Вы находитесь на странице: 1из 6

Rheumatology 2004;43:924929 doi:10.

Advance Access publication 18 May 2004

Risk of coronary heart disease and stroke in a large

British cohort of patients with systemic lupus
R. Bessant, A. Hingorani1, L. Patel, A. MacGregor2, D. A. Isenberg
and A. Rahman

Objectives. Patients with systemic lupus erythematosus (SLE) are at increased risk of myocardial infarction and stroke. We
sought to determine how much of this risk was dependent on recognized cardiovascular risk factors.
Methods. Initially a software package Cardio-Risk-Manager, which utilizes Framingham data, was used to calculate a 10-yr
risk of coronary heart disease (CHD) and stroke for 202 patients with SLE (Group 1) in comparison with hypothetical age- and
sex-matched comparators. Subsequently 47 patients who had been followed since 1991 (Group 2) were studied to compare their
predicted risks in 1991 with the actual number of cardiovascular events that occurred during the subsequent decade.
Results. Patients in Group 1 had a higher predicted 10-yr risk of stroke (P<0.0001), but not of CHD, than their comparators.
However, following age stratication, traditional risk factors predicted a higher risk of CHD (P<0.0001) and of stroke
(P<0.0001) in patients under 40 with SLE compared with age-matched comparators. The predicted 10-yr risks of CHD and
stroke for patients aged 40 and above were not signicantly different from those of their comparators. Predicted risks, however,
were lower than the true 10-yr event rate for CHD and stroke in patients in Group 2. In this group, during the 10 yr of follow-up
four patients (8.5%) suffered a CHD event and ve patients (10.6%) had a stroke, signicantly more than were predicted by
the presence of conventional risk factors (P<0.001 for CHD and P<0.001 for stroke, respectively).
Conclusions. Conventional risk factors predicted an increased risk of stroke and CHD in younger patients. They do not,
however, fully explain the high risk of cardiovascular disease in patients with SLE. Although it is important to address the
management of orthodox risk factors for cardiovascular disease in patients with SLE, other causes must be sought to explain
the increased incidence of CHD and stroke, especially in those aged over 40.
KEY WORDS: Systemic lupus erythematosus, Risk, Coronary heart disease, Stroke, Cardiovascular disease.

In spite of considerable improvements in treatment, morbidity and (SPECT) scanning has revealed coronary disease in 40% of such
mortality in patients with systemic lupus erythematosus (SLE) are patients [15]. Autopsy studies have demonstrated a prevalence of
substantial, with mortality rates of 510% at 5 yr and 1530% at 4153% for moderate to severe atherosclerosis affecting the aorta,
10 yr [16]. Patients with SLE have a nearly 5-fold increased risk of coronary, renal and cerebral vessels [1, 8].
death compared with the general population. The standardized Stroke occurs in up to 15% of patients with SLE, with some
mortality ratio is particularly high, 9.16, in patients aged less than suffering multiple events. The majority of these events occur within
55 yr [1]. A bimodal pattern of mortality has been described, with the rst 5 yr of the diagnosis of SLE [16, 17].
early deaths predominantly due to active SLE and intercurrent Although there is an increased relative risk for atherosclerotic
infection and late deaths principally due to atherosclerotic disease CVD in patients with SLE, there is less agreement about the
[1, 7, 8]. Cardiovascular disease (CVD) currently accounts for relative contributions of orthodox risk factors, such as hyperten-
between 20 and 30% of deaths in patients with SLE, but is sion, diabetes, smoking and hypercholesterolaemia and of lupus-
responsible for a proportionately greater percentage of deaths in specic risk factors (Table 1). This conict may partly arise from
late disease [1, 8, 9]. As the anti-inammatory/immunosuppressive the differences between study cohorts. Studies have suggested that
treatment of patients with SLE continues to improve [10], the not all of the increased risk of CVD seen in patients with SLE is
contribution of cardiovascular disease to morbidity and mortality attributable to orthodox risk factors [18, 19]. Esdaile et al. [18]
is likely to increase. showed that, after removing the effect of these known risk factors,
Jonsson et al. [11] showed that the myocardial infarction (MI) there is still a 7.9-fold increase in the risk of stroke and a 10.1-fold
rate in Swedish patients with SLE was increased 9-fold. Whilst increase in the risk of non-fatal myocardial infarction (MI) in
clinical evidence of coronary disease is found in 6.7 to 8.3% of patients with SLE.
patients with SLE [9, 12], subclinical atherosclerosis is far more Most studies of this nature have been carried out in North
common. Atheromatous carotid plaques have been demonstrated America and a calculation of risk of CVD in a European
by ultrasonography in 2540% of patients with SLE [13, 14]. population has not been published to date. We carried out an
Myocardial single photon emission computed tomography analysis in a large British cohort. We calculated the predicted 10-yr

Centre for Rheumatology, University College London Hospitals, London, 1Centre for Clinical Pharmacology, BHF Laboratories, University College
London Hospitals, London and 2School of Medicine, Health Policy and Practice, University of East Anglia, Norwich, UK.

