Q J Med 2013; 106:647–658 doi:10.1093/qjmed/hct083 Advance Access Publication 5 April 2013

Independent and conjoint associations of gout and hyperuricaemia with total and cardiovascular mortality A.G. STACK1,2, A. HANLEY3, L.F. CASSERLY1, C.J. CRONIN1, A.A. ABDALLA1, T.J. KIERNAN1, B.V.R. MURTHY4, A. HEGARTY2, A. HANNIGAN2 and H.T. NGUYEN2 From the 1Divisions of Nephrology and Cardiology, Department of Medicine, University Hospital Limerick, Limerick, Ireland, 2Graduate Entry Medical School, University of Limerick, Limerick, Ireland, 3Department of Cardiology, Mater University Hospital Dublin, Dublin, Ireland and 4Michael E. DeBakey Department of Surgery, Baylor College of Medicine, Houston, TX 77030, USA Address correspondence to Dr A.G. Stack, Department of Academic Medicine, Graduate Entry Medical School (GEMS), University of Limerick, Clinical Academic Liaison Building, St Nessans Road, Limerick, Ireland. email: [email protected] Received 21 February 2013 and in revised form 6 March 2013

Summary Background: Gout and serum uric acid are associated with mortality but their simultaneous contributions have not been fully evaluated in the general population. Purpose: To explore the independent and conjoint relationships of gout and uric acid with mortality in the US population. Methods: Mortality risks of gout and serum uric acid were determined for 15 773 participants, aged 20 years or older, in the Third National Health and Nutrition Examination Survey by linking baseline information collected during 1988–1994 with mortality data up to 2006. Multivariable Cox proportional hazards regression determined adjusted hazard ratios (HRs) and 95% confidence intervals (CIs) for each exposure and all analyses were conducted in 2011 and 2012. Results: Compared with subjects without a history of gout, the multivariable HR for subjects with gout

were 1.42 (CI 1.12–1.82) for total and 1.58 (CI 1.13–2.19) for cardiovascular mortality. Adjusted HRs per 59.5 mmol/l (1 mg/dl) increase in uric acid were 1.16 (CI 1.10–1.22) for total and cardiovascular mortality and this pattern was consistent across disease categories. In the conjoint analysis, the adjusted HRs for mortality in the highest two uric acid quartiles were 1.64 (CI 1.08–2.51) and 1.77 (CI 1.23–2.55), respectively, for subjects with gout, and were 1.09 (CI 0.87–1.37) and 1.37 (CI (1.11– 1.70), respectively, for subjects without gout, compared with those without gout in the lowest quartile. A similar pattern emerged for cardiovascular mortality. Conclusions: Gout and serum uric acid independently associate with total and cardiovascular mortality. These risks increase with rising uric acid concentrations.

Introduction

considered a possible target for therapeutic intervention.1–5 Nevertheless, there are a few published studies which have failed to demonstrate positive relationships between uric acid and cardiovascular and total mortality.6–8 It is also unclear whether the reported relationships between uric acid and mortality differ between men and women, among race groups and by baseline cardiovascular risk status.

It is widely considered that elevated uric acid levels contribute to increased cardiovascular risk and death. Most, but not all, studies to date have demonstrated a positive independent association between hyperuricaemia and total and cardiovascular mortality suggesting that this biomarker should be

! The Author 2013. Published by Oxford University Press on behalf of the Association of Physicians. All rights reserved. For Permissions, please email: [email protected]

648

A.G. Stack et al.

