Article Text

Download PDFPDF

Extended report
Cardiovascular risk of patients with gout seen at rheumatology clinics following a structured assessment
  1. Mariano Andrés1,2,
  2. José Antonio Bernal1,
  3. Francisca Sivera3,
  4. Neus Quilis3,
  5. Loreto Carmona4,
  6. Paloma Vela1,2,
  7. Eliseo Pascual1,2
  1. 1 Sección de Reumatología, Hospital General Universitario de Alicante, Alicante, Spain
  2. 2 Departamento de Medicina Clínica, Universidad Miguel Hernández, Alicante, Spain
  3. 3 Sección de Reumatología, Hospital General Universitario de Elda, Alicante, Spain
  4. 4 Instituto de Salud Musculoesquelética, Madrid, Spain
  1. Correspondence to Dr Mariano Andrés, Sección de Reumatología, Hospital General Universitario de Alicante, C/Pintor Baeza 12, Alicante 03010, Spain; drmarianoandres{at}


Objectives Gout-associated cardiovascular (CV) risk relates to comorbidities and crystal-led inflammation. The aim was to estimate the CV risk by prediction tools in new patients with gout and to assess whether ultrasonographic carotid changes are present in patients without high CV risk.

Methods Cross-sectional study. Consecutive new patients with crystal-proven gout underwent a structured CV consultation, including CV events, risk factors and two risk prediction tools—the Systematic COronary Evaluation (SCORE) and the Framingham Heart Study (FHS). CV risk was stratified according to current European guidelines. Carotid ultrasound (cUS) was performed in patients with less than very high CV risk. The presence of carotid plaques was studied depending on the SCORE and FHS by the area under the curve (AUC) of receiver operating curves.

Results 237 new patients with gout were recruited. CV stratification by scores showed a predominance of very high (95 patients, 40.1%) and moderate (72 patients, 30.5%) risk levels. cUS was performed in 142 patients, finding atheroma plaques in 66 (46.5%, 95% CI 37.8 to 54.2). Following cUS findings, patients classified as very high risk increased from 40.1% up to 67.9% (161/237 patients). SCORE and FHS predicted moderately (AUC 0.711 and 0.683, respectively) the presence of atheroma plaques at cUS.

Conclusions The majority of patients presenting with gout may be at very high CV risk, indicating the need for initiating optimal prevention strategies at this stage. Risk prediction tools appear to underestimate the presence of carotid plaque in patients with gout.

  • Gout
  • Cardiovascular Disease
  • Atherosclerosis
  • Ultrasonography

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.


Until very recently, the presence of the monosodium urate (MSU) crystal deposit was considered to be innocuous—apart from causing the severe symptoms that characterise gout. In fact, the recommendations suggested not to treat at the time of the diagnosis but to wait until there is disease progression.1 Now, the risk of cardiovascular (CV) disease in gout is well established2 ,3 and likely relates to persistent inflammation.4 The new European League Against Rheumatism recommendations indicate the expedience of treating from the time of diagnosis to avoid further gout attacks and growing crystal load and to possibly prevent CV events.5 We recently found that patients with asymptomatic hyperuricemia with silent MSU deposits suffered from a more severe form of coronary atherosclerosis.6 The possibility that the CV risk in patients with gout may be higher than what is shown by the standard evaluation scores remains an open question, and the rationale for early and effective treatment for gout largely depends on its CV consequences.

The aim of this investigation was to establish the CV risk in patients with gout at the time of their first visit to rheumatology, following a structured assessment. As secondary end points, we analysed the discriminative value of CV risk prediction tools to identify individuals with carotid atheroma plaques, and assessed the association between gout features and the presence of carotid atherosclerosis.


This was an observational, cross-sectional study, performed in a hospital-based rheumatology unit with a population coverage of 300 000 inhabitants. The study was approved by the local ethics committee and written informed consent was obtained from participants. Consecutive new patients with crystal-proven gout seen in the unit between 1 January 2014 and 30 June 2016 were screened for enrolment. Source of patients was registered. Patients were excluded if they rejected participation or had been previously treated by a rheumatologist or other medical specialist; previous management by a general practitioner was allowed.

Enrolled participants underwent a structured CV consultation through interview, physical exam and records review regarding conventional risk factors (CRF), CV disease, blood pressure, anthropometry and lab variables (see full list in online supplementary table S1). Gout-related variables (see online supplementary table S1) were also registered.

