Search tips
Search criteria 


Logo of nihpaAbout Author manuscriptsSubmit a manuscriptHHS Public Access; Author Manuscript; Accepted for publication in peer reviewed journal;
J Am Coll Cardiol. Author manuscript; available in PMC 2010 December 15.
Published in final edited form as:
PMCID: PMC2829945

Clinical implications of JUPITER in a United States population: insights from the Atherosclerosis Risk in Communities (ARIC) Study

Eric Y. Yang, M.D.,*†,1 Vijay Nambi, M.D.,*†,1 Zhengzheng Tang, M.S., Salim S. Virani, M.D.,* Eric Boerwinkle, Ph.D.,§ Ron C. Hoogeveen, Ph.D.,* Brad C. Astor, Ph.D., Thomas H. Mosley, Ph.D, Josef Coresh, M.D., Ph.D., Lloyd Chambless, Ph.D., and Christie M. Ballantyne, M.D.*



To describe the proportion of “JUPITER-eligible” individuals and clinical outcomes of individuals based on high-sensitivity C-reactive protein (hs-CRP) and low-density lipoprotein cholesterol (LDL-C) strata in the Atherosclerosis Risk in Communities (ARIC) study.


Questions remain after the JUPITER study including whether the observed cardiovascular disease (CVD) event rates would persist with time and how these event rates would compare with other populations (lower hs-CRP and/or higher LDL-C levels).


After stratification into 4 groups based on LDL-C and hs-CRP levels, with cutoffs at 130 mg/dL and 2.0 mg/L, respectively incident CVD events were examined (mean follow-up 6.9 years) and compared.


Of 8,907 age-eligible participants, 18.2% (n=1,621) were “JUPITER-eligible” (hs-CRP ≥2.0 mg/L, LDL-C <130 mg/dL) and had an absolute CVD risk of ~10.9% over a mean follow-up of 6.9 years (1.57% per year). If JUPITER hazard ratios were applied to this group, the number needed to treat to prevent one CVD event would be estimated at 38 over 5 years and 26 over 6.9 years.


ARIC participants with elevated hs-CRP and low LDL-C had a CVD event rate of 1.57% per year over 6.9 years similar to the CVD event rate noted in the JUPITER study placebo group (1.36% per year over 1.9 years). The association of hs-CRP ≥2.0 mg/L with increased CVD risk and mortality regardless of LDL-C provides us a simple method of using age and hs-CRP for identifying higher risk individuals.

Keywords: hs-CRP, LDL-C, ARIC, lipids, cardiovascular disease


Recently, rosuvastatin therapy was shown to reduce cardiovascular disease (CVD) events in men and women (≥50 years and ≥60 years of age, respectively) with low-density lipoprotein cholesterol (LDL-C) <130 mg/dL and high-sensitivity C-reactive protein (hs-CRP) ≥2.0 mg/L without known CVD in the Justification for the Use of statins in Prevention: an Intervention Trial Evaluating Rosuvastatin (JUPITER) (1).

Questions remained at the conclusion of JUPITER including: 1) would the absolute event rates and risk reduction persist if the trial was continued for a longer time given the short (1.9 years) follow-up 2) what percentage of the population may be eligible for therapy based on the study and 3) would similar findings have been observed if individuals with hs-CRP <2.0 mg/L were included. Although randomized-controlled trials may be required to answer some of these question (such as risk reduction), population-based studies may be ideal to evaluate JUPITER’s clinical implications and the natural history/ event rates in individuals with elevated hs-CRP and low LDL-C.

One analysis using the National Health and Nutrition Examination Survey (NHANES) 1999–2004 data suggested 8.07 million individuals(13.9%) of age-eligible individuals would meet JUPITER eligibility criteria, (2) while another analysis suggested that ~6.5 million individuals may be eligible (3). However, these studies lacked longitudinal follow-up for outcomes and the ability to provide actual CVD risk in a population. Other primary prevention studies compared CVD event rates in individuals by hs-CRP and LDL-C groups, but these analyses were performed post-hoc on randomized patients in clinical trials making them prone to selection bias; moreover, they used very different cutoff levels for hs-CRP and LDL-C.

Therefore, we examined JUPITER’s clinical implications in the Atherosclerosis Risk in Communities (ARIC) study, a population-based study of CVD disease incidence. We compared outcomes in the “JUPITER eligible” group to those in the other LDL-C and hs-CRP level strata (i.e., hs-CRP <2.0 mg/L and/or LDL-C ≥130 mg/dL).


