Cardiovascular Profile in Ankylosing Spondylitis: A Systematic Review and Meta-Analysis

Arthritis Care & Research Vol. 63, No. 4, April 2011, pp 557 563 DOI 10.1002/acr.20364 2011, American College of Rheumatology SPECIAL THEME ARTICLE: VASCULAR COMORBIDITY IN THE RHEUMATIC DISEASES Cardiovascular Profile in Ankylosing Spondylitis: A Systematic Review and Meta-Analysis SYLVAIN MATHIEU, 1 LAURE GOSSEC, 2 MAXIME DOUGADOS, 2 AND MARTIN SOUBRIER 1 Objective. Rheumatoid arthritis is associated with increased cardiovascular risk. In ankylosing spondylitis (AS), there is a paucity of information concerning this risk. Our objective was to assess the incidence of myocardial infarction (MI) or strokes and the cardiovascular risk profile in AS patients. Methods. We performed a systematic literature review using PubMed, EMBase, and the Cochrane Library up to August 2009. Incidence of MI or stroke was calculated by metaproportion. For cardiovascular risk factors, differences between AS patients and controls were expressed by standardized mean differences using inverse of variance method. Results. For MI, 8 longitudinal studies were included. In controls (n 82,745), 1,318 MI cases were observed (4.6%; 95% confidence interval [95% CI] 1.2%, 10.0%). In AS patients (n 3,279), 224 MI cases were reported (incidence 7.4%; 95% CI 5.2%, 10.0%). The increase in MI cases in AS patients was not significant (risk ratio 1.88; 95% CI 0.83, 4.28). For stroke, 7 longitudinal studies reported 327 strokes in AS patients (n 31,949), which is an incidence of 2.2% (95% CI 1.3%, 3.4%). In controls (n 7,372), one study reported 170 strokes (2.3%; 95% CI 2.0%, 2.7%). For cardiovascular risk factors, 15 case control studies and 9 abstracts were included (n 1,214 for patients and n 1,000 for controls). AS patients were characterized by a higher weighted mean intima-media thickness and higher risk of metabolic syndrome. In AS patients, there was a significant decrease in triglycerides, total cholesterol, and high-density lipoprotein (HDL) cholesterol. Conclusion. AS patients appear to be at higher risk of MI, which could be due to low HDL cholesterol levels or to systemic inflammation. Management of cardiovascular risk factors and control of systemic inflammation should be taken into account in AS. INTRODUCTION An increase in cardiovascular (CV) mortality and morbidity is reported in chronic inflammatory rheumatic diseases (1). In rheumatoid arthritis (RA), the existence of a high CV Supported by Abbott France. 1 Sylvain Mathieu, MD, Martin Soubrier, MD, PhD: Clermont 1 University, Centre Hospitalier Universitaire Gabriel Montpied, Clermont-Ferrand, France; 2 Laure Gossec, MD, PhD, Maxime Dougados, MD, PhD: Paris-Descartes University, Centre Hospitalier Universitaire Cochin, Paris, France. Dr. Mathieu has received consultant fees, speaking fees, and/or honoraria (less than $10,000) from Abbott France. Dr. Gossec has received consultant fees, speaking fees, and/or honoraria (less than $10,000 each) from Abbott, BMS, Chugai, Pfizer, Roche, Schering-Plough, and UCB. Dr. Dougados has received consultant fees, speaking fees, and/or honoraria (less than $10,000 each) from Abbott, Pfizer, Roche, and UCB. Dr. Soubrier has received consultant fees, speaking fees, and/or honoraria (less than $10,000 each) from Abbott, Roche, and Schering-Plough. Address correspondence to Sylvain Mathieu, MD, Service de Rhumatologie, Hôpital Gabriel Montpied, 58 Rue Montalembert 63003, Clermont-Ferrand, France. E-mail: smathieu @chu-clermontferrand.fr. Submitted for publication April 4, 2010; accepted in revised form September 17, 2010. risk is now well recognized (2 7). In ankylosing spondylitis (AS), an overall mortality of 1.6 1.9 times that of the general population has been reported, as well as an excess mortality rate varying from 20 40% for circulatory or CV disease (8 10). However, largely owing to the lack of evidence, opinion still differs on the association between atherosclerosis and AS (11). There are far fewer reports of high CV risk in AS than in RA and, in addition, the CV risk profile of AS patients differs from one study to another (1). Some authors reported that the prevalence of traditional risk factors was high in AS (12). AS patients have been described more often as smokers and with higher body mass index (BMI) than age-matched healthy controls. After adjusting for smoking status and BMI, these patients also had significantly increased C-reactive protein (CRP), interleukin-6, and fibrinogen levels. This suggests an unfavorable risk profile associated with AS related to systemic inflammation, which may predispose to CV disease (13,14). However, there are not enough studies to draw conclusions. The interest of meta-analysis in such a case is to increase statistical power and thereby increase accuracy (15,16). The aim of the present study was to have a more accurate assessment of the incidence of CV events (myocardial infarction [MI] and stroke) in AS and to compare the 557