Risk prediction tools

The Systematic COronary Evaluation (SCORE)7 and the Framingham Heart Study (FHS),8 adapted for Spanish population,9 were applied to all patients. Patients with a very high risk at enrolment because of prior CV disease, diabetes mellitus (DM) with complications or severe renal failure did not have a score calculated.

Carotid ultrasound

It was performed in patients not stratified at very high risk (prior CV disease, DM with complications, severe renal failure or SCORE >9%) by a trained rheumatologist following the Mannheim consensus.10 Both common carotid arteries and their proximal branches were assessed to detect atheroma plaques and to measure intima-media thickness (IMT).

CV risk stratification

This was the primary study outcome. CV risk was stratified according to 2011 European guidelines11 as low (SCORE <1%); moderate (SCORE 1%–4%); high (uncomplicated DM, estimated glomerular filtration rate (eGFR) 30–59 mL/min, IMT >0.9 mm or SCORE 5%–9%) and very high risk (prior CV disease, carotid atheroma plaques, DM with complications, eGFR <30 mL/min, SCORE >9%). CV risk stratification was performed before and after carotid ultrasound (cUS): patients with increased IMT (IMT >0.9 mm) were classified as high risk and those with atheroma plaques as very high risk.11 This allowed to assess the impact of evaluating subclinical atherosclerosis in the CV risk assessment of patients with gout.

Statistical analysis

A previous study12 found a prevalence of 30% of carotid plaques in patients with gout, but in a pilot phase of the present investigation this rate reached up to 40%. For a statistical power of 80%, a minimal sample size of 78 patients was estimated to reliably assess the prevalence of carotid atheroma plaques in recently diagnosed patients with gout. A descriptive analysis was done using central tendency measures (mean or median) and dispersion measures (SD or 25–75 percentiles) for continuous variables, and frequencies for categorical variables. Prevalence with 95% CIs was estimated for IMT and carotid plaques. A subgroup comparison was performed according to whether patients were evaluated by cUS or not, using χ2 test for categorical variables and Student's t-test and Mann-Whitney U test for continuous variables. The χ2 trend test was used to compare the CV risk stratification before and after cUS.

To assess the discriminative value of SCORE and FHS tools to detect individuals with carotid plaque, 2×2 tables were built for each tool: in rows, tools scores leading to high or very high risk (SCORE >4%, FHS >9%) were taken as positive, and scores leading to low or moderate risks (SCORE ≤4%, FHS ≤9%) as negative; in columns, presence or absence of atheroma plaques at cUS was considered as a positive or negative result, respectively. Sensitivity, specificity and positive and negative predictive values, with 95% CI, were then calculated. Afterwards, receiver operating characteristic (ROC) curves were plotted, allowing an estimation of the area under the curve (AUC) with 95% CI for each tool.

In order to assess the association between gout features and the cUS atherosclerotic findings (increased IMT or atheroma plaques), logistic regression models were used.

All analyses were performed using SPSS V.19.0 software (IBM Statistics, Armonk, New York, USA). Statistical significance was established as p<0.050.


Two hundred and fifty patients were screened for enrolment, but 13 declined to participate, so a total of 237 participants were finally included (participation rate 94.8%). Over half of patients were referred from emergency department (see online supplementary table S2).

In the study population, middle-aged, obese males predominated (table 1). A total of 216 patients (92.3%) had at least one CV risk factor, mainly hypertension or smoking. More than 40% of participants were on diuretics and about a 30% suffered from chronic kidney disease (eGFR <60 mL/min), which was severe in 19 cases (8.0%); 29.2% had previously suffered from CV disease, especially from myocardial infarction and peripheral artery disease, while 14 patients (5.9%) had various forms of CV disease.

Table 1

General characteristics of included patients

Regarding gout characteristics, a total of 30 patients (14.2%) were seen at first gout attack (early gout), while median gout duration was of 4 years (p25–75 1–10, range 0–50). Participants reported a median of 4 attacks (p25–75 2–12, range 1–400) that had involved an average of 3.6 joints (SD±2.8). Presenting attack was monoarticular in 177 cases (75.3%), oligoarticular in 45 (19.1%) and polyarticular in 13 cases (5.5%). A total of 54 patients (22.9%) showed clinical tophi at physical exam.

CV risk stratification

Risk prediction tools were applied to 149 patients (62.9%), as all others were considered at very high risk at presentation. Average risk scores were 3.8% (SD±3.5%) with SCORE and 6.3% (±4.1%) with FHS, both equivalent to a moderate risk. Seven patients showed SCORE >9%, being then classified as very high risk. An initial CV risk classification after risk prediction tools was performed (figure 1), and demonstrated very high (95 patients; 40.1%) or moderate (55 patients; 30.4%) risk levels.

