adjustment for multiple risk factors.32 Similarly, in the Framingham Offspring Study, elevated CRP levels were independently associated with incident hypertension over a mean follow-up of 3 years.33
In a prospective study by Dauphinot et al.34 that selected participants who were normotensive by 24-h ambulatory blood pressure monitoring and were not on anti-hypertensive medications, the baseline CRP was associated with a 2-year risk for new-onset hypertension with the risk increasing by 18% per 1 mg l−1 increment in CRP (odds ratio 1.18, 95% CI 1.01–1.39). This relationship remained significant among participants who did not have white-coat hypertension (that is, casual blood pressure <140/90 mm Hg) with risk of incident hypertension increasing by 52% per 1 mg l−1 increment in CRP (odds ratio 1.52, 95% CI 1.17–1.96). In the Hong Kong Cardiovascular Risk Factor Prevalence Study, CRP was independently associated with prevalent hypertension and, in subjects normotensive at baseline, was independently associated with development of hypertension over a median follow-up period of 5 years.35 This relationship was also shown among 3543 MESA participants who were non-hypertensive at baseline, in whom CRP levels were associated with a significantly increased risk of future hypertension.36 Interestingly, the association was attenuated, but remained statistically significant, by adjustment for traditional risk factors for incident hypertension. Another marker of inflammation, fibrinogen, showed a weaker association with the risk of future hypertension in MESA, but this association disappeared after risk factor adjustment.
The CardioVascular Disease risk FACtors Two-township Study, a community-based follow-up study in Taiwan recently reported its findings on the association of CRP with incident hypertension.37 This study enrolled more than 2000 non-diabetics with normal blood pressure. During a median follow-up of 3 years, 145 (7%) participants developed hypertension, 21 per 1000 person-years. The incidence rates of hypertension by increasing CRP tertiles were 9, 19 and 33 per 1000 person-years (P for trend <0.0001). In a Cox regression model that adjusted for age, sex, number of components of the metabolic syndrome, general and central obesity, family history of hypertension and a high urine sodium, the upper tertile CRP accounted for a 70% increased risk of hypertension compared with the lowest tertile. CRP slightly, but significantly, improved the prediction of the onset of hypertension on top of known risk factors (IDI 0.9%, P<0.05). Importantly, baseline CRP was significantly associated with future SBP and pulse pressure even after adjustment for baseline blood pressure and other variables.
Thus, strong epidemiological data from multiple independent cohorts associate CRP levels with incident hypertension. This association could be due to the involvement of CRP or vascular inflammation in the development of hypertension or, alternatively, due to the elevated blood pressure (in the normal pre-hypertensive range) causing vascular inflammation.
CRP gene polymorphisms
In addition to the obvious effects of infectious and non-infectious acute inflammatory insults, circulating levels of CRP vary in response to multiple environmental and behavioral factors including age, gender, blood pressure, blood cholesterol level, body mass index, insulin resistance, physical activity, tobacco exposure and sleep deprivation.38, 39 Thus, CRP levels often associate with these factors in epidemiologic studies. Despite this variability, measured CRP levels has been found to be stable over time in the same individual to the same extent as other cardiovascular, biochemical and physical attributes, which are routinely used in clinical care such as blood cholesterol levels and blood pressure as discussed earlier.4 In clinical care, repeat testing is warranted if CRP level is in the acute-phase levels or if an acute injury is suspected. An observation often overlooked by clinicians is that LDL cholesterol is a negative acute-phase reactant, whose levels significantly decrease with an acute injury. Therefore, lipid testing should also be performed when patients are in their usual state of health or repeated after the resolution of the acute injury.40