Received 22 September 2003; revised version accepted 31 March 2004.

Correspondence to: R. Bessant, Centre for Rheumatology, University College London Hospitals, Arthur Stanley House, Tottenham Street, London
W1T 4NJ, UK. E-mail: r.bessant@ucl.ac.uk
Rheumatology Vol. 43 No. 7 British Society for Rheumatology 2004; all rights reserved
Coronary heart disease and stroke risk in SLE 925

TABLE 1. Conventional and SLE-specic risk factors

Conventional risk factors References SLE-specic risk factors References

Hypercholesterolaemia 9, 12, 14, 18, 45 Disease duration 8, 9, 11, 12, 14

Hypertension 9, 14, 18, 45 Age at diagnosis of SLE 9, 12
Age 14, 18 Corticosteroid usea 9, 11, 12, 14
Gender 46 Pericarditis/myocarditis 45
Diabetes mellitus 45 Anti-phospholipid antibody 38, 46
Past history of CVD 14 Renal insufciency 46
Post-menopausal status 12, 14 Homocysteine levels 38, 46
Body mass index 14 Fibrinogen levels 14
LVH 8 CRP 14
Triglycerides 38 SLICC DI score 14
SLAM-R activity 14
Studies have demonstrated an association with both cumulative dose and duration of steroid use.

risk of CHD and stroke for each member of a cohort of 202 if they were current smokers or had given up smoking in the
patients with SLE, using a software package that utilizes data from preceding 12 months.
the Framingham study, to investigate whether management The software package then calculates a 10-yr risk of coronary
designed to target orthodox cardiovascular risk factors would be heart disease (CHD) and CVD (stroke) for each patient, and for a
predicted to reduce risk of CVD signicantly in patients with SLE. hypothetical age- and sex-matched comparator. The program
We then sought to determine, in a smaller group of patients, how assumes that the hypothetical comparators are healthy individuals
the predicted 10-yr risk of CHD and stroke compared with the with mean population levels of blood pressure, total cholesterol
number of CHD and stroke events that were actually observed and HDL cholesterol for an age- and sex-matched population.
during a 10-yr follow-up period. They are also assumed to be non-smokers, free of diabetes, LVH
and previous CVD. The risk for the comparator thus shows the
possible reduction in the patients risk of CVD which might be
achieved by ideal management of orthodox risk factors.
Methods The predicted CHD and stroke risks for the patients and the
comparators were then compared using the Wilcoxon signed ranks
Cross-sectional study: comparison of patients with SLE
test for two related samples. Similar analyses were performed
with hypothetical comparators following age stratication of Group 1 into those under 40 yr of age
Group 1 consisted of 202 consecutive patients (186 female, 16 and those aged 40 and above.
male) attending a specialist SLE clinic over a period of 12 months
in 2001, all of whom fullled the revised American College of
Rheumatology (ACR) criteria for the classication of SLE [20]. Longitudinal study: comparison of predicted risk with actual
Four patients aged over 70 yr were not included because the cardiovascular events
software package used is not validated for patients older than 70.
Analysis of this group was designed to determine whether these Group 2 was drawn from a sample of 64 patients, also fullling the
individuals had a greater predicted risk of CVD than age- and revised ACR criteria, who had attended the same specialist SLE
sex-matched comparators, when prediction is based on orthodox clinic in 1991 and were originally recruited for a study of lipid levels
risk factors only. and anti-cardiolipin antibodies in patients with SLE [34]. Sufcient
data to permit the calculation of a 10-yr predicted risk of CHD
and stroke in 1991, together with continuous follow-up data for
the succeeding 10 yr (or to the time of death), were available for
Cardio Risk Manager program 47 of these patients. Four patients had moved away, and for
The Cardio Risk Manager computer program, which is based the remaining 13 baseline data were insufcient. One of these 13
upon Microsoft Excel Version 5.0, was developed by Hingorani patients suffered a myocardial infarction between 1991 and 2001.
and Vallance [21] for use in primary prevention of atherosclerotic The 10-yr risks of CHD and stroke predicted for the remaining
events. This interactive program makes risk calculations utilizing 47 patients were compared with the number of cardiovascular
logistic regression equations derived from the Framingham events that actually took place during this 10-yr period (or until the
populations [2224]. The program has been validated against time of death) using a HosmerLemeshow analysis. The great
published data for large intervention studies [2527], and against majority of patients from Group 2 remained as current patients in
the risk estimates provided by the Shefeld tables [28, 29], which 2001 and were, therefore, also included in Group 1.
are also based on Framingham data. The fact that the
Framingham data were used successfully to predict the event rate
in the placebo arm of the West of Scotland Coronary Prevention Results
Study [30] suggests that the Framingham equations are directly
applicable to a UK population. For each patient, data are entered Cross-sectional study: characteristics of patients
on the following risk factors: systolic and diastolic blood pressure, The characteristics of the 202 patients in Group 1 are shown in
total serum cholesterol and high-density lipoprotein (HDL) Table 2. Their mean values for blood pressure and serum
cholesterol, blood glucose, history of smoking, previous cardio- cholesterol are not elevated, but there was a wide range of values.
vascular disease, left ventricular hypertrophy (LVH) and diabetes. Thirty-four patients were considered to be hypertensive (systolic
The presence of left ventricular hypertrophy was determined by blood pressure (SBP) >140 mmHg and/or diastolic blood pressure
electrocardiography and dened by the sum of the R wave in lead (DBP) >90 mmHg), 65 had hypercholesterolaemia (serum total
V5 or V6 and the S wave in lead V1 or V2 being greater than cholesterol >5.2 mmol/l) and 45 smoked. One hundred and thirty-
35 mm. Patients were considered to have a positive smoking history four of these patients were of Caucasian origin, 35 were black
926 R. Bessant et al.