Moreover, the degree to which the relationships between uric acid and cardiovascular mortality are modified by existing traditional risk factors has not to our knowledge been fully addressed by existing studies. Epidemiological studies have also suggested but not confirmed that gout, a chronic inflammatory disorder derived from abnormal uric acid metabolism, is linked to increased all-cause and cardiovascular mortality.9–12 Male participants in the Health Professional Follow-up study with a diagnosis of gout and followed for over a 12-year period experienced a 38% higher total mortality and a 55% higher cardiovascular mortality. Similarly, analysis of prospective data from the Multiple Risk Factor Intervention Trial found that men with both hyperuricaemia and gout experienced higher rates of acute myocardial infarction.10 Although these studies have contributed to our knowledge base, there are several unanswered questions. First, it is unclear from these studies whether hyperuricaemia and gout are simultaneous independent predictors of total and cardiovascular mortality. Second, it is unknown whether the presence of hyperuricaemia modifies the clinical impact of gout on cardiovascular outcomes. Third, existing studies have not explored the impact of these potential risk factors in cohorts of sufficient size, diversity and length of follow-up such that the results can be largely applied to the general population to ensure external validity. The purpose of this study was to investigate the nature and magnitude of the relationships between serum uric acid and gout with total and cardiovascular mortality in a representative sample of the general population.

Methods The Third National Health and Nutrition Examination Survey study The Third National Health and Nutrition Examination Survey (NHANES III) was a population-based survey conducted by the National Center for Health Statistics that assessed the health status of a representative sample of non-institutionalized persons living in the USA between 1988 and 1994.13 A complex stratified multistage probability design was used with over-sampling of elderly and minority populations. The data collection instrument consisted of a standardized questionnaire administered during a home interview followed by a detailed physical examination that included collection of blood specimens at a mobile

examination centre or at the participant’s home. The vital status of almost all NHANES III participants (99.9%), 17 years of age or older, was determined through matching to the National Death Index through 31 December 2006.14 Ethical approval for the study was obtained from the Ethics Review Board of the National Centre for Health Statistics and further ethical approval for use of NHANES data that is freely available on the web is not required as it is anonymized.

Sample The current analysis was restricted to adult participants, 20 years of age or older (n = 18 825), who had valid serum uric acid and creatinine measurements (n = 15 838) and had vital status through 31 December 2006. There were 15 773 subjects in the final study population.

Baseline measurements Blood samples were obtained from non-fasting persons and frozen serum sent to the Centers for Disease Control and Prevention for analysis. A detailed description of the assay methods for measuring serum uric acid levels has been published previously.1 Serum uric acid was measured using the automated colorimetric phosphotungstic acid procedure, which had been validated against a uricase assay, on a Technicon SMA12-60 (Technicon instruments, Tarrytown, NY, USA). Serum creatinine concentrations were measured by the modified kinetic Jaffe reaction using a Hitachi 737 analyzer (Boehringer Mannheim Corp., Indianapolis, IN, USA) and recalibrated to the Cleveland Clinic.15 Glomerular filtration rate was estimated from the abbreviated Modification of Diet in Renal Disease (MDRD) Study formula.16 The NHANES III provided data that were collected from face-to-face interviews and direct physical examinations. The presence of gout was based on self-report that required a physician’s diagnosis. Additional self-reported variables were extracted from the interview portion: age, gender, race/ethnicity, myocardial infarction, hypertension, diabetes, chronic lung disease, heart failure, arthritis, smoking status, physical inactivity and family history of premature myocardial infarction. Prescription data on medication for the management of gout were also available for NHANES III participants who were interviewed within a month of the baseline survey. Physical examination data included height, weight, waist circumference and the average of at least three blood pressure readings. Hypertension was also assessed by averaging at least three separate blood pressure readings made at interview visit

Gout, uric acid and mortality and categorized according to the Seventh Report of the Joint National Committee on Prevention, Detection, Evaluation and Treatment of High Blood Pressure (JNC VII).

Assessment of all-cause and cardiovascular mortality Mortality data were provided through linkage with the National Death Index available up to the 31 December 2006.14 Cardiovascular causes of death were identified from the International Classification of Diseases, 10th Revision (ICD-10) diagnosis codes in the NHANES-linked mortality files and included: deaths from acute myocardial infarction (121–122), other acute ischaemic heart disease (124), atherosclerotic cardiovascular disease (125.0), all other forms of chronic ischaemic heart disease (120, 125.1, 125.9) and cerebrovascular disease (160– 169). Deaths were analysed for all causes and cardiovascular causes (ICD-10 codes 120–169).