558 Mathieu et al classic and newer CV risk profile of AS patients with that of healthy controls. MATERIALS AND METHODS Literature search. Medline PubMed, EMBase, and the Cochrane Central Register of Controlled Trials were searched to identify all reports of interest published up to August 2009. All observational studies monitoring MI or stroke and all case control studies considering classic (blood pressure, glycemia, metabolic syndrome, BMI, and lipid profile) and newer (intima-media thickness) CV risk factors in AS patients and healthy controls were included. The following search terms were used: spondylarthropathies [MeSH] AND ( cardiovascular [tiab] OR intima [tiab] OR lipid [tiab] OR endothelial [tiab]). Our search concerned only articles published in English or French. A hand search of references was also performed. Finally, we collected data from electronic abstract databases of the annual scientific meetings of the French Society of Rheumatology congress, the European League Against Rheumatism (EULAR) congress, and the American College of Rheumatology from 2007 2009 using the terms ( cardiovascular OR lipid OR intima OR myocardial infarction OR stroke AND ankylosing spondylitis ). Trial selection. One author (SM) selected potentially relevant articles after reading the title, keywords, and abstract. Inclusion criteria for the full text were: articles published in English or French before August 2009, observational and case control studies, study population of AS patients according to the New York classification criteria (17), data giving the number of MIs or strokes, or the CV risk factor profile. The exclusion criteria were: commentary or discussion papers, case reports and studies including 5 patients, no data about CV risk factors or CV diseases, no AS patients, full text not available, and data not usable for statistical analysis (no SD or no interquartile range [IQR]). Data extraction. One reviewer (SM) extracted all of the data using a standardized data abstraction form. Disease characteristics. For each article, we collected (when available) the age and sex of the patients, the disease duration, the visual analog scale for pain or for disease activity, the Bath Ankylosing Spondylitis Activity Index (BASDAI) score (18), the Bath Ankylosing Spondylitis Functional Index score (19), and the number of patients presenting with HLA B27. For all extracted data, a central value (mean or median) and variability (SD or IQR) were obtained. Extraction of the CV outcomes. In observational studies, we extracted the number of MIs and of strokes, fatal or not. We also recorded the duration of the followup of these AS patients or controls. In case control studies, we collected the recognized CV risk factors: systolic and diastolic blood pressure, smoking status, glycemia, lipid profile (total cholesterol, low-density lipoprotein [LDL] cholesterol, highdensity lipoprotein [HDL] cholesterol, triglycerides, and atherogenic index [total cholesterol/hdl cholesterol]), BMI, number of patients with metabolic syndrome, and homocysteinemia. Intima-media thickness, a marker of subclinical atherosclerosis, was also recorded (20,21). Statistical analysis. Continuous variables were expressed as the weighted mean SD. Incidences of MI and stroke or prevalence of metabolic syndrome were calculated by meta-analysis of proportions (inverse of variance method) and for 100 person-years of exposure. The Mantel-Haenszel procedure was used to determine the risk ratio (RR) of MI, stroke, and metabolic syndrome in AS patients versus healthy controls. This method provided a common RR estimate and 95% confidence interval (95% CI). For traditional CV risk factors and intima-media thickness (continuous variables), the differences between AS patients and controls were expressed by the standardized mean difference using inverse of variance method: 0.2 0.8 for moderate and 0.8 for large. Statistical heterogeneity between results was assessed using I 2, which is the most common metric for measuring the magnitude of betweenstudy heterogeneity and is easily interpretable. I 2 values range between 0% and 100% and are typically considered low for 25%, modest for 25 50%, and high for 50%. This statistical method generally assumed the presence of heterogeneity when the P value of the I 2 test was less than 0.05. Random-effects models were used in the case of heterogeneity; otherwise, we used a fixed-effects model. Statistical analysis was conducted using Review Manager Software, version 5.0 (Cochrane). RESULTS Eligible studies. Figure 1 shows the flow chart of publications identified by the literature search and those fi- Figure 1. Flow chart of manuscript selection. AS ankylosing spondylitis.