African or Afro-Caribbean, 21 were Asian, four were Chinese and Comparison of predicted risk with actual
eight were from other ethnic groups or of mixed race. cardiovascular events
The median predicted 10-yr risks of CHD (1.4%, IQR 0.2 to 3.4)
and stroke (0.6%, IQR 0.4 to 1.3) for the patients with SLE in this
Comparison of patients with SLE with hypothetical group were similar to the median CHD (1.0%, IQR 0 to 3.0) and
comparators stroke (1.0%, IQR 0.1 to 1.0) risks predicted for the hypothetical
matched comparators. Analysis by the Wilcoxon signed rank test
The median predicted 10-yr risk of CHD for the 202 patients with
showed no signicant difference in risk of CHD or stroke between
SLE, calculated by the Cardio Risk Manager program, was 1.2%
the patients with SLE and their comparators. This is, however,
(interquartile range (IQR) 0.1 to 4.2). This risk was similar to the
probably due in part to the small sample size of Group 2.
median CHD risk of 1.0% (IQR 0 to 4.0) predicted for the However, during the 10 yr of follow-up four patients (8.5%)
hypothetical matched comparators (Table 3). In contrast, the SLE suffered a CHD event and ve patients (10.6%) had a stroke.
cohorts median risk of stroke was 0.8% (IQR 0.3 to 1.7), while Using the binomial distribution, we calculate that the observed
the median risk for the comparators was 1.0% (IQR 0.1 to 1.0). incidences of both CHD and stroke over this 10-yr period are
Analysis by the Wilcoxon signed rank test showed that patients signicantly greater than would be predicted by the Cardio Risk
with SLE had signicantly higher predicted risk of stroke Manager program (P<0.001 for CHD and P<0.001 for stroke).
(P<0.0001), but not of CHD, than their comparators. Four patients died during this 10-yr period: two following a stroke,
Since previous studies had suggested a particularly increased one as a result of an infection and one due to a malignancy.
relative risk of CVD in younger patients with SLE, Group 1 was Figure 1 illustrates the excess of observed CHD and stroke over
also stratied by age into patients aged under 40 (n 93) and those the number of expected events based on the risk predictions using
aged 40 or over (n 109). This analysis revealed that traditional data collected in 1991. The solid line indicates the number of
risk factors predicted a higher risk of CHD (median 0.1% observed events over the 10-yr period, whereas the broken line
compared with 0; P<0.0001) and of stroke (median 0.4% shows the number of expected events.
compared with 0.1%; P<0.0001) in patients with SLE aged
under 40 than in age- and sex-matched comparators. The 10-yr
risks of CHD and stroke for older patients were not signicantly
different from those of the comparators. Characteristics of patients who suffered a
cardiovascular event
The characteristics of these patients are summarized in Table 4.
Longitudinal study: characteristics of patients Two of the nine patients who suffered CHD or stroke were aged
under 40 yr in 1991. Two were hypertensive, ve had raised
The characteristics of this group were very similar to those of cholesterol and one was a smoker in 1991. The predicted risk of
Group 1. The mean age of the 47 patients in Group 2 was 43 yr suffering the event that actually occurred in these patients was
(range 2467 yr). Forty-ve were female and two male. Nine higher than that of the comparator in ve cases, and lower in four
patients smoked. cases. The mean age of these nine individuals was 48.0 yr,
compared with 40.2 yr for the remaining 38 patients.
TABLE 2. Patient data for Group 1 The mean duration of SLE for the 47 patients in Group 2 was
7.3 yr. The mean disease duration was greater in the nine
Risk factor Mean (standard deviation) Range individuals who suffered a cardiovascular event (11.3 yr) than in
the remaining 38 (6.3 yr).
Age (yr) 42.2 (12.1) 1769 Fifteen of the 47 patients in Group 2 (32%) had either IgG or
Systolic blood pressure (mmHg) 123 (16) 91205 IgM anti-cardiolipin antibodies. Anti-cardiolipin antibodies were
Diastolic blood pressure (mmHg) 79 (9) 60102
Total cholesterol (mmol/l) 5.0 (1.1) 2.68.7
measured by enzyme-linked immunosorbent assay (ELISA)
HDL cholesterol (mmol/l) 1.6 (0.5) 0.53.5 (Shield Diagnostics, Dundee) [34]. Whilst only two of the ve
Number patients who suffered a stroke had anti-cardiolipin antibodies
History of smoking 45 (21.2%) (40%), these were present in three out of the four patients who had
Diabetic patients 2 (1.0%) a CHD event (75%). Lupus anticoagulant was present in eight
ECG changes of LVH 15 (7.4%) patients, only one of whom had a cardiovascular event.
History of CVD 13 (6.4%) Eleven of the 47 patients had evidence of lupus nephritis (grades
Hypertensivea 34 (16.8%) 3 to 5) on renal biopsy. Of these 11 patients three suffered a
Hypercholesterolaemiab 65 (32.2%)
cardiovascular event (all of whom had grade 4 nephritis).
Gender Female: 186 (92.1%)
Male: 16 (7.9%) The mean total cumulative corticosteroid dose at enrolment for
the nine individuals who suffered a cardiovascular event (19.7 g)
Hypertension was dened as a systolic blood pressure >140 mmHg was slightly greater than that for the remaining 38 subjects (16.0 g).
and/or a diastolic blood pressure >90 mmHg. Four of these nine individuals (44.4%) had been on hydroxy-
Hypercholesterolaemia was dened as a serum total cholesterol >5.2 chloroquine or chloroquine treatment at or prior to enrolment,
mmol/l. compared with 25 of the remaining 38 (65.8%).

