Background Even though well-designed literature-based meta-analyses have confirmed the link between hyperuricemia and the risk for cardiovascular disease (CVD), it is known that study-level analyses can lead to biased assessments, and use of aggregated summary values has some limitations for explaining the heterogeneity. An individual patient data (IPD) meta-analysis offers several advantages including more accurate adjustment for confounders and better accounting of study heterogeneity.
Objectives To perform unadjusted and risk-adjusted IPD meta-analysis of the association between hyperuricemia and prevalent CVD using data from diverse epidemiologic populations in the United States.
Methods Baseline subject level data from 15 epidemiologic studies evaluating CVD risk conducted by the US National Heart Lung and Blood Institutes were pooled. Individuals who reported gout and using gout medications were excluded. Unadjusted and adjusted logistic regression analyses were performed. CVD was defined by self-reported physician diagnosis of myocardial infarction or stroke. Multivariable analyses adjusted for age, sex, sex-urate interaction, body mass index, smoking, total cholesterol, blood pressure, and diabetes mellitus.
Results Overall, 50,966 men (66.1%) and 26,082 women (33.9%) were studied. The mean (standard deviation) age and serum urate (sUA) concentration were 49 (14) years and 5.9 (1.6) mg/dL, respectively. The overall proportion of participants with sUA >6.0 mg/dL was 44.3% and those with sUA >7 mg/dL was 22%. The prevalence of CVD was 7.7% among those with sUA ≤6 mg/dL and 9.8% among those with sUA >6 mg/dL (p<0.001). The unadjusted (77,045 observations) and adjusted (age-gender adjusted, 77,045 observations; multivariable adjusted, 66,234 observations) logistic regression models indicated significant association between hyperuricemia and the risk for CVD. For sUA >6.0 mg/dL compared to ≤6.0 mg/dL, odds ratios (95% confidence intervals) were 1.31 (1.24, 1.37), 1.37 (1.20, 1.56), and 1.60 (1.42, 1.81) in the adjusted, age-gender adjusted, and multivariable adjusted models, respectively. For each mg/dL increase in sUA, odds ratios were 1.13 (1.12, 1.15), 1.24 (1.20, 1.28), and 1.20 (1.15, 1.24), respectively.
Conclusions These IPD meta-analyses confirm the results of literature-based meta-analyses that hyperuricemia is associated with elevated risk for CVD. Although the results of individual epidemiologic studies may vary due to differences in the risk profile, age, gender, and other risk factors, overall, the association between hyperuricemia and CVD appears to be robust.
Acknowledgement The authors would like to acknowledge Eswar Krishnan, MD, for his valuable contributions to the development of this analysis. This study was funded by AstraZeneca; editorial support was provided by PAREXEL and was funded by AstraZeneca.
Disclosure of Interest S. Baumgartner Employee of: Ardea Biosciences, a member of the AstraZeneca Group, R. Morlock Employee of: Ardea Biosciences, a member of the AstraZeneca Group