Uric acid and risk of pre-eclampsia: results from a large case-control study and meta-analysis of prospective studies

Abstract

To quantify the association between maternal uric acid levels and pre-eclampsia risk in a large collection of primigravid women. A case-control study (1365 cases of pre-eclampsia and 1886 normotensive controls) was conducted. Pre-eclampsia was defined as blood pressure ≥ 140/90 mmHg and proteinuria ≥ 300 mg/24 h. Sub-outcome analysis included early, intermediate, and late pre-eclampsia. Multivariable analysis for pre-eclampsia and its sub-outcomes was conducted using binary and multinomial logistic regression, respectively. Additionally, a systematic review and meta-analysis of cohort studies measuring uric acid levels < 20 weeks of gestation was performed to rule out reverse causation. There was a positive linear association between increasing uric acid levels and presence of pre-eclampsia. Adjusted odds ratio of pre-eclampsia was 1.21 (95%CI 1.11-1.33) for every one standard deviation increase in uric acid levels. No differences in the magnitude of association were observed between early and late pre-eclampsia. Three studies with uric acid measured < 20 weeks' gestation were identified, with a pooled OR for pre-eclampsia of 1.46 (95%CI 1.22-1.75) for a top vs. bottom quartile comparison. Maternal uric acid levels are associated with risk of pre-eclampsia. Mendelian randomisation studies would be helpful to further explore the causal role of uric acid in pre-eclampsia.

Figure 1

Association between uric acid quintiles and presence of pre-eclampsia after adjustment for maternal age. Mean (range) for quintiles given in μmol/L.

Figure 2

Serum uric acid levels and presence of early, intermediate and late pre-eclampsia in adjusted models* (*Multinomial regression model for early, intermediate and late pre-eclampsia adjusted for maternal age, recruitment centre, ethnicity, recruitment date, multiple pregnancy, smoking, socioeconomic position and infections during pregnancy. Gestational age was not included in the model because this variable defines the sub-outcomes. Results from the logistic model independent of the gestational age of onset is presented for comparison (hollow square). Odds ratios are reported on a logarithmic scale.).

Figure 3

Summary estimates for a comparison between top and bottom quartiles of uric acid for prospective studies with uric acid measured in early pregnancy compared to GenPE results. The p value corresponds to heterogeneity hypothesis testing among studies included in the meta-analysis.

Figure 4

Proposed scenarios for the role of uric acid in pre-eclampsia aetiology. In (A), increased uric acid is associated with pre-eclampsia but does not contribute to the pathogenesis of the disease, while in (B) uric acid acts as a causal factor for pre-eclampsia. In (C), the effect of the uric acid-lowering drug allopurinol^,,,,,, interrupts the causal pathway (as evidenced in non-pregnant individuals).