TABLE 3. Median predicted 10-yr CHD and stroke risks for Group 1 (cross-sectional study)

Median CHD risk (%) (IQR) Median stroke risk (%) (IQR)

SLE patients Comparators SLE patients Comparators

All patients (n 202) 1.2 (0.14.2) 1.0 (04.0) P 0.22 0.8 (0.31.7) 1.0 (0.11.0) P<0.0001
Age 40 (n 93) 0.1 (00.6) 0 (00) P<0.0001 0.4 (0.20.6) 0.1 (0.10.1) P<0.0001
Age >40 (n 109) 3.5 (1.76.0) 4.0 (1.07.0) P 0.26 1.5 (0.82.7) 1.0 (0.12.0) P 0.18
Coronary heart disease and stroke risk in SLE 927

(a) Only one patient, who went on to have a CHD event, was using
1.0 a statin in 1991. Statins were introduced in three further patients
after they had suffered a cardiovascular event.
Proportion of patients event free

Due to the small sample size of Group 2 there were insufcient

0.95 data to make any statistically signicant conclusions regarding
the effects of lupus-specic risk factors, such as disease duration
and anti-cardiolipin antibodies, on the risk of suffering a
cardiovascular event.

The Cardio Risk Manager program predicts the reduction in the
risk of CHD or stroke that could be achieved in individual patients
0 12 24 36 48 60 72 84 96 108
by successfully reducing orthodox risk factors. Our analysis
suggests that managing these risk factors could reduce the
predicted risk of CHD and stroke signicantly in patients with
(b) SLE under the age of 40, but not in older patients.
1.0 SLE has been shown to result in the premature onset of both
stroke and CHD [12, 35, 36]. Bruce et al. [36] found that the mean
Proportion of patients event free

age at a rst coronary event was 49 yr in patients with SLE

compared with 6574 yr in the general population. Manzi et al.
0.95 [12] investigated cardiovascular disease in 498 women with SLE,
and found that the risk of MI relative to age-matched controls was
highest, at 52.4, in those aged 3544 yr. Two thirds of all
0.9 cardiovascular events in this cohort occurred in patients aged
less than 55 yr. Ward [35] also found the relative risk of MI,
congestive heart failure and stroke to be greatest in young patients
with SLE (aged 1844 yr).
The most likely explanation for the results obtained by age
stratication in the current study is that risk factors such as
0 12 24 36 48 60 72 84 96 108 120 hypertension and hypercholesterolaemia are relatively uncommon
Months in people under 40. The increased incidence of these risk factors in
FIG. 1. KaplanMeier plots illustrating the excess of observed young patients with SLE therefore makes a signicant difference to
CHD (a) and stroke (b) over the number of expected events their risk of CVD compared with other people of the same age. The
based on the risk predictions using data collected in 1991. The same is not true of patients over 40, because those risk factors are
solid line indicates the number of observed events over the 10-yr also fairly common in the general population of that age group
period, whereas the broken line shows the number of expected and this is reected in the values that the computer program uses
events. to generate the level of risk for the age-matched hypothetical
One limitation of our analysis is that the comparators are
assumed to be free of risk factors such as smoking, diabetes and
previous history of CVD. Therefore the predicted risk of the
Five of the nine individuals who suffered a cardiovascular event hypothetical comparators will be lower than an actual population
were using aspirin (55.6%), compared with only four of the of people without SLE. Given this fact, the predicted risks in our
remaining 38 (10.5%). One of the nine (11.1%) and four of the patients over 40 are surprisingly low. Long-term management in a
remaining 38 (10.5%) were anticoagulated with warfarin. hospital setting may have resulted in the recognition and treatment

TABLE 4. Clinical details at enrolment of patients who subsequently suffered a cardiovascular event

Patient Age Event Ethnic Blood Total ACL Renal Steroid Risk Predicted risk of
number (yr)a Sex type origin pressure cholesterol status diseaseb dose (g)c Hydroxychloroquined factors event (%)e

1 37 F CHD Caucasian 120/80 5.2 IgG 4 15.45 Hx CVD 0.5 (0)

2 46 F CHD Caucasian 115/75 6.0 19.8 Yes Hx smoking 7.7 (3)
3 37 F CHD Indian 135/85 7.9 IgG 4 18.3 2.0 (0)
4 54 F CHD Caucasian 120/80 6.1 IgM 2.1 Yes 4.1 (7)
5 45 F Stroke Caucasian 120/70 4.0 4 61.65 0.8 (1)
6 58 F Stroke Caucasian 125/75 3.1 10.86 Yes 1.9 (2)
7 65 F Stroke West Indian 155/85 5.0 15.66 Yes 5.8 (5)
8 42 F Stroke Indian 110/65 5.9 IgG/IgM 0.5 (1)
9 48 F Stroke Caucasian 155/95 5.4 IgG/IgM 13.89 Hx CVD 2.8 (1)
Age is age at enrolment in 1991.
Number indicates grade of lupus nephritis.
Total cumulative corticosteroid dose at enrolment.
Yes indicates usage at or prior to enrolment.
Predicted risk of event relates to the 10-yr risk of the actual event suffered by the subject, namely CHD or stroke. The patients risk is given,
followed by the comparators risk (in brackets).
Abbreviations: ACL status, anti-cardiolipin antibody status; Hx CVD/smoking, previous history of CVD/smoking.
928 R. Bessant et al.