Study Population

The design and objectives of the ARIC study have been previously described (4). Briefly, the ARIC study is a prospective, biracial, observational study of CVD in 15,792 individuals between ages 45–64 years at the initial visit (1987–1989). Our analysis used Visit 4 (1996–1998) data which had hs-CRP levels available in 11,343 of 11,656 individuals. We applied JUPITER exclusion criteria (Figure 1) except that peripheral vascular disease (PVD) and current hormone replacement therapy (HRT) use in women were not excluded since adequate information for these variables were not available during ARIC Visit 4.. After exclusions, except for LDL-C and hs-CRP, there were 5,513 individuals, who were further stratified by LDL-C (<130 mg/dL vs. ≥130 mg/dL) and hs-CRP (<2.0 mg/L vs. ≥2.0 mg/L) into 4 groups (Figure 1).

Figure 1
Individuals meeting JUPITER criteria after exclusions in the ARIC population (1996–1998)

Laboratory Methods

Plasma levels of hs-CRP were measured using an immunonephelometric assay (Dade Behring Marburg GmbH, USA) (5) (reliability coefficient 0.99 based on 421 blinded replicates).

Definition and Ascertainment of Outcomes

Our primary outcome of first major adverse cardiovascular event (MACE) was similar to JUPITER’s, which included nonfatal myocardial infarction (MI), nonfatal stroke, hospitalization for unstable angina, coronary arterial revascularization procedure (coronary arterial bypass graft/percutaneous coronary intervention), or confirmed cardiovascular death (fatal CHD or fatal stroke) except that fatal stroke (not previously defined and adjudicated in the ARIC study) was not included and fatal CHD was defined differently. Additionally, we evaluated a secondary composite endpoint of nonfatal MI, fatal CHD, nonfatal stroke and coronary revascularization (i.e., primary endpoint excluding unstable angina). Finally, like JUPITER, we evaluated deaths due to any cause or malignancy (ICD9 Codes 140-208 or ICD10 Codes C00-C97). The methods for ascertaining the various outcomes in ARIC have been previously published (4,68).

Statistical Analysis

The four groups (Figure 1) were compared for adjusted means, proportions, or incidence rates using linear, logistic, or Poisson models. Model adjustments included age, race, and gender. Each group had 10-year CHD risk calculated using ARIC coronary risk score (ACRS) and Framingham risk score (FRS), both calibrated to the ARIC initial examination (1986–1989) (911). Tests of contrasts of the beta coefficients from the linear or logistic regression models were used to calculate ‘p’-values. Hazard ratios (HR) were calculated using Cox proportional hazard models to compare groups within LDL-C and hs-CRP strata. Kaplan-Meier event free survival from any MACE was also examined.

We also tested whether significant cardiovascular risk due to elevated hs-CRP persisted within each LDL-C stratum after further adjustments for smoking status, family history of premature CHD, LDL-C levels, and Adult Treatment Panel III-defined metabolic syndrome (12).


Of the total Visit 4 participants (n=11,656), 1,621 (683 men and 938 women) were age-eligible, met all screening criteria, and had LDL-C <130 mg/dL and hs-CRP ≥2.0 mg/L (i.e., “JUPITER-eligible”). These 1,621 individuals (Group 2) constituted 29.4% of the screened group (n=5,513) and 18.2% of the age-eligible ARIC individuals (n=8,907) (Figure 1).

In both LDL-C categories (i.e., LDL-C <130 and ≥130 mg/dL), ~50% of the individuals had hs-CRP ≥2.0 mg/L (Figure 1). Individuals with hs-CRP ≥2.0 mg/L were more likely to be older, women, black, current smokers, and on aspirin or antihypertensive therapy, regardless of LDL-C category and had a higher prevalence of metabolic syndrome, which was supported by higher waist circumference, body mass index, systolic blood pressure, triglyceride level, and glucose level (Table 1).

Table 1
Clinical characteristics by LDL-C/ hs-CRP categories, adjusted for age, gender, and race

The average predicted 10-year CHD risk of all groups using FRS or ACRS was “low” (i.e., 10-year CHD risk ≤10%) (Table 2). Individuals in the “JUPITER-eligible” group had the lowest average 10-year predicted CHD risk using either model (Table 2).