Risk of Myocardial Infarction and Stroke in AS Patients 559 patients were men, and 83.6% were positive for HLA B27. The included AS patients had an active disease characterized by a weighted mean SD BASDAI score of 4 (4.2 1.3) and systemic inflammation with a high mean SD erythrocyte sedimentation rate (ESR; 21.1 8.6 mm/hour) and a high mean SD CRP level (10.2 6.3 mg/dl). Figure 2. Comparison of myocardial infarction risk between ankylosing spondylitis (AS) patients and controls. M-H Mantel- Haenszel procedure; 95% CI 95% confidence interval. nally retained. A total of 921 citations were detained after research in the 3 databases. After reading the title, abstract, and full text, we obtained 16 eligible studies, plus 13 by hand searching and 9 by searching in abstract databases, for a total of 34,132 AS patients and 83,705 controls. Study characteristics. Of the 38 publications, 9 were abstracts, 15 were case control studies, and 14 were observational cohort studies. Eight longitudinal cohort studies assessed the occurrence of MI and 7 studies assessed that of stroke in AS patients. Of these, only 3 reported the occurrence of MI in both patients and controls and only 1 study reported that of stroke. The 30 other studies analyzed the CV risk profile in AS patients and sometimes in controls and gave information about the traditional and newer CV risk factors (see Supplementary Table 1, available in the online version of this article at http://online library.wiley.com/journal/10.1002/(issn)2151-4658). Characteristics of the included AS patients. Characteristics of the AS patients are given in Supplementary Table 2 (available in the online version of this article at http://onlinelibrary.wiley.com/journal/10.1002/(issn)2151-4658). The weighted mean SD age of the AS patients was 43.6 5.4 years, with a weighted mean SD disease duration of 11.3 8.9 years. Approximately 81% of AS CV events. MI. In the studies included, 224 MIs were reported in AS patients (n 3,279) over a mean followup period of 22 years (incidence 7.4% [95% CI 5.2%, 10.0%], i.e., 0.35/100 person-years). Three studies provided the numbers of MIs in the control group (n 82,745) and revealed 1,318 MIs (incidence 4.6% [95% CI 1.2%, 10.0%]). Meta-analysis of the 3 longitudinal studies comparing the occurrence of MI in AS patients (146 of 2,266) and healthy controls (1,318 of 82,745) showed an increase in MI in AS patients compared with controls, but not to a level of significance (RR 1.88; 95% CI 0.83, 4.28) (Figure 2). Stroke. In 7 longitudinal studies (n 31,949 patients), 327 strokes were reported in AS patients over 20.8 years of followup (incidence 2.2% [95% CI 1.3%, 3.4%], i.e., 0.07/100 person-years). One study alone reported 170 strokes in controls (n 7,372), an incidence of 2.3% (95% CI 2.0%, 2.7%). In this study, the increase in stroke occurrence in AS patients (72 of 1,843) compared with controls (170 of 7,372) was significant (risk difference 2%; 95% CI 1%, 3%). CV risk profile. In AS patients. Approximately onethird of the AS patients were smokers. The weighted mean SD BMI was 24.9 3.9 kg/m 2. These patients had no hypertension and normal systolic and diastolic blood pressure (weighted mean SD 120.4 12.7 mm Hg and 76.3 7.8 mm Hg, respectively). The weighted mean SD level of glycemia was normal (0.88 0.16 gm/liter), i.e., 1.1 gm/liter. Weighted mean SD levels of total cholesterol (1.80 0.14 gm/liter), LDL cholesterol (1.08 0.15 Table 1. Comparison of cardiovascular risk in case control studies* Characteristics Studies, no. AS patients, weighted mean SD Controls, weighted mean SD SMD (95% CI) P I 2,% Erythrocyte sedimentation rate, mm/hour 3 19.4 18.1 5.2 4.7 0.96 (0.51, 1.41) 0.001 65 C-reactive protein level, mg/dl 7 10.0 9.8 1.9 1.9 1.06 (0.79, 1.33) 0.001 53 Systolic blood pressure, mm Hg 10 120.4 12.8 119.1 13.1 0.08 ( 0.25, 0.40) 0.65 79 Diastolic blood pressure, mm Hg 10 76.3 7.8 74.4 10.5 0.13 ( 0.11, 0.37) 0.29 60 Glycemia, gm/liter 3 0.88 0.16 0.89 0.17 0.02 ( 0.36, 0.33) 0.92 74 Total cholesterol, gm/liter 10 1.77 0.36 1.93 0.36 0.32 ( 0.56, 0.07) 0.01 68 Low-density lipoprotein cholesterol, gm/liter 10 1.08 0.35 1.14 0.34 0.10 ( 0.23, 0.03) 0.14 36 High-density lipoprotein cholesterol, gm/liter 10 0.48 0.14 0.55 0.14 0.45 ( 0.78, 0.13) 0.007 82 Triglycerides, gm/liter 10 0.93 0.49 0.98 0.62 0.15 ( 0.28, 0.02) 0.03 4 Atherogenic index 3 3.69 1.14 3.38 1.16 0.35 ( 0.20, 0.89) 0.21 87 Homocysteinemia, mmoles/liter 3 11.8 9.9 10.4 3.3 0.11 ( 0.09, 0.31) 0.30 0 Body mass index, kg/m 2 11 25.0 4.0 24.8 3.8 0.03 ( 0.10, 0.15) 0.69 39 Intima-media thickness, mm 6 0.61 0.12 0.54 0.10 0.45 (0.12, 0.78) 0.008 66 *AS ankylosing spondylitis; SMD standardized mean difference; 95% CI 95% confidence interval. Fixed or random effects. Total cholesterol/high-density lipoprotein cholesterol.

