Mendelian Randomisation: Inferring the causal relationship between urate and associated biochemical and phenotypic parameters
The biological function of urate has been widely debated, with mounting evidence supporting its role as an antioxidant. Contradictory findings surround the detrimental properties of urate. Increased urate is the central cause of gout and is theorised to play a detrimental role in renal and metabolic-related pathology. However, controlling for confounders influencing serum urate associations is challenging, meaning that a primary limitation of epidemiological studies is the inability to draw causal conclusions, restricting inferences regarding the role urate plays in pathological and physiology states. A key question regarding the putative detrimental role urate plays in growth, renal and metabolic conditions remains to be answered; does urate have a causal role, is it compensatory or merely a consequence? A novel genetic approach, Mendelian randomisation, has the ability to overcome confounding issues associated with epidemiological research, thus, allowing causal relationships to be determined. Mendelian randomisation uses genetic variants (i.e. instrumental variables) to serve as a proxy for an exposure, enabling causal inferences to be concluded with an outcome of interest. Due to random allocation during gamete formation, the use of genetic variants is not confounded by environmental influences. By using Mendelian randomisation, the aim of this study was to disentangle the causal relationship between urate and renal function, adiposity and stature. The Mendelian randomisation approach, two-stage least squares regression, was used to examine causality. Two cohorts were used, the Atherosclerosis Risk in Communities and Framingham Heart longitudinal studies. A literature search was conducted to select genetic variants which significantly associated with the respective exposure. For example, a uric acid transporter genetic risk score was used as an instrumental variable for serum urate exposure to test for a causal role of urate on renal function. Two-stage least square and auxiliary regressions were carried out in STATA version 8.0 (StataCorp, TX, USA) using the ivreg and regress functions, respectively. Important findings were a male-specific causal role of increased serum urate on improved renal function (P=0.023). Results obtained when examining serum urate and adiposity revealed the complexity surrounding this interaction. A causal role of increasing adiposity mediating elevations in serum urate (P=0.001) was revealed, while conversely, evidence supported a causal relationship in the reverse, where increasing serum urate mediated decreases in adiposity (P=0.046). Lastly, investigating the causal relationship between serum urate and height showed no causative association in either direction; height did not mediate serum urate alterations (P=0.319), while similarly, serum urate did not mediate variations in height (P=0.479). Using a Mendelian randomisation approach this study revealed a causal role of serum urate beneficially influencing renal function, contrary to the established relationship of increased serum urate with reduced renal function. A complex interaction with adiposity was evident. It was proposed unconfounded, explained serum urate, predicted through uric acid transporters, may beneficially influence renal function and adiposity, possibly through its antioxidant properties. Consequently, previous findings using simple regression models concluding detrimental relationships may have been a result of residual confounding and/or mediator effects, with fructose proposed as a possible culprit, obscuring the true beneficial properties of serum urate on renal function and adiposity, while producing spurious associations with height.
Advisor: Merriman, Tony
Degree Name: Master of Science
Degree Discipline: Genetics
Publisher: University of Otago
Keywords: Urate; Mendelian; Randomisation; causal; kidney; weight; height
Research Type: Thesis