of orthodox risk factors in a proportion of patients. It is clear, Even aggressive management of these potentially reversible
however, that many of our patients do have potentially reversible conventional risk factors, however, will not reduce the incidence of
risk factors for CVD. Twenty-two per cent of Group 1 smoked, CHD and stroke to that of subjects without SLE. It is important to
17% were hypertensive and 32% had raised cholesterol. continue the international effort to identify other risk factors,
In Group 2 consideration of orthodox risk factors alone did not which contribute to the high prevalence of CVD in patients
predict the signicantly increased risk of CHD and stroke that was with SLE.
actually observed in this cohort during the 10-yr follow-up period. Ethical approval and informed consent were not required
Our results in a British cohort of patients with SLE are consistent because assessment of risk factors for cardiovascular disease is
with those previously obtained in North America. part of the standard management of all patients in our cohort of
Esdaile et al. [18] carried out a retrospective study of 263 patients with SLE and is not a research procedure. We are
patients with SLE at two Canadian centres. After excluding presenting the risks of a whole cohort of patients without any
patients with a history of CVD pre-dating the baseline visit from identiable information about single patients.
the analysis, this study reported an increased relative risk of non-
fatal MI of 10.1, of death due to CHD of 17.0 and of stroke of 7.9
in patients with SLE [18]. Although conservative assumptions were
made, this study suffered from missing data on the presence of Acknowledgements
LVH in 26% and smoking history in 9% of patients. LVH, an The authors would like to thank Dr Gareth Ambler, of the
important independent factor associated with an increase in Medical Statistics Unit, R&D, University College London
cardiovascular mortality [39], occurs six times more frequently in Hospitals NHS Trust, for the invaluable advice regarding the
patients with SLE than in the general population [40], and may statistical analysis used in this manuscript.
amplify the global and regional myocardial ischaemia caused by
any degree of coronary artery disease. The authors have declared no conflicts of interest.
Rahman et al. [37] found that amongst patients with premature
cardiovascular disease, those with SLE had fewer and less severe
conventional risk factors (hypertension, hypercholesterolaemia,
diabetes mellitus, smoking status and family history) than those References
without SLE. However, this study used only total cholesterol 1. Abu-Shakra M, Urowitz MB, Gladman DD, Gough J. Mortality
measurements, and it is possible that altered individual cholesterol
studies in systemic lupus erythematosus. Results from a single centre.
fractions, such as HDL cholesterol, were not detected amongst the
I. Causes of death. J Rheumatol 1995;22:125964.
patients with SLE in this study [34, 41].
2. Jacobsen S, Petersen J, Ullman S et al. A multicentre study of 513
Our results support the hypothesis raised by these authors
Danish patients with systemic lupus erythematosus. II. Disease
and others [18, 37, 38] that there are additional risk factors
mortality and clinical factors of prognostic value. Clin Rheumatol
for increased rates of CVD in patients with SLE which are not
considered by the Framingham equations.
3. Uramoto K, Michet C, Thumboo J, Sunku J, OFallon W, Gabriel S.
A large prospective multicentre international study, coordinated
Trends in the incidence and mortality of systemic lupus erythema-
from Toronto, is currently attempting to identify these factors by
tosus, 19501992. Arthritis Rheum 1999;42:4650.