560 Mathieu et al DISCUSSION Figure 3. Comparison of intima-media thickness between ankylosing spondylitis (AS) patients and controls. IV inverse of variance; 95% CI 95% confidence interval. gm/liter), and triglycerides (0.78 0.33 gm/liter) were normal, according to guidelines in France and the UK (22,23). The weighted mean SD level of HDL cholesterol (0.48 0.05 gm/liter) was decreased and 0.5 gm/liter, but the weighted mean SD atherogenic index (3.82 1.38) remained 4, which corresponds to a low risk of CV disease (22). Comparison of AS patients and controls. Fifteen case control studies and 9 abstracts were included (n 1,214 patients and n 1,000 controls). Table 1 shows the comparison between AS patients and controls concerning usual and new CV risk factors. The following were found in AS patients compared with healthy controls: a higherweighted mean SD intima-media thickness (0.61 0.12 versus 0.54 0.10 mm; P 0.008) (Figure 3), a lowerweighted mean SD level of triglycerides (0.93 0.49 versus 0.98 0.62 gm/liter; P 0.03), a lower-weighted mean SD cholesterol total (1.77 0.36 versus 1.93 0.36 gm/liter; P 0.01), a lower-weighted mean SD HDL cholesterol (0.48 0.14 versus 0.55 0.14 gm/liter; P 0.007), and an increased weighted mean SD atherogenic index without reaching statistical significance (3.69 1.14 versus 3.38 1.16; P 0.21). The risk of metabolic syndrome was higher in AS patients (52 of 150 versus 44 of 271 control patients; RR 2.13 [95% 1.48, 3.06], P 0.001) but without differences in glycemia levels or in BMI (Figure 4). There was also no difference in weighted mean systolic and diastolic blood pressure, or in levels of homocysteinemia and of LDL cholesterol (Table 1). As expected, there was an increase in the parameters of systemic inflammation in AS patients: weighted mean SD ESR (19.4 18.1 versus 5.2 4.7 mm/hour; P 0.001) and weighted mean CRP level (10.0 9.8 versus 1.9 1.9 mg/dl; P 0.001). Figure 4. Comparison of prevalence of metabolic syndrome between ankylosing spondylitis (AS) patients and controls. M-H Mantel-Haenszel procedure; 95% CI 95% confidence interval. In this meta-analysis, a higher risk of MI was found in AS patients compared with controls (RR 1.38; 95% CI 1.16, 1.65). A significant increase in intima-media thickness, a marker of subclinical atherosclerosis, was also found in AS. AS patients had a higher frequency of metabolic syndrome, lower triglyceride levels, and lower total cholesterol and lower HDL. As expected, a significant rise in ESR and CRP level was observed in AS patients. The presence of increased CV risk in RA is now clearly recognized. In 2008, Lévy et al reported that the risk of myocardial ischemia in RA patients was 1.63 higher compared with the general population (odds ratio [OR] 1.63; 95% CI 1.34, 2.00) (24). In AS, opinions still differ on the association with atherosclerosis. This meta-analysis gives some supplementary indication of increased CV risk in AS. A higher carotid intima-media thickness was found in AS patients. Gonzalez-Juanatey et al recently arrived at the same conclusion in a survey comparing 68 patients and 68 healthy controls (25). Likewise, Peters et al recently reported that the overall prevalence for MI was 4.4% in patients with AS versus 1.2% in the general population, resulting in an age- and sex-adjusted OR of 3.1 (95% CI 1.9, 5.1) for patients with AS (11). The reasons for this increased CV risk are not fully understood. Biologic inflammation seems to play a pivotal role, as attested by the higher ESR and higher CRP level found in AS patients in this study. The increased CV risk in AS may be secondary to systemic inflammation, which plays an important role in all phases of atherosclerosis from its onset to thrombosis, and could then potentialize the effect of traditional cardiac risk factors (26). In the general population, epidemiologic studies have shown that CRP level is an independent predictor of CV events (27). However, the AS patients also had a worse CV profile with a higher frequency of metabolic syndrome and a decrease in HDL cholesterol level, and the latter seems to be an important cardiac risk factor. A decrease in HDL cholesterol could also be partly responsible for the occurrence of coronary events (28). Conversely, an increase in HDL cholesterol of 1% is accompanied by a decrease in cardiac events of 2% (29). The mechanisms by which HDL cholesterol protects against atherosclerosis are numerous. In addition to its role in the process of reverse cholesterol transport (from the atheroma plaque to the liver), HDL cholesterol has antioxidant (reducing oxidation of LDL, making them less atherogenic) and antiinflammatory properties (inhibition of the expression of adhesion molecules on endothelial cells) (29). The decrease in HDL cholesterol is correlated with the intensity of inflammation (12). Therefore, an increase of 30 mm/hour in ESR is associated with decreased total cholesterol of 6% and HDL cholesterol of 11%. Likewise, an increase of 30 mg/dl of CRP level is associated with a decrease of total cholesterol of 3.8% and HDL cholesterol of 5.7%. This decrease in HDL cholesterol is also accompanied by functional changes, such as the replacement of apolipoprotein A1, which is