In addition to these environmental effects, it has been estimated that as much as 20–50% of the inter-individual variability in circulating CRP levels may be linked to genetic variation.38 Several single-nucleotide polymorphisms (SNPs) in the CRP gene have been shown to influence CRP levels at steady state.38, 39 Genome-wide association studies have also identified multiple loci in distant genes involved in inflammatory and metabolic pathways as important determinants of CRP levels.41 Similar to the association of CRP serum levels with cardiovascular disease, previous studies suggested that CRP SNPs may also associate with various aspects of cardiovascular diseases.38, 39 More recent larger studies have not confirmed these associations. For example, the CRP Coronary Heart Disease Genetics Collaboration compiled data on 194 418 participants, including 46 557 patients with prevalent or incident CHD and failed to show an association between CRP gene variants and the risk of CHD.42
Few studies have examined the association of CRP SNPs with hypertension. The Turkish Adult Risk Factor Study43 analyzed CRP gene polymorphisms in 1987 randomly selected subjects from Turkey. Although no association was found between CRP polymorphisms and the metabolic syndrome, CRP haplotypes were associated with hypertension (defined as SBP140 mm Hg and/or diastolic blood pressure90 mm Hg and/or use of anti-hypertensive medications) in both men and women. In a logistic regression model adjusting for several factors including circulating CRP level, one haplotype showed an odds ratio for hypertension of 0.29 (95% CI 0.24–0.95, P=0.03) in women. An earlier report from the British Women’s Heart and Health Study44 that genotyped more than 3500 women for a single CRP SNP (1059G/C) showed that, although this SNP was associated with a robust difference in serum CRP and that serum CRP was associated with SBP, SBP and the prevalence of hypertension were not higher among those carrying the genetic variant associated with higher CRP levels.
A recent study evaluated the association of eight CRP SNPs, circulating CRP levels and hypertension in a random sample of 2000 subjects (909 with hypertension at study entry) of the Han Chinese descent.45 CRP levels were associated with increasing systolic and diastolic blood pressures, with prevalent hypertension (adjusted odds ratio for hypertension per CRP quartile 1.39, 95% CI 1.22–1.58, P<0.001), with the change in SBP from baseline to follow-up and with the development of hypertension (71 participants free of hypertension at study entry developed hypertension during 2 years of follow-up) in the normotensive population (adjusted odds ratio for incident hypertension per CRP quartile 1.64, 95% CI 1.18–2.26, P<0.001). Some of the CRP SNPs were associated with plasma CRP levels, but none of the SNPs were significantly associated with prevalent hypertension, systolic or diastolic blood pressure, the change in blood pressure over the follow-up period or incident hypertension. However, an independent association between the two CRP SNPs and hypertension was demonstrated in another study of Han Chinese that included 1331 patients with hypertension and 1400 controls.46
The divergent findings of these studies point to the need for larger studies evaluating the association of CRP SNPs with blood pressure and hypertension and highlight the limitation of genetic studies by the phenotype examined (for example, what is the definition of hypertension used).47
The role of CRP in the hypertensive vascular disease process
In addition to acting as a biomarker of vascular health, CRP may have a direct role in the vascular disease process (Figure 3). Mechanistic studies have shown that CRP activates inflammatory cells including monocytes and promotes their uptake of LDL to form foam cells, activates endothelial cells to increase their expression of adhesion molecules and inflammatory signals and to decrease the production of vasodilating substances such as nitric oxide and activates vascular smooth muscle cells to become more proliferative.3 Importantly, CRP has been shown to be locally present in the vasculature at the site of disease activity.48 In vitro studies, although compelling, cannot answer the question of whether CRP is causally related to vascular disease. This has largely been addressed in animal models in which exogenous CRP is administered or transgenic animals are used that express elevated levels of CRP. In this regard, mice are attractive animals to study the pathogenic effects of CRP as mouse CRP is not an acute-phase reactant but is constantly synthesized in trace amounts irrespective of injurious stimuli. Thus, genetically modified mice that overexpress CRP or, alternatively, express CRP as an acute-phase reactant, and thus behave in a manner that parallels the human experience, allow for the in vivo study of the biological activities of CRP.