Table 2
Categorization of individuals* by 10-year CHD risk calculated by Framingham risk score and ARIC risk score

Over a mean follow-up of 6.9 years, the “JUPITER-eligible” group had a MACE rate of 15.73 events (primary outcome) per 1,000 person-years (adjusted for age, race and gender) (Table 3) which was elevated (hazard ratio [HR] of 1.65; 95% confidence interval [CI] 1.29, 2.11) when compared to Group 1 (LDL-C <130 mg/dL, hs-CRP <2.0 mg/L). Group 4 (LDL-C ≥130 mg/dL, hs-CRP ≥2.0 mg/L) had a MACE rate of 21.96 events per 1,000 person-years, which was the highest of all 4 groups, significantly higher than Group 3 (LDL-C ≥130 mg/dL, hs-CRP <2.0 mg/L; HR=1.80, 95% CI 1.39 to 2.33), and nonsignificantly higher than Group 2 (HR=1.27, 95% CI 0.96 to 1.67) (Table 4). However, all-cause mortality and deaths due to cancer were highest in the JUPITER group (14.82 and 6.30 deaths per 1,000 person-years, respectively). All event rates were higher in the groups with the elevated hs-CRP (i.e., Groups 2 and 4) for both LDL-C strata. Lower probability of event-free survival in these groups compared to their counterparts was also shown by Kaplan-Meier models (Figure 2). Men with elevated hs-CRP had greater event rates for all outcomes in all groups (Table 3 and Table 4). The number of clinical events in each group is shown in Supplemental Table 1.

Figure 2
Kaplan-Meier cardiovascular event-free survival curves of the four groups analyzed
Table 3
Event rates per 1,000 person-years adjusted for age, race, and gender.
Table 4
Hazard ratios (HR) (95% confidence intervals [CI]) adjusted for age, race, and gender comparing hs-CRP groups (≥2.0 mg/L to <2.0) within each LDL-C strata and comparing LDL-C groups (≥130 mg/dL to <130) within each hs-CRP ...

After additional adjustments as detailed in the Methods section, the hazard ratios comparing hs-CRP ≥2.0 mg/L to <2.0 mg/L (i.e., Groups 2 vs. 1 and Groups 4 vs. 3) for MACE, mortality and cancer deaths remained significant (Supplemental Table 3); however, the hazard ratios comparing LDL-C ≥130 mg/dL to <130 mg/dL (i.e., Groups 3 vs. 1 and Groups 4 vs. 2) for the same outcomes, except mortality, became nonsignificant (Supplemental Table 3).


At the end of the JUPITER study, questions remained (13) as detailed in the Introduction. Our analysis examined some of these questions in the ARIC study and compared the characteristics, predicted 10-year CHD risk, and actual cardiovascular event rates over a long-term follow-up of those who were and were not “JUPITER-eligible.”

Cardiovascular Risk in JUPITER-eligible ARIC Participants over Longer Follow-up

The MACE rates in the JUPITER-eligible group in our study over a mean follow-up of 6.9 years was similar to the JUPITER placebo arm (Table 5) (1) suggesting that the event rates observed in JUPITER are likely to persist over longer observation periods. When the event reduction observed with rosuvasatin in JUPITER was applied to our “JUPITER-eligible” group, we determined the number needed to treat (NNT) to prevent one first MACE was ~38 individuals over 5 years and ~26 over 6.9 years using the Altman-Andersen method.

Table 5
Comparison of ARIC JUPITER-eligible group, JUPITER placebo arm, JUPITER treated arm, and their respective event rates.

Comparison of 10-year Cardiovascular Risk by LDL-C and hs-CRP Groupings and its clinical implications

Although individuals in the “JUPITER-eligible” group numerically had the lowest average 10-year predicted CHD risk, they had a significantly higher MACE rate compared to Group 1, and a numerically higher MACE rate compared to individuals of Group 3. Group 4 individuals numerically had the highest average 10-year predicted CHD risk and the highest MACE rates of the four groups, confirming prior analyses from primary prevention trials which showed that the highest incidence of CVD occurred in “high” LDL-C and “high” hs-CRP groups (14,15).

Based on the known benefit of statins in those with LDL-C ≥130 mg/dL, it would be ethically difficult to conduct a placebo-controlled study in patients with LDL-C ≥130 mg/dL and hs-CRP ≥2.0 mg/L. Clinicians, therefore, may have to extrapolate the available data and consider all patients with hs-CRP ≥2.0 mg/L for statin treatment for the primary prevention of CVD.