Statistical analysis Serum uric acid concentrations were modelled as a continuous variable and in quartiles and baseline characteristics were compared across quartile groups. Similarly, the characteristics of subjects were compared according to the presence or absence of gout. For continuous variables, differences across quartiles were tested with analysis of variance whereas for dichotomous variables, comparisons were conducted using the chi-square test and all comparisons were weighted according to the complex sampling strategy. Total and cardiovascular mortality rates were calculated for the entire cohort expressed as deaths per 1000 person-years. Mortality rates were calculated for subjects according to the presence or absence of gout and by quartiles of serum uric acid. Cox proportional hazard regression was used to model relationships of gout and serum uric acid with total and cardiovascular mortality adjusting for baseline characteristics. Covariates for adjustment included age, sex, race, cardiovascular conditions (history of coronary disease, heart failure and stroke) and cardiovascular risk factors; namely a history of diabetes and hypertension, systolic and diastolic blood pressure at baseline (mm Hg), physical inactivity, and smoking behaviour (never, past and current), serum cholesterol and body mass index. In extended models, adjustments were further made for measures of systemic inflammation (C-reactive protein and serum ferritin). To determine whether the associations of gout and uric acid with mortality differed by disease and lifestyle categories, we tested for

649

interactions, defined as P < 0.01. For each covariate, the unadjusted hazard ratios (HRs) and adjusted HR for death were calculated with 95% confidence intervals (CIs). Weighted Cox regression accounted for the complex survey design, the unequal probability of subject selection and non-response rates. Model fit was assessed by the Taylor method and 2 log likelihood ratio (SAS v 9.3; Cary, NY, USA).

Results Baseline characteristics of the population The mean characteristics of all study participants aged 20 years and older are illustrated in Table 1. The weighted mean age of the study population and standard error (SE) was 44.7 years (0.5), 77% were White, 10.4% were Black and 5.1% were Mexican-American. The mean serum uric acid level was 315.3 (1.19) and 2.7% (n = 4 544 496) participants had a recorded history of gout.

Characteristics of population by quartile of serum uric acid and gout Higher quartiles of serum uric acid were associated with higher prevalence of cardiovascular conditions. Mean uric acid concentrations increased from 214 to 434 mmol/l from the lowest to the highest quartiles. Higher uric acid concentrations correlated with higher levels of inflammatory markers, Creactive protein and serum ferritin levels (P < 0.001). In contrast, higher uric acid levels correlated with lower glomerular filtration rates (P < 0.001). The prevalence of gout was significantly higher in men (3.9  0.33%) than women (1.7  0.18%) and increased with advancing age. Compared with subjects without gout, subjects with gout were older, had higher prevalence of cardiovascular disease and Framingham risk factors. Similarly, subjects had significantly higher levels of uric acid, and greater proportions treated with diuretics and uric acid lowering therapy (Table 2).

Serum uric acid and all-cause mortality During 10 years of follow-up, 2699 (10.5 %) died and 1276 were ascribed to cardiovascular disease. The crude all-cause mortality rate was 11.1 per 1000 person-years and increased from 7.1 per 1000 person-years to 16.8 person-years from the lowest to the highest quartiles (Table 3). In the unadjusted analysis, a 59.5-mmol/l (1 mg/dl in conventional units) increase in uric acid was associated with a 28% increase in mortality (95% CI 1.24– 1.33). With adjustment for demographic factors,

Demographics Age at interview (years)b Sex (%) Men Women Race/ethnicity (%) White non-Hispanic Black non-Hispanic Mexican-American Other Clinical condition (% present) Gout Arthritis History of myocardial infarction Chronic bronchitis Congestive heart failure Stroke Diabetes (by history) Hypertension (by history) JNC VII staging of hypertension Normal