monitoring a cohort of patients with newly diagnosed SLE
4. Blanco FJ, Gomez-Reino JJ, de la Mata J et al. Survival analysis of
306 European Spanish patients with systemic lupus erythematosus.
Further possible risk factors include the presence of anti-
cardiolipin antibodies. Anti-cardiolipin antibodies were present Lupus 1998;7:15963.
in more than half of the nine patients from Group 2 who actually 5. Cervera R, Khamashta MA, Font J et al. Morbidity and mortality in
suffered a CVD event during the 10-yr follow-up period. The systemic lupus erythematosus during a 5-year period. A multicenter
number of patients involved is too small to permit further prospective study of 1,000 patients. European Working Party on
statistical analysis of such risk factors. Systemic Lupus Erythematosus. Medicine (Baltimore) 1999;
Other suggested risk factors for CVD in patients with SLE 78:16775.
include raised homocysteine levels, reduced physical activity due to 6. Stahl-Hallengren C, Jonsen A, Nived O, Sturfelt G. Incidence studies
fatigue, persistent inammation and impaired endothelial func- of systemic lupus erythematosus in Southern Sweden: increasing age,
tion. Impaired endothelium-dependent vasodilatation (EDD), decreasing frequency of renal manifestations and good prognosis. J
which was inversely correlated with total cholesterol levels, has Rheumatol 2000;27:68591.
been demonstrated in patients with SLE [42]. Endothelial function 7. Urowitz MB, Bookman AAM, Koehler BE, Gordon DA, Smythe
is also impaired in patients with primary systemic necrotizing HA, Ogryzlo MA. The bimodal mortality pattern of systemic lupus
vasculitis, but can be normalized following treatment of this erythematosus. Am J Med 1976;60:2215.
condition [43], suggesting that early suppression of disease activity 8. Rubin LA, Urowitz MB, Gladman DD. Mortality in systemic lupus
in chronic inammatory rheumatic disorders may reduce long- erythematosus: the bimodal pattern revisited. Q J Med 1985;55:8798.
term vascular damage. Another more recently identied risk factor 9. Petri M, Perez-Gutthann S, Spence D, Hochberg MC. Risk factors for
is the reduced activity of paraoxonase in patients with SLE [44]. coronary artery disease in patients with systemic lupus erythematosus.
The physiological role of paraoxonase is to prevent low-density Am J Med 1992;93:5139.
lipoprotein oxidation with its attendant atherogenic effects. 10. Urowitz MB, Abu-Shakra M, Gladman DD, Gough J, Farewell VT.
It is now accepted that cardiovascular disease is a common cause Improved survival in SLE. J Rheumatol 1997;24:10615.
of morbidity and mortality in patients with SLE. As a reduction in 11. Jonsson J, Nived O, Sturfelt G. Outcome in systemic lupus erythe-
the risk of CVD may be obtained by addressing the presence of matosus: a prospective study of patients from a dened population.
orthodox risk factors, especially in those aged under 40, we suggest Medicine 1989;68:14150.
that all patients with SLE should be screened for these factors 12. Manzi S, Meilahn EN, Rairie JE et al. Age specic incidence rates of
and offered appropriate intervention if indicated. Patients with myocardial infarction and angina in women with systemic lupus
hypercholesterolaemia, for example, may benet from a reduc- erythematosus: comparison with the Framingham study. Am J
tion in the dose of steroid where possible and the introduction of Epidemiol 1997;145:40815.
steroid-sparing agents, especially anti-malarials, which have a 13. Petri M, Civelek C. Discordance of myocardial SPECT and carotid
lipid-lowering effect. Dietary advice and lipid-lowering drugs duplex in systemic lupus erythematosus [abstract]. Arthritis Rheum
should be commenced early where indicated. 1998;41(Suppl):S218.
Coronary heart disease and stroke risk in SLE 929

