In the gout cohort, we demonstrated a statistically significant but modest elevation of risk of admission for subsequent MS, PD and MND. After excluding cases of MS, PD or MND observed in the first year after hospitalisation with gout, there was no elevated RR for these conditions. These results do not confirm findings from previous studies suggesting a reduced risk of neurological disease following gout as a consequence of a neuroprotective antioxidant effect of uric acid, or the corollary that individuals with low levels of uric acid have an impaired ability to counteract free radical toxicity [19,20]. It is possible that the inflammatory state associated with the development of symptomatic gout or recurrent inflammatory arthritis (rather than simply hyperuricemia in previous work) counteracts any potential protective antioxidant effect of uric acid. The observed slightly increased risk in early years might be a reflection of surveillance bias, whereby an individual seen in hospital for one condition is more likely to have a second detected. Another possible explanation for the small increase in risk in the early years, that was not sustained long-term, is that individuals admitted to hospital with gout may undertake measures successfully to lower their levels of uric acid (and reduce future gout attacks), which may involve both dietary changes and the use of uric acid reducing medication [24]. Further investigation of the comparison of risk of MS, PD or MND in treated vs untreated hyperuricemia would have enabled better characterisation of the association, but such data was not available in this study.
Further, from hospital admission statistical records alone, we cannot be sure that, in the gout cohort, the onset of gout really did invariably precede the onset of MS, PD and MND. The date of first admission for a neurodegenerative disease is likely to have been preceded by a period of sub-clinical/undiagnosed disease, and thus, the disease state itself may have resulted in somewhat elevated levels of uric acid and an increased risk of gout. The fact that the raised risk was not sustained over time means that we cannot be confident that the findings support a causal role for uric acid in altering neurological disease risk. It needs further investigation to establish whether uric acid levels alter neurological disease risk (and/or disease progression), given the potential therapeutic approach of altering uric acid levels (through either dietary or pharmacological measures) in altering risk of neurological diseases.
In the MS, PD and MND cohorts, we found a prolonged reduced risk of subsequent admission for gout. Notably, for the MS and PD cohorts a reduced risk of gout was seen following the first year of initial admission for neurological disease. In the MND cohort, this reduced risk only appeared five years following initial MND admission. Although, we can only speculate about pathophysiological mechanisms, our findings on gout before and after the neurological diseases might indicate that the causal pathway is that the neurological disease states result in lowered uric acid levels. For example, central nervous system inflammation which is a typical feature of MS has been associated with increased production of the strong oxidant peroxynitrite and free radicals, which may in turn lead to reduced levels of uric acid, which is a peroxynitrite scavenger [25]. A cross-sectional study which found markedly reduced levels of serum urate in MND patients with bulbar onset, and with a longer duration of disease, suggested that the reduced levels of serum urate may be a consequence of malnutrition secondary to the impact of MND, particularly in its later stages [26].
This study’s strengths include its large size and the fact that it is a national study and therefore generalizable as the cohorts are likely to relatively accurately reflect the national prevalence of these neurological disorders. Further, it demonstrates what can be done with the analysis of readily available electronic medical records.
The study also has limitations. Importantly, it is not a cohort study with follow-up from the date of “first ever” diagnosis, but just from “first recorded” diagnosis in a hospital day case or inpatient record. Data are not available on patients who migrate out of England. The dataset is limited to people who were admitted to hospital, or who received day case specialist care, and thus there exists the potential for selection bias in omitting milder cases of gout. Further, given that only a minority of individuals with elevated uric acid levels develop gout, using gout as a proxy for high uric acid levels may underestimate the association between uric acid and neurological diseases. Uric acid levels appear positively correlated with risk of gout, and the gout cohort is therefore more likely to capture those with more severe uric acid elevation in which, over time, urate crystals have deposited in or around joints. There is very limited information on potential confounding factors such as detailed socioeconomic characteristics, beyond the IMD index, ethnicity and smoking.
Studies of alternative design would be costly and complex undertakings. For example, if funds allowed, “shoe-leather” studies of individual patients could be undertaken by personal follow-up, such as following large numbers of people with gout to await development of these three uncommon neurological diseases, or studies could be undertaken of people with MS, PD or MND to identify clinical gout or to measure their serum uric acid levels. The latter would need to be designed in ways that allow the investigators to determine the time sequence of ‘high levels of uric acid before the neurological condition’ or ‘neurological condition before changes in uric acid levels’.