1. Introduction
Gout is an acute sterile inflammation primarily triggered by the deposition of monosodium urate crystals in joints [
1,
2]. Historically, gout was thought to be a disease of only the wealthy or royal class, and it was called the “disease of the kings.” However, the incidence of gout has largely increased in multiple countries and regions as rapid economic development has provided the general population with richer diets [
3]. From 2010 to 2020, the global incidence of gout rose from 0.08% to 2–4% [
4,
5]. In addition to severe pain symptoms, gout itself may contribute to an increased risk of conditions such as cardiovascular disease [
6], chronic kidney disease [
7], and rheumatoid arthritis [
8]. Currently, gout remains a challenging condition to manage. Its progressively rising incidence, coupled with various associated comorbidities, poses a considerable economic burden on both individuals and society. A survey indicated that, compared with non-gout individuals, those with gout experience a substantial increase in annual medical expenses, amounting to several thousand or even tens of thousands of dollars [
9]. Additionally, reports suggest that in the United States alone, the annual economic burden caused by gout exceeds billions of dollars [
10]. Consequently, gout emerges as a noteworthy issue in the global public health domain.
The pathogenic mechanisms of gout are still quite complex [
11], and the deposition of serum urate is widely recognized as the primary risk factor for gout [
12]. Currently, pharmaceutical treatments are widely employed to address the issue of elevated serum urate levels causing gout [
13]. While these medicines show good short-term efficacy, long-term use may induce gastrointestinal reactions, skin rashes, systemic complications, and even renal failure [
14,
15]. Furthermore, such medications cannot prevent, halt, or reverse the progression of this complex disease [
16]. Therefore, there is an urgent need for non-pharmacological preventive and therapeutic approaches at present. In behavioral medicine, improving diet, physical activity, and alcohol consumption habits is considered beneficial for the prevention and treatment of gout [
17]. Among these, physical activity may play a particularly crucial role. On one hand, there is a close relationship between physical activity and obesity (one of the risk factors for gout); on the other hand, skeletal muscles can produce anti-inflammatory cytokines during exercise, helping to break the vicious cycle of chronic inflammation and thereby reducing the impact of gout [
18]. To explore the preventive effect of physical activity on gout, many scholars have conducted large-scale surveys, such as cross-sectional studies observing the association between physical activity and serum urate levels as well as gout symptoms [
19,
20]. Their research primarily supports the positive role of physical activity.
To address the limitations of observational studies, randomized controlled trials (RCTs), considered the “gold standard” for causal inference, are crucial [
21]. However, high-quality RCTs often require substantial sample sizes and rigorous randomization processes to effectively balance confounding factors. Therefore, alternative research methods have considerable value, and Mendelian randomization (MR) has emerged as a novel technique in this context. MR uses genetic variants as instrumental variables to explore causal relationships between other external factors, thus allowing us to re-evaluate the findings from observational studies. MR is based on the principle that if genetic variation (or genes) influences a modifiable risk factor, and this risk factor, in turn, affects the risk of a certain disease, then genetic variation should be associated with the risk of that disease [
22]. Eventually, we can estimate the causal impact of modifiable risk factors on disease risk based on certain assumptions [
22]. Since an individual’s genotype is fixed at the formation of the zygote, theoretically the association between genes and traits should not be confounded by environmental factors encountered during an individual’s later life history [
23]. Furthermore, the random allocation of genetic variation during meiosis and random mating within populations can further balance confounding factors [
23]. These advantages make MR an analog of RCTs and thus play a crucial role in causal inference.
As mentioned above, since the previous findings on the association between physical activity and gout were obtained from observational studies, they can be threatened by confounding, and the causality cannot be confirmed due to the nature limitation of this research design [
24], especially considering that gout symptoms can reduce overall activity levels (which implies reverse causality) [
25]. Moreover, due to the often long and unpredictable intervals between gout attacks, exploring the impact of physical activity through large-sample and well-controlled trials is challenging and resource-intensive. For these reasons, using the MR approach to identify the causal effect of physical activity on gout and its risk factor serum urate is essential and may offer additional evidence to form a triangulation with existing studies of different designs [
26]. So far, there has been only very limited exploration in this regard [
27], and the intensity of physical activity has not been specifically studied. Since the intensity of physical activity is believed to alter its benefits on gout, our study, in response, is designed to explore the differences. In general, this study aims to use MR design to validate the impact of different intensities of physical activity on serum urate concentration and the incidence of gout.