14. Manzi S, Selzer F, Sutton-Tyrrell K et al. Prevalence and risk factors 31. Grover SA. Role of WHO-MONICA Project in unravelling of the
of carotid plaque in women with systemic lupus erythematosus. cardiovascular puzzle. Lancet 2000;355:6689.
Arthritis Rheum 1999;42:5160. 32. Grover SA, Coupal L, Hu XP. Identifying adults at increased risk
15. Bruce I, Burns R, Gladman D, Urowitz M. Single photon emission of coronary disease: How well do the current cholesterol guidelines
computed tomography dual isotope myocardial perfusion imaging work? J Am Med Assoc 1995;274:8016.
in women with systemic lupus erythematosus. I. Prevalence and 33. Haq IU, Ramsay LE, Yeo WW, Jackson PR, Wallis EJ. Is the
distribution of abnormalities. J Rheumatol 2000;27:23727. Framingham risk function valid for northern European populations?
16. Futrell N, Millikan C. Frequency, etiology, and prevention of stroke A comparison of methods for estimating absolute coronary risk in
in patients with systemic lupus erythematosus. Stroke 1989;20:58391. high risk men. Heart 1999;81:406.
17. Kitagawa Y, Gotoh F, Koto A, Okayasu H. Stroke in systemic lupus 34. MacGregor AJ, Dhillon VB, Binder A et al. Fasting lipids
erythematosus. Stroke 1990;21:15339. and anticardiolipin antibodies as risk factors for vascular
18. Esdaile JM, Abrahamowicz M, Grodzicky T et al. Traditional disease in systemic lupus erythematosus. Ann Rheum Dis 1992;51:
Framingham risk factors fail to fully account for accelerated 1525.
atherosclerosis in systemic lupus erythematosus. Arthritis Rheum 35. Ward MM. Premature morbidity from cardiovascular and cerebro-
2001;44:23317. vascular diseases in women with systemic lupus erythematosus.
19. Abrahamowicz, M, Grodzicky T, Senecal JL et al. Myocardial Arthritis Rheum 1999;42:33846.
infarction (MI) and stroke in SLE: markedly increased incidence after 36. Bruce IN, Gladman DD, Urowitz MB. Premature atherosclerosis in
controlling for risk factors. Lupus 1998;7(Suppl 1):69. systemic lupus erythematosus. Rheum Dis Clin North Am 2000;
20. Tan EM, Cohen AS, Fries JF et al. The 1982 revised criteria for the 26:25778.
classication of systemic lupus erythematosus. Arthritis Rheum 37. Rahman P, Urowitz MB, Gladman DD, Bruce IN, Genest J Jr.
1982;25:12717. Contribution of traditional risk factors to coronary artery disease in
21. Hingorani AD, Vallance P. A simple computer program for guiding patients with SLE. J Rheumatol 1999;26:23638.
management of cardiovascular risk factors and prescribing. Br Med J 38. Svenungsson E, Jensen-Urstad K, Heimburger M et al. Risk factors
1999;318:1015. for cardiovascular disease in systemic lupus erythematosus.
22. Anderson KM, Odell PM, Wilson PWF, Kannel WB. Cardiovascular Circulation 2001;104:188793.
disease risk proles. Am Heart J 1990;121:2938. 39. Levy D, Garrison RJ, Savage DD, Kannel WB, Castelli WP.