Figure 1

Cardiovascular (CV) risk stratification of included patients with gout at presentation. Grey bars show the risk before performing the carotid ultrasound (cUS), while black bars show the final CV risk after the carotid evaluation for subclinical atherosclerosis. Data shown as number of patients per each subgroup and percentage of total sample.

cUS was performed in 142 patients (59.9%) not initially classified as very high CV risk after interview and scores. In comparison with those without cUS, they were younger and male predominant, had less CV risk factors (except for smoking), use of diuretics and chronic kidney disease (table 1). Lipids levels were significantly lower in those without cUS, while no differences on other lab variables were found. cUS revealed increased IMT in 64 patients (45.1%, 95% CI 36.8% to 53.3%) and carotid atheroma plaques in 66 (46.5%, 95% CI 37.8% to 54.2%), which were bilateral in 27 cases (19.0%, 95% CI 12.5% to 25.5%). Forty-four patients (31.0%) showed both increased IMT and atheroma plaques at cUS.

After cUS assessment, the risk of 80 patients (56.3% of those with cUS) was upgraded (one from low risk, 42 from moderate risk and 37 from high risk). Figure 1 shows the final CV risk stratification, which significantly differed from the initial stratification (p<0.001 for trend test); 67.9% of the included patients were finally considered as being at very high CV risk level.

Discriminative value of the CV risk prediction tools for carotid plaques

Table 2 shows the results of the calculated discriminative value of SCORE and FHS, and figure 2 displays the plotted ROC curves. The 2×2 tables built to estimate the discriminative scores are supplied in online supplementary tables S3 and S4. The estimated AUC was 0.711 for SCORE and 0.683 for FHS. These indicate a moderate discriminative capacity of risk tools for predicting carotid atheroma plaques, as a randomly selected individual from the plaque group would have a score value higher than that for a randomly chosen individual from the non-diseased group only 68%–70% of the time.

Table 2

Discriminative value of cardiovascular risk tools for the presence of carotid plaques at ultrasound

Figure 2

Receiver operating characteristic curves of the Systematic COronary Evaluation (SCORE) (solid line) and the Framingham Heart Study (FHS) (dotted line) prediction tools for the presence of carotid atheroma plaques at ultrasound.

Association analysis between gout features and carotid atherosclerosis

The results of the logistic regression analysis are shown in table 3. No association was found between individual gout-related variables and the presence of carotid atherosclerosis findings (increased IMT or atheroma plaques).

Table 3

Logistic regression with presence of atherosclerotic findings at carotid ultrasound (increased IMT or atheroma plaques) as dependent variable and gout features (one variable at a time) as covariates


Atherosclerotic CV disease is the leading cause of morbidity and mortality worldwide. CV prevention acting on CRF has led to a progressive decrease in mortality.13 CV preventive strategies are less costly than treating the complications,14 especially on subjects at risk, which support efforts to identify high-risk populations through risk prediction tools15 or subclinical atherosclerosis screening.16 This is the first study assessing the CV profile of patients with gout at presentation incorporating these strategies. Here, two out of every three patients with gout were classified at a very high CV risk at presentation, comparable to having already suffered from a myocardial infarction. This should lead to optimisation of preventive strategies, such as intensive statins therapy or stricter lipid levels targets.11 ,17 ,18

Almost half of the patients with gout who underwent cUS had atheroma plaques. A previous, smaller study found a rate of around 30%;12 the difference is likely explained by patient selection, ethnicity or methodology. For the present study, no control group was selected, but two large studies assessing subclinical atherosclerosis in Spanish middle-aged populations are available for comparison:19 ,20 in both studies, the prevalence of carotid plaques ranged from 25.1% to 31.0% of subjects. Thus, the prevalence of carotid plaques in patients with gout appears greater than the general population and similar to other chronic inflammatory diseases such as rheumatoid arthritis21 or systemic lupus erythematosus.22

Both SCORE and FHS tools performed moderately in identifying patients with gout at very high CV risk. This finding, not previously reported for gout, follows the results in other inflammatory arthritis.23 ,24 Risk assessment tools are recommended for identifying patients at risk,15 but considering their low accuracy to detect carotid plaques, they merit cautious use in gout. In the present study, tool-based risk level was later upgraded in 56.1% of patients after cUS. In the study by Coll et al,20 the risk level was only upgraded in 25% after cUS, while in the Progression of Early Subclinical Atherosclerosis study19 the proportion was close to our study, but used multivessel ultrasonography and coronary CT.