Risk of Myocardial Infarction and Stroke in AS Patients 561 responsible for the antiinflammatory properties, with serum amyloid A in AS patients with high CRP levels (30). This study has several limitations. First, only 3 studies have assessed the incidence of MI both in AS patients and in controls. Furthermore, the sample size of these studies and their design are different, which has an effect on results interpretation and should discourage the drawing of definitive conclusions. However, these few studies, which represent the followup of 2,266 AS patients and of 82,745 controls, are the only source of information available in the literature. Second, an important limitation is related to the principle of meta-analysis: that the reader has only the data available in the published articles. In this meta-analysis comparing MI incidence between AS patients and controls, a major question is the comparability of these groups, particularly in age and sex (but also in other CV disease risk factors), since these can be strongly associated with CV events and may differ between patients with AS and controls. Because of the lack of this information, we have no sense of the extent to which confounders such as age, sex, and other CV risks may be present and have a role in the increased risk reported for AS patients. Third, we did not assess the methodologic quality of the included trials. We initially decided to include only prospective longitudinal studies, i.e., the ideal means of assessing the incidence of CV events. Because of the low number of these studies, we had to include cross-sectional studies, case control studies, and even congress abstracts. However, this limitation is necessary in meta-analyses, which aim to obtain a larger number of patients and to improve the evaluation of whether statistically reliable differences exist between comparison groups. Large prospective studies addressing the same clinical question of CV risk in AS are needed to confirm our results. We found a significantly higher proportion of metabolic syndrome in AS than in controls, but no differences in the parameters defining metabolic syndrome (i.e., blood pressure, glycemia, level of triglycerides). These results are surprising but could be explained by the methodology of the meta-analysis, which assesses only the available data. In our analysis, studies assessing blood pressure, glycemia, and lipid profile in AS patients give no information about the metabolic syndrome. Another possible explanation could be publication bias. Indeed, the chance of an article being published can depend on whether the results are positive or negative (31,32). Chan et al reported that articles with statistically significant efficacy outcomes had more than 2-fold greater odds of being fully reported as compared with those with nonsignificant efficacy outcomes (33). As well, authors may be less likely to submit manuscripts reporting negative results (34). It is likely that only studies reporting a significant difference in metabolic syndrome between AS patients and controls were published. In meta-analysis, another bias is to pool data about an outcome at different moments with perhaps a different definition of the outcome. In our meta-analysis, the definition of the metabolic syndrome in 3 studies derived from the updated Third Report of the National Cholesterol Education Program s Adult Treatment Panel (23). In conclusion, AS patients appear to be at higher risk of MI compared with controls. This could be due to an atherogenic lipid profile or to systemic inflammation. As recently reported in the EULAR evidence-based recommendations (2), management of CV risk factors and control of systemic inflammation should be taken into account in AS. ACKNOWLEDGMENT The authors wish to gratefully acknowledge the help of Mme. Guillemette Utard, Librarian (BUIM, Paris Descartes University), for the literature search. AUTHOR CONTRIBUTIONS All authors were involved in drafting the article or revising it critically