23. Anderson KM, Wilson PWF, Odell PM, Kannel WB. An updated Prognostic implications of echocardiographically determined left
coronary risk prole. A statement for health professionals. ventricular mass in the Framingham Heart Study. N Engl J Med
Circulation 1991;83:35662. 1990;322:15616.
24. Wolf PA, DAgostino RB, Blanger AJ, Kannel WB. Probability of 40. Roman MJ, Salmon JE, Sobel R et al. Prevalence and relation to risk
stroke: arisk prole from the Framingham study. Stroke 1991; factors of carotid atherosclerosis and left ventricular hypertrophy in
22:31218. systemic lupus erythematosus and antiphospholipid antibody syn-
25. Scandinavian Simvastatin Survival Study Group. Randomised trial of drome. Am J Cardiol 2001;87:6636, A11.
cholesterol lowering in 4444 patients with coronary heart disease: the 41. Borba EF, Bonfa E. Dyslipoproteinemias in systemic lupus erythema-
Scandinavian simvastatin survival study (4S). Lancet 1994;344: tosus: inuence of disease, activity, and anticardiolipin antibodies.
13839. Lupus 1997;6:5339.
26. Sacks FM, Pfeffer MA, Moye LA et al. The effect of pravastatin on 42. Piper MK, Heaton S, Gardner-Medwin J, Townend J, Bacon PA,
coronary events after myocardial infarction in patients with average Gordon C. A study of endothelial function in systemic lupus
cholesterol levels. N Engl J Med 1996;335:10019. erythematosus (SLE). Lupus 2001;10(Suppl 1):S25.
27. Shepherd J, Cobbe SM, Ford I et al. Prevention of coronary heart 43. Raza K, Thambyrajah J, Townend JN et al. Suppression of
disease with pravastatin in men with hypercholesterolaemia. N Engl J inammation in primary systemic vasculitis restores vascular endothe-
Med 1995;333:13017. lial function: lessons for atherosclerotic disease? Circulation 2000;
28. Haq U, Jackson PR, Yeo WW, Ramsay LE. Shefeld risk and 102:14702.
treatment table for cholesterol lowering for primary prevention of 44. Delgado Alves J, Ames P, Donohue S et al. Antibodies to high-density
coronary heart disease. Lancet 1995;346:146771. (Correction: Lancet lipoprotein and 2-glycoprotein 1 are inversely correlated with
1996;348:12512.) paraoxonase activity in systemic lupus erythematosus and primary
29. Ramsay LE, Haq IU, Jackson PR, Yeo WW, Pickin CM, Payne JN. antiphospholipid antibody syndrome. Arthritis Rheum 2002;
Targeting lipid-lowering drug therapy for primary prevention of 46:268694.
coronary heart disease: an updated Shefeld table. Lancet 1996; 45. Gladman DD, Urowitz MB. Morbidity in systemic lupus erythema-
348:3878. tosus. J Rheumatol 1987;14(Suppl 13):2236.
30. West of Scotland Coronary Prevention Study Group. Inuence of 46. Petri M. Detection of coronary artery disease and the role of
pravastatin and plasma lipids on clinical events in the West of Scotland traditional risk factors in the Hopkins Lupus Cohort. Lupus
coronary prevention study (WOSCOPS). Circulation 1998;97:14405. 2000;9:1705.