Approximately 40% of patients with diabetes develop diabetic kidney disease (DKD) and is the main cause of chronic kidney disease (CKD) worldwide (Alicic et al., 2017; Tuttle et al., 2021). DKD and its complications, including diabetes, heart failure, and obesity, are interrelated conditions that increase the risk of kidney failure and cardiovascular mortality and increase the cost of healthcare (Li et al., 2020). Therefore, modulating the risk factors for DKD has an important role in reducing the risk of CKD and its comorbid conditions. Obesity is associated with a high risk for both diabetes and cardiovascular complications (Melmer et al., 2018). The risk largely depends on the distribution of adipose tissue (Santilli et al., 2017). Trunk fat and waist circumference (WC) are associated with type 2 diabetes, cardiovascular disease, and multiple metabolic risk factors (Xu et al., 2015; Yang et al., 2019; Yang et al., 2021). However, these findings may have been confounded by unmeasured confounding risk factors, including objectively measured physical activity, cardiorespiratory fitness, or other underlying medical conditions that were not captured appropriately. Consequently, it is unclear whether the observed associations are causal.

Mendelian randomization (MR) is an epidemiological approach designed to evaluate causality that utilizes the principle that genotypes are free from confounding and reverse causation bias (Davey Smith and Hemani, 2014; Davies et al., 2018; Liu et al., 2021). It resembles the conditions of randomized controlled trials (RCTs), although MR may be performed retrospectively (Supplementary Figure S1). Given that all inherited genetic variants that occur prior to disease onset were determined at conception, MR has recently emerged as a tool to assess causal relationships (Emdin et al., 2017; Tan et al., 2022).

In this study, we examined whether the body mass index (BMI), trunk fat mass, and WC were causally associated with increased DKD risk using a two-sample summary MR.

In this population-based cohort study, MR was conducted to investigate the causal impact of BMI, trunk fat mass, and WC on DKD. In agreement with previous observational studies, our findings suggested that genetically mediated BMI, trunk fat mass, and WC are causally associated with an increased risk of DKD, where a 1 SD increase in the above-mentioned genetic IVs conferred a 99%, 80%, or 148% increased risk of DKD in the European population, respectively.

Each SD with higher genetic BMI (1.99; 1.47–2.69), trunk fat mass (1.80; 1.28–2.53), and WC (2.48; 1.40–4.42) was related to increased DKD risk. Human epidemiological studies suggest that obesity is a major cause of DKD, and the growing incidence of obesity contributes to diabetes associated with CKD prevalence (Winocour, 2018; Yang et al., 2020). An MR analysis in individuals of European ancestry indicated that each SD in higher BMI conferred an increased risk of DKD (1.33; 1.17–1.51). Our results are very similar to those of previous observational studies and MR analyses (Todd et al., 2015; Wan et al., 2020; Polemiti et al., 2021). However, certain research do not support this conclusion (Keane et al., 2003; Huang et al., 2014). A study showed that a high BMI (≥25 kg/m²) was a protective factor against worsening renal function in patients with type 2 diabetes mellitus complicated by CKD at stages 3 or 4 (Huang et al., 2014). Inconsistent results in observational studies can be explained by reverse causal bias or may be confounded by unmeasured confounding risk factors. MR can compensate for the limitations of conventional epidemiological studies and can be used to assess causal relationships. However, only one method was utilized in the previous MR (Todd et al., 2015). The consistency of the three methods we used to render our study more reliable.

Accumulation of trunk fat mass is associated with higher glucose levels, post-load glucose levels, and diabetes risk in observational epidemiological studies (Snijder et al., 2004; Tatsukawa et al., 2018). In addition, it is correlated with both visceral and subcutaneous fat accumulation as evaluated by computed tomography and is considered an inflammatory marker in non-dialyzed CKD patients (Axelsson et al., 2004; Sanches et al., 2008). Choi et al. (2017) confirmed that an increase of 1 kg in trunk fat mass is associated with a 15% increase in men with DM and a 19% increase in women after adjustment for several confounding factors. Using the MR method in current European populations, our results are in accordance with those of the above study (1.80; 1.28–2.53). WC has mostly been employed as a surrogate marker of abdominal adiposity. A number of studies involving CKD and non-CKD subjects have confirmed a strong correlation between WC and trunk fat, as evaluated using dual-energy X-ray absorptiometry (Orsatti et al., 2010; Bazanelli et al., 2012). Xu et al. (2015) suggested that a higher WC is associated with greater odds of developing CKD in type 2 diabetes mellitus (1.019 and 1.002–1.006). Our study provided by MR arrives at a similar conclusion: each SD with higher genetic WC (2.48; 1.40–4.42) was found to have a positive causal effect on DKD. These results revealed that higher overall and body fat distributions are causal risk factors for DKD in European populations. Moreover, our MR studies suggested that the causal effect of WC on DKD risk was greater than that of overall obesity. Obesity can stimulate the formation of lipid metabolites, cytokines, and hormones, which involves changes in the insulin signaling pathway and accelerates progression of insulin resistance.

Obesity is a condition in which the number and size of adipocytes increase, which further increases the total fat mass (Balsan et al., 2015). Adiponectin, which is mainly secreted by adipocytes, is considered to have anti-inflammatory, anti-atherosclerotic, and insulin-sensitizing properties. It regulates energy homeostasis and glucose and lipid mechanisms via the activation of adenosine monophosphate-activated protein kinases (Yamauchi et al., 2002). Snijder et al. (2009) revealed that trunk fat is negatively correlated with adiponectin levels. This may explain the adverse effect of trunk fat mass on DKD. Moreover, the causal effects of overall obesity, trunk fat mass, WC on DKD, and WC were slightly greater than those of generalized obesity. Thus, we emphasize that both the mass and distribution of body fat play a causal role in DKD.

Observational studies are more susceptible to reverse causality or confounding than the MR analysis, which can provide the best evidence for assessing causal relationships involving BMI, trunk fat mass, WC, and DKD etiology. Statistical differences were evaluated using the two-sample MR approach. To minimize potential pleiotropy, three different methods were performed to evaluate directional pleiotropy. A strength of this study was that the causal effects of BMI, trunk fat mass, and WC on DKD risk are robust and unbiased, owing to the consistency of the three methods. Another strength is that we simultaneously assessed the relationship between different risk factors and DKD, the most common micro-vascular complication of diabetes. Therefore, we were able to investigate the differences across disease risk factors.

One potential limitation of our study is that only European descent were recruited. However, regarding disease heterogeneity, studies have demonstrated different genetic profiles in different cohorts of patients, which are attributed to ethnic diversity, different risk factors, and different epigenetic profiles. Therefore, the reliability of the causal associations should be verified in other non-European populations (Tan et al., 2021b).

In short, overall and body fat distributions have a causal effect on DKD risk, and WC may exert greater effects. These results reveal that individuals can substantially reduce DKD risk for diabetes by reducing body fat mass and modifying body fat distribution.