for important intellectual content, and all authors approved the final version to be submitted for publication. Dr. Mathieu had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis. Study conception and design. Mathieu, Gossec, Dougados, Soubrier. Acquisition of data. Mathieu, Dougados, Soubrier. Analysis and interpretation of data. Mathieu, Gossec, Dougados, Soubrier. ROLE OF THE STUDY SPONSOR Abbott France Pharmaceutical Company provided a grant to organize a systematic literature review methods workshop, but played no further role in the project. Abbott France had no direct role in the study design, data collection, analysis, and interpretation of the data, writing of the manuscript, approval of manuscript content, or in the decision to publish this work. Neither the submission nor the publication of this article was contingent on the approval of Abbott France. REFERENCES 1. Choy E, Sattar N. Interpreting lipid levels in the context of high-grade inflammatory states with a focus on rheumatoid arthritis: a challenge to conventional cardiovascular risk actions. Ann Rheum Dis 2009;68:460 9. 2. Peters MJ, Symmons DP, McCarey D, Dijkmans BA, Nicola P, Kvien TK, et al. EULAR evidence-based recommendations for cardiovascular risk management in patients with rheumatoid arthritis and other forms of inflammatory arthritis. Ann Rheum Dis 2010;69:325 31. 3. Lindqvist E, Eberhardt K. Mortality in rheumatoid arthritis patients with disease onset in the 1980s. Ann Rheum Dis 1999;58:11 4. 4. Del Rincon I, Williams K, Stern MP, Freeman GL, Escalante A. High incidence of cardiovascular events in a rheumatoid arthritis cohort not explained by traditional cardiac risk factors. Arthritis Rheum 2001;44:2737 45. 5. Turesson C, Jarenros A, Jacobsson L. Increased incidence of cardiovascular disease in patients with rheumatoid arthritis: results from a community based study. Ann Rheum Dis 2004; 63:952 5. 6. Maradit-Kremers H, Crowson CS, Nicola PJ, Ballman KV, Roger VL, Jacobsen SJ, et al. Increased unrecognized coronary heart disease and sudden deaths in rheumatoid arthritis. Arthritis Rheum 2005;52:402 11. 7. Maradit-Kremers H, Nicola PJ, Crowson CS, Ballman KV, Gabriel SE. Cardiovascular death in rheumatoid arthritis: a population-based study. Arthritis Rheum 2005;52:722 32. 8. Lautermann D. Ankylosing spondylitis-cardiac manifestations. Clin Exp Rheumatol 2002;20 Suppl 28:S11 5. 9. Lehtinen K. Mortality and causes of death in 398 patients

562 Mathieu et al admitted to hospital with ankylosing spondylitis. Ann Rheum Dis 1993;52:174 6. 10. Kaprove RE, Little AH, Graham DC, Rosen PS. Ankylosing spondylitis: survival in men with and without radiotherapy. Arthritis Rheum 1980;23:57 61. 11. Peters MJ, Visman I, Nielen MM, van Dillen N, Verheij RA, van der Horst-Bruinsma IE, et al. Ankylosing spondylitis: a risk factor for myocardial infarction? Ann Rheum Dis 2010; 69:579 81. 12. Divecha H, Sattar N, Rumley A, Cherry L, Lowe G, Sturrock R. Cardiovascular risk parameters in men with ankylosing spondylitis in comparison with non-inflammatory control subjects: relevance of systemic inflammation. Clin Sci (Lond) 2005;109:171 6. 13. Van Halm VP, van Denderen JC, Peters MJ, Twisk JW, van der Paardt M, van der Horst-Bruinsma IE, et al. Increased disease activity is associated with a deteriorated lipid profile in patients with ankylosing spondylitis. Ann Rheum Dis 2006;65: 1473 7. 14. Ross R. Atherosclerosis: an inflammatory disease. N Engl J Med 1999;340:115 26. 15. Egger M, Smith DG, Altman D. Systematic reviews in health care: meta-analysis in context. London: BMJ Books; 2001. 16. Moore RA, Gavaghan D, Tramer MR, Collins SL, McQuay HJ. Size is everything: large amounts of information are needed to overcome random effects in estimating direction and magnitude of treatment effects. Pain 1998;78:209 16. 17. Van der Linden S, Valkenburg HA, Cats A. Evaluation of diagnostic criteria for ankylosing spondylitis: a proposal for modification of the New York criteria. Arthritis Rheum 1984; 27:361 8. 18. Garrett S, Jenkinson T, Kennedy LG, Whitelock H, Gaisford P, Calin A. A new approach to defining disease status in ankylosing spondylitis: the Bath Ankylosing Spondylitis Disease Activity Index. J Rheumatol 1994;21:2286 91. 19. Calin A, Garrett S, Whitelock H, Kennedy LG, O Hea J, Mallorie P, et al. A new approach to defining functional ability in ankylosing spondylitis: the development of the Bath Ankylosing Spondylitis Functional Index. J Rheumatol 1994;21: 2281 5. 20. Boutouyrie P, Tropeano AI, Asmar R, Gautier I, Benetos A, Lacolley P, et al. Aortic stiffness is an independent predictor of primary coronary events in hypertensive patients: a longitudinal study. Hypertension 2002;39:10 5. 