Vongpatanasin et al.49 examined the effect of CRP on blood pressure in transgenic mice expressing rabbit CRP linked to a promoter/regulatory region to allow for manipulation of transgene expression by dietary interventions. Mice overexpressing CRP had significantly elevated blood pressure as measured by radiotelemetry (122±4 vs 110±1 mm Hg, P<0.05). Manipulation of transgene expression of CRP resulted in a marked alteration in blood pressure within days of an increase or decrease of CRP that paralleled the CRP change. Furthermore, CRP transgenic mice experienced exaggerated elevation in blood pressure in response to angiotensin II most likely secondary to the downregulation of angiotensin receptor type 2 expression in the vasculature due to a CRP-induced decrease in nitric oxide. An earlier report had demonstrated that CRP can also increase angiotensin receptor type 1 expression in vascular smooth muscle cells providing an alternative mechanism by which CRP can increase the sensitivity to angiotensin II.50 Another study found that CRP can enhance the effects of aldosterone on the mechanical properties of the endothelium by increasing its stiffness and decreasing its permeability, ultimately leading to an increased vascular resistance.51 These experiments provide direct evidence for the hypertensive effects of CRP.
A recent report confirmed these findings in Sprague–Dawley rats that were injected with an adeno-associated virus to induce overexpression of human CRP.52 Elevated levels of human CRP in these rats led to a significant and prolonged elevation in SBP (∼20 mm Hg higher compared with control). Interestingly, rosuvastatin treatment attenuated the CRP-induced increase in blood pressure despite persistently elevated CRP levels. Detailed studies revealed that the beneficial effects of rosuvastatin may have been mediated by countering CRP-induced endothelial dysfunction and oxidative stress.
Irrespective of the induction of hypertension, extensive evidence demonstrates that vascular remodeling after injury is an inflammatory process that directly involves CRP.3 Using CRP transgenic mice that carry a transgene containing the entire human CRP promoter and thus express human CRP in a manner that parallels the human condition, our laboratory has shown that CRP transgenic mice subjected to ligation of the common carotid artery developed nearly twofold greater neointima formation compared with control non-transgenic mice.53, 54, 55 This was associated with an extensive human CRP messenger RNA expression and a protein deposition in the injured vessels of the transgenic mice. CRP amplified the inflammatory response to acute vascular injury by binding to immunoglobulin G Fc receptors (FcγRs) on the surface of macrophages resident in the peri-adventitia of the injured vessel wall, and initiating an inflammatory cascade that resulted in complement component C3 activation and deposition, culminating in exacerbation of vascular remodeling. Other investigators have shown in CRP transgenic mice that human CRP promotes cardiac fibrosis and the reduction in left ventricular ejection fraction seen in response to angiotensin II infusion56 and exacerbates pressure overload-induced cardiac remodeling induced by transverse aortic constriction.57
There is a large body of evidence showing CRP to be a strong, precise, accurate and independent marker of cardiovascular disease risk. CRP also associates with prevalent hypertension and with markers of arterial stiffness and end-organ damage in hypertensive patients, and serum CRP is a useful biomarker that predicts overall vascular health in these patients. Multiple cohorts have now identified CRP as a predictor of incident hypertension in individuals without hypertension at baseline. Whether CRP gene variants that result in elevated serum levels of CRP confer an increased risk of hypertension is still controversial, and the data available so far are limited by the small sample size and problems with identifying the phenotype examined, as hypertension involves a disease process that is far more complex than a blood pressure number. Elegant mechanistic studies have demonstrated that CRP induces hypertension in rodents and has started to delineate the mechanisms of this intricate process. Furthermore, CRP in animal models exacerbates vascular remodeling in response to injury and worsens end-organ damage caused by hypertension. If these findings are true in humans, this may open the door to therapies that target inflammatory pathways for the treatment of vascular pathology. Two on-going randomized clinical trials58, 59, 60 are testing the hypothesis that anti-inflammatory therapies reduce cardiovascular events and will provide valuable data that may impact the treatment of patients at high risk due to a pro-inflammatory vascular profile.
This work was supported, in part, by a Veterans Affairs Biomedical Laboratory Research & Development Service Merit Award and an American Heart Association NCRP Scientist Development Grant 0930098N. Grant support from Novartis Pharmaceuticals.
About this article
European Journal of Clinical Investigation (2019)