Despite being highly prevalent in gout,25 CRFs do not fully explain the increased CV risk in gout: population-based studies constantly show an extra risk in gout after adjusting for CRFs,2 ,3 and in the present study a moderate discriminative value of CRFs-based scores was found. Other factors such as crystal-led inflammation (acute and persistent), hyperuricemic state or intake of non-steroidal anti-inflammatory drugs may likely explain this poor CV outcome.26 Therefore, as well as controlling CRFs, MSU crystals dissolution by reducing serum uric acid levels might control this increased CV risk, although data are still controversial.27–29 Also, initiating urate-lowering agents at the first gout attack could contribute to CV prevention.5

In the present study, no significant association between gout characteristics and carotid atherosclerosis was noted, probably owing to the heterogeneity of the patients and the small sample size for this secondary objective. Long disease duration, if untreated, leads to high crystal load and long exposure to crystal-led inflammation, as crystals are taken as danger signals by the innate immune system through the nucleotide-binding oligomerization domain-like receptor P3 (NLRP-3) inflammasome pathway30—shared with cholesterol crystals, which play a relevant role in the inflammatory component of atherosclerosis.31 ,32 Persistent inflammation has been linked to the accelerated atherosclerosis seen in patients with inflammatory disorders.33 In fact, patients with tophi, with an increased inflammatory state,34 suffer from more severe myocardial infarctions35 and increased mortality.36 To date, the mechanism by which inflammation leads to atherosclerotic complications remains unclear. C reactive protein levels, a well-established independent CV risk factor,37 correlate poorly with extension of atherosclerotic disease,38 suggesting that inflammation may be mainly involved in plaque rupture, thrombus formation and the occurrence of a CV event, rather than in atheroma plaque formation.39 This might also explain the absence of significant association between gout features and carotid atherosclerosis found in the present study.

One of the strengths of the present study was to only include patients with crystal-proven gout, avoiding enrolment of misclassified cases that may bias the results. All participants underwent a structured CV assessment that can be easily reproduced in further studies or clinical practice. Incorporating subclinical atherosclerosis screening such cUS or coronary artery calcium to the assessment strengthened the CV risk stratification in patients with gout, also shown in general population.16

There are a number of limitations in our study. Patients were seen in a hospital-based rheumatology unit where severe, refractory cases are more likely to be referred; thus, results might not apply to patients with gout managed in primary care or self-treated, where a lower prevalence of comorbidities and CV risk might be expected. Most patients were seen from emergency or during intercurrent hospitalisations, reducing the selection bias of severe gout forms; however, the results of this study should be replicated at a primary care setting. The absence of a control group might limit the impact of the results, especially regarding the prevalence of carotid plaques found. However, finding that two out of three patients with gout might suffer from a very high CV risk, with the subsequent implications on management, is itself remarkable. In addition, the sonographer was not blinded to clinical data and this may have an impact on cUS results, although this was minimised by the fact that this is a standardised, objective procedure. Finally, the cross-sectional study design precludes evaluating the impact of the present CV assessment in patients with gout; a prospective follow-up analysis of this inception cohort is expected to provide further insight.

In summary, most patients with gout presenting for first time at rheumatology clinics might be at very high CV risk. A structured CV assessment revealed a notable prevalence of subclinical atherosclerosis, and a poor-to-moderate accuracy of the SCORE and the FHS tools to detect carotid plaques. These data support the need for initiating optimal CV prevention strategies at the very first visit.


The authors thank their colleagues from the Rheumatology Department for their remarkable effort in patient recruitment for the purpose of this study. The authors thank Professor Antonio Picó MD, PhD; Agustín Martínez MD, PhD; Santos Castañeda MD, PhD; Professor Víctor Martínez-Taboada MD, PhD and Juan Miguel Ruiz-Nodar MD, PhD, for their thoughtful criticism and valuable suggestions that significantly improved the present manuscript. We also thank Philip Courtney, MD, PhD for revising the use of English.


View Abstract


  • Handling editor Tore K Kvien

  • Contributors MA wrote the first draft of the manuscript. All authors contributed to the drafting of the manuscript and approved the final version.

  • Competing interests MA has received speaking fees from Menarini labs. EP has received consulting, speaker fees or grants from the following companies: Menarini, AstraZeneca, Savient, Procaps and Novartis.

  • Patient consent Obtained.

  • Ethics approval Ethics Committee of Hospital General Universitario de Alicante (Spain).

  • Provenance and peer review Not commissioned; externally peer reviewed.