Prevalence and Characteristics of the ARIC Population Meeting JUPITER Criteria (Group 2)

The “JUPITER-eligible” individuals in our study were similar to the treated and placebo arms of JUPITER at baseline; however, our study had more women and greater aspirin use (Table 5) (1). Our study showed that ~18.2% of age eligible individuals met LDL-C and hs-CRP criteria which were similar to the 19.8% reported in JUPITER. However, a NHANES analysis showed only ~13.9% of the age-eligible American population to be eligible (2). Therefore, the actual prevalence of “JUPITER-eligible” Americans may range between ~14% to ~20%. However, ~50% of either LDL-C strata (i.e., LDL-C <130 and ≥130 mg/dL) had hs-CRP ≥2.0 mg/L in our study. Therefore, if all individuals with hs-CRP ≥2.0 mg/L were considered eligible for statins (although not tested in JUPITER), the number would be significantly higher.

Other Observations

Groups with elevated hs-CRP (Groups 2, 4) had significantly higher all-cause and cancer mortality than their counterparts with low hs-CRP (Groups 1, 3). Groups with high LDL-C (Groups 3, 4) had nonsignificantly higher all-cause and cancer mortality than their counterparts with low LDL-C (Groups 1,2) (Table 4). Malignancy and inflammatory disorders were not excluded since they were not previously defined in the ARIC study. Regardless, it appears that those with elevated hs-CRP may have increased mortality which may merit further investigation.

Higher proportions of women were observed in groups with hs-CRP ≥2.0 mg/L (Groups 2, 4). Overall, women with hs-CRP ≥2.0 mg/L had higher event rates compared to counterparts with hs-CRP <2.0 mg/L. Importantly, despite women being generally considered at lower CVD risk than men, they were the majority in groups with hs-CRP ≥2.0 mg/L, which had the highest event rates.

Lastly, CVD risk and mortality remained significantly increased in those with elevated hs-CRP (Groups 2, 4) even after further adjustments for other risk factors. This finding suggests an association of elevated hs-CRP with increased cardiovascular risk independent of “traditional” risk factors. Likewise, the increased cancer death risk associated with elevated hs-CRP remained significant within each LDL-C stratum after these adjustments. Therefore, based on JUPITER and our analysis, the mere utilization of age (men ≥50 years, women ≥60 years) and hs-CRP ≥2.0 mg/L identified a population at higher CVD and mortality risk, which may be decreased with statin therapy. This simple way of identifying higher risk individuals using these 2 variables will make it easier for clinicians without having to use risk equations.


The ARIC population consists predominantly of whites and blacks, limiting the generalizability to other ethnicities. HRT use in women (an exclusion criteria in JUPITER) and PVD presence were not excluded (see Methods section). JUPITER excluded only those with a “history” of symptomatic PVD and did not measure ABI’s on all participants. Other JUPITER exclusion criteria (e.g., immunosuppressant use, rheumatologic diseases, malignancy, inflammatory bowel disease, elevated creatinine kinase >3 times the upper limit of normal, or thyroid disorders) were not available for the ARIC participants.


Our study showed that ~18.2% of the age-eligible ARIC population would have been eligible for JUPITER, and they had an absolute CVD risk of ~10.9% over 6.9 years,,suggesting the CVD risk observed in the JUPITER study may persist over an extended period of time. Our study in concert with JUPITER supports a simple method of using age and hs-CRP for identification of higher risk individuals.

Supplementary Material


Supplemental Figure 1: Event rates in 1,000 person-years adjusted for age, race, and gender over a mean follow-up of 6.9 years:

Event rates for each of the 4 groups are shown for outcomes of interest. Absolute event counts and event counts and rates by gender can be found in the accompanying Supplemental Tables. *All major CV events include fatal CHD, hospitalized MI, hospitalized stroke, unstable angina, and coronary revascularization


We thank staff and participants of ARIC for their contribution and Joanna Brooks, B.A., for editorial assistance.

Financial Support:The Atherosclerosis Risk in Communities Study is a collaborative study supported by National Heart, Lung, and Blood Institute contracts N01-HC-55015, N01-HC-55016, N01-HC-55018, N01-HC-55019, N01-HC-55020, N01-HC-55021, and N01-HC-55022 from the National Heart, Lung, and Blood Institute (NHLBI), Bethesda, MD.