21. O Leary DH, Polak JF, Kronmal RA, Manolio TA, Burke GL, Wolfson SK Jr, et al. Carotid-artery intima and media thickness as a risk factor for myocardial infarction and stroke in older adults. N Engl J Med 1999;340:14 22. 22. French Health Products Safety Agency. French recommendations for the management of dyslipidemia in France. Saint- Denis (France): AFSSAPS; 2005. 23. National Cholesterol Education Program (NCEP) Expert Panel on Detection, Evaluation, and Treatment of High Blood Cholesterol in Adults (Adult Treatment Panel III). 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). Circulation 2002;106:3143. 24. Levy L, Fautrel B, Barnetche T, Schaeverbeke T. Incidence and risk of fatal myocardial infarction and stroke events in rheumatoid arthritis patients: a systematic review of the literature. Clin Exp Rheumatol 2008;26:389 95. 25. Gonzalez-Juanatey C, Vazquez-Rodriguez TR, Miranda-Filloy JA, Dierssen T, Vaqueiro I, Blanco R, et al. The high prevalence of subclinical atherosclerosis in patients with ankylosing spondylitis without clinically evident cardiovascular disease. Medicine (Baltimore) 2009;88:358 65. 26. Sattar N, McCarrey DW, Capell H, McInnes IB. Explaining how high-grade systemic inflammation accelerates vascular risk in rheumatoid arthritis. Circulation 2003;108:2957 63. 27. Buckley DI, Fu R, Freeman M, Rogers K, Helfand M. C-reactive protein as a risk factor for coronary heart disease: a systematic review and meta-analyses for the U.S. Preventive Services Task Force. Ann Intern Med 2009;151:483 95. 28. Gordon DJ, Rifkind BM. High-density lipoprotein: the clinical implications of recent studies. N Engl J Med 1989;321: 1311 6. 29. Davidson MH, Toth PP. High-density lipoprotein metabolism: potential therapeutic targets. Am J Cardiol 2007;100:32 40. 30. Van Eijk IC, Peters MJ, Serne EH, van der Horst-Bruinsma IE, Dijkmans BA, Smulders YM, et al. Microvascular function is impaired in ankylosing spondylitis and improves after tumour necrosis factor blockade. Ann Rheum Dis 2009;68: 362 6. 31. Dickersin K. How important is publication bias? A synthesis of available data. AIDS Educ Prev 1997;9 Suppl:15 21. 32. Johnson RT, Dickersin K. Publication bias against negative results from clinical trials: three of the seven deadly sins [letter]. Nat Clin Pract Neurol 2007;3:590 1. 33. Chan AW, Krleza-Jeric K, Schmid I, Altman DG. Outcome reporting bias in randomized trials funded by the Canadian Institutes of Health Research. CMAJ 2004;171:735 40. 34. Olson CM, Rennie D, Cook D, Dickersin K, Flanagin A, Hogan JW, et al. Publication bias in editorial decision making. JAMA 2002;287:2825 8. 35. Spanakis E, Sidiropoulos P, Papadakis J, Ganotakis E, Katsikas G, Karvounaris S, et al. Modest but sustained increase of serum high density lipoprotein cholesterol levels in patients with inflammatory arthritides treated with infliximab. J Rheumatol 2006;33:2440 6. 36. Sari I, Okan T, Akar H, Cece H, Altay C, Secil M, et al. Impaired endothelial function in patients with ankylosing spondylitis. Rheumatology (Oxford) 2006;45:283 6. 37. Kang E, Choi S, Lee S, Park Y. Atherogenic lipid profile in ankylosing spondylitis [abstract]. Ann Rheum Dis 2007;66 Suppl 2:399. 38. Okan T, Sari I, Akar S, Cece H, Goldeli O, Guneri S, et al. Ventricular diastolic function of ankylosing spondylitis patients by using conventional pulsed wave Doppler, myocardial performance index and tissue Doppler imaging. Echocardiography 2008;25:47 56. 39. Mathieu S, Joly H, Baron G, Tournadre A, Dubost JJ, Ristori JM, et al. Trend towards increased arterial stiffness or intimamedia thickness in ankylosing spondylitis patients without clinically evident cardiovascular disease. Rheumatology (Oxford) 2008;47:1203 7. 40. Smith PG, Doll R. Mortality among patients with ankylosing spondylitis after a single treatment course with X-rays. Br Med J 1982;13:449 60. 41. Radford EP, Doll R, Smith PG. Mortality among patients with ankylosing spondylitis not given X-ray therapy. N Engl J Med 1977;11:572 6. 42. Han C, Robinson DW Jr, Hackett MV, Paramore LC, Fraeman KH, Bala MV. Cardiovascular disease and risk factors in patients with rheumatoid arthritis, psoriatic arthritis, and ankylosing spondylitis. J Rheumatol 2006;33:2167 72. 43. Karoli N, Rebrov A. Arterial stiffness in patients with ankylosing spondylitis [abstract]. Ann Rheum Dis 2009;68 Suppl 3:645. 44. Choi ST, Kim JH, Kang EJ, Lee SW, Park MC, Park YB, et al. Osteopontin might be involved in bone remodelling rather than in inflammation in ankylosing spondylitis. Rheumatology (Oxford) 2008;47:1775 9. 