4. Discussion
This study aimed to employ MR approach to examine the impact of different intensities of physical activity on serum urate and gout. We obtained suggestive evidence from our core estimator (IVW) supporting the role of moderate-intensity physical activity in reducing the incidence of gout (OR = 0.628,
p = 0.034). Moreover, none of the methods employed for detecting pleiotropy, including the MR-Egger intercept test, MR-PRESSO global test, funnel plots, and Cochran Q heterogeneity test, demonstrated evidence of horizontal pleiotropy that could distort this result. The Rucker framework used in our sensitivity analysis and a MR strategy by others [
41] also support this result returned by IVW. Notably, after removing an instrumental variable that could lead to controversial results, this association reached the significance level set in our study (OR = 0.555,
p = 0.006), further emphasizing the protective effect of moderate-intensity physical activity against gout. On the other hand, although MR-Egger regression suggested that moderate physical activity could reduce serum urate levels and vigorous physical activity could reduce the risk of gout, our funnel plot (based on MR-Egger) and pleiotropy diagnostics indicated that these results are possibly subject to horizontal pleiotropy. Therefore, we refrain from discussing these two associations further, focusing cautiously on the “moderate physical activity—gout” association only.
Our study results are generally consistent with some previous observational studies. For example, a cross-sectional survey in Sweden recruited 868 gout patients and compared them with randomly selected general participants, revealing lower levels of physical activity reported by male gout patients [
19]. Another study showed that gout patients engaging in physical exercise had fewer annual episodes and reduced pain compared with those who did not exercise [
70]. These studies collectively emphasize the protective role of physical activity against gout. However, since these studies only investigated a single time point, they cannot confirm the causal relationship between physical activity and gout. A study utilizing wearable devices for tracking physical activity identified a significant reduction in walking volume during gout attack periods, implying that patients with more frequent gout attacks may be less engaged in physical activity [
71]. This uncertainty underscores the value of MR analysis since the causal direction is pre-determined in such a research design.
In recent years, scholars have increasingly employed this method to explore associations whose causality is challenging to determine through controlled experiments, such as the impact of physical activity on COVID-19 mortality [
66], the influence of physical activity on cancer risk [
72], and the effect of computer gaming on mental health [
73]. Regarding the topic of gout, one relevant MR study indicated that the overall level of physical activity measured by accelerometers had no effect on the incidence of gout [
27]. This finding contrasts with our “moderate physical activity—gout” association. We speculate that the different measurement methods of the independent variable (exposure) may be the primary reason for this discrepancy. The other study focused on accelerometer-measured physical activity, which is more objective and accurate, but it is difficult to distinguish the form and intensity of the physical activity. Thus, the measured physical activity might include portions unfavorable for reducing or even increasing the risk of gout, thereby weakening the strength of the association.
We did not find solid evidence to support the impact of physical activity on serum urate. Therefore, inflammation may be an important clue to explain our findings regarding gout. Research indicates that inflammatory factors such as interleukin (IL)-1β, IL-8, IL-17, NLRP3 inflammasome, and tumor necrosis factor-alpha (TNF-α) are involved in the inflammatory processes of gout, and immune cells, including neutrophils, monocytes/macrophages, and lymphocytes, play a crucial role in the onset of gout [
74]. As mentioned in our introduction, myokines produced by physical activity help break the vicious cycle of chronic inflammation, thereby reducing the detrimental effects of gout [
18]. Physical activity may exert anti-inflammatory effects by lowering IL-6, reducing C-reactive protein, and inhibiting TNF-α, thereby reducing the risk of gout [
75,
76,
77]. However, the role of physical activity intensity in this regard is less understood. A controlled experiment on mice suggested that moderate-intensity exercise produced anti-inflammatory effects, while high-intensity exercise showed no significant difference in inflammation compared with the non-exercise control group [
78]. This anti-inflammatory effect is believed to be achieved by physical activity through the downregulation of TLR2 on circulating neutrophils and inhibition of serum CXCL1 [
78]. This result somewhat aligns with our study findings, indicating that the intensity of physical activity can modulate the protective effects of physical activity against gout. Nevertheless, since our study only analyzed associations and no other mediators were investigated, the specific mechanisms remain to be explored in future research.
This study has several advantages. Firstly, the MR approach allowed us to overcome the confounding effects between the exposure and outcome, regardless of whether the confounders were measured [
79]. Secondly, we used distinct samples, avoiding the bias introduced by sample overlap in previous studies [
29,
80].
However, our study has some limitations. Firstly, the two-sample MR approach requires data from the same underlying population but different samples [
29]. This is the reason why the STROBE-MR statement recommends a justification of the similarity of the genetic variant—exposure associations between the exposure and outcome samples [
31]. However, since our exposure phenotypes (physical activity of different intensities) were not measured in the outcome database, we could not conduct such a comparison. Secondly, geographic clustering of genetic variation may introduce false genetic associations between our exposure and outcome. For example, mating behaviors (e.g., assortative mating) may be influenced by spatial factors. Therefore, individuals living in geographically close areas may have similar genes, leading to associations between genes and cultural, economic, social, political, and other environmental factors [
81]. To address this issue, stratified analyses by region/area are necessary. However, the nature of the used summary data forbids such sensitivity analyses, which warrants further research.