Abbreviations List

high-sensitivity C-reactive protein
cardiovascular disease
coronary heart disease
low-density lipoprotein cholesterol
ankle-brachial index
myocardial infarction
ARIC coronary risk score
Framingham risk score
peripheral vascular disease
major adverse cardiovascular event


Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

Relationship with industry:

Vijay Nambi: Research collaboration with General Electric

Salim S. Virani: Speakers bureau for Abbott, Daiichi-Sankyo

Christie M. Ballantyne: Consultant for Abbott, AstraZeneca, Bristol-Myers/Squibb, GlaxoSmithKline, KOWA, Merck, Merck/Schering-Plough, Metabasis, Novartis, Pfizer, Sanofi-Synthelabo, Schering-Plough, Takeda; grant/research support from: Abbott, AstraZeneca, GlaxoSmithKline, Merck, Sanofi-Synthelabo, Schering-Plough, Takeda; honoraria from Abbott, AstraZeneca, GlaxoSmithKline, Merck, Merck/Schering-Plough, Novartis, Pfizer, Sanofi-Synthelabo, Schering-Plough, Takeda; and speakers bureau for AstraZeneca, GlaxoSmithKline, Merck, Merck/Schering-Plough, Pfizer, Schering-Plough.

Remaining authors have nothing to disclose.

The study has been registered in (, ID: NCT00005131.


1. Ridker PM, Danielson E, Fonseca FA, et al. Rosuvastatin to Prevent Vascular Events in Men and Women with Elevated C-Reactive Protein. N Engl J Med. 2008;359:2195–2207. [PubMed]
2. Spatz ES, Canavan ME, Desai MM. From Here to JUPITER: Identifying New Patients for Statin Therapy Using Data From the 1999–2004 National Health and Nutrition Examination Survey. Circ Cardiovasc Qual Outcomes. 2009;2:41–48. [PubMed]
3. Michos ED, Blumenthal RS. Prevalence of Low Low-Density Lipoprotein Cholesterol With Elevated High Sensitivity C-Reactive Protein in the U.S.: Implications of the JUPITER (Justification for the Use of Statins in Primary Prevention: An Intervention Trial Evaluating Rosuvastatin) Study. Journal of the American College of Cardiology. 2009;53:931. [PubMed]
4. The Atherosclerosis Risk in Communities (ARIC) Study: design and objectives. The ARIC investigators. Am J Epidemiol. 1989;129:687–702. [PubMed]
5. Roberts WL, Moulton L, Law TC, et al. Evaluation of nine automated high-sensitivity C-reactive protein methods: implications for clinical and epidemiological applications. Part 2. Clin Chem. 2001;47:418–425. [PubMed]
6. Toole JF, Chambless LE, Heiss G, Tyroler HA, Paton CC. Prevalence of stroke and transient ischemic attacks in the Atherosclerosis Risk in Communities (ARIC) study. Ann Epidemiol. 1993;3:500–503. [PubMed]
7. White AD, Folsom AR, Chambless LE, et al. Community surveillance of coronary heart disease in the Atherosclerosis Risk in Communities (ARIC) Study: methods and initial two years' experience. J Clin Epidemiol. 1996;49:223–233. [PubMed]
8. Rosamond WD, Chambless LE, Folsom AR, et al. Trends in the incidence of myocardial infarction and in mortality due to coronary heart disease, 1987 to 1994. N Engl J Med. 1998;339:861–867. [PubMed]
9. Wilson PW, D'Agostino RB, Levy D, Belanger AM, Silbershatz H, Kannel WB. Prediction of coronary heart disease using risk factor categories. Circulation. 1998;97:1837–1847. [PubMed]
10. Chambless LE, Folsom AR, Sharrett AR, et al. Coronary heart disease risk prediction in the Atherosclerosis Risk in Communities (ARIC) study. J Clin Epidemiol. 2003;56:880–890. [PubMed]
11. Folsom AR, Chambless LE, Duncan BB, Gilbert AC, Pankow JS. Prediction of coronary heart disease in middle-aged adults with diabetes. Diabetes Care. 2003;26:2777–2784. [PubMed]
12. Third Report of the National Cholesterol Education Program (NCEP) Expert Panel on Detection, Evaluation, and Treatment of High Blood Cholesterol in Adults (Adult Treatment Panel III) final report. Circulation. 2002;106 3143-421. [PubMed]
13. Hlatky MA. Expanding the orbit of primary prevention--moving beyond JUPITER. N Engl J Med. 2008;359:2280–2282. [PubMed]
14. Ridker PM, Rifai N, Pfeffer MA, et al. Inflammation, pravastatin, and the risk of coronary events after myocardial infarction in patients with average cholesterol levels. Cholesterol and Recurrent Events (CARE) Investigators. Circulation. 1998;98:839–844. [PubMed]
15. Ridker PM, Rifai N, Clearfield M, et al. Measurement of C-reactive protein for the targeting of statin therapy in the primary prevention of acute coronary events. N Engl J Med. 2001;344:1959–1965. [PubMed]