45. Choe JY, Lee MY, Rheem I, Rhee MY, Park SH, Kim SK. No differences of carotid intima-media thickness between young patients with ankylosing spondylitis and healthy controls. Joint Bone Spine 2008;75:548 53. 46. Ariza-Ariza R, Hernandez-Cruz B, Lopez-Antequera G, Navarro-Sarabia F. Frequency of atherosclerosis in patients with ankylosing spondylitis: a case-control study [abstract]. Ann Rheum Dis 2009;68 Suppl 3:646. 47. Khan MA, Khan MK, Kushner I. Survival among patients with ankylosing spondylitis: a life-table analysis. J Rheumatol 1981;8:86 90. 48. Baskan BM, Sivas F, Aktktin LA, Dogan YP, Ozoran K, Bodur H. Serum homocysteine level in patients with ankylosing spondylitis. Rheumatol Int 2009;29:1435 9. 49. Malesci D, Niglio A, Mennillo GA, Buono R, Valentini G,

Risk of Myocardial Infarction and Stroke in AS Patients 563 La Montagna G. High prevalence of metabolic syndrome in patients with ankylosing spondylitis. Clin Rheumatol 2007; 26:710 4. 50. Calistan M, Erdogan D, Gullu H, Yilmaz S, Gursoy Y, Yildirir A, et al. Impaired coronary microvascular and left ventricular diastolic functions in patients with ankylosing spondylitis. Atherosclerosis 2008;196:306 12. 51. Sari I, Kebapcilar L, Alacacioglu A, Bilgir O, Yildiz Y, Taylan A, et al. Increased levels of asymmetric dimethylarginine (ADMA) in patients with ankylosing spondylitis. Intern Med 2009;48:1363 8. 52. Van Denderen JC, Peters MJ, van Halm VP, van der Horst- Bruinsma IE, Dijkmans BA, Nurmohamed MT. Statin therapy might be beneficial for patients with ankylosing spondylitis [letter]. Ann Rheum Dis 2006;65:695 6. 53. Van Eijk IC, de Vries MK, Levels JH, Peters MJ, Huizer EE, Dijkmans BA, et al. Improvement of lipid profile is accompanied by atheroprotective alterations in high-density lipoprotein composition upon tumor necrosis factor blockade: a prospective cohort study in ankylosing spondylitis. Arthritis Rheum 2009;60:1324 30. 54. Demiralp E, Kardesoglu E, Kiralp MZ, Cebeci BS, Keskin I, Ozmen N, et al. Aortic elasticity in patients with ankylosing spondylitis. Acta Cardiol 2004;59:630 4. 55. Peters MJ, van Eijk IC, Smulders YM, Serne EH, van der Horst-Bruinsma IE, Dijkmans BA, et al. Accelerated atherosclerosis in ankylosing spondylitis [abstract]. Ann Rheum Dis 2009;68 Suppl 3:159. 56. Sukenik S, Pras A, Buskila D, Katz A, Snir Y, Horowitz J. Cardiovascular manifestations of ankylosing spondylitis. Clin Rheumatol 1987;6:588 92. 57. Papadakis JA, Sidiropoulos PI, Karvounaris SA, Vrentzos GE, Spanakis EK, Ganotakis ES, et al. High prevalence of metabolic syndrome and cardiovascular risk factors in men with ankylosing spondylitis on anti-tnf treatment: correlation with disease activity. Clin Exp Rheumatol 2009;27:292 8. 58. Court-Brown WM, Doll R. Mortality from cancer and other causes after radiotherapy for ankylosing spondylitis. Br Med J 1965;2:1327 32. 59. Wilkinson M, Bywaters EG. Clinical features and course of ankylosing spondylitis; as seen in a follow-up of 222 hospital referred cases. Ann Rheum Dis 1958;17:209 28. 60. Huizer EE, van Eijk IC, Peters MJ, de Koning MH, van de Stadt RJ, van der Horst-Bruinsma IE, et al. The effect of etanercept treatment on the lipid profile in patients with ankylosing spondylitis [abstract]. Ann Rheum Dis 2007;66 Suppl 2:398. 61. Gonzalez-Lopez L, Sanchez-Hernandez JD, Aguilar-Chavez EA, Cota-Sanchez AR, Lopez-Olivo MA, Villa-Manzano AI, et al. Hyperhomocysteinemia in ankylosing spondylitis: prevalence and association with clinical variables. Rheumatol Int 2008;28:1223 8. 62. Lehtinen K. Cause of death in 79 patients with ankylosing spondylitis. Scand J Rheumatol 1980;9:145 7. 63. Poddubnyy DA, Nikitina NM, Gaydukova IZ, Rebrov AP. Relationship between arterial stiffness parameters, cardiovascular risk factors and activity of systemic inflammation in patients with ankylosing spondylitis [abstract]. Ann Rheum Dis 2009;68 Suppl 3:638. 64. Karvounaris SA, Sidiropoulos PI, Spanakis EK, Papadakis JA, Lanara SD, Dimaka A, et al. High prevalence of metabolic syndrome and cardiovascular risk factors in men with ankylosing spondylitis [abstract]. Ann Rheum Dis 2007;66 Suppl 2:401. 65. Poddubnyy DA, Rebrov AP. Conventional and novel cardiovascular risk factors in patients with ankylosing spondylitis [abstract]. Ann Rheum Dis 2007;66 Suppl 2:407. 66. Unverdi S, Goker B, Oktar S, Ozturk M, Haznedaroglu S, Ureten K, et al. Assessment of the morphologic evidence of subclinical atherosclerosis by carotid B-mode ultrasonography in patients with ankylosing spondylitis [abstract]. Ann Rheum Dis 2007;66 Suppl 2:411.